Public Disclosure AuthorizedPublic Disclosure AuthorizedPublic Disclosure AuthorizedPublic Disclosu.
2025-02-14
61页
5星级
The economic impact of the University of Bristol THE ECONOMIC IMPACT OF THE UNIVERSITY OF BRISTOL AN.
2025-02-10
44页
5星级
1st Quarter 2025SVB Asset Management views on economic and market factors affecting global markets a.
2025-02-10
31页
5星级
The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile Prepared for CTIA 22 January 2025 Project Team Dr.Hector Lopez Julien Martin The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile Contents NERA Contents Executive Summary.i 1.The importance of the wireless industry to the American economy.1 2.The unique role of licensed mid-band spectrum.5 3.The economic impact of allocating mid-band spectrum to mobile.8 3.1.Continued improvements to mobile service to millions of Americans.9 3.2.Improving broadband with FWA.12 3.3.Supporting industries that rely on mobile connectivity.17 3.4.Supporting industries that serve the wireless industry.24 4.Allocating additional spectrum to Mobile vs Wi-Fi.28 5.Conclusion.32 The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile Executive Summary NERA i Executive Summary The wireless industry has become a cornerstone of the American economy,influencing nearly every aspect of daily life and business operations.As technology continues to advance,the demand for wireless communication continues to surge,with Americans consuming an astounding 100 billion gigabytes of data in the past year alone.This growing reliance on wireless networks underscores their role as critical infrastructure,essential for facilitating economic transactions,maintaining personal connections,and maintaining national security.The industrys significant contributions to economic output and job creation further highlight its importance.Wireless enables vast swathes of economic activity,both directly through investments in communication infrastructure and indirectly by enabling new services and improving worker productivity.Over the past decade,wireless has contributed over$5 trillion of GDP and 3 million jobs to the U.S.economy.Approximately 1.1 GHz of licensed spectrum below 6 GHz has supported this economic growth and employment.However,the wireless industry is rapidly approaching a spectrum deficit that will result in network congestion,thereby hindering the continued growth fueled by the wireless industry.Projections indicate that wireless operators will need at least 400 MHz of additional spectrum by 2027 to meet the needs of the U.S.economy,a deficit that will continue to grow to over 1400 MHz by 2032.Additional wireless spectrum is fundamental to so many aspects of the U.S.economy.In particular,this study focuses on the economic activity and consumer benefits generated by:continued improvements to mobile service to millions of Americans;improved fixed broadband coverage and penetration via fixed wireless access(FWA);support for industries that rely on mobile connectivity,such as video streaming and cutting-edge VR/AR;and support for industries that serve the wireless industry,such as construction and electronic maintenance.All this economic activity and wireless industry investment enabled by additional licensed spectrum will contribute significantly to the American economy.We estimate that each additional 100 MHz of mid-band spectrum to mobile will generate$264 billion of GDP,about 1.5 million new jobs,and about$388 billion in consumer surplus.The impact of 400 MHz of mid-band spectrum would be$1.1 trillion of GDP,6.18 million new jobs,and about$1.5 trillion in consumer surplus.Beneficial effects would continue to accumulate beyond 400 MHz,and we estimate that by 2028 even 400 MHz of new 5G spectrum will not be enough to keep up with consumer demand.The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile Executive Summary NERA ii Table 1:Summary of the economic impact of allocating each additional 100 MHz of mid-band spectrum to mobile GDP($B)Employment(M)Consumer Surplus($B)Continued improvements to mobile service to millions of Americans 385 Improving broadband with FWA 40 0.30 3 Supporting industries that rely on mobile connectivity 188 0.93 Supporting industries that serve the wireless industry 36 0.32 Total 264 1.55 388 Note:Effect of each 100 MHz up to 400 MHz The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The importance of the wireless industry to the American economy NERA 1 1.The importance of the wireless industry to the American economy The wireless industry is an integral part of the U.S.economy,with wireless communications deeply integrated into how we live and work.Americans consumed 100 billion GBs of data last year,and data traffic continues to grow.1 From enabling economic transactions to connecting families and defending the nation,wireless networks are not just a convenience,but critical infrastructure.By providing fast,efficient communication,the wireless industry enables vast swaths of economic activity.The core wireless industry,which includes mobile and wholesale network operators,contributes significantly to the American economy.Figure 1 shows the consistent investments wireless providers make in the U.S.s communications infrastructure.Since the launch of the first commercial cellular networks,wireless operators have invested over$700 billion in capital expenditures to build and deploy networks throughout the nation.2 In 2023 alone,the wireless industry invested$30 billion;and this decade,the industry has been the second-largest source of direct investment in the United States.3 These significant capital investments in infrastructure have allowed wireless providers to expand network coverage,improve service quality,and introduce advanced technologies such as 5G,all of which have translated into substantial economic output.Over the last decade,the core wireless industry generated$270 billion in gross output and$133 billion in GDP annually.4 Figure 1:Cumulative wireless industry capital expenditure in the United States,2011-2024 Source:CTIA Annual Wireless Industry Surveys 1 Timothy Tardiff,“Wireless Investment and Economic Benefits,”AACG(Apr.2024),available at https:/www.ctia.org/news/wireless-investment-and-economic-benefits 2 Ibid.3 Ibid.4 Compass Lexecon,2022,The Importance of Licensed Spectrum and Wireless Telecommunications to the American Economy,available at:https:/api.ctia.org/wp-content/uploads/2022/12/Compass-Lexecon-Licensed-Spectrum-Report.pdf The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The importance of the wireless industry to the American economy NERA 2 Beyond its direct economic contribution to output and GDP,the wireless industry also generates significant indirect benefits through its infrastructure deployment and wider supply chain,as it purchases goods and services like network equipment and software from other industries.And as the wireless industrys workers spend their incomes,they generate additional induced economic effects,multiplying the industrys impact across the economy.Once indirect and induced effects are taken into account,the core wireless industry is estimated to have driven$650 billion in gross output and$376 billion in GDP a year over the last decade.5 Figure 2 shows that the GDP contribution of the core wireless industry has been steadily growing at a compound annual growth rate of about 5%.By any account,the wireless industry has been a critical driver of the American economy.Figure 2:GDP contribution of the core wireless industry in the United States,2011-2020 Source:Adapted from Compass Lexecon Report Note:The core wireless industry includes network operators and MVNOs and does not include other downstream or upstream industries.The wireless industrys impact across the economy extends further still.As a general-purpose technology,wireless communication networks fuel the broader economy by enabling innovation and economic activity in other,downstream industries that rely and build upon the services they provide.For example,a recent study by Accenture estimates that 5G networks will contribute an additional$159 billion to the American manufacturing industrys GDP over a 5-year period.6 Compass Lexecon,meanwhile,find that in 2020 alone the wireless industry broadly contributed$825 billion in GDP;including the effects of the core mobile industry and the broad mobile ecosystem which includes social networking sites,mobile gaming,smartphone apps,search engines and digital advertising.7 Through network expansion and new services,most notably mobile Fixed Wireless Access(FWA),the wireless industry also expands broadband connectivity in under-served areas,creates new competition 5 Ibid.6 Accenture,available at https:/ 7 Compass Lexecon,2022 The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The importance of the wireless industry to the American economy NERA 3 to wired broadband,and promotes inclusive access to the economy.FWA has been pivotal in closing the digital divide and bringing Americans online.In 2022,90%of new home broadband connections were 5G FWA,and many of these connections were in areas that are underserved by fixed networks such as cable.8 Numerous studies have established the economic benefits associated with increasing broadband penetration.For example,The World Bank estimates that in developed economies,a 10-percentage point increase in broadband penetration boosts GDP growth by 1.2%.9 Other studies have found effects of similar magnitudes for developed economies,with concomitant benefits on wages and employment.By employing workers and enabling up and downstream economic activity,the wireless industry is also a significant driver of employment in the United States.Directly,wireless providers enable a large number of high-quality jobs,employing engineers,technicians,customer service representatives and administrative staff to run their operations.The ongoing expansion of 5G networks is expected to continue to create jobs well into the decade.Indirectly,the wireless industry also supports jobs upstream,as it fuels demand for goods and services in other sectors like manufacturing and software development.Over the last decade,the core wireless industry is estimated to have enabled nearly 2 million jobs per year.10 In the first half of this decade,meanwhile,5G is projected to create or transform up to 16 million jobs.11 Importantly,the wireless industry also plays a crucial role in closing the digital divide.Studies in the U.S.have found that improving the quality of broadband has a disproportionately positive impact on reducing unemployment in rural areas.According to one study by Lobo et al.(2020),unemployment rates are about 0.26 percentage points lower in counties with access to high quality broadband than in counties with lower quality services.12 This highlights the vital role that the wireless industry plays in levelling the playing field and promoting economic opportunity and inclusive access to the economy.In addition to its direct contribution to the economy,the wireless industry also generates substantial consumer welfare by connecting Americans and offering ever-improving services at lower costs.Today,Americans pay$0.006 per MB of data.13 This represents a 93crease from a decade ago.14 At the same time,the services offered to consumers are hugely improved.Average mobile broadband 8 CTIA,2024,CTIA Response to FCC Communications Market Report 2024,available at:https:/api.ctia.org/wp-content/uploads/2024/06/240606-FINAL-CTIA-Comments-for-2024-Communications-Marketplace-Report.pdf 9 https:/documents1.worldbank.org/curated/zh/178701467988875888/pdf/102955-WP-Box394845B-PUBLIC-WDR16-BP-Exploring-the-Relationship-between-Broadband-and-Economic-Growth-Minges.pdf 10 CTIA,2024,Annual Survey Highlights,available at:https:/api.ctia.org/wp-content/uploads/2024/09/2024-Annual-Survey-Highlights.pdf 11 Accenture,2021,The Impact of 5G on the United States Economy,available at:https:/ 12 Lobo,Alam and Whitacre,2020,Broadband speed and unemployment rates:Data and measurement issues,available at:https:/ 13 Cable.co.uk,2024,The cost of 1GB of mobile data in 237 countries,available at:https:/www.cable.co.uk/mobiles/worldwide-data-pricing/14 Nielsen,2011,Average U.S.Smartphone Data Usage Up 89%as Cost per MB Goes Down 46%,available at:https:/ Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The importance of the wireless industry to the American economy NERA 4 speeds in the U.S.are now over 100 Mbps a thirty-fold increase relative to the North American average of 3 Mbps at the start of the decade.15 And these benefits are widespread too:some 95%of U.S.adults say they use the internet and 15cess the internet solely via their smartphone.16 In 2015,the 645 MHz of spectrum licensed for wireless broadband networks was estimated to generate between$5 trillion and$10 trillion in savings for consumers.17 Given the improvements in service quality and declines in prices,savings should be higher today.Again,this underscores the critical role that the wireless industry plays in supporting Americas society and its economy.15 Cisco,2011,Cisco Global Cloud Index:Forecast and Methodology,20102015,available at:https:/ 16 Pew Research,2024,Internet,Broadband Fact Sheet,available at:https:/www.pewresearch.org/internet/fact-sheet/internet-broadband/17 The Brattle Group,2015,Mobile Broadband Spectrum:A Vital Resource for the U.S.Economy,available at:https:/ The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The unique role of licensed mid-band spectrum NERA 5 2.The unique role of licensed mid-band spectrum Radio frequency spectrum has been described as one of the“Nations most important national resources”,owing to its role supporting the wireless industry.18 The large sums invested by wireless operators on acquiring spectrum licenses demonstrates their vital importance to their networks.Figure 3 shows the cumulative revenues raised for the Treasury by wireless operators from spectrum licenses acquired in FCC auctions since 2014.Figure 3:Cumulative revenues raised for the Treasury in licensed spectrum by wireless operators since 199419 Source:FCC Different frequencies within the usable range of wireless spectrum have different characteristics and use cases:Low-band frequencies below 1 GHz are ideal for providing wide-area coverage owing to their long-range propagation,and also penetrating deep indoors,but they are in limited supply so cannot support significant capacity;Lower mid-band spectrum lies between 1.0 and 3.0 GHz.This set of frequencies comprises core mobile frequencies like the PCS and AWS bands,which have been an integral part of 18 https:/www.whitehouse.gov/briefing-room/presidential-actions/2023/11/13/memorandum-on-modernizing-united-states-spectrum-policy-and-establishing-a-national-spectrum-strategy/19 FCC,2022,Fiscal Year 2023 Budget Estimates to Congress,available at:https:/docs.fcc.gov/public/attachments/DOC-381693A1.pdf.Values for 2023 and 2024 are the same as 2022 as there have not been any auctions in those years.$54$95$95$114$114$117$129$210$233$233$233 20142015201620172018201920202021202220232024H-BlockAWS-3600 MHz28 GHz24 GHz3.5 GHzmmWave3.7 GHz2.5 GHz3.45 GHzThe Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The unique role of licensed mid-band spectrum NERA 6 wireless operators networks for a long time.These bands offer superior capacity to low-band frequencies;Mid-band spectrum,which comprises frequencies between 3.0 and 8.5 GHz,holds a unique position in wireless providers spectrum portfolios,offering a combination of high bandwidth for capacity and good propagation;and High-band spectrum above 10 GHz offers exceptionally large capacity,but signals only travel for short distances,making it best suited for use in dense urban areas and venues like stadiums.The C-Band,with frequencies around 3500 MHz,has emerged as the key 5G band worldwide.The outsized share of U.S network traffic now carried over this spectrum evidences its importance.For instance,Verizon,which invested heavily into acquiring C-Band licenses,reports that half of its network traffic is now carried by C-Band spectrum,and it expects this share will grow further.20 Mid-band spectrums critical importance is also reflected in its wide-spread deployment.Ericsson estimates that 85%of the population in North America is covered by mid-band spectrum only 5low the reported coverage of low-band spectrum.21 To maximize the economic benefits that flow from this high capacity and speed,it is essential that sufficient spectrum is available so that wireless network capacity can keep up with the growing demand for data and provide for new innovative services.Currently,380 MHz of mid-band is licensed for mobile in the U.S.22 While this bandwidth may currently be sufficient to meet data demand,recent studies estimate that the U.S.will need between 400 MHz and 2 GHz of additional spectrum to support future traffic growth.For example,Brattle estimates that the U.S.will need an additional 400 MHz by 2027 and 1.4 GHz by 2032.23 Similarly,the GSMA estimates that densely populated American cities,such as New York,will require between 1 and 2 GHz of additional mid-band spectrum by the end of the decade.24 Releasing additional mid-band spectrum suitable to power mobile connectivity is crucial if the U.S.is to retain its status as the global leader in wireless connectivity.Yet,the U.S.is at risk of falling behind,as other countries continue to award spectrum for mobile.China,for example,has already allocated 20 Verizon Communications Inc,2024,Q2 2024 Earnings Call,available at:https:/ 21 Ericsson,2023,5G Network Coverage Outlook,available at:https:/ does provide a definition for mid-band in this document 22 FCC,2022,2022 Communications Marketplace Report,available at:https:/docs.fcc.gov/public/attachments/FCC-22-103A1.pdf 23 Brattle 2023,How much licensed spectrum is needed to meet future demand for network capacity?24 Coleago,2021,Estimating the mid-band spectrum needs in the 2025-2030 time frame,available at:https:/ The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The unique role of licensed mid-band spectrum NERA 7 700 MHz of spectrum in the 6 GHz band to meet the demand for mobile data.25 As a result,China now leads the way in terms of spectrum allocations for mobile,and by 2027,is expected to have released up to1560 MHz of exclusive-use,high-power mid-band spectrum to commercial wireless operators.26 Meanwhile,the U.S.is expected to lag four of the G7 member countries by 2027 if no additional mobile mid-band spectrum is released.Figure 4:Mid-band spectrum allocated to mobile in different countries by 2027 Source:Adapted from Analysys Mason,2022,Comparison of total mobile spectrum in different markets.Notes:We have removed any spectrum that is not available on an exclusive use,full-power basis.25 CTIA,2023,China Commits to 5G Mid-Band Spectrum with 6 GHz Allocation:U.S.Needs Clear Response,available at:https:/www.ctia.org/news/china-commits-to-5g-mid-band-spectrum-with-6-ghz-allocation-u-s-needs-clear-response#:text=The 6 GHz band China,5G technology in the band 26 Analysys Mason,2022,Comparison of total mobile spectrum in different markets,available at:https:/api.ctia.org/wp-content/uploads/2022/09/Comparison-of-total-mobile-spectrum-28-09-22.pdf 1,560 800 600 450 390 390 380 326 300 ChinaJapanS.KoreaCanadaUKFranceUSItalyGermany1,180 MHzdeficit vs ChinaThe Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 8 3.The economic impact of allocating mid-band spectrum to mobile Allocating additional mid-band spectrum to mobile in the form of full-power flexible-use licenses will create value throughout the economy.Although our analysis may not capture all of the benefits over time,in this paper we focus on the impact on mobile consumers,FWA consumers,industries that rely on mobile connectivity,and industries that support the wireless industry:Continued improvements to mobile service to millions of Americans.FWA consumers will benefit from additional coverage,increased penetration,higher speeds,higher additional data consumption,and lower prices.Industries that rely on mobile connectivity,such as video streaming,mobile games,and cutting-edge virtual reality and AI,will enjoy a more robust platform to deliver their services.Industries that support the wireless industry will benefit from the additional capital expenditure(capex)required to deploy the spectrum and the additional operational expense(opex)required to maintain the network.We measure the economic impact of allocating each additional 100 MHz to mobile by estimating its impact on three metrics:gross domestic product(GDP),employment,and consumer surplus.GDP is a measure of value added to the economy,that is,the value of the gross output of an industry minus the value of the intermediate inputs required to produce the output.The additional GDP produced by the spectrum allocation is a measure of the value of the additional goods and services that can be consumed by final demand.Employment represents the number of additional one-year jobs.Consumer surplus is a measure of consumer benefit and is the difference between what consumers would be willing to pay and what they actually pay.For all metrics,we estimate the impact of allocating mid-band spectrum by comparing a situation with additional spectrum against a counterfactual in which the spectrum is not allocated.Table 2 shows the metrics estimated for each channel affected by the spectrum allocation and deployment.The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 9 Table 2:The sources of the economic value of allocating mid-band spectrum to mobile Channel GDP Employment Consumer Surplus Better mobile service at no additional cost *Improving broadband with FWA*Supporting industries that rely on mobile connectivity*Supporting industries that serve the wireless industry*We estimate the effect of each additional 100 MHz by first estimating the impact of 400 MHz and then dividing by four.Therefore,throughout the paper,we report the marginal impact of adding 400 MHz of spectrum and the average effect of 100 MHz(up to 400 MHz).Conceptually,we would expect the first 100 MHz to have a larger effect than the last 100 MHz,but we do not estimate the impact of 100 MHz blocks individually.However,the sum of the first four 100 MHz blocks would equal the effects reported in this paper.Beneficial effects would continue to accumulate beyond 400 MHz,and we estimate that by 2028 even 400 MHz of new 5G spectrum will not be enough to keep up with consumer demand.However,estimating the economic benefits beyond 400 MHz is outside this papers scope.3.1.Continued improvements to mobile service to millions of Americans Mobile consumers will be the prime beneficiaries of the additional spectrum.Wireless operators will deploy and operate the spectrum to meet future growth in mobile data consumption.Figure 5 presents historical data on mobile data consumption from 2010 to 2023,measured in GB per month.The graph reveals significant growth in mobile data usage,reaching an average of 24 GB per month in the United States in 2023.The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 10 Figure 5:Historical mobile data traffic per capita in the U.S.Source:TeleGeography.Note:Includes FWA.For example,15 years ago,data-intensive applications such as video streaming and video calls on mobile were not commonplace,whereas nowadays,they represent a typical consumer experience.For the most part,consumers have not had to pay additional dollars for the ever-increasing capabilities of mobile networks.One way of estimating the value to consumers of the increased data consumption possible with the additional mid-band spectrum is to estimate the consumer surplus produced by the spectrum.The consumer surplus is the difference between the value consumers would be willing to pay and what they actually pay.Without additional spectrum,networks would eventually become congested.Plans tiered by consumption would likely come back,and those consumers wishing to add additional data would pay more.This difference,what consumers would be willing to pay without additional spectrum and what they pay with spectrum,is the consumer surplus produced by the spectrum.To estimate the impact of the spectrum on consumer surplus,we use previous research that links the price of the spectrum and consumer surplus.Conceptually,the price is inherently linked to the consumer surplus because both are based on comparing a future with and without additional spectrum.Table 3 shows the results of previous research linking spectrum prices and consumer surplus.Specifically,these papers show that the spectrum produces between 0.9 and 1.35 annual dollars of consumer surplus for every dollar of auction price.Table 3:Consumer surplus to price multipliers Paper Consumer Surplus to Price Multiplier Hazlett&Munoz 2004a 0.9 Hazlett&Munoz 2009b 1 Rosston 2003c 1.35 Source:(a)Hazlett and Munoz,2004,A Welfare Analysis of Spectrum Allocation Policies,Joint Center:AEI-Brookings Joint Center for Regulatory Studies.(b)Hazlett and Munoz,2009,A welfare analysis of spectrum allocation policies.RAND Journal of Economics Vol.40 No.3:424-454.05101520253020102011201220132014201520162017201820192020202120222023GB/monthThe Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 11(c)Rosston,2003,The long and winding road:the FCC paves the path with good intentions.Telecommunications Policy 27:501515.We use the two most recent auctions to estimate the price of 400 MHz of mid-band spectrum:the C-Band(3.7 GHz)and the 3.45 GHz auctions.The C-Band achieved a total price of$94 billion for 280 MHz,and the 3.45 GHz achieved a total price of$22 billion for 100 MHz.27 Table 4 shows the prices paid at the time of the auctions.The prices paid are not directly comparable because the volume of MHz available varied,and the C-Band prices are based on the 2010 census population and the 3.45 GHz on the 2020 census population.To address this,we calculate the price per MHz pop in 2025 US dollars by(a)using inflation to adjust the prices paid to 2025 prices28;and(b)applying a 2025 population projection to obtain 2025 prices per MHz-Pop.29 Table 4:Recent mid-band spectrum prices Total price paid MHz Price paid$MHz-Pop 2025 Price$MHz-Pop C-Band 94.17 280 1.10 1.17 3.45 GHz 22.51 100 0.68 0.79 MHz-Pop Weighted Average 1.07 Source:NERA Economic Consulting Combining the consumer multipliers and the average price for the mid-band spectrum,we obtain an implied annual consumer surplus of between$128 billion and$192 billion.We use three different discount rates typically used to discount spectrum consumer surplus to calculate the cumulative impact of allocating the spectrum.30 Table 5 shows the present value for discount rates of 5%,7.5%,and 10%using the multipliers identified in the three research papers on this subject.27 The C-band price includes gross proceeds,accelerated relocation payments to satellite companies,and relocation payments.28 IMF Data Portal,Inflation rate,average consumer prices(Annual percent change),last retrieved November,2024,available at:https:/data.imf.org/?sk=4FFB52B2-3653-409A-B471-D47B46D904B5&sId=1485878855236 29 The 2025 sticker price per MHz-Pop is weighted based on each Partial Economic Area(PEA)population.To estimate the population for each PEA in 2025,we first calculated the population for each county in 2025 using data from the U.S.Census Bureau,available here:https:/www.census.gov/data/tables/time-series/demo/popest/2020s-counties-total.html.We then aggregated the county populations to determine the total population for each PEA.30 See,for example,Bazelon and McHenry,2015,Mobile Broadband Spectrum:A Vital Resource for the U.S.Economy.Available at:https:/api.ctia.org/docs/default-source/default-document-library/brattle_spectrum_051115.pdf The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 12 Table 5:Present value of the consumer surplus Surplus to price multiplier Implied annual consumer surplus($B)PV 5%($B)PV 7.5%($B)PV 10%($B)Hazlet&Munoz 2004a 0.9 128 2,561 1,707 1,280 Hazlet&Munoz 2009b 1 142 2,845 1,897 1,423 Rosston 2003c 1.35 192 3,841 2,561 1,920 Source:Ibid and NERA Economic Consulting Given the high uncertainty in future spectrum prices in this band,we use the 10%discount rate to be conservative in our estimation.The future mid-band spectrum may be allocated at lower prices than the C-Band or 3.45 GHz since 400 MHz would more than double the stock of the mid-band spectrum.However,the alternative could be true,and the new spectrum could be as valuable,or even more valuable,given mounting capacity pressures and operators desire to expand FWA and other services.We use the most recent auction data as we do not estimate future spectrum prices in this paper.Table 6 shows our selected consumer surplus.We show the marginal impact of adding 400 MHz and the average effect for each piece of 100 MHz.Table 6:Consumer surplus associated with a better mobile service at no additional cost 400 MHz 100 MHz Concept Consumer Surplus($B)Consumer Surplus($B)Hazlet&Munoz 2004a 1,280 320 Hazlet&Munoz 2009b 1,423 356 Rosston 2003c 1,920 480 Average 1,541 385 Source:Ibid.3.2.Improving broadband with FWA 5G Fixed Wireless Access(FWA)is the fastest-growing terrestrial broadband technology.31 In 2024,FWA accounted for nearly all of the net broadband additions and one of the largest terrestrial footprints.32 Figure 6 shows the FWA adoption in the U.S.Starting around the fourth quarter of 2020,5G FWA has accounted for the large majority of net adds in the fixed broadband market.This 31 Singer and Urschel,2023,Competitive Effects of Fixed Wireless Access on Wireline Broadband Technologies.Available at:CTIA-Competitive Effects of Fixed Wireless Access on Wireline Broadband Technologies 32 Opensignal,5G Fixed Wireless Access(FWA)Success in the US:A Roadmap for Broadband Success Elsewhere?,Available at:5G Fixed Wireless Access(FWA)Success in the US:A Roadmap for Broadband Success Elsewhere?|Opensignal The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 13 tremendous growth has been achieved despite the capacity limitations faced by network operators.For example,T-Mobile reports a waiting list of over 1 million to become fixed wireless customers.33 Figure 6:FWA adoption growth in the U.S.Source:Opensignal34 This expansion has generated great consumer benefits through lower prices and additional broadband coverage.35 Additional mid-band spectrum for mobile will further improve FWAs capacity and economics.In particular,adding 400 MHz of mid-band to the existing 380 MHz of licensed spectrum will essentially double the mid-band capacity available for FWA.The additional capacity will increase the benefits of competition and penetration.In terms of competition,Singer and Urschel estimate that the consumer benefits associated with more choice and price competition in the fixed market owing to the availability of mobile FWA are around$6 billion annually.Table 7 shows the breakdown of benefits by market type and benefit channel.At current prices,some consumers switch from cable or fiber to FWA,whereas others choose to stay with their existing technologies.The switchers benefit from an improved match between the service they need and the price they pay.Those who decide to stay benefit from the lower prices offered by cable and fiber providers in response to the FWA providers offerings.33 Fierce Network,2024.The 1 million people on T-Mobiles fixed wireless waiting list will get a little help from fiber,Available at:https:/www.fierce- 34 Opensignal,5G Fixed Wireless Access(FWA)Success in the US:A Roadmap for Broadband Success Elsewhere?,Available at:5G Fixed Wireless Access(FWA)Success in the US:A Roadmap for Broadband Success Elsewhere?|Opensignal 35 https:/www.ctia.org/news/fcc-shows-how-wireless-is-delivering-much-needed-home-broadband-competition-closing-the-digital-divide The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 14 Table 7:Yearly consumer benefits of FWA in markets with existing service Market type Benefits from switching per annum($M)Price savings owing to competition($M)Total($M)Cable 369 5,735 6,104 Cable/Fiber 27 219 246 Total 396 5,954 6,350 Source:Singer and Urschel(2023)A critical assumption in Singer and Urschels research is that FWA providers obtain all the spectrum they need to compete effectively with other terrestrial technologies.They do not estimate the spectrum requirements to materialize the benefits they estimate.However,they mention that GSMA estimates that an additional 2 GHz of mid-band spectrum is needed to sustain FWA delivering a download data rate of 100 Mbps in rural communities in the longer term.36 Based on these estimations,we attribute 20%of the total benefits to the additional 400 MHz in consideration in this paper.In addition,we estimate the preset value of these benefits based on a discount rate between 5%and 10%.Our calculations are presented in Table 8.Table 8:Present value of the consumer benefits associated with 400 MHz of additional mid-band spectrum Market type Total($M)Benefit of 400 MHz($M)PV 5%($M)PV 7.5%($M)PV 10%($M)Cable 6,104 1,221 24,416 16,277 12,208 Cable/Fiber 246 49 982 655 491 Total 6,350 1,270 25,398 16,932 12,699 Source:NERA Economic Consulting We estimate the benefits of additional penetration in three steps.First,we identify the impact of the additional spectrum on broadband coverage.Second,we estimate the impact of the marginal coverage on national broadband penetration.Finally,we use previous research to identify the impact of increases in penetration on GDP and employment.According to the latest national broadband map released by the FCC,96.2%of the country is covered by terrestrial technologies.37 Terrestrial technologies include cable,fiber,FWA,and others.Figure 7 shows terrestrial coverage by county density decile.We ordered all counties by residential unit density and created buckets containing 10%of the countrys total residential units.We show the coverage of 36 Singer and Urschel,2023,Competitive Effects of Fixed Wireless Access on Wireline Broadband Technologies.Available at:CTIA-Competitive Effects of Fixed Wireless Access on Wireline Broadband Technologies 37 https:/broadbandmap.fcc.gov/data-download.November 13 2024.The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 15 terrestrial technologies for each bucket.The data shows that more densely populated counties have greater coverage.Figure 7:Terrestrial broadband penetration by density decile Source:NERA analysis of FCC data Note:Each column represents 10%of the residential locations in the U.S.The first column represents the least-dense 10%,and the last column the most-dense 10%.We first focus on the impact of the spectrum on coverage.In our estimation,we consider a location covered by FWA if the base stations that can serve the location have enough capacity to provide the service.Therefore,additional spectrum increases coverage by expanding the capacity of existing base stations and increasing the profitability of new ones.We use county unit density to estimate the impact of additional spectrum on additional coverage.In particular,we assume that with an additional 400 MHz of mid-band spectrum,FWA coverage in a given country would be similar to todays coverage of a county with 2.05 times its density an increase proportional to the spectrum holdings.Previous research by the GSMA has found that the number of units that can be served from a single base station is proportional to the spectrum holding.In particular,a base station can support 90 users with 400 MHz,315 with 1.4 GHz,and 540 with 2.4 GHz.38 We estimate that adding 400 MHz of mid-band spectrum will increase the total number of residential units covered by 1.1 million,or 0.7%of the total residential units.Figure 8 shows the increase by county decile.Based on our estimation,the increase will be more pronounced in the sparser counties where terrestrial deployment with other technologies is more expensive.Our calculation also assumes a modest increase in coverage in some of the top 10%more densely populated counties.While these counties enjoy a relatively high terrestrial coverage,FWA providers still have the potential to increase coverage if they can secure the capacity needed to serve their customers.38 GSMA and Coleago Consulting,2021.Estimating the mid-band spectrum needs in the 2025-2030 time frame(Global Outlook),available at:Estimating-Mid-Band-Spectrum-Needs.pdf.The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 16 Figure 8:Increase in coverage associated with an additional 400 MHz of mid-band spectrum Source:NERA Economic Consulting To estimate the impact on penetration,we assume that other terrestrial technologies eventually would cover the same residential units by 2035.That is,the additional spectrum would create an initial increase in penetration but gradually fade out when compared to a counterfactual in which other terrestrial technologies would eventually reach the same coverage and penetration.We assume that the additional penetration caused by the spectrum would reach a peak of 50%of the newly covered residential units and slowly fade out by 2035.Figure 9 shows the increased broadband penetration caused by the additional spectrum.Figure 9:Increase in penetration associated with an additional 400 MHz of mid-band spectrum Source:NERA Economic Consulting Finally,we use employment and GDP forecasts and results from the literature to estimate the impact of the increased penetration on GDP and employment.The Bureau of Economic Analysis estimated that GDP was 27.36 trillion in 2023.39 We use a constant 2%annual growth rate to project GDP until 39 Source:Bureau of Economic Analysis,2024,Gross Domestic Product(Second Estimate),available at:https:/www.bea.gov/sites/default/files/2024-11/gdp3q24-2nd.pdf The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 17 2035.Regarding jobs,the Bureau of Labor Statistics projects that total jobs will grow from 169.9 million in 2023 to 176.6 million in 2033.40 We complement these forecasts with literature estimations.In particular,the ITU has estimated that a 1%increase in penetration increases GDP by 0.1856%.41 Similarly,Crandall et al.2007 estimated that a 1%increase in penetration increases jobs by between 0.2%and 0.3%.42 Our estimated economic impact is shown in Table 9.Table 9:Economic impact of increased FWA penetration associated with 400 MHz 2025 2026 2027 2028 2029 2030 2031 2032 2033 2034 2035 GDP($B)12.1 24.3 24.3 21.7 19.0 16.3 13.6 10.9 8.2 5.5 2.7 Employment(000)96.3 190.0 187.2 164.1 141.5 119.6 98.2 77.4 57.3 37.6 18.5 Source:NERA Economic Consulting 3.3.Supporting industries that rely on mobile connectivity The wireless industry provides essential services that support a wide range of other industries and economic activity.Originally,the wireless industry supported basic communications via text messaging and voice calls.Today,it enables much richer forms of interaction,allowing people to share not only messages and voice calls,but also photos,videos,and other forms of content through social media platforms.This has given rise to companies like Meta and Snapchat,which collectively generated$139 billion in output in the U.S.in 2024 and account for 35%of total mobile network traffic.Likewise,the motion picture and sound recording industry has been transformed by allowing people to consume content on their mobile devices,wherever they are.Video and audio streaming generated 32%of network traffic in 2024 and services like Netflix and Spotify generated$70 billion in output in the U.S.While social media and content streaming account for the majority of mobile network traffic and generate billions in output,the impact of the wireless industry is far broader.Search engines like Google and Bing have seen significant increases in usage owing to the availability of wireless services.Google estimates that 63%of its organic search traffic in 2023 originated from mobile devices,highlighting the key role that wireless networks play in enabling access to information and e-commerce.Today,over three-quarters of U.S.adults report buying things online using a smartphone.43 40 Source:Bureau of Labor Statistics,2024,EMPLOYMENT PROJECTIONS 20232033,available at:https:/www.bls.gov/news.release/pdf/ecopro.pdf 41 Source:ITU,2021,The economic impact of broadband and digitalization through the COVID-19 pandemic,available at:https:/www.itu.int/dms_pub/itu-d/opb/pref/D-PREF-EF.COV_ECO_IMPACT_B-2021-PDF-E.pdf 42 Crandall et al.,2007,The Effects of Broadband Deployment on Output and Employment:A Cross-sectional Analysis of U.S.Data,Issues in Economic Policy,The Brookings Institution.43 Pew Research Center,2022,For shopping,phones are common and influencers have become a factor,available at:https:/www.pewresearch.org/short-reads/2022/11/21/for-shopping-phones-are-common-and-influencers-have-become-a-factor-especially-for-young-adults/The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 18 The mobile gaming industry,too,generates vast amounts of output and leverages the fast connection speeds and low latencies available on modern wireless networks to allow users to play games on their mobile devices.The file-sharing industry,meanwhile,has taken advantage of wireless networks to allow secure,mobile access to files that would traditionally only have been accessible on a home or office fixed network.In doing so,the file sharing industry accounts for 7%of mobile network traffic.In Table 10 below,we outline the output and mobile network traffic generated by the five key uses for mobile networks discussed above.We focus on these use cases as they account for most of mobile network traffic today and significantly contribute to the American economy.However,we note that these use cases do not account for all network traffic nor all of the economic activity attributable to wireless networks.Nor can they account for new uses of the network that could develop when innovators are assured of the capacities necessary to support those uses.Additionally,5G and future wireless technologies are increasingly serving as the foundation for enterprise connectivity innovations,and the economic benefits of those should expand over time assuming sufficient spectrum is made available to account for those use cases.Therefore,the figures in this section should be interpreted as a conservative estimate of the total impact on downstream industries that an additional mobile allocation of 400 MHz of mid-band spectrum would have on the American economy.Table 10:Output and network traffic generated by selected use cases for wireless connectivity Social media Video and audio streaming Device and cloud gaming General web apps File sharing Share of total mobile network traffic 352%7%5%7%Output generated in 2024($B)139.0 73.0 55.4 287.5 7.3 Share of traffic taking place on mobile 18%8%7%8%Industry output attributable to mobile($B)24.6 5.9 5.4 21.5 0.6 Output CAGR 16.7%6.1%3.4.4%0.4%Source:Network usage data from Sandvine Global Internet Phenomena Report 202444;industry output data from IbisWorld Market Research Reports45.Output attributable to mobile is equal to the produce of industry output share of traffic taking place on mobile networks.The output for general web apps is assumed to be from search only.This economic activity generates an ever-increasing demand for data as it flows through wireless networks.However,the demand for data is not evenly distributed throughout the day,with peak usage generally occurring in the evening,when people consume the most data as they stream content,engage with social media,or browse the internet.Figure 10 displays the amount of video 44 Sandvine,2024,The Global Internet Phenomena Report,available at:https:/ 45 IbisWorld,2024,Market Research Reports,available at:https:/ Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 19 streaming traffic taking place over mobile networks by hour.Peak usage occurs between 5 and 6 p.m.,whereas usage is significantly lower between 1 and 5 a.m.,when most people are sleeping.Data traffic demand during the busiest hours places wireless networks under significant strain,reducing customer service quality.The reduction in speeds,higher latencies,or interrupted connectivity affects wireless subscribers and hampers the ability of industries that rely on this connectivity to drive their businesses.For example,if a consumer is unable to watch a streaming video owing to buffering as a result of network strain,they will inevitably churn away from the streaming platform to do something else.Consequently,the streaming platform loses out on ad revenue that it would have otherwise obtained had the network enabled the user to stream content smoothly.Figure 10:Video streaming usage on mobile networks by hour Source:Adapted from Sandvine Mobile Internet Phenomena Report 2021 Today,network congestion only affects a relatively low proportion of total network traffic and only constrains economic activity at the busiest hours.However,the demand for data continues to grow as the quality of content streamed over the internet improves,online games require more bandwidth and more commerce takes place over the internet.As use cases continue to demand more data,the strain placed on networks will intensify,not just at the busiest hour but starting to affect more hours throughout the day.Without additional spectrum,mobile networks will become more congested across wider swathes of the day,a problem that may only be partially alleviated by expensive investments in densifying networks.This lost traffic will directly translate into lost sales and foregone revenue across the many industries that depend on reliable wireless communications.To estimate the economic impact that an additional allocation of 400 MHz of spectrum to mobile would have,we estimate the value of the output that would be foregone between 2025 and 2040 if no additional spectrum was allocated to mobile.Volume of mobile traffic that would be foregone without an additional 400 MHz of spectrum The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 20 We begin by projecting mobile data traffic from 2023 to 2040 by applying a 16GR to the current mobile traffic volume.This is the rate at which Ericsson forecasts mobile data traffic will grow until the end of the decade.46 We then forecast the unconstrained,desired network usage for each hour of the day using the estimated share of traffic occurring at each hour from Sandvine hourly network usage estimates.Next,we estimate the volume of traffic that could be served at each hour under two spectrum allocation scenarios:no additional spectrum is allocated to mobile;and an additional 400 MHz of spectrum is allocated to mobile.Under the first scenario,we assume that without additional spectrum,mobile data network capacity at a given hour cannot exceed present-day traffic at the busiest hour to forecast the volume of traffic that can be served in each hour given spectrum constraints.For the second scenario,we assume that the maximum hourly traffic capacity that can be carried is proportional to the quantity of spectrum that is allocated to mobile.An additional 400 MHz of spectrum would therefore increase the capacity of the network at the busy hour by 36%relative to the first scenario.47 We assume that the pace of network densification is identical in each of the two scenarios.In reality,network operators are likely to build cell sites and densify their networks at a faster rate under the first scenario to minimally compensate for a shortage of spectrum.However,the rate in data growth exceeds the pace at which operators densify their networks to such a large degree that this simplifying assumption has little impact on results.Between 2013 and 2023,the number of cell sites deployed in the U.S.grew at an average of 3.6%a year.48 Over that same period,wireless data grew at a rate of 42%per year,from 3 trillion MB in 2013 to over 100 trillion MB in 2023.49 The difference in traffic between these two scenarios allows us to identify the volume of traffic that would be foregone were no additional spectrum to be allocated to mobile.We note that because our model examines the difference between the two traffic-constrained scenarios,our estimated economic impact is robust to the baseline level of the total unconstrained traffic used and the growth rates assumed over the projected period.To see this,observe that in Figure 11,the rapid growth in the unconstrained traffic(in yellow)does not lead to unbound growth in traffic differences between the two constrained traffic scenarios(in blue and purple).Moreover,the more aggressive the traffic growth assumption,the smaller the estimated economic impact of an additional 400 MHz allocation because network congestion becomes severe early on,and the marginal spectrum alleviates a 46 Ericsson,Mobile data traffic outlook,available at:https:/ 47 There is currently 1123 MHz of spectrum allocated to mobile in the U.S.An additional 400 MHz of spectrum represents an increase in the allocation of 36%.48 CTIA,2022,Summary of CTIAs Annual Wireless Industry Survey,available at:https:/api.ctia.org/wp-content/uploads/2022/09/Summary-of-CTIAs-Wireless-Industry-Survey-2022.pdf;and CTIA,2024,2024 Annual Survey Highlights,available at:https:/www.ctia.org/news/2024-annual-survey-highlights#:text=By the end of 2023,reforms were enacted in 2018.49 Ibid.The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 21 relatively small proportion of the lost traffic.This counsels for a pipeline of additional spectrum with longer foresight,accounting for spectrum needs beyond these initial 400 MHz benefits.Figure 11:Mobile data traffic with and without additional 400 MHz of spectrum Source:NERA analysis based on Ericsson Mobility Visualizer traffic data for North America Traffic associated with mobile network use cases To attribute total mobile traffic across the five use cases identified in Table 10 we multiply total network traffic forecasts by the share of mobile traffic accounted for by each application.We assume that the relative proportion of mobile traffic generated by each application remains constant through to 2040.Output associated with mobile network use cases We forecast the output of each of the five use cases listed in Table 10 until 2040 using CAGRs obtained from IbisWorld Market Research Reports.We then estimate the share of this output that is attributable to mobile under the assumption that output is proportional to traffic.We multiply total output by the share of traffic associated with that application that takes place over mobile.For many applications,output is closely related to the volume of traffic.For example,the advertising revenue of a social media platform is closely linked to the traffic that platform serves.Value of additional traffic served with additional 400 MHz We recognize that an applications output will not necessarily have a linear relationship with the traffic it generates.For instance,some of the additional bandwidth consumed by say,Instagram,will be associated with serving higher quality video content rather than being generated by new users.And while higher definition content will attract more users and improve user retention,this effect will likely The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 22 exhibit diminishing marginal returns.To account for this,we estimate each applications marginal output that is generated with each additional exabyte(EB)of data per year.This means for social media applications,for example,each additional EB of data generates about 30%less revenue in 2030 than it does in 2025.We then obtain the marginal output that would be generated by each application with an additional 400 MHz allocation to mobile by multiplying the extra traffic generated by that application with additional spectrum by the marginal output generated by each incremental EB of data.Obtaining estimated GDP and employment impacts We use input-output multipliers to estimate the total impact of the additional direct output generated by each application on GDP and employment.We use type II multipliers to account for the direct,indirect,and induced effects of the additional direct output.We identify the input-output multipliers for each application based on previous research by Compass Lexecon as described in the table below.50 We use the latest data from the Compass Lexecon report,which examines the period 2011-2020.To obtain the GDP multiplier we divide total GDP generated by the application by its direct output.We benchmark the implied GDP multipliers against those in the Regional Input-Output Modeling System(RIMS)dataset produced by the Bureau of Economic Analysis(BEA).We observe that the multipliers in Table 11 are slightly higher than the RIMS multipliers for industries in the Information sector.However,we note that this is to be expected given these industries deep integration within the economy and substantial spillover effects,which are not captured in RIMS multipliers owing to their regional nature.To obtain the employment multiplier we divide total jobs enabled by the application by its direct output.We benchmark the implied employment multipliers against those in the RIMS dataset as well as employment multipliers produced by the Economic Policy Institute.We observe that the multipliers in Table 11 for device and cloud gaming,general web apps and file sharing are consistent with the multipliers published by BEA and the Economic Policy Institute.However,the employment multiplier we obtain from the Compass Lexecon paper for social media is 18.7.The figure is higher than expected,and we therefore use a more conservative multiplier of 9.3 from the Economic Policy Institute.51 50 Compass Lexecon,2022 51 Economic Policy Institute,2019,Updated employment multipliers for the U.S.economy,available at:https:/www.epi.org/publication/updated-employment-multipliers-for-the-u-s-economy/.The value of 9.3 is the total employment multiplier associated with the software publishers industry.The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 23 Table 11:Input-output multipliers associated with each application Application GDP multiplier Employment multiplier Social media 1.82 9.3 Video and audio streaming 1.82 9.3 Device and cloud gaming 0.51 8.5 General web apps 2.00 8.2 File sharing 2.00 8.2 Source:Compass Lexecon(2022)and Economic Policy Institute(2019).For video and audio streaming,we use the same multipliers as for the social media industry.For general web apps and file sharing,we use the same multipliers as for search engines.The employment multiplier is scaled to produce jobs per$1M in direct output.Economic impact Table 12 shows the total impact on U.S.GDP and employment of an additional 400 MHz by selected applications between 2025 and 2040.Over this period,we estimate that the social media,video and audio streaming,mobile gaming,general web apps and file sharing industries will generate around 750 billion dollars in GDP.At the same time,these industries will generate close to 4 million jobs.The allocation of additional mid-band spectrum for mobile may give rise to new applications and business models that are not feasible today owing to network constraints.In our calculation,we only estimate the impact of network constraints on foregone revenue from existing applications and businesses.With more bandwidth and faster speeds,new applications like VR gaming or improved telehealth services may emerge,driving the creation of further output,GDP,and jobs.In particular,AI is expected to greatly accelerate data growth.AI mobile data is expected to grow at a 55GR between 2023 and 2033,increasing the need and value of the spectrum.AI traffic is expected to be bursty and unpredictable,require low latency,and increase the demand for the uplink(for applications using cloud computing).52 Table 12:Total economic impact generated by an additional 400 MHz for selected applications,2025-2040 Application GDP($B)Employment(M)Social media 342 1.7 Video and audio streaming 74 0.4 Device and cloud gaming 18 0.3 General web apps*311 1.3 File sharing 8 0.0*Total 753 3.7 Source:NERA Economic Consulting.*General web apps includes 5%traffic share plus residual traffic share of 14%*0.031 52 Harri Holma,The AI revolution:Preparing for a surge in 5G uplink traffic The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 24 3.4.Supporting industries that serve the wireless industry The wireless industry will spend billions of dollars deploying and operating 400 MHz of additional mid-band spectrum increasing the demand for equipment,construction,power,and other industries that serve the wireless industry.To estimate the economic impact of the additional capex and opex associated with the deployment and operation of the spectrum,we follow two steps:1.We estimate the additional capex and opex required to deploy and operate 400 MHz of mid-band spectrum nationwide.2.We use an input-output model and data from the Bureau of Economic Analysis to estimate the economic impact of the marginal capex and opex on GDP and employment.Capex and opex required to deploy and operate 400 MHz of mid-band We estimate that the capex required to deploy 400 MHz of mid-band spectrum is approximately$35 billion over seven years.The capex will be used to install new equipment and infrastructure,requiring additional expenses to operate and maintain.We estimate the associated additional operating expenses will be around$9 billion annually.We estimate the required capex to deploy 400 MHz of mid-band spectrum based on the capex spent deploying 280 MHz of C-Band spectrum(3.7 GHz).We estimate the relevant opex based on previous studies establishing that wireless operators typically spend 25%of the capex as yearly opex.53 Based on previous studies,we project that deployment would take about seven years and that equipment would be operated for ten years.54 55 Based on company filings,we estimate that deploying 280 MHz of C-band required about$20 billion between 2021 and 2024.AT&T has reported spending$7 billion in deploying C-Band and Verizon has reported$10 billion.56 57 Based on these figures,we estimate that the total capital expenditure required to deploy 280 MHz of C-Band was about$20 billion.Table 13 shows the MHz-Pop weighted holdings for each operator and its associated capex.53 GSMA and Coleago Consulting,2021.Estimating the mid-band spectrum needs in the 2025-2030 time frame(Global Outlook),available at:Estimating-Mid-Band-Spectrum-Needs.pdf 54 Prieger,2020.An Economic Analysis of 5G Wireless Deployment:Impact on the U.S.And Local Economies,available at:Microsoft Word-ACT Report-An Economic Analysis of 5G(Feb 2020).docx 55 Sosa and Rafert,2019.The Economic Impacts of Reallocating Mid-Band Spectrum to 5G in the United States,available at:The Economic Impacts of Reallocating Mid-Band Spectrum to 5G in the United States 56 T,2021,AT&T to spend less than Verizon on C-band 5G rollout,available at:https:/ 57 T,2023.Verizon confirms climb-down from C-band capex peak,available at:Verizon confirms climb-down from C-band capex peak The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 25 Table 13:Implied capex associated with deploying 280 MHz of C-Band Operator MHz Disclosed Capex($B)Average Capex per 100 MHz Implied CAPEX AT&T 79.8 7 8.77 7.00 T-Mobile 27.4 1.94 UScellular 4.9 0.35 VZW 160.7 10 6.22 10.00 Other 7.1 0.50 Total 280 7.07 19.79 Source:NERA Economic Consulting based on public filings Based on the C-Band capital expenditures and previous studies that have established the timeline and capex,we estimate the marginal opex and capex associated with deploying the spectrum.Table 14 shows the deployment schedule.For purposes of our analysis,we have assumed that spectrum becomes available in four tranches of 100 MHz and that total deployment will take seven years.We believe this is a reasonable assumption given past trends,but different bands have different timelines,and buildout would vary depending on how much spectrum is made available at what time.We assume ten years of opex are triggered any time capex is spent.We adjust the 2021 C-Band reference numbers by inflation with respect to their initial deployment years but make no additional adjustments on projected capex or opex.Table 14:Estimated capex and opex associated with four tranches of 100 MHz of mid-band spectrum Year of initial deployment Total capex($B)Yearly opex($B)2025 8.57 2.142 2026 8.73 2.182 2027 8.91 2.228 2028 9.10 2.275 Total 35.31 8.83 Source:NERA Economic Consulting There is uncertainty regarding the quantity and availability dates for mid-band spectrum,and the scenario presented here is a hypothetical.If quantities change,our calculations will scale down proportionally for blocks of 100 MHz up to 400 MHz.Spectrum capex and opex depend more on the number of carriers than the bandwidth.In mid-band spectrum,the maximum current carrier is 100 MHz which means that our calculations are proportional in blocks of 100 MHz but would not be proportional for smaller increments.If dates change,inflation-adjusted figures will change in proportion to the delay or acceleration with respect to the base schedule.However,inflation has a The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 26 minor impact on these calculations,provided that projected inflation for the relevant period is about 2.1%per year.58 I-O GDP and employment multipliers We use input-output multipliers to estimate the impact of the marginal capex and opex on industries that serve the wireless industry.First,we identify the industries where the marginal capex and opex are spent.Second,we distribute the total annual expenditure to each industry.Finally,we identify the multipliers associated with those industries to compute the effect of the marginal expenditure.We identify the industries and their corresponding weights based on previous research.For capex,we use the industries and weights identified by Sosa and Rafert(2019).59 For opex,we obtained the weights from an Analysis Mason paper identifying the destination of opex,and we matched their categories with NAICS industries.60 We obtained the multipliers for each industry from the Bureau of Economic Analysis.61 We use type II multipliers to account for the direct,indirect,and induced effects of the expenditures on GDP and employment.Table 15 shows the industries affected by the marginal capex and opex,their GDP and employment multipliers,and their weight in each category.Table 15:Multipliers associated with the additional capex and opex Industry Gross Domestic Product1 Employment1 CAPEX weight2 Opex weight3 334210 Telephone apparatus manufacturing 1.03 7.1 1534220 Broadcast and wireless communications equipment 1.11 7.4 4735920 Communication and energy wire and cable manufacturing 0.80 6.3 10#3240 Power and communication structures 1.17 9.8 29D1100 Electric power generation,transmission,and distribution 1.02 4.7 81200 Electronic and precision equipment repair and maintenance 1.31 14.9 38%Total Capex 1.10 8.03 Total Opex 1.21 11.12 58 IMF,2024,World Economic Outlook,October 2024 59 Sosa and Rafert,2019.The Economic Impacts of Reallocating Mid-Band Spectrum to 5G in the United States,available at:The Economic Impacts of Reallocating Mid-Band Spectrum to 5G in the United States 60 Anil Rao,2020.Network automation is key to delivering significant opex reduction and increasing agility in the 5G era 61 RIMS II multipliers,Bureau of Economic Analysis(BEA),see:RIMS II multipliers|U.S.Bureau of Economic Analysis(BEA)The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile The economic impact of allocating mid-band spectrum to mobile NERA 27 Source:(1)RIMS II multipliers,Bureau of Economic Analysis(BEA)(2)Sosa and Rafert,2019.The Economic Impacts of Reallocating Mid-Band Spectrum to 5G in the United States (3)Anil Rao,2020.Network automation is key to delivering significant opex reduction and increasing agility in the 5G era Economic impact We use the total capex over seven years and the total opex over ten years to estimate the total effect on the economy on GDP and employment.Table 16 shows the total capex and opex we estimate are needed to deploy 400 MHz of mid-band spectrum.The table also shows the total years we are considering in our economic impact.Our opex estimation is likely conservative as operators may maintain the equipment longer than 10 years.Table 16:Incremental capex and opex required to deploy and operate 400 MHz of mid-band spectrum Total($B)Annualized($B)Years Capex 35.3 5.0 7 Opex 88.3 8.8 10 Total 123.6 13.9 Source:NERA Economic Consulting Table 17 shows our economic impact estimation in terms of GDP and jobs.As with other estimations in this paper,we report the marginal impact of 400 MHz,and the average impact of each 100 MHz.Table 17:Economic impact of deploying and maintaining 400 MHz of mid-band spectrum 400 MHz 100 MHz GDP(B)New Jobs GDP(B)New Jobs Capex 38.8 283,626 9.7 70,906 Opex 106.5 981,520 26.6 245,380 Total 145.3 1,265,146 36.3 316,287 Annualized Capex 5.5 40,518 1.4 10,129 Opex 5.3 49,076 1.3 12,269 Total Annualized 10.9 89,594 2.7 22,399 Source:NERA Economic Consulting The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile Allocating additional spectrum to Mobile vs Wi-Fi NERA 28 4.Allocating additional spectrum to Mobile vs Wi-Fi Spectrum is a scarce natural resource with many competing uses.Two key conflicting uses for spectrum are mobile telecommunications,which benefits most from high-power exclusive-use licenses,and unlicensed use,which includes Wi-Fi,where users share spectrum.To ensure the U.S.remains a leader in wireless telecommunications,and for spectrum to deliver maximum economic value,both use cases require access to sufficient spectrum bandwidths.Figure 12 shows that while the U.S.has ensured Wi-Fis spectrum needs are met by allocating more mid-band frequencies to unlicensed use than any other country,its spectrum policy has not delivered the mid-band spectrum mobile operators need.As a result,the U.S.has fallen behind the likes of China,Japan,and the UK,all of which have allocated more mid-band spectrum to commercial wireless use than the U.S.Of course,wireless operators can,to a degree,substitute between spectrum bands to compensate for a shortfall of a particular type of spectrum.However,even when looking at total spectrum allocations across all bands,the U.S.,which has allocated 1,123 MHz of spectrum below 6 GHz to mobile62,still lags well behind China,which has allocated over 1,800 MHz to mobile.63 Spectrum requirements As Wi-Fi technology and applications evolved,so too have their bandwidth requirements.In 2020,the FCC addressed this by opening 1200 MHz of spectrum in the 6 GHz band for Wi-Fi and other unlicensed applications.This allocation was in addition to the spectrum already allocated in the 2.4 and 5 GHz bands.In a white paper published by Intel,Akhmetov et al.estimate that Wi-Fi 7,the latest iteration beginning deployment in the U.S.,would need three non-overlapping channels of 320 MHz to ensure optimal long-term Wi-Fi performance for bandwidth-demanding future applications.64 Even ignoring the constraints on allocations imposed in the context of competing uses for spectrum,Wi-Fis spectrum needs appear to be met.Moreover,the U.S.already leads its peers in terms of unlicensed spectrum allocations,as shown in Figure 12.American wireless operators,meanwhile,currently only have access to 380 MHz of full-power mid-band spectrum five times less than that allocated to Wi-Fi.And while the U.S.may lead in terms of unlicensed allocations,it has already fallen behind several of its peers,both in terms of mid-band and total spectrum allocations for commercial wireless use.The last auctions of spectrum for commercial wireless use in the U.S.took place in 2022,when 100 MHz of spectrum in the 3.45 GHz band was assigned,followed by an auction for 2.5 GHz overlay licenses which amounted to approximately 68 MHz of spectrum nationwide.Since then,wireless operators have had to cope with the exponential 62 FCC,2024,2024 Communications Marketplace Report,Fig.II.B.11,available at:https:/docs.fcc.gov/public/attachments/FCC-24-136A1.pdf 63 Analysys Mason,2022,Comparison of total mobile spectrum in different markets,available at:https:/api.ctia.org/wp-content/uploads/2022/09/Comparison-of-total-mobile-spectrum-28-09-22.pdf.To compute the current holdings for China,we sum the 682 MHz of spectrum below 3.0 GHz,the 460 MHz of mid-band spectrum already released,and the 700 MHz of spectrum in the 6 GHz band that was allocated after the Analysys Mason report was published.The total is 1842 MHz.64 Akhmet et al.,2022,6 GHz Spectrum Needs for Wi-Fi 7,available at:https:/ The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile Allocating additional spectrum to Mobile vs Wi-Fi NERA 29 rise in demand for data using their existing spectrum allocations.Of greater concern,however,is the absence of a clear pipeline of future spectrum allocations for commercial wireless use in the U.S.(and even the authority for the FCC to auction that spectrum),especially in light of the continuing increase in demand for wireless data.Figure 12 shows that by 2027,the U.S.s mid-band spectrum allocation will fall behind all but one of its G7 peers.By then,it is estimated that the U.S.could face a mobile spectrum deficit of 400 MHz,which could more than triple to over 1400 MHz by 2032.65 Given the long lead times required before spectrum that has been designated for study for a particular use can be deployed,it is imperative that federal policymakers prioritize identifying additional spectrum for commercial wireless services.The discrepancy in mid-band spectrum allocations to unlicensed use and commercial wireless use does not align with how these technologies use the spectrum.Wi-Fi and other unlicensed users operate at low power in localized environments,like homes and offices,where spectrum is readily available for re-use,even in close proximity.With Wi-Fi,a relatively small amount of bandwidth is capable of serving high volumes of traffic in a given geographic area.In contrast,mobile networks must serve users spread across wide geographic areas,travelling at differing speeds,and often outdoors.The wide coverage areas that need to be served mean that wireless operators need access to full-power,dedicated spectrum licenses to avoid interference and maintain network quality.Therefore,the scope for frequency re-use is far more limited than with unlicensed use,leading to wireless operators needing much more spectrum per GB of traffic served than unlicensed users.65 The Brattle Group,2023,How Much Licensed Spectrum is Needed to Meet Future Demands for Network Capacity?,available at:https:/api.ctia.org/wp-content/uploads/2023/04/Network-Capacity-Constraints-and-the-Need-for-Spectrum-Brattle.pdf The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile Allocating additional spectrum to Mobile vs Wi-Fi NERA 30 Figure 12:Mid-band spectrum allocations for unlicensed use and commercial wireless use in selected countries by 2027 Source:Adapted from Analysys Mason,2022,Comparison of total mobile spectrum in different markets.Notes:We have removed any spectrum that is not available on an exclusive use,full-power basis.Economic impact In assessing how to allocate spectrum,policymakers should consider the economic impact of allocations alongside the technical needs of different use cases.In 2024,the WiFiForward coalition commissioned a paper to establish the economic benefits of Wi-Fi and an additional allocation of 125 MHz of mid-band spectrum for unlicensed use.66 This study identified eight sources of GDP and consumer benefit that would derive from an additional 125 MHz of spectrum in the 7 GHz band.In Table 18,we summarize the economic value generated by source.Although Wi-Fi as a whole generates tremendous value,the results in Table 18 suggest that given how much spectrum Wi-Fi already has,an additional 125 MHz allocation would produce relatively little incremental economic value.66 Telecom Advisory Services,2024,Assessing the economic value of Wi-Fi in the United States,available at:https:/wififorward.org/wp-content/uploads/2024/09/Assessing-the-Economic-Value-of-Wi-Fi.pdf The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile Allocating additional spectrum to Mobile vs Wi-Fi NERA 31 Table 18:Economic value of an additional 125 MHz allocation to Wi-Fi Source Metric 2025 2026 2027 Benefit to consumers of free Wi-Fi traffic offered in public sites GDP($B)0.36 Deployment of free Wi-Fi in public sites CS($B)0.35 Benefit to consumers of faster free Wi-Fi under Wi-Fi 6E&above CS($B)0.01 0.01 0.01 Closing the digital divide:use of Wi-Fi to increase coverage in rural&isolated areas GDP($B)0.36 0.84 1.72 Wide deployment of IoT GDP($B)2.05 4.90 11.20 Capex&Opex savings Cellular offloading PS($B)0.08 0.17 0.31 Revenues of service providers offering paid Wi-Fi access in public places GDP($B)0.02 0.04 0.05 Aggregated revenues of WISPs GDP($B)0.01 0.02 0.04 Source:Telecom Advisory Services,2024,Assessing the economic value of Wi-Fi in the United States Note:CS:consumer surplus,PS:producer surplus In contrast to unlicensed users,commercial wireless operators face a looming spectrum shortage.As we highlight in Table 19,the impact on GDP,employment and consumer surplus of allocating additional spectrum to mobile is substantial.Direct comparisons between the two studies are challenging owing to the Wi-Fi studys limited forecast horizon to 2027.However,we note that Wi-Fi has already been allocated an additional 1200 MHz in the 6 GHz band,which is only just recently seeing equipment for.There is a less immediate need for additional unlicensed spectrum than commercial wireless users,who only have access to 380 MHz of full-power mid-band spectrum.Furthermore,some use cases can be equally or better served by mobile than Wi-Fi,for example,the wide deployment of IoT.Consequently,the economic impact of adding 100 MHz to a base of 380 MHz for mobile can reasonably be expected to have a larger impact than adding 100 MHz to a base of 1900 MHz already allocated for unlicensed use.Table 19:Summary of the economic impact of allocating each additional 100 MHz of mid-band spectrum to mobile GDP($B)Employment(M)Consumer Surplus($B)Better mobile service at no additional cost 385 Improving broadband with FWA 40 0.30 3 Supporting industries that rely on mobile connectivity 188 0.93 Supporting industries that serve the wireless industry 36 0.32 Total 264 1.55 388 Note:Effect of each 100 MHz up to 400 MHz The Economic Impact of Each Additional 100 MHz of Mid-band Spectrum for Mobile Conclusion NERA 32 5.Conclusion The wireless industry serves as a vital pillar of the American economy,significantly contributing to innovation,economic growth,and job creation.Its extensive investments in infrastructure and technology not only enhance communication capabilities but also support a wide range of industries that rely on wireless connectivity for their operations.As the demand for data continues to escalate,additional mid-band spectrum will be more needed than ever to continuing fueling the American economy.NERA 2112 Pennsylvania Avenue NW 4th Floor Washington,DC 20037 QUALIFICATIONS,ASSUMPTIONS,AND LIMITING CONDITIONS This report is for the exclusive use of the NERA client named herein.This report is not intended for general circulation or publication,nor is it to be reproduced,quoted,or distributed for any purpose without the prior written permission of NERA.There are no thirdparty beneficiaries with respect to this report,and NERA does not accept any liability to any third party.Information furnished by others,upon which all or portions of this report are based,is believed to be reliable but has not been independently verified,unless otherwise expressly indicated.Public information and industry and statistical data are from sources we deem to be reliable;however,we make no representation as to the accuracy or completeness of such information.The findings contained in this report may contain predictions based on current data and historical trends.Any such predictions are subject to inherent risks and uncertainties.NERA accepts no responsibility for actual results or future events.The opinions expressed in this report are valid only for the purpose stated herein and as of the date of this report.No obligation is assumed to revise this report to reflect changes,events,or conditions,which occur subsequent to the date hereof.All decisions in connection with the implementation or use of advice or recommendations contained in this report are the sole responsibility of the client.This report does not represent investment advice nor does it provide an opinion regarding the fairness of any transaction to any and all parties.In addition,this report does not represent legal,medical,accounting,safety,or other specialized advice.For any such advice,NERA recommends seeking and obtaining advice from a qualified professional.CONFIDENTIALITY Our clients industries are extremely competitive,and the maintenance of confidentiality with respect to our clients plans and data is critical.NERA rigorously applies internal confidentiality practices to protect the confidentiality of all client information.Similarly,our industry is very competitive.We view our approaches and insights as proprietary and therefore look to our clients to protect our interests in our proposals,presentations,methodologies,and analytical techniques.Under no circumstances should this material be shared with any third party without the prior written consent of NERA.NERA
2025-02-08
37页
5星级
Halal Markets Unveiled:Opportunities in Southeast Asias Economic PowerhousesASEAN Halal Industry An .
2025-02-08
20页
5星级
E U R O P E A N D C E N T R A L A S I A S T U D I E SGREATER HEIGHTSGrowing to High Income in Europe.
2025-02-08
28页
5星级
BCGs Guide to Cost and GrowthExecutive PerspectivesStrategic Insights for Navigating Economic Uncert.
2025-02-07
24页
5星级
Unlocking the Social Economy:Towards Equity in the Green and Digital TransitionsI N S I G H T R E P .
2025-02-07
33页
5星级
H1 20253Introduction and Key Findings5Perspectives on The Innovation Economy7Macro12VC Fundraising 17VC Investment25VC-Backed Tech Benchmarks30ExitsSTATE OF THE MARKETS H1 20252Nearly every investor we speak to sees AI as a platform shift analogous to the steam engine,the internet or the rise of mobile.Pick your parallel.Even conceding the potentially inflated expectations and valuations this sentiment will drive investment.”STATE OF THE MARKETS H1 20253Marc CadieuxPresident SVB Commercial B Mark GallagherHead of Investor CoverageSVB Commercial B Venture investment is booming,companies are raising colossal rounds and valuations are flexing to all-time highs.That is,if youre an AI company.For the rest of the innovation economy,the times are slower.Deal activity is stagnant,valuations are low and exits are limited.AI is the explosive,albeit unproven,fuel of the innovation economy.The positive is that new technologies spur recoveries.In 2009,mobile a vertical turned horizontal spurred the global financial crisis(GFC)recovery that took under three years to return to peak levels and fueled durable growth.Nearly every investor we speak to sees AI as a platform shift analogous to the steam engine,the internet or the rise of mobile.Pick your parallel.Even conceding the potentially inflated expectations and valuations this sentiment will drive investment.At the company level,the innovation economy is recovering in a healthy way.Efficiency reigns supreme.Companies that successfully raised capital in 2024 managed their burn,and companies across life stages are closer to profitability.With the focus on efficiency,revenue growth has been slow to improve.Growth rates reached a floor in 2024 and are no longer falling,but they arent improving either.Lower interest rates may be the accelerant that kick-starts growth.Both the market and Federal Open Market Committee(FOMC)members expect the federal funds rate(FFR)to fall below 4%by year-end,which could help spur additional spending on new technologies from public companies and lead to higher growth rates.But the strong December jobs report and continued wage growth at 4%may put the breaks on future cuts,and if inflationary policies such as tariffs are implemented,the rate-cutting cycle could end.If a low interest rate outcome prevails,it may provide the last push to crack the exit window,and provide the much-needed liquidity to end the three-year exit drought.A handful of notable companies like Chime,xAI,Stripe and others are well positioned to go public in 2025.Furthermore,a less aggressive anti-trust policy at the Federal Trade Commission(FTC)may send big tech on a shopping spree especially for AI companies with talent and tech that can scale in-house offerings.Most facets of the innovation economy found market bottom in 2024 and are transitioning to growth in the year ahead.While hype cycles come and go,advances in one sector spur innovation in ways we cannot yet anticipate.What we know for sure is that investment and innovations today scaffold the foundation of future growth.STATE OF THE MARKETS H1 20254AI drives the next wave of growth in venture investment.Jump to PageDemand for venture outpaces the supply,throwing prices on ice.Jump to PageCompanies raising VC have controlled their burn,leading to low growth.Jump to PageLarge funds dominate fundraising,changing long-term venture dynamics.Jump to PageSTATE OF THE MARKETS H1 20254VC fund life cycles extend,changing GP and LP1 expectations.Jump to PageMost unicorns are stuck in the stable without metrics to go to the IPO race track.Jump to PageA growing number of VC-backed companies are running out of runway.Jump to PageM&A remains scarce and increasingly reserved for the most troubled companies.Jump to PageNotes:1)General Partner(GP).Limited Partner(LP).5“Many VCs went all-in for crypto in 2020.Unfortunately,by 2022 plans for decentralized financial systems were shaken when the crypto industry collapsed,crushing many startups and the funds that backed them.AI use cases are clearer and less speculative,but likely will play out very differently than VCs expect today.The crypto boom and bust is training data that every investor should include in their AI projection models.”Eric Paley General PartnerSource:SVB Interviews.STATE OF THE MARKETS H1 2025 RETURN TO TABLE OF CONTENTS“Conversations have picked up with companies looking to go public.Overall Fed policy is positive.People like the clarity of the new administration.But you have a high bar in the tech market.To IPO,companies need high ARR(more than$300M-$400M)and a good Rule of 40.But more than that,you need to be able to predict the next 12 months of revenue.”“As hiring remains competitive,weve seen public companies leverage liquidity to attract and retain top talent.Private companies increasingly want to tap into the same benefits.As a result,private companies are approaching Forge to better understand how they can adopt liquidity programs on a similar scale.”Jordan SaxeSr.ManagingDirector,Listings:AmericasEric ThomassianHead of Private Company RelationsSTATE OF THE MARKETS H1 20256Marc CadieuxPresidentSVB Commercial BankSilicon Valley BMark Gallagher Head of Investor CoverageSVB Commercial BankSilicon Valley BMarc Cadieux is president of Silicon Valley Banks commercial banking business where he focuses on the needs of innovation companies at all stages of development,including the investors who back them.Mark Gallagher is the co-head of the investor coverage practice.He and his team provide tailored services,industry insights and strategic guidance to top investors in the innovation economy.Eli OftedalSenior Analytics ResearcherSVB Market InsightsSilicon Valley BJosh PherigoSenior Analytics ResearcherSVB Market InsightsSilicon Valley BAndrew Pardo,CFASenior Analytics ResearcherSVB Market InsightsSilicon Valley BThe SVB Market Insights team leverages SVBs proprietary data,deep bench of subject-matter experts and relationships with world-class investors and founders to develop a holistic view of the innovation economy for our State of the Markets Report.We partnered with lead authors Marc Cadieux and Mark Gallagher,who bring over a half century of industry knowledge and experience working with many of the top companies and investors across the innovation economy.Together,were proud to present this 29th edition of SVBs State of the Markets Report.To learn more about the lead authors see page 37.Jake Ledbetter,CFASenior Analytics ResearcherSVB Market InsightsSilicon Valley B 7STATE OF THE MARKETS H1 2025 RETURN TO TABLE OF CONTENTSSTATE OF THE MARKETS H1 20258US VC Fundraising1Notes:1)For funds headquartered in the US by date closed.2)For funds that have a reported focus.Only half of funds have a reported focus.3)Limited partner(LP).4)Tech defined broadly as VC excluding healthcare.5)Late-stage defined by PitchBook Data,Inc.as Series C or a round that occurs more than five years after a company is founded.6)Nasdaq and New York Stock Exchange(NYSE).Source:Preqin,PitchBook Data,Inc.,S&P Capital IQ and SVB analysis.2025 OutlookUS venture funds outperformed our 2024 outlook to the tune of$16B fueled by large funds and AI.The top 10%of funds accounted for 64%of venture fundraising in 2024,and half of funds closed reported a focus in AI.2 With the same trends likely to persist in 2025 fueled by a continued lower rate environment and potential distributions to LPs3 we anticipate a growth in fundraising this year.US Series A Tech Deals42025 OutlookSeries A tech deals underperformed our expectation,hitting the lowest level since 2012.But the backlog of seed companies looking to raise a Series A remains,thus we expect moderate growth in Series A tech deal activity to reach 1,450 deals.While this would represent an inflection point toward growth,activity levels are still lower than they were a decade ago.US Late-Stage Tech Valuations4,52025 OutlookLate-stage valuations rebounded quickly,and we expect continued expansion.But the absolute increase obfuscates the reality.AI deals are the primary driver of this jump.For example,AI has a 100%valuation premium to non-AI at Series C.Secondly,late-stage deals are often structured through instruments like liquidation preferences and ratchets.That said,strong valuations reflect improving sentiment.US VC-Backed Tech IPOs on Major US Exchanges62025 OutlookWe jumped the gun on our 2024 IPO outlook;anticipated exits did not materialize.We expect the IPO window may tentatively open for a select group of top companies that are profitable or have a clear path to profitability.Several tech companies such as Chime,xAI and Stripe are all positioned to go public after ServiceTitans strong performance in Q4.$66B$50B$80B2025 OutlookActual2024 Outlook1,500 1,450 1,370 Actual2024 Outlook2025 Outlook$83M$80M$95MActual2024 Outlook2025 Outlook15710Actual2024 Outlook2025 Outlook$0B$10B$20B$30B$40B$50B$60B$70B$80B$90B$100BQ1 16Q3 16Q1 17Q3 17Q1 18Q3 18Q1 19Q3 19Q1 20Q3 20Q1 21Q3 21Q1 22Q3 22Q1 23Q3 23Q1 24Q3 245.50%4.50%lllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllllll2025 Year End2026 Year End2027Year EndLonger run0%2%4%6%8%Jan.20Jul.20Jan.21Jul.21Jan.22Jul.22Jan.23Jul.23Jan.24Jul.24The Federal Reserve has made great progress in its fight against inflation,but VC continues to reel from the consequences of high interest rates.Inflation settled around 2.5%-3.0%at the end of 2024,and consumer survey respondents reckon 2025 will show similar readings.The downward trend in expectations will be especially reassuring to Fed officials who have stressed the importance of anchoring inflation expectations.With this progress has come a normalizing of rates,with 100 basis points of rate cuts in the back half of 2024.In his December speech,however,Chair Jerome Powell downplayed potential future rate cuts.Indeed,the dot plot suggests rates just under 4%for the remainder of 2025 and slightly below that for 2026.Markets tend to agree,at least through year end.Of course,expectations are just that:predictions about an uncertain future.Shocks like economic downturns or tariff policies could lead to a reversal in interest rate policy.Lower interest rates are a potential tailwind for VC.Over the past several years,VC investment has been highly correlated to interest rates.As rates decrease,VC activity may be expected to get a boost.But this is not a return to the market peak seen during the zero interest rate policy(ZIRP)period.In a moderate interest rate environment,rates will likely play a smaller role in determining VC levels,similar to the 2017-2019 period.Notes:1)Dot plot based on the December 17-18 FOMC meeting.Market-implied rates as of January 9,2025;each observation represents one FOMC meeting through 2026.2)Upper end of target range.3)Count of articles in major news sources by year,indexed to 100 at 2019 levels.Source:University of Michigan Surveys of Consumers,Bureau of Labor Statistics,FOMC,Bloomberg,PitchBook Data,Inc.,Federal Reserve Economic Data,St.Louis Fed,Factiva and SVB analysis.STATE OF THE MARKETS H1 20259Median75th Percentile25th PercentileActual RateExpectation,MedianExpectation,RangeFOMC ExpectationsMarket ExpectationQuarterly Deal Value AnnualizedFed Funds Rate2Current Level4.54.03.53.02.52.0llllllllllMost FOMC members believe the FFR will be between 3.75%and 4.00%at the end of 2025.0100200300400500600192021222324Inflation and VCInterest Rates and VCIn 2024,there were still 3x the number of articles on inflation and VC as in 2019.lThe Fed is in a balancing act.While inflation has stabilized,it is at a level above the Feds target.The cost:A weakening labor market.“If you have a job,youre doing very well,”noted Jerome Powell in his December 18 press conference.“If you are looking for a job,though,the hiring rate is low.”The headline unemployment number has remained stable around the 4seline,but other metrics show signs of softening,such as the 23-percentage-point drop in the share of workers saying that jobs are plentiful.The tech job market was perhaps the tip of the spear in terms of white collar job weakening,as venture dollars dried up.Since then,job growth in professional services,IT and finance have underperformed other industries.Similarly,there is a bifurcation in consumer spending.From 2018 through 2021,consumer spending increased similarly for both upper-and lower-income consumers.Starting mid-2021,however,there has been a bifurcation.Those whove kept their high-earning jobs continue to spend more.Lower-income earners,meanwhile,are increasing their spending far less,with the bulk of the spending increase going to combat inflation.Flat spending delivered another blow to a challenged consumer sector.Revenue growth rates among VC-backed consumer companies fell 40-percentage points since 2021 on top of a 60cline in VC investment.1Notes:1)Decline in revenue growth rates at the median for US VC-backed companies.2)Current employment statistics industry code 50(“Information”)includes tech.Source:Bloomberg,Bureau of Labor Statistics,Federal Reserve and SVB analysis.STATE OF THE MARKETS H1 2025101.5%2.0%2.5%3.0%3.5%Q1 18Q3 18Q1 19Q3 19Q1 20Q3 20Q1 21Q3 21Q1 22Q3 22Q1 23Q3 23Q1 24Q3 24Credit CardsConsumer LoansUnemployment RatePercentage Reporting Jobs Are PlentifulAll IndustriesFinanceProfessional and Business ServicesInformation/Tech2-4%-2%0%2%4%6%8%Jan.22Apr.22Jul.22Oct.22Jan.23Apr.23Jul.23Oct.23Jan.24Apr.24Jul.24Oct.24In 2022,information/tech sector jobs grew more than other industriesin 2024,they grew less.2018201920202021202220232024-15%-10%-5%0%5%High IncomeLow IncomeMid Income0 0P%0%3%6%9%Jan.18Jun.18Nov.18Apr.19Sep.19Feb.20Jul.20Dec.20May.21Oct.21Mar.22Aug.22Jan.23Jun.23Nov.23Apr.24Sep.24Job Plentiful PercentageUnemployment RateUS tech layoffs have slowed as most of the companies that could shed jobs to save money have already done so.For those in the job market,there are fewer positions to choose from.Tech hiring has stagnated for the last two years.Tech companies are hiring at half the pace they were in the 2022 peak,the weakest level since at least 2016,according to LinkedIn data.US tech salaries are showing signs of weakening due to lower demand and pressure from lower-cost and growing talent pools overseas.Emerging markets in Asia,Africa and Latin America are adding millions of coders every year,quickly closing the software skills gap to the US.Startups are taking note.About 60%of companies already outsource app development.India is expected to overtake the US in number of developers by 2030.AI is having a greater impact on programming work in the US,though it doesnt appear to be replacing human developers(yet).According to an annual survey from Stack Overflow,63%of developers now use AI in their work,up from 44%last year.Most use it to directly write code,find answers or debug.Complex coding tasks are still best left to the humans,a sentiment reflected in lukewarm responses gauging developers trust in the accuracy of results.This may explain why only 12%of developers said they view AI as a threat to their job.Notes:1)Hiring rate is the percentage of LinkedIn members in the technology,information and media industry who added a new employer to their profile in the same month the new job began.The hiring rate is indexed to the average rate in 2016.Layoffs in thousands.2)Annual survey most recently conducted in May 2024 with 65,000 respondents.3)Among countries with at least 1M developers.4)Based on global HackerRank scores,last updated in 2016.Source:LinkedIn Workforce Report,Stack Overflow Developer Survey,GitHub Octoverse Report and SVB analysis.STATE OF THE MARKETS H1 2025110K30K60K90K120K150K180K4050607080901002021202220232024US tech hiring has slowed to half its peak pace in 2022.26()0222369Ac%MexicoTurkeyArgentinaBrazilJapanIndonesiaPhilippinesVietnamIndiaNigeriaSingaporePakistanBangladesh42wDCrc 24Currently Using AI in Development:Favorable Opinion of AI Tools for Development:Top Uses for AI:1.Writing Code 82%2.Finding Answers 68%3.Debugging/Testing 57%4.Documenting Code 40%5.Generating Content 35%Trusts the Accuracy of AI Outputs:Developers15M22MYoY Growth33!%Median Salary$21K$130KTalent Rank431st28thLinkedIn Tech Hiring IndexNumber of Tech Job Layoffs AnnouncedAsiaAfricaLatin AmericaEurope$100K$120K$140K$160K$180K$200K$220KDeveloper:QAData AnalystDeveloper:DesktopDeveloper:Front-EndDeveloper:GraphicsDeveloper:Emb.AppsDeveloper:Full StackDevOps SpecialistEngineer:DataData ScientistDeveloper:Back-EndEngineer:Cloud Infr.Engineer:Site ReliabilityEngineer:ManagerSalaries for full-stack developers are down 7.1%since 22.23 Median24 Median(Decrease)24 Median(Increase)202312STATE OF THE MARKETS H1 2025 RETURN TO TABLE OF CONTENTSNotes:1)Assessed over the trailing 24 months to smooth data given significant swings caused by large top-end outliers.Source:Preqin,PitchBook Data,Inc.and SVB analysis.Rank20202021 202320241Tiger GlobalGeneral Catalyst2Andreessen HorowitzNew Enterprise Associates3LightspeedAndreessen Horowitz4AccelKhosla Ventures5New Enterprise AssociatesARCH Venture Partners 6Flagship PioneeringNorwest Venture Partners7ARCH Venture Partners Flagship Pioneering8Khosla VenturesTiger Global9Norwest Venture PartnersGreenoaks Capital10General CatalystOrbiMedIf Bernie Sanders were a venture economist,he would undoubtedly draw attention to the growing inequality in venture fundraising.The bottom 90%of venture firms have raised as much capital as the top 2%,highlighting a significant skew towards the largest funds.Since 2020,the VC industry has been increasingly dominated by large firms,with the top 10 firms alone capturing 22%of all fundraising.This concentration of capital is leading to entrenchment,with the elite group of top 10 VC fundraisers changing little from year to year.This dominance of large funds is marginalizing mid-sized funds.There is a clear bifurcation in the market,where the biggest funds focus on making large investments and,in some cases,nearly“index”the venture market.On the other hand,small funds carve out niches,targeting specific sectors or stages.This leaves mid-sized funds in a precarious position.Their role is less clear.Most are neither giants able to compete in mega-deals nor niche funds in hyper-specialized markets.This could lead to consolidation and a less competitive market,with capital and talent increasingly concentrated among a few top firms.STATE OF THE MARKETS H1 202513$0B$20B$40B$60B$80B$100B$120B$140B$160B$180B$200B$220B$240B$260B$280B0%5 %05EPUep 00200120022003200420052006200720082009201020112012201320142015201620172018201920202021202220232024US Venture Capital Fundraising:Trailing Twelve Months Share of Venture Capital Fundraising Going to$500M Fund1$500M funds consistently capture 35-45%of venture fundraising The top 10 firms capture 22%of all fundraisingTop 10%of Firms:$258BBottom 90%of Firms:$148BColors illustrate change in rankingDeal Size,Top 20 VCsDeal Size,All VCsSeries ASeries BSeries C$0M$10M$20M$30M201920202021202220232024$0M$20M$40M$60M201920202021202220232024$0M$30M$60M$90M201920202021202220232024$0M$25M$50M$75M$100M201920202021202220232024$0M$75M$150M$225M$300M201920202021202220232024$0M$125M$250M$375M$500M201920202021202220232024Series ASeries BSeries CValuation,Top 20 VCsValuation,All VCsBased on the data,lessons from past downturns have not been fully absorbed.Namely,scaling is hard!There is only so much capital that can be effectively deployed in each company without driving inefficient burn.For VCs investing in early stages,fund sizes are difficult to scale.Larger funds will naturally have big portfolios of small bets that begin to mirror the market,limiting outperformance potential.This venture pitfall persists.The largest VCs are deploying more capital per deal and paying more per deal compared to the median VC fund.Overpaying can lead to underperformance,which is particularly evident in the top quartile of large funds.Concentration of capital and power can drive up prices unnecessarily,leading to outsized valuations during peak times valuations that the industry is still struggling to come to terms with today.This trend poses significant challenges for the industry.Larger funds simply have more capital to deploy,and those that invested early can dominate later-stage deals.Together,this can effectively squeeze out smaller VCs.Nevertheless,the incentives for individual firms to grow remain compelling,making it difficult to reverse course without LP pressure.Should muted returns become the norm,however,fund sizes may decrease and LPs may increasingly opt for other asset classes.STATE OF THE MARKETS H1 202514Notes:1)Top 20 VCs defined as US-based VCs that have raised the most during their life,calculated by the funds aggregate VC fund size.2)Internal rate of Return(IRR);For each vintage,large and small funds are those that have fund sizes above or below the median,respectively.3)Big funds are those$750M ,small are less than$250M.Analysis assumes the top quartile return of each group for vintage years 2010-2019.Carry net of an 8%hurdle rate.Source:Preqin,PitchBook Data,Inc.and SVB analysis.Management Fees Carried Interest$50M$150M$336M$389M0 0%SmallBigFund SizeMiddle 50%of Small FundsMiddle 50%of Large FundsMedian Small FundsMedian Large Funds-10%0 0 002005201020152020VintageAfter 2014 over half of returns are unrealized.The old rule of an 8-to 12-year fund life cycle is not a reality for most funds.Top quartile funds dont actually return capital for 16-20 years.To reflect this new reality,some funds are changing the language in their limited partner agreements(LPAs)to reflect longer fund life cycles but cut off the fee period.With large funds investing at the latest stages,companies are able to stay private longer,and the trend toward large funds is only likely to continue.As a result,the average age of a US VC-backed unicorn is now 10.3 years,just four months less than the average age of tech IPOs.The vast majority of those unicorns do not have the metrics to make a compelling IPO(see pg.32).Despite longer time frames,LPs are still investing in venture.During peak times,we saw record-low time between funds.While it is increasing,it is still historically low;however,this time will likely continue to grow as fewer funds have come back to market since 2022 and investment rates remain below peak levels.What this data misses is the VCs that may not raise capital again after investing their first fund at the peak of the market and having marginal returns to show for it.But a venture firm doesnt disappear overnight unless they sell their portfolio in a secondary.It takes 16-20 years to liquidate their investments and close their doors.STATE OF THE MARKETS H1 2025150P00 0%000500E0P0 05 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024Median DistributionsMedian Residual Value of Investment(Unliquidated/Unreturned)1Fund BreakevenVintage Yr.Fund Age2019171615141312111098765432118100%An 8-12 yr.fund cycle is not realistic:Most of the fund is not distributed.Majority is distributed to LPs with limited residual value:True fund life cycle is 16-20 yrs.Notes:1)For top quartile multiple on invested capital(MOIC)funds.Distributions are Distributions to paid in capital(DPI)and Residual Value is the residual value of paid in capital(RVPI)both expressed as a percent of capital paid in.Source:Preqin and SVB analysis.413226282729322930242324231515107912244952364241464745483634322929312324202934586664787064716169625045474444404034434120052006200720082009201020112012201320142015201620172018201920202021202220232024Middle 50%of FundsMedianFunds More Than Double the PaceDeployment rate slows and the time between funds increases amid venture contraction.Consistent,Predictable Fundraising CyclesStartup launch programs have become the front door for the venture ecosystem,welcoming in thousands of startups each year that might otherwise be overlooked or go unfounded.Deals from incubators and accelerators are typically small dollar values relative to seed deals,but they comprise a significant share of overall VC activity,accounting for a quarter of all deals in 2024.1 Incubators are a stabilizing force in early-stage formations.Not only do they act as a quality screen for investors,theyre also less fickle in downturns.When VCs apply the brakes during market lulls,incubators tend to continue churning out new cohorts at a steady rate.The era of startup programs took root during the Global Financial Crisis when programs such as Y Combinator,Plug and Play and Techstars helped launch iconic companies like Stripe,AirBnB and DoorDash.More than 13,000 companies from these programs across the country have raised over$200B in VC over the last 15 years.As the model has spread nationwide,the impact from startup launch programs has been more pronounced in non-tech hubs,where supportive local governments,corporations and universities give these programs a concentrating effect,attracting as much as 40%of all VC deals in some states.Here,incubators fill the market gap by finding and supporting founders outside of the main innovation hubs.Notes:1)Incubator and accelerator deals are presented here as a share of overall VC deals to show their scale,however,we exclude these deals in our analysis of VC activity elsewhere in the report.2)Includes companies that received an incubator or accelerator deal,as classified by PitchBook.Premium as compared to companies with no incubator/accelerator deal.Source:PitchBook Data,Inc.and SVB analysis.STATE OF THE MARKETS H1 20251643%0%5 %0 052006200720082009201020112012201320142015201620172018201920202021202220232024Accelerators and Incubators accounted for 24%of US VC deals in 24.$0B$50B$100B$150B$200B2005200720092011201320152017201920212023In VC Raised by AlumsNotable Alum:Companies IncubatedSince 05Founded:2005Founded:200625w%SeedEarly StageLate StageCompanies from incubators have a valuation premium at every stage.23%C)1$A )8%STATE OF THE MARKETS H1 202517 RETURN TO TABLE OF CONTENTSSimple supply and demand models go a long way in describing the current state of the innovation economy.We assessed demand by looking at the number of companies that need to raise in the next six months and how much those companies would need to finance operations at current burn rates.The supply of capital is simply a function of US VC fundraising and investment.As company fundraising boomed in 2020 and 2021,the demand for capital fell because fewer companies needed to raise at any given time.At the same time,supply increased,pushing prices for companies higher as measured by revenue multiples.Fast-forward to 2022 and the trend flipped.Demand began to rise and supply began to fall,pushing multiples down.Not only are valuations lower,but the speed of valuation growth is slower.It now takes the typical Series A company over two years to increase its valuation as much as companies in 2021 did in a single year.While this is partially attributable to the supply and demand in venture,it is important to note that growth rates for VC-backed companies have also slowed substantially(see pg.26).This slower growth further drives down multiples as high growth is one of the main reasons for investing in a company with a high multiple.STATE OF THE MARKETS H1 202518Median Series C Revenue Multiple(Trailing 4 Quarters)Spread Between Supply and Demand Indexes for US VC1 0 0000%Series ASeries BSeries CSeries D201920202021202220232024Notes:1)Demand for venture is a function of the number of companies that need funding in the next six months and the amount those companies are burning.Supply is a function of fundraising and investment(equal weighted index of the two).A baseline for the index was established between 2017-2019;the percentage point variance is expressed in relation to that baseline.2)Calculated at the valuation increase between rounds divided by the years between rounds for the given year a company closed a deal.Source:SVB proprietary data,PitchBook Data,Inc.and SVB analysis.Middle 50%of DealsMedianQ1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q4Q1Q2Q3Q42018201920202021202220232024-100pp-50pp0pp50pp100pp150pp200pp250pp5x10 x15x20 x25xDemand:Company Funding Needs Next 6 Months Supply:Venture Fundraising and Investment Supply significantly outpaces demand.Demand outpaces supply.Multiples Contracts as Oversupply Ends0 x10 x20 x30 x40 x50 x60 x70 x80 x90 x100 xQ1Q3Q1Q3Q1Q3Q1Q3Q1Q3Q1Q3202320242023202420232024Series ASeries BSeries CAnd just like that,VC is back.US VC investment totaled$204B in 2024,a 30%year-over-year increase and the third-highest annual total on record.The recovery marks an about-face for the venture ecosystem.This year started at a low-point,after eight straight quarterly declines in annual VC investment,and ended on a hot streak with three quarterly increases.What happens next depends on the prospects for the one technology most responsible for the turnaround:generative AI(GenAI).Exclude AI investment and the story changes.There is no meaningful investment uptick for companies not leveraging AI.Investment for this group is essentially flat for the last year.AI has gobbled up VC market share in the last two years.At the peak of the last cycle,only one in four companies getting VC deals had AI as a vertical.Now,its half of all companies.And a handful of these are controlling a huge portion of the VC dollars.For the first time,more mega-deal dollars went to AI companies($73B)than to non-AI companies($47B).This inflection marks a turning point like we havent seen since the rise of mobile technology after the GFC.The emergence of the iPhone and the App Store kicked off a wave of innovation that at first was confined to a core group of mobile-focused companies.VC flowed disproportionately to this group for the first few years,sparking a general VC recovery as mobile spread to all companies.Could we see a similar trend with AI?Notes:1)What-if projections simulate investment levels if AI company investment follows the same path as the mobile tech vertical post-GFC,indexed to the investment peak prior to the decline.Our forecast picks up when mobile VC returned to its pre-GFC peak,which is where we are with AI now.Source:PitchBook Data,Inc.and SVB analysis.STATE OF THE MARKETS H1 202519Non-AI VC InvestmentHorizontal PlatformsAI grows 50%-60%in 2025 before leveling off.All other investment grows 20%in 2025.02550751001251501752002021202220232024202520262027$13B$17B$25B$61B$34B$39B$73B$65B$66B$79B$203B$111B$73B$47B201820192020202120222023202461%Of VC mega-deals went to AI companies.$0B$100B$200B$300B$400B200720082009201020112012201320142015201620172018201920202021202220232024202520262027AI/ML:Vertical AppsAI Chips and Machines48%of VC went to AI-leveraged companies in 2024.What if AI investment follows the trajectory of mobile investment after the GFC?161100105Other VCGFC RecoveryCore AI VCNon-AI VCCore AI VCNon-AI VCCombinedIts hard to comprehend the advancements in computing that have led us to GenAI.In 1969,the Apollo Guidance Computer calculated 14,000 math operations per second to deliver astronauts to the moon.Today,we measure compute in quadrillions of operations per second(called a PetaFLOP).ChatGPT 1 took a full day of PetaFLOP computing to train its 100-million parameter model.But even PetaFLOP-days arent cutting it anymore.Metas latest model,Llama 3.1,required a staggering 1,200 PetaFLOP-years to train on over a trillion words.All of that compute doesnt come cheap.Every new large language model(LLM)costs hundreds of millions of dollars to develop,and foundational AI companies are churning these out several times per year,releasing multiple versions that are optimized for developers to build upon.The metric that may best capture this activity is NVIDIAs revenue.The AI chipmaker has cornered the market on semiconductors needed to train new models,and its sales are rising in proportion to public adoption of AI.Corporations are increasingly investing in AI products and tooling.Research by the VC firm Menlo Ventures shows that US companies spent at least$16B on AI products in 2024,a 7x increase from 2023.1 Thats only expected to grow as the costs come down and apps get better.STATE OF THE MARKETS H1 202520Notes:1)According to Menlo Ventures analysis of dollars spent on foundation models,model training and deployment,AI-specific data infrastructure and new GenAI applications from startups and established corporations.2)This is an illustrative example with model capability and inference costs approximated based on estimated data such as the number of parameters to train models and subjective factors like iterative improvements in models.Source:SEC filings,Google trends,company websites and SVB analysis.0255075100$0B$5B$10B$15B$20B$25B$30B$35B$40B201720182019202020212022202320241101001,00010,000100,0001,000,0002018201920202021202220232024PetaFLOP-DaysGPT-1Llama 3.1BERTXLNetGPT-2GPT-3LLaMA 2OPTPaLMNemotron-4(NVIDIA)OtherGPT-4Inference CostModel CapabilityGPT-3(2020-22)GPT-4(2023-?)OpenAI-o1(2024-?)GPT-4o MiniGPT-3 AdaGPT-3 DavinciGPT-4 TurboGPT-3 BabbageGPT-4oo1-PreviewModel CapabilityInference CostFull-Scale,LegacyLightweight,UtilitiesOptimized,General UseFull-Scale,PremiumPhase 1:Most-capable,highest-cost model is released.Phase 2:Less-powerful model with more efficient inference costs.Phase 3:Further optimized for specific tasks or faster use.Phase 4:Full-scale models,obsolete by new architecture.Ex)GPT-3 series(now)Ex)GPT-4 TurboLife Cycle of LLM Model DevelopmentEx)GPT-4oEx)GPT-4o minio1-MiniWith AI driving nearly all of the growth in VC investment,its not surprising that the sectors benefiting most are those where the AI hype is peaking:enterprise software(LLMs)and frontier tech(autonomous machines).Attention on these sectors is at an all-time high and so is investment.Companies at the core of GenAI,such as xAI,Databricks and OpenAI,are generating massive VC deals,pushing enterprise software investment up 47%from 2023.Much of the capital for these deals is consumed by the high cost of training models.A single new LLM released to market takes hundreds of millions of dollars in compute to train,and the pace of new releases is only growing.Then there are the machines.Autonomous vehicles are driving a large share of the investment in frontier tech,which has jumped from the fourth-most heavily invested sector in 2022 to the second-most favored sector in 2024.Defense technology is also emerging as a growth area for frontier tech investors,with notable deals for several defense tech unicorns such as Anduril among the largest deals of the year.Consumer tech is still struggling to find its footing in the era of AI.Only 25%of consumer companies have AI as a key vertical,yet those that are building AI products have a much higher valuation over those that dont(4x premium for later-stage companies and 2x for early-stage).STATE OF THE MARKETS H1 202521Notes:1)Based on SVBs proprietary taxonomy of PitchBook deals.2)xAI closed two$6B deals in 2024.3)Anthropic closed three deals for$9.2B raised in 2024.At least$3B of this was convertible debt.They closed another$1B in January 2025 and are in-progress to close$2B more,according to PitchBook.Source:PitchBook Data,Inc.and SVB analysis.20192020202120222023202412345Enterprise SoftwareFrontier TechFintechConsumer InternetClimate Tech47P%-41%YoY Change in US VC-25%-12%-7%-23I0%-65i61 614%ClimateTechEnterpriseSoftwareFintechFrontierTechConsumerInternetThe valuation premium for companies with an AI vertical is greatest at the later stage.SeedEarly StageLater Stage0 0Pp 1620172018201920202021202220232024Enterprise SoftwareFrontier Tech FintechConsumer InternetClimate TechCompanyVC in 24SectorFocusxAI2$12.1BEnterpriseFoundational AIDatabricks$10.0BEnterpriseAI InfrastructureAnthropic3$9.2BEnterpriseFoundational AIOpenAI$6.6BEnterpriseFoundational AIWaymo$5.6BFrontier TechAutonomous VehiclesAnduril$1.5BFrontier TechAerospace and DefenseCoreweave$1.1BEnterpriseAI InfrastructureMistral AI$1.1BEnterpriseFoundational AIWayve$1.0BFrontier TechAutonomous VehiclesScale AI$1.0BFrontier TechAerospace and DefenseDefense tech is emerging from the shadows to claim a more prominent role in the venture ecosystem.The wars in Ukraine and Israel have drawn stark awareness to the impact that technologies such as drones have on the modern battlefield.VC investment in US defense technology has ticked higher as a result,jumping 2x in 2023 and staying at that level in 2024.The largest deals are dominating with the top 10 deals accounting for about 80%of VC dollars in the last two years,a 20-percentage-point jump from 2022.At least seven defense tech unicorns received later-stage deals in 2024,positioning the cohort well for potential exits in the year ahead.More VCs are mentioning defense tech as a specific focus area than ever before.General Catalyst named defense a key strategy for their recent$8B fund(though it wasnt clear how much of that was earmarked for defense).Follow-on investors could further increase the demand for what is still a niche segment of the venture ecosystem.Defense tech companies have steeper capital requirements than other sectors.Later-stage deal sizes were 4x higher for defense tech than other technologies.Machines are expensive(and complicated)to build,which can be a deterrent for investors.However,the companies that do find product market fit,tend to achieve exit velocity,given the large government contracts that tend to be lucrative and dependable.STATE OF THE MARKETS H1 202522Notes:1)Terms include“defense,”“instability,”“war,”and exclude“financial instability.”2)Defense tech includes all of PitchBooks analyst curated vertical:“Aerospace and Defense”as well as an SVB-curated list of VC-backed defense contractors.3)Post-money valuations for all disclosed deals.Source:CB Insights,PitchBook Data,Inc.and SVB analysis.5010015020025030020202021202220232024US corporate attention on defense and security is up 2x from what it was in 2020.$1.6B$2.1B$3.5B$5.4B$2.5B$5.2B$5.0B59EcQdy 18201920202021202220232024Total VCTop 10 Largest Deals$21.4M$29.5M$6.5M$7.5MEarly StageLater StageDefense TechOther VC$0B$2B$4B$6B$8B$10B$12B$14B2015201620172018201920202021202220232024Exited UnicornsActive UnicornsThe 15 active US defense tech unicorns are valued at$50B.Seed extensions are capturing the highest percentage of seed deals and capital ever witnessed.Starting with the 2015 seed cohort,extension rates(i.e.,the share of seed companies that raised an additional seed round)ticked up year by year,peaking for those that raised a seed in 2021.Graduation rates moved similarly to extension rates up until the 2021 cohort.Following the 2021 class,graduation and extension rates started to tick down.On a relative basis,graduation rates fell faster than extension rates.This shift occurred for a number of reasons.First,seed cohorts from 2020-2021 raised in a growth-at-all-costs environment,whereas more recent seed cohorts were forced to be capital efficient from day one.Second,the venture landscape recalibrated as investors pulled back,pushing graduation rates down and leading companies to depend on extension rounds.Third,the cohorts that raised in 2020-2021 need more time to reach the higher Series A benchmarks expected of them.Lastly,seed companies are using extensions to kick the can down the road in hopes of raising a Series A at a better valuation.As a result of these trends,seed extension deal sizes and valuations continue to climb.Until those older cohorts work through the system,expect graduation and extension rates to drudge along.STATE OF THE MARKETS H1 202523Notes:1)Seed extension defined as any seed round after the first seed for the specific startup.Source:PitchBook Data,Inc.and SVB analysis.20202021202220232024MonthsMonths2019Cumulative Seed Extension Rates 2015-2024Cumulative Graduation Rates 2015-20242015-20180 0Pp0 152016201720182019202020212022202320240 0Pp0 152016201720182019202020212022202320241st Seed2nd Seed3rd SeedDeal CountCapital Invested$0.0M$0.5M$1.0M$1.5M$2.0M$2.5M$3.0M$3.5M$4.0M2015201620172018201920202021202220232024$0M$2M$4M$6M$8M$10M$12M$14M$16M$18M2015201620172018201920202021202220232024Median Deal SizeMedian Pre-Money Val.1st Seed2nd Seed3rd Seed0%5 %0691215182124273033360%5 %069121518212427303336Extension rates increase from 2015-2021 before declining.2015-2020 graduation rates moved in lock step with extension rates.2021 marked a divergence in that trend despite extension rates being above graduation rates on a relative basis.$35BCore-weave$9B$12B$11B$12B$23B$27B$29B$41B$36B$24B$46B1,1161,4291,3941,4191,5391,7391,8152,3392,0891,7001,32105001000150020002500$0B$5B$10B$15B$20B$25B$30B$35B$40B$45B$50B20142015201620172018201920202021202220232024Rising interest rates in 2022-2023 sent ripples through the capital markets,curbing the appetite for debt among public tech companies.Yet in the startup world,venture debt is a key lever,compensating for a slowdown in VC funding and providing critical runway extension.In the past,later-stage venture debt was a complement to equity.When it was a replacement to equity,it was due to the companys strong fundamentals,such as reducing burn.This could become a problem for companies and their lenders if the financing was insufficient to achieve the milestones necessary to raise the next round,or if new investors are unwilling to see their new dollars go to repay debt.Venture-backed companies are also finding new ways of using debt.CoreWeave,for instance,turned to a collateral-backed facility for financing compute.Historically,lenders pulled back during downturns,as those who invested heavily during the peak times realized losses.During this cycle,however,the opposite has occurred.New entrants,such as deep-pocketed private credit funds,are further increasing the competitive pressure,offering sweetheart deals to gain market share.Whats clear is that venture debt is no longer just a stopgap measure.STATE OF THE MARKETS H1 202524Notes:1)Sample includes companies listed on major US exchanges with a primary industry of“information technology.”Calendar years and quarters are shown.Averages use data winsorized at the 5th and 95th percentile.2)Q4 2024 data is extrapolated based on average quarterly data for Q1-Q3.3)The majority of companies in the dataset are later-stage.4)Data for 2024 includes Q1-Q3 only.Source:S&P Capital IQ,PitchBook-NVCA Venture Monitor(Q3 2024),PitchBook Data,Inc.,SVB proprietary data and SVB analysis.Average Median Deal ValueDeal CountDeal Value Linear TrendMedianAverage101111141213Q1 23Q2 23Q3 23Q4 23Q1 24Q2 2420242023202220212020201920182017201620152014$0M$1M$2M$3M$4M$5M$0M$10M$20M$30M$40M$50M20242023202220212020201920182017201620152014Median Deal SizeAverage Deal Size23!%$54275551 192020202120222023Q1 24Q2 24Q3 24Extrapolation Average Driven Up by Large AI DealsSTATE OF THE MARKETS H1 202525 RETURN TO TABLE OF CONTENTSOne of the most common questions we hear from founders is“what are the benchmarks for raising capital?”Unsurprisingly,the answer has changed over time.Revenue growth is no longer as important as it once was.In fact,the typical company raising a Series A is growing at 69%today.This is down from 171%YoY in 2021.Managing burn is of utmost importance today.Among companies that raised capital in 2024,the typical Series B company only increased its burn 8%YoY.This means that companies are growing,but they arent growing their burn.Companies that are raising are increasingly efficient.This is vastly different from companies raising in 2021 and 2022 that rapidly grew burn YoY in an environment where capital was easier to come by.At the Series A we have also seen a significant increase in the median revenue companies have at the time of raise.The median Series A company now has a whopping$2.5M in annual revenue 75%higher than companies had in 2021.This has coincided with more companies raising multiple seed rounds and a bottleneck of seed-stage companies seeking to raise a Series A.There were fewer Series A tech deals done in 2024 than at any point in the last decade those that are being done are the exception.STATE OF THE MARKETS H1 202526Notes:1)YoY growth rate comparing annualized quarterly values;does not include extension rounds.2)The annualized current run rate;does not include extension rounds.Source:SVB proprietary data,PitchBook Data,Inc.and SVB analysis0 00000 1920202021202220232024$1.4M$5.0M$14.2M$2.5M$6.0M$13.8MSeries ASeries BSeries C75 %-3%Series ASeries BSeries C20212024Series ASeries BSeries C-50%0P00 0%00050 18201920202021202220232024Increase in Burn the Year Following the DealIncrease in Burn the Year Leading up to the Deal Starting Point(1 year Before Deal)Year of Venture RoundUnhealthy,Inefficient Increases in Burn Fueled by Too Much CapitalCompany burn stagnates as many companies continue to grow into their burn rates.-286%-171%-105%-71%-45%-21%-12%-3%-450%-400%-350%-300%-250%-200%-150%-100%-50%0%$1.0M$2.5M$5.0M$7.5M$10.0M$20.0M$50.0M$100.0MThe long and winding road that leads to profitability may be shorter today than in 2021.More companies are approaching profitability and doing so sooner in their life cycle.This is not to say early-stage companies are profitable far from it.The median VC-backed tech company with$1M in revenue has a profit margin of negative 286%.But as the YoY increases in burn settle near zero and revenue growth rates continue(albeit slower),companies continue to trend toward profitability.In fact,the median VC-backed tech company with$1M in revenue saw margins improve 119 percentage points since 2021.The trade-off of lower burn and higher profitability is slower growth.When companies burn less,they spend less on marketing and expansion that drive the top-line growth.Therefore,in addition to exogenous factors like a slower economy and lower spending on new tech,growth rates have fallen.Balancing growth and profitability is a tightrope all companies walk,but many have been falling off.The median Rule of 40 fell in 2022 and 2023,as growth rate declines outpaced the improvements in profit margin.But 2024 marked an inflection point;growth rates leveled out and profitability continued to improve,which means companies are generally operating with better Rule of 40 metrics.STATE OF THE MARKETS H1 202527Notes:1)Year over year revenue growth.2)Revenue corresponds to bins:$1M-$2.5M,$2.5M-$5M,$5M-$7.5M,$7.5M-$10M,$10M-$20M,$20M-$50M,$50M-$100M.3)Rule of 40 is equal to revenue growth rate plus profit margin.Source:SVB proprietary data and SVB analysis.201920202021202220232024$1M$2.5M$5M$7.5M$10M$20M$50M$100M$1M$2.5M$5M$7.5M$10M$20M$50M$100M0 0 1920202021202220232024-250%-200%-150%-100%-50%0P%Company RevenueProfitability has improved,but back at 2019-2020 levelsCash has always been king.But right now,most startups cash reserves would be lucky to be a prince.As investment remains subdued,companies are feeling the pinch.Most have cut where and what they can,but without investing in growth or being able to raise another round,startups have started to see their reserves dwindle.Median runway for US tech startups has settled at 12 months in 2024 the lowest level since 2019.Furthermore,61%of startups saw their cash runway decline from the previous year,the second highest share since 2016.For those that have been fortunate enough to raise cash,theyre raising far fewer months of runway compared to previous years.On a median basis,startups are raising nine months less of runway compared to the boom times of 2021.To be sure,some of this is supply driven with late-stage capital fleeing the ecosystem.It may also be demand driven,as startups have realized that all capital is not created equal,and there is such a thing as too much capital.However,there are a mounting number of startups that need to raise in the coming months.Its estimated that half of cash-burning US tech startups will need to raise in the next year similar to 2019 levels.While 2025 has brought more optimism that checkbooks will open,some companies are still likely to be grounded on the runway.STATE OF THE MARKETS H1 202528Notes:1)Data for 2024 based on Q4 data where applicable.If Q4 is not available,then Q3 is used.2)Cash runway raised determined by using current burn rates for companies with 100%increase in cash balance from the previous quarter and the company raised an equity round.Source:PitchBook Data,Inc.,SVB proprietary data and SVB analysis.0%5 %0HMonths of Cash Runway Bucket2019202020212022202320240 Mos.5 Mos.10 Mos.15 Mos.20 Mos.25 Mos.30 Mos.35 Mos.40 Mos.25th50th75th-14 Months-4 Months201920202021202220232024Share of Startups with Decreasing Runway YoYPercentile0 0Pp069121518212427303336201920202021202220232024Months Until Need to Raise55BTcXa%Q419Q420Q421Q422Q423Q424STATE OF THE MARKETS H1 202529Notes:1)Q4 used for each year except 2024 where Q4 is not available for some companies.In those instances,Q3 is used instead.Source:SVB proprietary data and SVB analysis.5776661215141215122333282327292019 2020 2021 2022 2023 20246998771318181514122534342524252019 2020 2021 2022 2023 2024710109971418191614142736372420222019 2020 2021 2022 2023 2024577644101313131082025292218162019 2020 2021 2022 2023 2024577755111414131192125282218162019 2020 2021 2022 2023 2024917131214151834262327293674514568812019 2020 2021 2022 2023 2024688876121617141312243333242424201920202021202220232024Consumer InternetFintechEnterprise SoftwareFrontier Tech$0-$10M$10M-$25M$25M-$50M$50M Median75th Percentile25th PercentileUS VC-Backed Startups711119991320221518212438392431352019 2020 2021 2022 2023 202471086981420181314142538322327262019 2020 2021 2022 2023 2024STATE OF THE MARKETS H1 202530 RETURN TO TABLE OF CONTENTSThe VC slowdown is testing startups resilience,particularly when it comes to managing debt.With less funding available,some companies are finding it harder to stay on track with repayments a trend that evokes parallels to the early pandemic period.Data suggests that at the end of 2023 and into 2024,more startups began having difficulty with debt repayments,a sign of financial trouble for these companies.While this peak has since eased,levels remain elevated,reflecting the ongoing adjustments many companies face in todays funding environment.For startups encountering financial strain,options are more constrained than in previous years.Acquisitions,whether full buyouts or tech-focused deals,have become less frequent.An increasing number of companies are winding down entirely.Bankruptcy filings in Silicon Valley are on the rise,underscoring the harsh realities of operating in a capital-constrained environment.Macro headwinds in the funding environment are creating a critical turning point for many companies.An increasing share of VC-backed startups is showing no growth or profit,forcing many to confront hard choices about their future.As the startup ecosystem contends with this wave of financial fragility,the question remains:How many will sink before the tide turns?STATE OF THE MARKETS H1 202531Notes:1)Two-quarter moving average.2)Data for 2024 includes Q1-Q3 only.“Other”outcomes are excluded,so each year does not sum to 100%.3)Silicon Valley includes San Francisco,San Mateo,Santa Clara and Alameda counties.Data includes bankruptcies across industries.Source:US courts,SVB proprietary data and SVB analysis.Borrower Winds DownBorrower Is AcquiredDistressed Debt ResolutionRefinanceLoan Amortizes FullyBorrower Raises Equity9095100105110115120Q1 22Q2 22Q3 22Q4 22Q1 23Q2 23Q3 23Q4 23Q1 24Q2 24Q3 24IndexLinear Trend Since Q1 220%5 %Q4 2018 Q4 2019 Q4 2020 Q4 2021 Q4 2022 Q4 2023 Q4 2024050100150200250300350400Q1 20Q2 20Q3 20Q4 20Q1 21Q2 21Q3 21Q4 21Q1 22Q2 22Q3 22Q4 22Q1 23Q2 23Q3 23Q4 23Q1 24Q2 24Q3 24Chapter 7(Liquidation)Chapter 11(Reorganization)No Growth and No ProfitNo Growth,No Profit and Less than 12 Months of Runway15#2)#!%9%7%2%4%7%5%7%8%5!$ 2020212022202320242018201920202021202220232024The herd of US VC-backed tech unicorns continues to grow,with few exiting,closing their doors,or taking a down round below$1B post-money.With the growth of the herd,so too comes growing demand for liquidity.Secondary markets and M&A activity may provide some liquidity to unicorns,but IPOs will need to play a key role as well.But the IPO bar is higher today,and few unicorns surpass it.According to Jordan Saxe,who oversees Nasdaqs listings in the Americas,to IPO“companies need high ARR(more than$300M-$400M NTM ARR)and good Rule of 40.”Many US tech unicorns are simply too small to be likely IPO candidates.Thirty percent of US tech unicorns have less than$100M in annual revenue.An even greater percentage are growing too slowly to be a compelling IPO.Nearly half of US VC-backed tech unicorns are growing slower than 15%annually.Profitability is also an important factor.“You need to be profitable or have a clear path to profitability.If not,you will not get a warm reception from investors,”Saxe said.Even if we consider IPO benchmarks to be relatively low:over$100M run rate for revenue,at least 15%YoY growth,and greater than negative 25%margin,only one quarter of the unicorn cohort are IPO hopefuls.But it is hard to know exactly what the benchmarks are today.They are certainly higher than they were in 2021,but few have exited to establish new benchmarks.Notes:1)Unicorn value is the last known valuation.Value of tech IPOs is the value at IPO.Source:SVB proprietary data,PitchBook Data,Inc.and SVB analysis.STATE OF THE MARKETS H1 202532Too Small:Revenue$100MToo Low Growth:Revenue growth 15%AnnuallyIPO HopefulsYears1%2%3%3%5%8%9%9%8%6%5%3%2%2%1%1%33456789 10 11 12 13 14 15 16 17 18 19 2010.3 Yrs.Average Age of UnicornsUnicorns Created in 2021All Other Unicorns10.6 Yrs.Average Age of a Tech IPO$2.4TCurrent Value of All US VC-Backed Tech Unicorns$1.3TValue of US VC-Backed Tech IPOs at Time of IPO Since 201015 yrs.of IPOs released less value than the current value of US VC-backed tech unicorns.75%pAd6Unicorns CreatedUnicorns Exited,Fallen,FailedTotal2015201620172018201920202021202220232024TotalEBITDA-25%Unprofitable:50100150200250300350400450500-900901802703604505406307208109009901080-100%00 0000P0%Down Round IPOsDespite most investors calling for a thawing of the exit market(including us),the IPO window barely cracked open a relative surprise considering US public markets were up 20% in 2024.1 While the 2024 IPO cohort wasnt mighty in numbers,it was mighty in clout.Seemingly forever-private social media platform Reddit finally went public after eyeing an IPO for years.Notable startups like Rubrik,Pony.AI and ServiceTitan also went public.So,what gives?Notably,of the seven US VC-backed tech IPOs in 2024,four of them were down rounds a popular narrative among the investor community of why some startups dont want to exit.While down rounds seem unpleasant,theyre not uncommon.Additionally,it is far from the whole story.Successful companies(such as Block)have taken down round IPOs only to soar past previous private high-water marks.Its also worth noting that with war chests still fairly full,most late-stage startups might not need the capital(even though investors would benefit from the liquidity).Despite this,still look for tech startups to test public waters should markets remain favorable.One additional wrinkle that may pressure startups to go public is IPO ratchet structures,which put startups on the clock to go public to minimize dilution impact should they trade below the hurdle price.Notably,both Block and ServiceTitan(both down round IPOs)had ratchet provisions.STATE OF THE MARKETS H1 202533Notes:1)Based on the S&P 500 price return from 12/31/2023-12/31/2024.2)Company names in order of left to right:Astera Labs,Rubrik,Ibotta,Reddit,Pony.ai,zSpace and Service Titan.Performance data as of 12/31/2024.3)Last private valuation.4)Metric as of 1,080 days post-IPO.Company names in order of appearance:Coupa,Cardlytics and Block.Source:PitchBook Data,inc.,S&P Capital IQ,S-1 filings and SVB analysis.IPO/LPV3Current MV/IPOMedian:Low PeriodMedian:High PeriodMedian OverallAverage OverallShare of Down Round IPOsIPO DayLPV3989x42,695x4738x4-30%-100P 1020112012201320142015201620172018201920202021202220232024Low PeriodHigh PeriodHigh PeriodLow PeriodSTATE OF THE MARKETS H1 202534Notes:1)Revenue growth determined using latest annualized quarterly revenue at time of IPO.If quarterly data is not provided,then the available time frame provided by the company is used.Earnings before interest,taxes,depreciation and amortization(EBITDA)margins determined using latest quarterly data at time of IPO.2)VC-backed determined using SVB analysis of previous equity rounds.3)Tech determined using SVB analysis and taxonomy.4)Revenue size determined using revenue level provided by PItchBook Data,Inc.Source:PitchBook Data,Inc.,S&P Capital IQ,S-1 filings and SVB analysis.201920202022-202320242021Bubbles Scaled to Revenue Size4-100%-50%0P00 0%000500%-150%-125%-100%-75%-50%-25%0%P%Revenue Growth YoYEBITDA MarginUberRobinhoodDoorDashMaplebearLyftAirbnbPalantirToastCompassAppLovinPelotonCoinbaseTuSimpleCompanies are scraping to the bone when it comes to exhausting all options before exploring an acquisition.At least thats what it seems.To start,companies are being acquired closer to the end of their runway.Median cash runway at time of acquisition fell 35%to just below six months,dropping for the first time since 2019.Financials tell a similar story.Pre-pandemic revenue growth hovered around 10%to 20%and EBITDA margins -80%to-100%at time of acquisition.Those figures(on a median basis)have slipped lower the past two years.In fact,revenue growth trends downward leading up to an acquisition.This is in stark contrast compared to the frothier times of 2020-2022.Revenue growth held fairly steady leading up to an acquisition,potentially suggesting that more deals were strategic rather than done out of necessity.In todays climate,more startups are likely forced to look for a new home and subsequently lose revenue sources leading up to that.Another data point that supports this thesis is the share of deals that report a valuation.Out of nearly 900 US VC-backed M&A deals done in 2024,only 18%disclosed a purchase price.1 While data may be backfilled as more information becomes available,its unlikely this number will reach peaks of previous years.In part,this is attributed to the fact startups are getting acquired for much less than what they raised and were valued at.See our previous analysis from last years State of the Markets here.STATE OF THE MARKETS H1 202535Notes:1)VC-backed determined using SVB analysis of previous equity rounds.2)Revenue growth determined by annualizing a companys revenue on its most recent statement.Source:PitchBook Data,Inc.,SVB proprietary data and SVB analysis.82uved%4.1 Mos.4.5 Mos.2.7 Mos.6.1 Mos.7.4 Mos.8.4 Mos.5.5 Mos.2017201820192020202120222023-2024-140%-120%-100%-80%-60%-40%-20%0%0 0 172020202120222023-202420182019Revenue Growth YoYEBITDA MarginRunway at PurchaseShare with Less than 12 Mos.Runway51PF%3371%4 Quarters Prior3 Quarters Prior2 Quarters Prior1 Quarter Prior2018-20192020-20222023-202420192024202220232020202131&06%$0.0T$0.2T$0.4T$0.6T$0.8T$1.0T$1.2T$1.4T200020022004200620082010201220142016201820202022202412s%With the growth of unrealized returns and limited exit activity,GPs,LPs and employees are hungry for liquidity.“Momentum is building within the ecosystem for alternative paths to liquidity,”says Eric Thomassian,Head of Private Company Relations at Forge Global.Enter secondary markets.GPs of venture firms are selling down positions to reduce exposure to bets placed in 2021,and to boost DPI before their next fund.In some instances,smaller GPs are selling off their entire portfolios.GPs at asset managers and hedge funds are using secondaries as an off-ramp for private exposure and reducing growth investing.Some LPs like family offices,pension funds and endowments are selling co-investments and fund interests.While employees are selling options.But the secondary market is challenging,with its limited price transparency,inefficient price discovery,extended settlement cycles,high transaction costs and stock transfer restrictions.That said,it is increasingly more transparent and accessible with the rise of secondary exchanges like Forge Global,Nasdaq Private Markets and others.Furthermore,the growth of VC-specific secondary funds creates more opportunities for transaction.While secondary transaction volumes are elevated,73%of investors have not participated in secondary markets.There is still a lot of opportunity for growth as markets become more liquid and efficiency improves.STATE OF THE MARKETS H1 202536Notes:1)Total unrealized returns in the US innovation economy as of year end;2024 data as of March 2024.2)Data smoothed using trailing six months.3)Pitchbook Data,Inc.survey of global venture investors from 2024.Source:Forge Global,Preqin,PitchBook Data,Inc.and SVB analysis.No,we do not participate in secondaries.Yes,we are buying through secondaries.Yes,we are selling through secondaries.050100150200250Mar.20Jun.20Sep.20Dec.20Mar.21Jun.21Sep.21Dec.21Mar.22Jun.22Sep.22Dec.22Mar.23Jun.23Sep.23Dec.23Mar.24Jun.24Sep.24-70%-60%-50%-40%-30%-20%-10%0 %Jan.21 Jul.21 Jan.22 Jul.22 Jan.23 Jul.23 Jan.24 Jul.2475%of funds since 2015 have a DPI less than 25%.37Marc CadieuxPresidentSVB Commercial BankSilicon Valley BEli OftedalSenior Analytics ResearcherSVB Market InsightsSilicon Valley BMark Gallagher Head of Investor CoverageSVB Commercial BankSilicon Valley BJosh PherigoSenior Analytics ResearcherSVB Market InsightsSilicon Valley BAndrew Pardo,CFASenior Analytics ResearcherSVB Market InsightsSilicon Valley BMarc Cadieux is president of Silicon Valley Banks commercial banking business where he focuses on the needs of innovation companies at all stages of development,including the investors who back them.Marcs career at Silicon Valley Bank,a division of First Citizens Bank,began in 1992.In the three decades since,he has held a variety of top credit and sales roles serving some of the worlds most innovative companies.Most recently,he served as chief credit officer,appointed in 2013,and oversaw credit policy and process,credit underwriting,loan approval and portfolio management activities.He is a strong advocate of bank initiatives to expand opportunities for those who are underrepresented in the innovation economy.He serves as an executive sponsor for the companys employee resource group focused on women employees.Mark Gallagher is the co-head of the investor coverage practice.He and his team provide tailored services,industry insights and strategic guidance to top investors in the innovation economy.Mark has served as a financial partner to venture capital firms and technology and life science companies for the majority of his career.During his 22-year tenure with SVB,he has been involved in a number of strategic projects and initiatives,most recently leading the corporate venture capital practice.Hes held numerous leadership roles including head of the Northeast technology banking practice,head of business development in New England and several years running the Northeast life science practice.A supporter and champion of the New England technology community,Mark serves as a board member for BUILD Boston and was formerly on the board of overseers for The Mass Technology Leadership Council(MTLC).STATE OF THE MARKETS H1 2025Jake Ledbetter,CFASr.Analytics ResearcherSVB Market InsightsSilicon Valley BSilicon Valley Bank(SVB),a division of First Citizens Bank,is the bank of some of the worlds most innovative companies and investors.SVB provides commercial banking to companies in the technology,life science and healthcare,private equity and venture capital industries.SVB operates in centers of innovation throughout the United States,serving the unique needs of its dynamic clients with deep sector expertise,insights and connections.SVBs parent company,First Citizens BancShares,Inc.(NASDAQ:FCNCA),is a top 20 U.S.financial institution with more than$200 billion in assets.First Citizens Bank,Member FDIC.Learn more at .Silicon Valley B See complete disclaimers on following page.2025 First-Citizens Bank&Trust Company.Silicon Valley Bank,a division of First-Citizens Bank&Trust Company.Member FDIC.38STATE OF THE MARKETS H1 202538The views expressed in this report are solely those of the authors and do not necessarily reflect the views of SVB.This material,including without limitation to the statistical information herein,is provided for informational purposes only.The material is based in part on information from third-party sources that we believe to be reliable but which has not been independently verified by us,and,as such,we do not represent the information is accurate or complete.The information should not be viewed as tax,accounting,investment,legal or other advice,nor is it to be relied on in making an investment or other decision.You should obtain relevant and specific professional advice before making any investment decision.Nothing relating to the material should be construed as a solicitation,offer or recommendation to acquire or dispose of any investment,or to engage in any other transaction.All non-SVB named companies listed throughout this document,as represented with the various statistical,thoughts,analysis and insights shared in this document,are independent third parties and are not affiliated with Silicon Valley Bank,division of First-Citizens Bank&Trust Company.Any predictions are based on subjective assessments and assumptions.Accordingly,any predictions,projections or analysis should not be viewed as factual and should not be relied upon as an accurate prediction of future results.Investment Products:Are not insured by the FDIC or any other federal government agencyAre not deposits of or guaranteed by a bankMay lose value2025 First-Citizens Bank&Trust Company.Silicon Valley Bank,a division of First-Citizens Bank&Trust Company.Member FDIC.39STATE OF THE MARKETS H1 202539
2025-02-07
39页
5星级
Public Disclosure AuthorizedPublic Disclosure AuthorizedPublic Disclosure AuthorizedPublic Disclosu.
2025-02-07
90页
5星级
Public Disclosure AuthorizedPublic Disclosure AuthorizedPublic Disclosure AuthorizedPublic Disclosu.
2025-02-06
33页
5星级
OECD Economic Surveys:Chile 2025January 2025Volume 2025/1versionlaunchOECD Economic Surveys:Chile202.
2025-02-05
130页
5星级
Photo by Rachael Rinchiuso Kearney,ChicagoSlow and steady?Global Economic Outlook 202420282H 2024 up.
2025-01-23
13页
5星级
August 2024Cabinet Office,Government of JapanProvisional TranslationEconomic White Paper 2024Annual Report on the Japanese Economy and Public Finance Toward a Vibrant New Economic StageSummaryContentsThis material has been tentatively prepared to explain the Annual Report on the Japanese Economy and Public Finance.For quotations and other purposes,please refer to the text of the Annual Report on the Japanese Economy and Public Finance.Chapter 1.Macroeconomic Trends and ChallengesSection 1.Trends in the MacroeconomySection 2.Establishing an Economic Structure That Will Not Go Back to Deflation Chapter 2.Overcoming Growth Constraints Due to Labor ShortagesSection 1.Status of Labor Shortages and Measures TakenSection 2.Issues Concerning Labor MobilitySection 3.Issues Concerning Foreign Workers in Japan Chapter 3.Utilizing Assets Toward an Enriched Socio-EconomySection 1.Household Financial Assets and Investments Section 2.Prospects for and Issues Concerning Housing StockSection 3.Utilizing Human Capital of Elderly Workers Nominal GDP increased to a record high level of 597 trillion yen(P1,Fig 1).Corporate profits hit a record high,and appetite for businessinvestment remained strong(P5,Figs 1&3).In contrast,private consumption lacks strength(P2,Fig 1)Cash and deposit holdings of the corporate sector stand out internationally.Underinvestment so far has contributed to the low potential growth rate(P6.Figs 1&2).CPI has risen in the range of 2%(YoY)since last fall(P9,Fig 1).As for the wage growth deflated by price level(i.e.real wage),hourly wages of part-time workers have increased since mid-2023,and the rate of decrease in monthly wages of full-time workers has been consistently slowing(P11,Fig 3).The nominal wage hike in the 2024 annual negotiation(shunto)hits the highest in the last 33 years.More enterprises decided on higher wage increases than in the last year(P10,Figs 1&2).Higher growth in payrolls is expected in the months ahead.Pass-through from purchase to sales prices almost returned to the pre-deflation period(1980s-early 1990s)(P12,Fig 1).Promoting pass-through of labor costs in SMEs remains important for realizing a virtuous cycle of wages and prices.Service trade deficit increases primarily in digital-related areas.Domestic demand surges in areas where foreign companies have an advantage(P7,Figs 1&2).Strengthening the earning power in potentially competitive areas is a challenge.Number of users of spot-work apps increased approx.70-fold in 4 years.Recruiting channels in the labor market are diverse partially thanks to DX(P4,Figs 2&4).Chapter 1.Macroeconomic Trends and Challenges Labor shortages in corporations reached a record high level,with the increase in career-changers and the fiercer competition for talent(P15,Fig 3).Corporations strengthen wage hikes and labor-saving investment against labor shortages(P15,Fig 4).Labor-saving investment contributes to improved labor productivity(P17,Fig 2).Job opening to application ratios are higher for construction workers and caregivers amid the labor shortage,while the ratio is 0.4 for clerical workers(P19,Fig 2).Reskilling is increasingly important since clerical workers jobs are potentially substituted by AI(P19,Fig 3)Foreign workers have increased,accounting for 3.4%of total employees.The wage gap with Japanese workers is reduced to 7%by controlling various attributes.Wages of skilled workers living in Japan for a long period is higher than those of Japanese counterparts.(P22,Figs 1,4&5)Chapter 2.Overcoming Growth Constraints Due to Labor Shortages Households financial assets have increased by 60%since 2000,though the majority is explained by cash and deposits(P23,Figs 1&2).The elderly dont dissave the assets due to longevity risk.Meanwhile,younger generations are increasing their risk appetite(P25.Figs 1&2).Secondhand housing acquisition increases regardless of income group,now accounting for a quarter of total housing acquisition.(P29,Fig 1).Further promoting the circulation of existing homes is important for realizing a better QoL.Labor participation rates of the elderly in Japan,both male and female,stand out among the major advanced economies(P31,Fig 1).People willing to work beyond 65 are increasing(P32,Fig 2).It is essential to inform people of how not adjusting employment affects lifetime income and to review various systems.The ratio of companies that set the wage level after mandatory retirement age(teinen)above 80%of the pre-retirement wage increased to around 40%(P33,Fig 1).Chapter 3.Utilizing Assets Toward an Enriched Socio-EconomyFigure 1.Trend of GDP597 555 480500520540560580600 20121314151617181920212223 24(s.a.,tril.yen)NominalRealQuarterCY7.08.520253035401471014710147101462021222324(10,000 units)3-month moving averageMonthlyMonthCY70758085909510010511011520192021222324(Q4 2019=100)QuarterCYU.S.JapanU.K.FranceGermanyChapter 1 Section 1:Trends in the Macroeconomy(1)-GDP Sources:Cabinet Office,Japan Automobile Dealers Association,Japan Light Motor Vehicle and Motorcycle Association,U.S.Department of Commerce,U.K.Office for National Statistics,German Federal Statistical Office,and INSEENominal GDP is on an increasing trend and reached 597 trillion yen,while real GDP decreased in 2024 Q1 due to special factors like the suspension of production and shipment by some automotive manufacturers.Special factors in 2024 Q1(real growth rate:0.7%)Figure 2.New automobile salesNew automobile sales are picking up after the suspension of production and shipment last December.Figure 3.Real private consumption in major advanced economiesHowever,private consumption lacks strength compared to other major advanced economies.M/MSpecial FactorsEstimated EffectsImpact of the 2024 Noto Peninsula EarthquakeAbout 0.1%pt(1)Suspension of production and shipment by some automotive manufacturersOn consumption(durable goods)Almost all of 0.5%pt(2)On investment(transport equipment)About 0.1%pt(3)Reactionary fall in service exportAbout 0.6%pt(4)1:Direct effect on GDP,estimated by Cabinet Office2:Contribution of durable goods(some comes from decline in cell phone)3:Contribution of transport equipment4:Contribution of service export,excluding direct domestic purchase by non-resident householdsFigure 2.Consumer sentiment and expected inflationFigure 4.The effect on consumption by 1%increase in asset value held by households(wealth effect)-40-20020406080100120 202021222324(10,000yen)(2)elderly unemployed householdsQuarterCYSources:Cabinet Office,Bank of Japan,Ministry of Internal Affairs and Communications,and MIC.Accumulated excess savings during the Covid-19 pandemic remains at a high level for workers households,especially high-income householders,while it is depleted in elderly,low-income,and unemployed households.240250260270280290300310320330201718192021222324(s.a.,tril.yen)Estimated level from real disposable income,etc.Actual levelQuarterCYIn 2024.Q1,the actual level decreased mainly because of the effects of suspension of production and shipment by some automotive manufacturers.Figure 1.Real private consumptionFigure 3.Excess savings by household attributes(with 2 or more people)0.00.51.01.52.02.53.03.54.04.520253035404550201012141618202224Consumer confidence indexConsumers expectation of inflation(right scale)()(CY)Private consumption is gradually going back to a level that can be explained by disposable income and other factors following Covid-19s categorization as Class 5.During the recent phase of price increase,consumer confidence index shows a strong inverse correlation with households inflation expectation.The effect by capital gain increases after 2013.Encouraging a shift from saving to investment is expected to stimulate consumption.050100150200250 202021222324(10,000yen)(1)working householdsQuarterCYAverage1st quartile2nd quartile3rd quartile4th quartile5th quartileAverageLow incomeHigh income00.010.020.03TotalSelectiveexpenditureBasicexpenditure2003-20122013-2023(%)*Chapter 1 Section 1:Trends in the Macroeconomy(2)Private Consumption-30-20-1001020301 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5 9 1 5201415161718192021222324Transition probabilityfor Male,15-54 years oldNot in labor forceparticipation levelTransition probability for older people,over 55 years oldTransition probabilityfor Female,15-54 years oldTotal number of transitions from Not in labor force to Labor force(broken line)(12-month moving average,Difference from Jan 2014,10,000 people)Chapter 1 Section 1:Trends in the Macroeconomy(3)-Labor MarketSource:Ministry of Internal Affairs and CommunicationsEven amid the decreasing trend in the working agepopulation,the number of employees has been increasing,led by the increase in female employees.Figure 2.Transitions from a population not in the labor force to a population in the labor forceMonthCY562-200-100010020030040050060070017171717171717171717171 520131415161718192021222324(Difference from Jan 2013,10,000 people)MonthCYMale(Non-regular employee)Female(Non-regular employee)Male(Regular employee)Female(Regular employee)Employee,excluding executive of company or corporation(broken line)Attention should be given to the flat trend after early 2024 in the flow from“not in labor force”to“labor force”in female,15-54 years old.Figure 1.Trends in the number of employees010203040506070Job-huntingwebsiteHelloworkOwnedmediaReferralInformationsessionHead huntingSocial recruitingAlumni(%)5 years agoPresent-1,00001,0002,0003,0004,0005,0006,0007,00017171732021222324Sales and marketingCustomer service and servingTransportationCleaning,etc.OtherTotal(12-month moving average,difference from Apr 2020,%)MonthCY4060801001201401601802002202401718192021222324(CY)(3rd week of February 2017=100)Job openings for part-time workers(Private employment agency)Job openings for part-time workers(Public employment security office and private employment agency)Figure 2.Channels for recruiting full-time workers(mid-career)important channels for firms(based on survey)Chapter 1 Section 1:Trends in the Macroeconomy(4)-Labor MarketSources:Cabinet Office,Ministry of Health,Labor and Welfare,Nowcast Inc.,and Timee Inc.7080901001101201301401501601718192021222324(CY)(3rd week of February 2017=100)Job openings for full-time workers(Public employment security office and private employment agency)Job openings for full-time workers(Public employment security)Figure 1.Job openings for full-time workersFigure 4.Total users for spot-work apps Figure 3.Job openings for part-time workersAs firms diversify recruiting channels,the number of job openings for full-time workers observed by big data shows steady performance compared to the official statistics collected by“Hello Work.”Job openings for part-time workers using channels by private firms also greatly increased.In addition,the number of users of spot-work apps increased approx.70-fold in 4 years.With DX,matching in the labor market is entering a new phase.201720177.49.410.60481216MarchJuneSeptemberDecemberForcastActual result(Nominal,YoY,%)FY2022FY2023FY20240123456789100510152025301985909520000510152024(tril.yen)Ordinary profitsOperating profitsRatio of ordinary profits to sales(right scale)Ratio of operating profits to sales(right scale)(%)(CY)Chapter 1 Section 1:Trends in the Macroeconomy(5)-Business SectorSources:Ministry of Finance,Bank of Japan,and Cabinet OfficeDue to labor force shortages,many construction firms adjust the amount of receiving orders.Attention should be given to the fact that such supply constraints can restrain investment in buildings and structures.Figure 4.Ratio of firms adjusting the amount of receiving orders due to labor force shortagesFigure 1.Corporate profit and profit margin10.810.69.4-24-20-16-12-8-404812162019909520000510152023（Nominal,change from previous FY,%）Business investment in tangible assets(incl.investment in land)Business investment including computer software and R&DFigure 2.Realized investment(Tankan survey,all sector,all size)Corporate profits hit a record high both in terms of amount of ordinary/operating profits and operating profit margin,showing a steady trend.Realized investment(in terms of YoY)in FY 2023 is at the highest level since FY 1991.Figure 3.Planned investment in FY 2024(Tankan Survey,all sector,all size)Planned investment in FY 2024 shows a two-digit increase in June,which shows firms continued strong motivation for investment.(CY)Business investment including computer software and R&D49.9 33.3 28.4 24.2 22.6 22.2 17.6 11.8 0102030405060ConstructionAccommodations,eating and drinking servicesAll industriesTransport andpostal activitiesManufacturingReal estate,goods rentaland leasingWholesalingRetailing(%of firms)Chapter 1 Section 1:Trends in the Macroeconomy(6)-Business SectorSources:Bank of Japan,Cabinet Office,and OECD.While corporate profits are showing steady performance,the amount of accumulated cash and deposits in the corporate sectors stand out compared to major advanced economies and is showing an increasing trend,reaching 60%of GDP.Figure 2.Potential growth rate59.317.3 20.5 32.5 38.4 0102030405060701995 9799010305070911131517192123(of GDP)JapanU.S.GermanyU.K.France(CY)The potential growth rate in Japan remains the lowest in major advanced economies,due to the negative contribution from labor and the shrinking contribution from capital.It is crucial to direct accumulated cash to investment and higher wages.-101234FranceGermanyU.K.U.S.JapanFranceGermanyU.K.U.S.JapanFranceGermanyU.K.U.S.JapanFranceGermanyU.K.U.S.Japan198699200009201019202023Labor InputCapital InputTotal Factor ProductivityPotential Growth Rate(%)(CY)Figure 1.Accumulated cash and deposits in corporate sector-10-8-6-4-2024620141618202223(Tril.Yen)Total service balanceComputer serviceProfessional&management consulting servicesCopyrights and other royaltiesInsurance and annuity servicesTravelOther(CY)2.1 3.3 0.7 1.2 0.8 2.8 0.00.51.01.52.02.53.03.54.0201920232019202320192023広告動画音楽配信電子書籍(Tril.Yen)83250Chapter 1 Section 1:Trends in the Macroeconomy(7)-Imports and ExportsSources:Ministry of Finance,Bank of Japan,and various sources for each industryFigure 3.Relationship between exports and exchange rate,effects from exchange rate shock on export prices and export volume*-2.0-1.5-1.0-0.50.00.51.01.52.02.5Real goods exportExport prices(contract currency)1975-20092010-2023(Cumulative change in the 6th months due to 10preciation,)58Figure 2.Domestic market size of selected digital services4050607080901001108090100110120147101471014710147101471014520192021222324(2020100,3MA)MonthCY（2020100）Yen appreciationYen depreciationReal effective exchange rate Real exportExports volume indexTrade deficit in services are increasing primarily in digital industries and insurance.Domestic demand is surging for digital products in which foreign companies have a comparative advantage.Measures to improve competitiveness in fields with advantages is essential.Exports of goods is pausing for increase even aimed the trend of the depreciating yen.Recently,the effects from the exchange rate shock have become smaller due to the expansion of overseas production and the change in price-setting patterns in foreign markets.Internet advertisementMovie and music streaming/e-booksPublic cloud serviceFigure 1.Net deficit of trade in service05101520255678910Average20s30s40s50sRatio to disposable income(right scale)Monthly Payment-3,000-2,000-1,00001,0002,0003,000Two-ormore-personhouseholdsWorkers householdsHouseholds payingback debts for housesDepositOther savings Liabilities for houses and lands(reverse sign)Other liabilities(reverse sign)(10,000 yen)As for the impact of interest hikes on households,average households will benefit from income gain,as their financial assets outweigh their loans.Households with a housing loan(approx.a quarter of households with two or more people)could suffer from higher loan repayments.The impact could be larger for younger households with a relatively high debt service ratio,though the share of younger households with housing loan payments is limited.-100%-80%-60%-40%-20%0 %Real effective exchange rateFactors of nominal effective exchange rateRelative price factor(Cumulative percent change since April 1995)Figure 1.Trend in real effective exchange rate-0.50.00.51.01.52.02.53.03.5(%)Uncollateralized overnight call rateMortgage loan floating Interest ratesShort-term prime rate0.020.190.000.100.200.300.400.500.600.70(%)Term deposit interest rateDeposit interest rateChapter 1 Section 1:Trends in the Macroeconomy(8)-Exchange and Interest RateSources:Ministry of Internal Affairs and Communications,Bank of Japan,and BloombergFigure 3.Balance sheet in households(with two or more people)(56%)(24%)(23%)(0.3%)(5.1%)(8.8%)(6.4%)Figure 4.Amount of repayment in housing loans(households with two or more people)The long-term trend of depreciation in the real effective exchange rate caused by domestic inflation has remained lower compared to trading partners,while,most recently,the widening gap between domestic and foreign interest rates is causing yen depreciation.(CY)After the regime change in monetary policy(March 2024),the short-term prime rate remains flat while the deposit interest rate is increasing.Figure 2.Trend in short-term interest rates(CY)(CY)(10,000 yen)(%)-2-1012345614710147101471014 52021222324ServicesElectricity&city gasAll itemsGasoline&keroseneFood,less fresh foodFresh foodOther goods(Contribution to yearly change,%)All items,less fresh food(core)All items,less fresh food and energy(Core-core)HighestJan.2023.4.3(All items)6MonthCYChapter 1 Section 2:Establishing an Economic Structure That Will Not Go Back to Deflation(1)Sources:Ministry of Internal Affairs and Communications;Bank of Japan;Teikoku Databank Ltd.Business View:Price revision trend survey of“195 major food and beverage makers”(July 2024).The dotted line in Figure 2 shows the YoY rate in which the impact of the revision of the Renewable Energy Surcharge Rate is excluded.The pass-through from yen depreciation to consumer goods price is strengthening.Attention should be paid to developments in domestic prices in the near future.Figure 1.Consumer price indexFigure 4.Accumulated response of CPI for goods to exchange rates(Depreciation of yen,one-standard deviation)Figure 2.Consumer price index(goods)and import price index0.000.040.080.120.160.20123456789101112(%)(Period)2000-20242000-2019Figure 3.Factors behind food price hike More firms increasingly point to the passing through of labor cost(due to wage hikes)and yen depreciation as the factors behind food price hikes.The rate of increase in goods prices is only moderately declining amid decreasing import prices.Since 2023 fall,consumer prices are increasing in a range of about 2%.-50510-30-20-10010203040506020111520222324（YoY,）IPI（6 months ahead,yen basis）CPI(Goods)（less fresh food,excluding policy factors）（right scale）(Year)（YoY,）020406080100Ingredient costsEnergy pricesMaterial and packaging costsLogistics costsDepreciation of YenLabor costs2023Jan.-Nov.2024Jan.-Nov.(%)5.663.585.102.123.563.7-3-2-1012345671990929496982000020406081012141618202224(YoY,%)(FY)Shunto wage increase rate(including regular salary increase)Base pay onlyCPI(All items,excluding the impact of consumption tax)Figure 1.Shunto(annual wage negotiation)wage increase rate and price increase rate00.10.20.30.40.5012345678(Kernel Density)FY2024FY2023(YoY,%)Sources:Japanese Trade Union Confederation,Central Labour Relations Commission,Japan,Ministry of Internal Affairs and Communications,and Payroll Inc.In 2024,in addition to a higher wage increase rate in the younger generation than last year,wages for 40s are also increasing,showing a more wide-spread trend for wage hike.Figure 2.Distribution of Shunto wage increase rates(base pay)Figure 3.Big data on wage increase rate by age groupIn the 2024 negotiations,higher wage increase rates are agreed in more firms compared to 2023.Surpassing the increase rate in 2023,the wage hike agreed in“Shunto”is 5.1%(3.56%for base pay only),the highest in 33 years.0.4 3.1 4.2 0.5 2.4 3.6 0.4-0.1 2.7 0.1-0.1 1.0-1.00.01.02.03.04.05.0202220232024(YoY,%)Under 2930-3940-4950-59Apr.-Jun.Apr.-Jul.Apr.-Jul.Chapter 1 Section 2:Establishing an Economic Structure That Will Not Go Back to Deflation(2)1.8-3-2-101234201012141618202224Ratio of part-time workersPart-time workerTotal cash earnings(line)Full-time workerCY(YoY,%)Jan.-May Average1520253035200205081114172023(%)Monthly Labor Survey(Survey answered by firms,based on number of jobs)Labor Force Survey(Survey answered by workers)Part-time jobLabor Force Survey(survey answered by workers)Employees who work less than 35 hours per weekChapter 1 Section 2:Establishing an Economic Structure That Will Not Go Back to Deflation(3)Sources:Ministry of Health,Labor and Welfare and Ministry of Internal Affairs and Communications.In figure 3,nominal wages are divided by CPI(all items).Partly due to the increase in side-work,the ratio of part-time workers tends to show an increasing trend in the survey answered by firms(Monthly Labor Survey),because one job is counted as one worker.Figure 1.Trend in nominal wage growthFigure 3.Real wage growth rate by employment type Figure 2.Various indicators of the ratio of part-time workers(CY)-0.20.1-0.31.2-4-3-2-10123147101471014520222324(YoY,%)Part-time workerhourly wage(Establishment withfive or more employees)Full-time workertotal cash earnings(Establishment withfive or more employees)Full-time workerContractual Cash Earnings(Establishment with five or more employees）Full-time workercontractual cash earnings(Establishment with 30 or more employees)MonthCYThe rise in the ratio of part-time workers puts downward pressure on the average wage increase rate.It is meaningful to see wages by employment type in evaluating trends.In terms of real wages(wages in which inflation is taken into account),hourly wages for part-time workers have been rising(YoY)after mid-2023.For full-time workers,the decreasing rate of monthly wages(contractual cash earnings)is shrinking.For those working at establishments with 30 or more employees,wages have increased for the first time in 26 months.-1.5-1.0-0.50.00.51.01.52.02.53.01471014710147101462021222324(Contribution to yearly change,%)Transportation&CommunicationCulture&RecreationServices with high labor cost ratio（line graph）Medical careEducationMonthCY90951001051101151201251471014710147101471015202021222324(CY2020=100)Japan,service sector wagesJapan,CPI for servicesU.S.,CPI for servicesU.S.,service sector wagesMonthCY-40-30-20-1001020304050-2502550752020-2024(present)201320191980-1996(Pre-Deflation)1997-2012(Deflationary Period)(Change in Purchasing Prices Diffusion Index,point)(Change in selling prices diffusion index,point)Chapter 1 Section 2:Establishing an Economic Structure That Will Not Go Back to Deflation(4)The extent of pass-through from purchasing prices to selling prices has almost returned to that of the pre-deflation era.Figure 2.Wages and prices in the service sector in Japan and the U.S.Figure 1.Pass-through from purchasing prices to selling prices(non-manufacturing)Figure 3.CPI for services with high labor cost ratioThe price increase rate for service sectors with a high labor cost ratio is steadily becoming larger.Promoting the pass-through of labor costs in SMEs remains important for realizing a virtuous cycle of wages and prices.Both wages and prices in the service sector have turned to a moderate upward trend.Sources:Ministry of Health,Labor and Welfare;Ministry of Internal Affairs and Communications,Bank of Japan,and U.S.Bureau of Labor Statistics.Figure 2.International comparison of public service prices901001101201301401 7 1 7 1 7 1 7 1 7 1 7 1 7 1 7 1 7 1 7 1 7 1 620131415161718192021222324(CY2015=100)U.S.E.U.JapanMonthCYChapter 1 Section 2:Establishing an Economic Structure That Will Not Go Back to Deflation(5)The increasing rates in public service prices show different trends depending on their types of revision processes.In the period to establish the norm that both prices and wages rise,it is crucial to strike a balance between appropriate pass-through for wage hikes in the public service sector and stabilizing daily life for the people.Figure 3.CPI for public services(By revision process type)0.3 0.0 3.8-8-6-4-20246147101471014620222324(YoY,%)General services,less house rent(private)&imputed rent【19.3(weight)】Public services,less house rent(public)【11.9(weight)】Rent【18.3(Weight)】MonthCYFigure 1.CPI for services96981001021041061081101471014710147101462021222324(CY2020100)Notified prices（Forwarding charges,airplane fares etc.）Licensed prices(Railway fares,taxi fares,etc.)Prices determined by central government or the Diet(Medical treatment,charges for nursing care,etc.)Prices determined by local governments(Nursery school fees,sewerage charges etc.)MonthCYAmong service prices,the prices of public services and rent are almost flat.Unlike in Japan,public service prices show increasing trends in other advanced countries.Sources:Ministry of Internal Affairs and Communications,U.S.Bureau of Labor Statistics,Eurostat,and Consumer Affairs Agency.Figure 2.Relationship between price inflation and GDP gap(1)Rent-2.0-1.00.01.02.03.04.0-10-9-8-7-6-5-4-3-2-1012345(YoY change in core-core,%)1984-20002001-2019(GDP gap(Lag 2、%)95100105110115120201718192021222324(Jan-Mar 2017=100)Japan,CPI for rent of primary residenceFrance,CPI for rent of primary residenceFrance,CPI(All items,less rent and cigarettes)France,rent reference index(IRL)Japan,CPI(All items,less rent)QuarterCYChapter 1 Section 2:Establishing an Economic Structure That Will Not Go Back to Deflation(6)Sources:Ministry of Internal Affairs and Communications,Institut National de Statistique et dEconomie,France,and Cabinet OfficeRent has been less responsive to the output gap compared to prices of other goods and services.Rent is expected to start increasing after income grows at a pace greater than price hikes.Figure 1.Rent in Japan and France-2.0-1.00.01.02.03.04.0-10-9-8-7-6-5-4-3-2-1012345(YoY change in rent,%)(GDP gap(Lag 2,%)2001-20191984-2000Rent in France is increasing while general price increase is capped by regulations.On the other hand,rent in Japan is almost flat amid the price hike in other goods.(2)All items,less fresh food and energy(core-core)020406080Improve working conditions for existing staffRecruit more staff(both new graduates and mid-career workers)Train staffHike mandatory retirement age and extend the scheme to re-hire after retirementLoosen the requirement for new staffInvestment for saving laborHire staff who once worked at the company but leftFind new counterparts for outsourcing0102030405060708090100020406080100Young(34 years old)Middle-aged(35-54 years old)Older(55 years old and up)(Current labor shortage rate,%)(Labor shortage rate 5 years ago,%)Labor shortage rate has risen Sources:Bank of Japan and Cabinet Office 0102030405060708090100-60-40-200204060(DI,“excessive”“insufficient”,%pt)(CY)Non-manufacturingManufacturingTotalFigure 1.Employment Condition DI(Labor shortage perceived by firms)Figure 2.Changes in labor shortages among businessesFigure 3.Factors preventing businesses from resolving labor shortagesFigure 4.Measures taken by businesses to cope with labor shortagesThe labor shortage among businesses,especially in the non-manufacturing sector,is at a historically high level since the bubble.Shaded areas indicates a recession.Compared to pre-COVID,labor shortages have risen among middle-aged workers in addition to younger workers.The increasing number of job transfers and the resulting intensification of recruiting competition are behind the labor shortage among businesses.Companies are providing better treatment to their workers in response to the labor shortage and more firms have increased labor saving investment.Chapter 2 Section 1:Status of Labor Shortages and Measures Taken(1)Survey in 2018Survey in 2024(%)Survey in 2019Survey in 2024(%)020406080Few applicants applyApplicants dont have sufficient skill/abilityApplicants get jobs at other firms with better offersNew staff leaves the firms within a short timeCannot afford to train new staffWorkload increases at a higher pace than the number of new staffOther reasonsChapter 2 Section 1:Status of Labor Shortages and Measures Taken(2)Sources:Ministry of Health,Labour and Welfare,Ministry of Internal Affairs and Communications and Recruit Works Institute-1.00.01.02.03.04.05.0-0.8-0.6-0.4-0.200.20.40.6(Active job opening ratio(difference from the average)(Scheduled cash earnings per hour(YoY,%)2013-2023(present)-4.0-2.00.02.04.06.0-0.6-0.4-0.200.20.40.60.8(Active job opening ratio(difference from the average)(Scheduled cash earnings per hour(YoY,%)2013-2023(present)1994-2000(before deflation)2001-2012(under deflation)Figure 1.Labor market tightness and wage growth rate(2)Full-time workers(1)Part-time workers1994-2000(before deflation)2001-2012(under deflation)Wages for part-time workers are rising in response to the tighter labor market and those for full-time workers are returning to the figure where the balance between labor supply and demand is determined.Figure 2.Non-labor force population and reservation wage of females-20-15-10-50510 20192021222324(Difference from Q1 2019,%)45-54 years old55-64 years old25-34 years old35-44 years oldQuarterCY102103104 20192021222324(Q1 2012=100)QuarterCY(1)Non-labor force population of females(2)Reservation wage of femalesThe non-labor force population is aging due to the labor participation of females and population changes.Consequently,the reservation wage of females is rising.020406080100High installation costsHigh running costsTraining for staff is requiredHiring new specialists is requiredEffect is uncertainNo idea what tools to be introducedConcern of quality decrease in products and servicesOther reasons(%)ManufacturingNon-manufacturingChapter 2 Section 1:Status of Labor Shortages and Measures Taken(3)(Sources)Bank of Japan,Cabinet Office and Ministry of Economy,Trade and Industry-505101520有形固定資産(YoY,%(average in FY2013-22)Excessive AppropriateInsufficient-505101520有形固定資産Excessive AppropriateInsufficient(YoY,%(average in FY2013-22)Figure 1.Business investment byemployment condition DIFigure 4.Barriers to labor-saving investmentFigure 2.Labor productivity improvement through labor saving investmentFigure 3.Impact of a fall in capital goods price on labor share(1)Large firms(2)SMEsFirms facing labor shortages are more actively investing in capital.Software investment is particularly active among SMEs.Firms that increased labor saving investment such as automation and AI experienced productivity improvementIn recent years,software investment has been labor-neutral while machine investment became labor-substitutional.In addition to cost,labor-saving investment is hampered by a lack of human resources capable of handling new technology.*0102030405060Labor saving investmentRobotics and automation of customer service,etc.RPAWeb and IT-related software and systems(%)Tangible fixed assetTangible fixed assetSoftwareSoftware-0.3-0.2-0.10.00.10.20.3機械設備(%)*FY2000-2010FY2011-2021Substitutive for laborComplementary for laborSoftwareMachinery0102030405060012345678910JapanGermanyU.S.FranceUnemployment rateRatio of the long-term unemployment(right scale)(%)(%)00.20.40.60.810.20.40.60.81.01.2(Number of new hires/Number of unemployed）JapanGermanyUnited States(Number of jobs available/Number of unemployed)Difficult to find employment0Chapter 2 Section 2:Issues Concerning Labor Mobility(1)(Sources)Ministry of Health,Labour and Welfare,OECD,Ministry of Internal Affairs and Communications,Eurostat and US Department of LaborFigure 3.International comparison of matching efficiency in the labor marketFigure 1.Unemployment rate and vacancy rate012345678012345678(Vacancy rate,%)Jan.2000Widening mismatchNarrowingJan.2005Jan.2015ImprovingWorsening in thesupply-demandJan.1990Jan.1995Jan.2010Jan.2021Jan.2020May 2024(Unemployment rate,%)Figure 2.International comparison between unemployment rates and ratios of the long-term unemploymentInternationally,the percentage of people who experienced unemployment for more than a year is high,whereas the unemployment rate is low in Japan.The mismatch in terms of the UV curve had widened significantly since the 1990s,but has partly narrowed since the mid-2010s.Easy to find employment The matching efficiency in Japan is lower than in the US and Germany.When the number of job openings and unemployed workers is the same,the probability of leaving unemployment in one month is 80%in US but 30%in Japan.Differences in the labor mobility smoothness from unemployment to employment are observed,posing a challenge to the efficiency of resource reallocation in the labor market.020406080100Routine paperworkLabor managementCoordination of schedules,etc.Accounting,finance and taxationCorporate legal affairsMarketingSales,telephone support,etc.Manufacturing&AssemblyDriving&Shipping-100-50050100150200HokkaidoAomoriIwateMiyagiAkitaYamagataFukushimaIbarakiTochigiGunmaSaitamaChibaTokyoKanagawaNiigataToyamaIshikawaFukuiYamanashiNaganoGifuShizuokaAichiMieShigaKyotoOsakaHyogoNaraWakayamaTottoriShimaneOkayamaHiroshimaYamaguchiTokushimaKagawaEhimeKochiFukuokaSagaNagasakiKumamotoOitaMiyazakiKagoshimaOkinawa(Deviation from efficient matching,%)20172022(2)Transport and machine operation workers-100-50050100150200HokkaidoAomoriIwateMiyagiAkitaYamagataFukushimaIbarakiTochigiGunmaSaitamaChibaTokyoKanagawaNiigataToyamaIshikawaFukuiYamanashiNaganoGifuShizuokaAichiMieShigaKyotoOsakaHyogoNaraWakayamaTottoriShimaneOkayamaHiroshimaYamaguchiTokushimaKagawaEhimeKochiFukuokaSagaNagasakiKumamotoOitaMiyazakiKagoshimaOkinawa(Deviation from efficient matching,%)20172022(1)Clerical workers4.73.72.41.50.4012345678Construction OccupationsNursing care-related occupationsAutomobile DrivingProductionOffice Work(Times)Overall average(1.24 times)Chapter 2 Section 2:Issues Concerning Labor Mobility(2)Sources:Ministry of Health,Labour and Welfare,Ministry of Internal Affairs and Communications,doda and Cabinet OfficeFigure 1.Mismatch rate by occupationShort supplyOversupply(1)Public employment security office)(2)Private employment agencyFigure 2.Effective job openings by occupationFigure 3.Operations replaced by AI and automation(%)Mostly or partiallyhave replacedWant to replacein the futureNot considering alternatives,Not applicable,etc.Looking at mismatch rates*by occupation in each prefecture,most prefectures are suffering an undersupply of transportation and machine operation workers,whereas an oversupply of clerical workers can be observed mainly in metropolitan areas.*The gap between the actual number of job placements by occupation and the potential number of job placements if efficient matching had been achieved.Short supplyOversupplyThe active job opening ratio for clerical jobs is lower than 1.0,whereas those for construction and nursing care are around 3.0 or 4.0.The low job opening is similar in the private employment agency.Clerical work may be replaced by AI and other technologies in the future.It is important to promote re-skilling to facilitate labor reallocation from over-supplied to under-supplied fields and to improve productivity through labor-saving investment in under-supplied fields.02468101214012345671 4 7101 4 7101 4 7101 4 7101 4 7101 520192021222324MonthYear(Times)(Times)SalesAdministrative,AssistantPlanning and ManagementTotalEngineer(IT&Telecommunications)(scale right)Marketing and Service-0.20-0.100.000.100.200.300.400.50(%)“Pure productivity effect”“Denison effect”“Baumol Effect”Labor productivity growth-0.15-0.10-0.050.000.050.100.15Agriculture,forestry and fishingManufacture of food productsManufacture of textiles,and related productsChemicalsManufacture of basic metals and fabricated metal productsComputer,electronic,electrical equipmentManufacture of transport equipmentManufacture of machinery and equipment n.e.c.ConstructionWholesale and retail tradeTransportation and storageAccommodation and food service activitiesInformation and communicationFinancial and insurance activitiesProfessional,scientific and technical activities,etc.Human health and social work activities(%)2011-19CY2001-10CY0.0100.0150.0200.0250.0300.0352001-0506-1011-1516-19CY(2)Contribution of the“Denison effect”Chapter 2 Section 2:Issues Concerning Labor Mobility(3)Source:EU KLEMS.-0.20.00.20.40.60.81.01.21.41.61.8200119200110201119(%)“Pure productivity effect”“Denison effect”“Baumol effect”Labor productivity growthCYFigure 1.Extent of shifts of employment between industries(Lilien Index)Figure 2.Breakdown of labor productivity growthFigure 3.Breakdown of labor productivity growth(by type of industry)(1)Contribution of labor productivity growth(Average in 2001-19CY)The decisive factor for macroeconomic productivity growth is the productivity increase in each industry.The positive contribution of the productivity boosting effect of inter-industry labor reallocation(Denison effect)declined from the 2000s to the 2010s.Looking at the contribution of the Denison effect by industry,the ones with the largest contributions,for both positive and negative,are those with relatively low productivity.From the 2000s to the 2010s,the contributions of industries with large positive contributions declined significantly,while the negative contribution continued to be large in health and social care services.The activeness of inter-industry labor mobility has been declining over the long term.Chapter 2 Section 2:Issues Concerning Labor Mobility(4)Source:Ministry of Health,Labour and Welfare食料品等繊維化学等生産用機械電子部品等情報通信機械輸送用機械-20-15-10-5051015200.00.51.01.52.0(Increasing rate in regular employees(20182023),%)(Wages,relative to all industries,average in 2018-23)建設情報通信運輸卸売小売金融、保険学術研究、専門技術宿泊、飲食学校外教育、学習支援等社会保険社会福祉、介護-20-15-10-5051015200.00.51.01.52.0Figure 1.Relative wages and permanent employment by industry(2018-2023)Even in recent years,both before and after COVID-19,labor migration to high-productivity sectors has not necessarily progressed.In the manufacturing sector,where relative wages are high,employment is shrinking and a labor-saving trend can be seen in many industries,except for some growing fields.In the non-manufacturing sector,although there is variation across industries,employment is expanding in response to rising social demand in many sectors with low relative wages,such as nursing care,accommodation and food services.FoodTextileTransport equipmentChemicalsElectrical parts and devicesProduction machineryICT machinery(Wages,relative to all industries,average in 2018-23)(Increasing rate in regular employees(20182023),%)Accommodation and restaurantsWholesaleEducation and learning SupportMedical,healthcare and welfareInformation and communicationsResearch and technical servicesConstructionTransportWholesaleFinancial and insurance services68 205 05010015020025020121416182022OthersSpecified skilled worker(SSW)StudentPermanent residentTechnical intern training(TIT)Skilled professionalTotal(10,000)(CY)23-28.3-7.1-35-30-25-20-15-10-50w/o control variablesw/control variables(Difference from Japanese workers,%)*Chapter 2 Section 3:Issues Concerning Foreign Workers in JapanSources:Ministry of Health,Labour and Welfare and Ministry of Internal Affairs and Communications00.010.020.030.040.050.060.07JapaneseForeign totalSkilled professionalSSW/TITPermanent resident(Age)(Density)Figure 1.Number of foreign workers by status of residenceMany foreign workers work in Tokyo,northern Kanto and Tokai regions and the manufacturing sector employ more foreign workers than others.Figure 2.Ratios of foreign workers by prefecture and industryFigure 3.Age distribution of foreign workersCompared to Japanese workers,foreign workers are younger and have shorter tenure.Figure 4.Wage gap between Japanese and foreign workersHighly skilled foreign workers with longer working experience in Japan tend to earn higher wages than Japanese workersFigure 5.Wage gap by occupation and status of residenceThe wage gap is 75%smaller when education,tenure,job experience,gender,type of employment and difference of establishments are controlled.The number of foreign workers has been increasing along with the expansion of the acceptance system and reached over 2 mil.,3.4%of total employees.While wages for foreign workers are 28%lower than Japanese,three-quarters of this gap is explained by age,tenure and other factors.0.02.04.06.08.0HokkaidoAomoriIwateMiyagiAkitaYamagataFukushimaIbarakiTochigiGummaSaitamaChibaTokyoKanagawaNiigataToyamaIshikawaFukuiYamanashiNaganoGifuShizuokaAichiMieShigaKyotoOsakaHyogoNaraWakayamaTottoriShimaneOkayamaHiroshimaYamaguchiTokushimaKagawaEhimeKochiFukuokaSagaNagasakiKumamotoOitaMiyazakiKagoshimaOkinawaConstructionManufacturingInformation and CommunicationWholesale and retailAccomodations,eating and drinkingEducation and learning supportMedical,health care and welfareServiceOthersTotal(%)-2.3-7.4-31.3 4.3-2.3-7.0-17.5 2.5-7.3-14.9-23.0-3.5-45-40-35-30-25-20-15-10-5051015SkilledprofessionalSSWTITPermanentresident*(Difference from Japanese workers%)*Professional and engineeringClericalManufacturing processFinancial assets in households have been increasing in both Japan and the U.S,although the growth in Japan was lower than the U.S.While risk assets such as stocks make a large contribution in the U.S.,cash and deposits make a large contribution in Japan.-10010203040506070(CY)Total(polyline)DepositStockInvestment trustsBondLife insuranceOther-50050100150200250300(CY)Total(polyline)OtherStockInvestment trustsBondLife insuranceDeposit(%,compared to March 31,2000)Chapter 3 Section 1:Household Financial Assets and Investments(1)Sources:Ministry of Internal Affairs and Communications,FRB,and Bank of Japan020406080100Over70 yearsold55-69yearsold40-54yearsoldUnder40yearsoldUnited States（2023）(%)DepositStockPensionBondLife insuranceOtherFigure 1.Financial assets held by households(2)United States(1)JapanFigure 3.Structure of household financial asset holdings by age0102030405060708090100日本(%)BondStockInvestment trustsDepositLife insuranceOtherFigure 2.Structure of financial assets held by households(Mar.2024)(%,compared to March 31,2000)Japanese households are more risk averse in the management of their financial assets than U.S.households.In asset composition ratios by age group,the structure is the same as above,with Japanese households preferring cash and deposits and U.S.households preferring risk assets.020406080100Over70yearsold55-69yearsold40-54yearsoldUnder40yearsold(%)Japan（2019）DepositStockLife insuranceInvestment trustsBondOtherJapanUSInvestment trusts0100200300400500600700800Under 2530-3440-4450-5460-6470-7480-84(10,000)OthersSecuritiesLiquid depositsFixed-term depositsLife insuranceFigure 1.Financial asset holdings per household(2019)50556065707580859095100Male(mode:88 years old)Female(mode:93 years old)Female life expectancy87.09 yearsMale life expectancy81.05 years(Age)Chapter 3 Section 1:Household Financial Assets and Investments(2)(Sources)Ministry of Internal Affairs and Communications,Ministry of Health,Labour and Welfare and the Central Council for Financial Services InformationFigure 3.Purpose of financial asset holdings by elderlyhouseholds010203040506030405060708090200710152023(%)Living funds after retirementPreparing for illness and disaster(CY)60-7960 and overFor travel and leisure(right scale)Bequest(right scale)(%)Figure 2.Distribution of deaths by age While financial assets per Japanese household have increased toward the peak in the early 60s,these assets do not decrease in the elderly years including oldest-old,as people secure funds for retirement against longevity risk.In Japan,the challenge is to encourage households to make effective use of their affluent stock of financial assets for economic activities.051015202530Less than3 mil.Yen3-5mil.Yen5-7.5mil.yen7.5-10mil.yen10-12mil.YenMore than12 mil.Yen(%)20232019201305101520253020s30s40s50s60s70 and over(%)2023201920130100200300400500201518202224(10,000 account)(Year)Under 20 years oldOver 80 years old60s70s40s50s30s024681012141618-500-1000-15002000 and over(%)（Annual income,10 thousand yen）2019survey2014survey2009survey2004survey1999surveyFigure 1.Number of NISA accounts by ageChapter 3 Section 1:Household Financial Assets and Investments(3)Figure 2.Ratios of households that have shifted their assets from cash and deposits to riskier assets Figure 3.Percentage of securities by annual income(1)By age of household head(2)By annual household income The number of NISA accounts has increased especially among the working-age population including younger groups.The percentage of households that have shifted their financial assets from cash and deposits to risky assets has increased significantly compared to 10 years ago,especially among younger groups.On the other hand,households with higher annual incomes tend to allocate their income to risky assets and hold a high share of such assets.From the perspective of promoting the“shift from savings to investment,”it is important to increase household income.Sources:Financial Services Agency,the Central Council for Financial Services Information,and Ministry of Internal Affairs and Communications.Chapter 3 Section 1:Household Financial Assets and Investments(4)A comparison of consumption by age of household head between Japan and the U.S.shows that(1)overall consumption peaks in the 55-64 age group in both,(2)the elderly group spends more on food and beverages in Japan,(3)education spending increases significantly in the 45-54 age group in Japan,(4)the spendings in entertainment and accommodation are suppressed in the 45-54 age group in Japan.Annex:Japanese and U.S.consumption patterns by age-15-10-50510152034years andyounger35-44years45-54years55-64years65-74years75 yearsand older(1)Consumption expenditure(Deviation from average,%)U.S.Japan-40-30-20-10010203034years andyounger35-44years45-54years55-64years65-74years75 yearsand older(Deviation from average,%)(2)Food and beverageJapanU.S.-150-100-5005010015020025034yearsandyounger35-44years45-54years55-64years65-74years75 yearsand older(Deviation from average,%)(3)EducationJapanU.S.-40-30-20-100102030405034years andyounger35-44years45-54years55-64years65-74years75 yearsand older(4)Entertainment and accommodation(Deviation from average,%)JapanU.S.Sources:Ministry of Internal Affairs and Communications and U.S.Bureau of Labor Statistics.Chapter 3 Section 2:Prospects for and Issues Concerning Housing Stock(1)Sources:Ministry of Land,Infrastructure,Transport and Tourism,Ministry of Internal Affairs and Communications,Ministry of Health,Labour and Welfare and US Census Bureau.05010015020025019517090201023Housing for rent,etc.Owner-occupied housing,etc.(Units of 10,000)(Year)Figure 1.New housing starts and household composition0510152025303540450246810121416199095200005101520(Units)(Units)Per 1,000 households(Japan)right scalePer 1,000 people(Japan)Per 1,000 people(U.S.)Per 1,000 households(U.S.)right scale(Year)Figure 2.Housing starts per population/household20.1 25.1 9.0 2.6 38.1 0102030405019809020001020Single-person householdsMarried couples with childrenSingle parent and childrenMarried couples onlyThree-generation households(%)(Year)(1)New housing starts(2)Change in household compositionThe number of new housing starts,especially that of owner-occupied housing,has declined to less than half of its peak.This is mainly due to an increase in the number of single-person households and a decrease in the number of married couples with children.Although the number of housing starts per population/household is declining,it is about twice that of the U.S.27.11020304020121416182022(FY)(%)0102030400-10km10-20km20-30km30-40km40-50km50-60km60-70kmFY2012FY2022(%)(Distance from the Center)Chapter 3 Section 2:Prospects for and Issues Concerning Housing Stock(2)Sources:Ministry of Internal Affairs and Communications,IPSS,Ministry of Land,Infrastructure,Transport and Tourism and Japan Housing Finance Agency-10-8-6-4-2024Under 2930-3940-4950-5960-69(Rate from 1993 to 2018),%)Change in homeownership rate for single-person householdsChange in homeownership rateChange in homeownership rate for household with two or more peopleChange in household compositionFigure 1.Contribution of factors to change in homeownership rate As the number of single-person households increases,the average percentage of owner-occupied houses and the number of owner-occupied units is expected to decrease.As housing prices rise,home acquisitions in the suburbs have spread over the past decade in three major metropolitan areas.As housing prices rise,the percentage of acquiring existing homes is also on the rise.Figure 2.Trends in homeownership rate and number of owner-occupied houses due to changes in household composition05001,0001,5002,0002,5003,0003,5004,0004,500545658606264198898200818283848(%)(10,000 households)EstimationHomeownership rateNumber of owner-occupied houses(right scale)(CY)Figure 3.Acquisition Status and Locations of Existing Houses(2)Estimated rate of acquisitions of existing housesThe percentage of owner-occupied houses has been declining for many age groups over the long-term,as the percentage of single-person households with low owner-occupied rates has increased.(1)Percentage of housing acquisitions by distance from the center in the three major metropolitan areas(Age group)Chapter 3 Section 2:Prospects for and Issues Concerning Housing Stock(3)Sources:Japan Housing Finance Agency and Ministry of Internal Affairs and CommunicationsThe acquisition of existing houses has spread across a wide range of groups over the past decade,regardless of income levels.Acquisition of existing houses is larger among older age groups,and the willingness to acquire them may be getting stronger among younger cohorts.Custom-built houseSpec houseCondominiumsExisting single-familyhouseExisting condominium0 0%0 0%Up to 300Income Range(10,000 Yen)1000-300-600600-1000(1)20120 0%0 0%Income Range(10,000 Yen)Up to 300300-600600-10001000-Figure 1.Acquisition status of existing house by annual income(2)2022Figure 2.Cohort analysis of acquiring existing houses-40-30-20-10010203040-2425-2930-3435-3940-4445-4950-5455-5960-6465-3334-3839-4344-4849-5354-5859-6364-6869-7374-7879-8384-8889-93Age EffectCohort Effect（：Deviation from the baseline）（Baseline）（Baseline）(Year)(Year of Birth)1.01.52.02.53.02004060810121416182022JapanCanadaFranceGermanyItalyU.K.U.S.(Year)Figure 4.International comparison of transparency in real estate transactionHigh transaction transparencyLow transaction transparencyChapter 3 Section 2:Prospects for and Issues Concerning Housing Stock(4)Sources:Japan Housing Finance Agency,Ministry of Land,Infrastructure,Transport and Tourism,Cabinet Office,US Bureau of Economic Analysis and JLLIn about the past decade,older existing homes have been valued in a longer term.To further develop the market for existing home transactions,efforts to promote renovation to extend the life of homes are important.The new construction premium may be disappearing in condominiums which are generally versatile.On the other hand,the new construction premium is still observed in detached houses which are more individualized.It is also important to increase transparency in real estate transactions.0102030405060708090100151015202530202220212019201720152013（10,000 Yen/）(Building Age)05101520253015101520253020222020201820162014（10,000 Yen/）(Building Age)(2)Detached houses(1)CondominiumsFigure 1.Changes in depreciation trends due to building age(Tokyo Area)Figure 3.Comparison of Home Renovation Market between Japan and the U.S.15.131.50.61.20.00.51.01.52.0010203040JapanU.S.Ratio to Housing InvestmentRatio to GDP(right scale)(%)(%)-10-8-6-4-20246820131517192122(%)(Year)With PremiumWithout PremiumFigure 2.Estimation of new construction premium(Tokyo Area)0246810121416201416182022(%)(Year)(1)Condominiums(2)Detached houses586062646668OECD-AverageJapanKoreaUnitedStatesCanadaGermanyUnitedKingdomItalyFranceAverage effective age of labor market exitCurrent retirement ages(Years old)Figure 2.Average effective age of labor market exit and current retirement ages5860626466687072OECD-AverageJapanKoreaUnitedStatesCanadaGermanyUnitedKingdomItalyFrance(Years old)Current retirement agesAverage effective age of labor market exitChapter 3 Section 3:Utilizing Human Capital of Elderly Workers(1)(1)Male(2)FemaleFigure 1.Labor participation rate of older people,65-74 years oldThe average retirement age for the elderly in Japan exceeds the age at which they start receiving their pension benefits,while the opposite is true in Europe and the United States.51.80102030405060JapanUnitedStatesUnitedKingdomGermanyItalyFranceKorea(%)(1)Male2022201333.10102030405060JapanUnitedStatesUnitedKingdomGermanyItalyFranceKorea(%)(2)Female20222013Source:OECD.Compared to other advanced countries,the labor participation rate of the elderly has been high and rising significantly over the past decade for both males and females.01020304050607080901002018Male2023Male2018Female2023Female(%)Over 70 years old66-70years old61-65years oldUnder 60years oldYearGenderChapter 3 Section 3:Utilizing Human Capital of Elderly Workers(2)Sources:Ministry of Internal Affairs and Communications,National Institute of Population and Social Security Research,Cabinet Office,Ministry of Health,Labour and Welfare,Japan Pension Service and National Tax AgencyDue to the measures to ensure continued employment,the number ofelderly workers is increasing and expected to expand over the next 10 years.Figure 2.Change in desired age for employment by genderFigure 1.Estimation of future number of elderly workers456-1000100200300400500600700Exceed the Income Barrier（150 million yen）(10,000 yen)Increment of disposable pensionIncrement of disposable wageDecrease of spousal allowanceof a husbandIncrement of lifetime disposable income.01,0002,0003,0004,0005,0006,0007,000就業調整(467万円)平均給与現行制度平均給与在老(10,000 yen)DisposablepensionDisposable wage435448Figure 3.Increase in lifetime disposable income for seniors working exceeding the“Income threshold”(1)Increase in income by working exceeding the“income barrier”for women aged 60-64 with spouses.(compared to earning 1 million yen annually)(2)Disposable wage and pension of university graduate male aged 65-74(with and without a pension deduction for working seniors)With“employment adjustment”（487 million yen）Earning average wage without“employment adjustment”(With pension deduction)(Without deduction)More elderly people are willing to work after 65 years-old for both males and females.It is important to encourage people to work to exceed the“income barrier”by informing them that doing so will increase their lifetime disposable income and to review the social security system to support seniors willingness to work.2,031 1,635 1,468 1,0001,2001,4001,6001,8002,0002,20020152025303540(10,000 people)(CY)Elderly workers(over 60s)Case2:If the employment rate remains at the current level.Case1:If the employment rate increases at the same pace as in recent years(2019-23CY average).5.5 9.6-3.2-9.7-15-10-5051015(%)15$E%6%1%8out 60-70out 40-50out 80-90out the sameLess than 30%OthersChapter 3 Section 3:Utilizing Human Capital of Elderly Workers(3)Figure 1.Current status,changes and background of wage reductions when rehiring elderly employees in companies(2)Percentage difference from 5 years ago(3)Changes in the range of wage declines and the abilities demanded of the elderly(1)Wage levels for older workers after retirement(2024)Although the majority of firms(45%)set wages for workers over the retirement age at 60-70%of pre-retirement wages,an increasing number of firms(40%)are paying wages more than 80%of the pre-retirement level.These firms expect their elderly employees to play a leading role by utilizing their knowledge and experience.-2.7-0.91.9-2.30.10.8-4-3-2-10123(Difference from average percentage,%)Ability to train and mentor other staffAppropriate managementSource:Cabinet office.About 80-90out 60-70out 40-50out the same ConstructionReal estate and goods rental and leasingWholesalingRetailingTransport and postal activitiesInformation services Services for businessesAccommodations,eating and drinking services-15-10-50510152010203040506070(Employment Conditions DI(inverse sign)(Difference from average percentage、%)Manufacturing76.5 32.0 29.1 21.5 18.4 12.5 6.3 020406080100HealthconsiderationsDecline inproductivityCreatingconvincingvaluationmethodsLowwillingnessand motivationDifficulty insecuringassignedtasksIncreased burdenof labor costsLack ofknowledgeand skills(Number of firms which raise each item as challenges,%)Chapter 3 Section 3:Utilizing Human Capital of Elderly Workers(4)Figure 1.Challenges for companies in hiring older workersFigure 3.Impact on labor cost rate and profit margin of companies raising retirement age(1)Working hours and percentage of“health considerations”(2)Labor shortage and“decline in productivity”Figure 2.Challenges and background to employment of the elderlyMany companies report“health considerations”and“declining productivity”as challenges in hiring elderly workers.Companies that have raised the retirement age experience high labor cost ratios,but profit ratios havent necessarily worsened.It is important to improve productivity by encouraging labor-saving investments,etc.Industries with long working hours report the health of the elderly as an issue for hiring.Industries with relatively low labor shortages report their low productivity as an issue.-1.5-1.0-0.50.00.51.01.52.0labor cost rateoperating income rate(Impact of raising retirement age,%)Impact on operating income rate is not significantStatistically significant labor cost rate increasesConstructionManufacturingInformation services Transport and postal activitiesWholesaling and retailingReal estate and goods rental and leasingAccommodations,eating and drinking services-20-15-10-5051015160165170175180185(Total hours worked per month for general workers)(Difference from average percentage,%)Sources:Cabinet office,Bank of Japan,and Ministry of health,labor and welfare.
2025-01-22
38页
5星级
Diversity,Equity and Inclusion Lighthouses 2025I N S I G H T R E P O R TJ A N U A R Y 2 0 2 5Centre .
2025-01-22
44页
5星级
A World Bank Group Flagship ReportGlobal Economic Prospects JANUARY 2025Global Economic Prospects JA.
2025-01-21
250页
5星级
Vanguard research December 2024Vanguard economic and market outlook for 2025:Beyond the landingThe g.
2025-01-17
26页
5星级
uuuDecember 20242025 OutlookCharting the economys next chapterInvestment and Insurance Products:NOT .
2025-01-14
36页
5星级
World Economic Situation and Prospects 2025The World Economic Situation and Prospects 2025 is a repo.
2025-01-13
212页
5星级
IPSOS Almanac 2025PrefaceWelcome to the Ipsos Almanac where we look back at the events that shaped 2.
2025-01-03
27页
5星级
罗兰贝格：预见2026：中国行业趋势报告（90页）.pdf
智源研究院：2026十大AI技术趋势报告（34页）.pdf
中国互联网协会：智能体应用发展报告（2025）（124页）.pdf
三个皮匠报告：2025银发经济生态：中国与全球实践白皮书（150页）.pdf
三个皮匠报告：2025中国商业航天市场洞察报告-中国商业航天新格局全景洞察（25页）.pdf
国声智库：全球AI创造力发展报告2025（77页）.pdf
中国电子技术标准化研究院：2025知识图谱与大模型融合实践案例集（354页）.pdf
三个皮匠报告：2025中国情绪消费市场洞察报告（24页）.pdf
中国银行：2026中国高净值人群财富管理白皮书（66页）.pdf
亿欧智库：2025全球人工智能技术应用洞察报告（43页）.pdf
0731-84720580
商务合作：really158d
友链申请 (QQ)：1737380874
最新报告
中英对照
全文搜索
报告精选
PDF上传翻译
多格式文档互转
入驻&报告售卖
会员权益
机构报告
券商研报
财报库
专题合集
英文报告
数据图表
会议报告
其他资源
研报
新质生产力
DeepSeek
低空经济
大模型
AI Agent
AI Infra
具身智能
自动驾驶
宠物
银发经济
人形机器人
企业出海
算力
微短剧
薪酬
白皮书
创新药
行业分析
个股研究
年报财报
IPO招股书
会议纪要
宏观策略
政策法规
其他
人工智能
信息科技
互联网
消费经济
汽车交通
电商零售
传媒娱乐
医疗健康
投资金融
能源环境
地产建筑
传统产业
英文报告
其它
行业聚焦
芯片产业
热点概念
全球咨询智库
人工智能
500强
新质生产力
会议峰会
新能源汽车
企业年报
互联网
公司研究
行业综观
消费教育
科技通信
医药健康
人力资源
投资金融
汽车产业
物流地产
电子商务
传统产业
传媒营销
其它
十五五规划系列报告合集（共48套打包）
2026低空经济/低空产业报告合集（共47套打包）
AI、科技与通信
广告、传媒与营销
消费、零售与支付
HR、文化与旅游
金融、保险与投资
能源、环境与工业
医疗制药与大健康
物流、地产与建筑
其他行业
AI ▪ 科技 ▪ 通信
数字化
金融财经
智能制造
电商传媒
地产建筑
医疗医学
能源化工
其他行业
收藏
下载
2026-02-02
AI查数
行业数据
政策法规
商业模式
产业链
竞争格局
市场规模
产业概述
其它
2026年
AI读财报
年报
一季报
半年报
三季报
IPO招股书
社会责任报告
A股
IPO申报
港股
美股&全球
新三板
微信扫码登录
手机快捷登录
账号登录
