Japanese Households’ Financial Asset Allocation: A Critical Study of the Government’s AssetRelocation Policy
（Executive Summary）
This study critically analyzes the Japanese government’s policy to shift households’ financial assets from bank deposits toinvestment instruments. There are three points for discussion; (1) whether U.S. households’ allocation pattern should be used asa benchmark, (2) the effects of long-term stagnation of the Japanese economy on households’ asset accumulation, and (3) themacroeconomic impact of asset relocation. Regarding the first point, the financial asset allocation patterns of U.S. households areseverely affected by the top 5 percent asset group, whose risk-asset ratio is significantly higher than average households.Therefore, great caution should be exercised when using the U.S. allocation pattern, derived from the aggregate householdfinancial assets, as a benchmark for policies aimed at relocating Japanese households’ financial assets toward risky assets.Regarding the second point, the growth rate of Japanese households’ financial assets are distinctly lower than that of U.S.households. This is mainly owing to long-term stagnation of the Japanese economy and sluggish stock prices. Additionally, the lowrisk-asset ratio among the Japanese younger generation is not caused by their risk aversion, but mainly because of liquidityconstraints resulting from holding risky nonfinancial assets, such as real estate. Regarding the third point, the corporate sectorremains a fund-surplus sector in Japan. This implies that its growth is not hampered by liquidity constraints and that shiftinghouseholds’ bank deposits to risky assets will not accelerate the Japanese economy’s growth rate. Conversely, as the economygrows faster, households will spontaneously increase investments in risky assets as disposable income increases and liquidityconstraints are relaxed.
1.Foreword
Financial assets held by the Japanese household sector in 2022 was US$ 14.9 trillion, or US$ 118,000 per head, which is muchless than that of the U.S. household sector (US$ 115.5 trillion/US$ 348,000 per head). The asset allocation patterns for bothcountries are shown in figure 1.
The Japanese Financial Services Agency (FSA), the government’s financial regulation body, and some economists argue that theportfolio allocation of Japanese households is significantly biased toward bank deposits (checking and time deposits) and cashcompared to that of U.S. households, citing the data shown in figure 1 as evidence. This figure was published by the Bank of Japan,as a part of the “Flow of Funds Statistics of Japan, the U.S. and the Euro area.”
According to this figure, Japanese households have a lesser share of stocks/investment trusts (Japan=15.7%, U.S.=52.4%) than U.S. households, while majority of financial assets are allocated to bank deposits (Japan=54.3%, U.S.=13.7%). The FSA and someeconomists contend that Japanese households must shift their financial assets from bank deposits to stocks and investmenttrusts based on U.S. households’ asset allocation. The FSA campaign to “shift from bank deposits to investment instruments”began in the late 1990s[1].
Despite the efforts of government and financial institutions, asset relocation toward investment instruments has barely realized inthe last 20 years. Figure 2 depicts the share of risky assets as a percentage of total financial assets beginning from 1994, a fewyears from the burst of the “bubble” in 1990?91. Although the risky asset ratio fluctuated as share prices moved, no upward trendwas observed during this period.
Several recent studies published in Japan have concluded that the shift from bank deposits to investment instruments has notoccurred in the last 20 years[2].
This study critically analyzes the government’s promotional policy to shift Japanese households’ asset allocations from bankdeposits to investment instruments by examining data. This study discusses three points; (1) issues related to benchmarking U.S.asset allocation, (2) the negative effects of the long-term stagnation of the Japanese economy, and (3) the macroeconomiceffects of asset relocation.
2. Issues related to benchmarking the U.S. asset allocation
This chapter explains the reasons against using the asset allocation of U.S. households as a benchmark for evaluating that ofJapanese households.
2.1. Problems with using the U.S. asset allocation pattern as a benchmark
First, it is important to note that considering the asset allocation pattern shown in figure 1 as representative of “typical” or“average” families would be a misunderstanding. The data used in figure 1 were derived from the flow of funds statistics, which aremacroeconomic statistics capturing financial transactions between major economic sectors in a country, such as the household,corporate, government, and overseas sectors, and their balance sheets. Figure 1 shows how each country’s household sectorallocated financial assets as a whole. Therefore, the allocation patterns of individual households could differ significantly fromthose shown in the figure.
In reality, financial asset allocation of individual households is determined by income level, age, asset/liability conditions, and otherfactors. Therefore, there is no unique optimal allocation pattern suitable for all households. In other words, the U.S. assetallocation pattern for the household sector, shown in figure 1, cannot be considered optimal.
Second, to evaluate the suitability of the allocation pattern of U.S. households, the allocation patterns of other developedcountries are shown in figure 3 for comparison. It clearly indicates that the allocation pattern of the U.S. household sector is anoutlier compared to those of other developed countries.
The unique asset allocation pattern in the U.S. is closely related to the uneven distribution of financial assets among households.To verify this phenomenon, household survey data are available for the U.S. and Japan. Figure 4 displays the outstanding amount offinancial assets held by households, categorized into five (Japan) and four (the U.S.) groups based on the average outstandingamount of financial assets. The wealthiest Japanese group owns 57 times as many assets as the poorest group’s average, while inthe U.S., the wealthiest group holds 240 times as many assets as the poorest group, indicating that financial assets are held quiteunevenly in the U.S. compared with Japan.
Third, it is a misunderstanding to assume that every household in the U.S. allocates 40% of their financial assets to risky assets.As discussed previously, financial assets in the U.S. are significantly unevenly distributed compared to those in other developedcountries, including Japan. This also holds true for the distribution of risky assets. Figure 5 illustrates the distribution of stockholdings in both Japan and the U.S. Although the asset class classifications differ between the two countries, the top 20 percentof Japan owns approximately 60 percent of total shares, while the U.S. top 25 percent holds 96 percent of total stocks.
Similarly, figure 6 shows the percentage of households holding risky assets (stocks and investment trusts) for each country.Although figure 1 indicates that the household sector as a whole allocated approximately 40 percent of their assets to stocks,only 15 percent of households actually hold stocks. Further, only 10 percent of U.S. households hold investment trusts directly.These findings are consistent with figure 5, indicating that the top 25 percent of the asset size class own almost all stocks.
In Japan, the share of households holding stocks and stock investment trusts is 20.9 percent, which is slightly higher than that ofU.S. households (in a Japanese survey, stocks and stock investment trusts cannot be separated). Similarly, the percentage ofhouseholds holding trust investments in Japan is 7.6 percent, which is slightly lower than that of U.S. households. Figure 6suggests that the share of households holding risky assets is at most 20 percent, which appears significantly low than the data infigure 1. The difference between the two figures indicates that figure 1 illustrates the aggregate data of all households and issignificantly influenced by the extremely biased allocation of financial assets among households[3].
2.2. The degree of risk aversion of Japanese households
Some Japanese economists and policymakers argue that Japanese households are too risk averse. This section assesses whetherthis statement can be supported by the data.
The first point to note is that total assets owned by households consist of not only financial assets, but also nonfinancial assets,such as real estate. Real estate value fluctuates widely because of frequent booms and busts in land prices. Additionally, owing tothe accounting laws in Japan, the value of houses depreciates swiftly. Therefore, households naturally regard nonfinancial assetsas risky, because of the volatility in real estate value.
Figure 7 illustrates the allocation of total assets (financial and nonfinancial assets) in both countries. Japanese householdsallocate approximately 60 percent to nonfinancial assets, while U.S. households’ share of nonfinancial assets is approximately halfthat of Japanese households.
If broader risky assets are defined as the sum of risky financial and nonfinancial assets, households in both countries allocateapproximately 60?70 percent of their total assets to broader risky assets. In Japan, although the share of risky financial assets issmaller than that of the U.S., the share of broader risky assets increases to the level of U.S. households because prices ofhousing and real estate are higher relative to the prices of other goods in Japan. A significant share of broader risky assetscreates liquidity constraints for Japanese households. Therefore, households seem to prioritize holding sufficient liquid assets,such as bank deposits, and to restrain risky assets, such as stocks and investment trusts.
Figure 8 shows that as financial asset holdings increase, the share of nonfinancial assets to total assets declines. Therefore,liquidity constraints are strong among low- to middle-class households, whereas high-class households can allocate significantamounts of risky financial assets as liquidity constraints would not be a concern.
In Japan, households with mortgage payments, which impose liquidity constraints, have lower risk-asset ratios than those withoutmortgage payments, as shown in figure 9.
Overall, figures 7 to 9 indicate that when households allocate their financial assets to risky assets, they consider not only thevolume of financial assets, but also nonfinancial assets, which are risky assets as well.
Second, prior studies in this field have directly compared the risk aversion of U.S. and Japanese households. One such study wasperformed by Ito, Takizuka and Fujiwara (2017) who measured risk aversion of both Japanese and U.S. households by questioningparticipants on the type of reward they prefer for a certain type of job. Their results show no significant difference between thetwo countries in terms of their risk aversion. They stated “we cannot conclude that Japanese households are more risk-aversethan the U.S.,” and that the difference in risk-asset ratio cannot be explained by risk aversion of Japanese households.
3.The effects of long-term stagnation of the Japanese economy
The second point that is examined is the negative effects of the long-term stagnation of the Japanese economy on assetaccumulation in Japanese households.
3.1. Stagnation of the Japanese economy
Figure 10 shows the long-term movement of the outstanding amount of financial assets held by Japanese and U.S. households.Owing to data constraints, the graph is plotted from 1994. In 2021, U.S. household assets grew 5.5 times compared to 1994,whereas those of Japan increased only 1.7 times during the same period. What are the main reasons for this large gap betweenthe two countries?
Financial assets are accumulated net savings. Savings on the other hand, is income minus consumption by definition, whichsuggests that the growth rates of financial assets and income are co-related, assuming the savings rate (saving/income) isconstant. Figure 11 displays the nominal GDP of both countries as a proxy for household income (year-end 1994=100).
Figure 11 shows the severity of the Japanese economy stagnation. The index for 2021 was 107.1, indicating that the Japaneseeconomy grew by only 7.1 percent in 30 years, or 0.25 percent annually. In contrast, the U.S. economy increased 3.3 times, or 4.4percent annually. Although the Japanese nominal GDP was significantly affected by Japanese deflation during this period, the largegap in GDP compared to the U.S. must have had a significant effect on financial asset accumulation in Japan.
Next, figure 12 and 13 include data from the Euro area for comparison. Owing to data constraints, the graphs are plotted from2003 as the benchmark, and show that the Euro area figures fall between that of the U.S. and Japan. These graphs also confirmthe close relationship between nominal GDP and household financial assets.
Figure 14 illustrates the stock price index movements in both countries as a proxy for the rate of return on risky assets.
The difference in the rates of return between the two countries is evident from figure 14. In case of the U.S., households couldeasily gain benefits from holding stocks as stock prices have soared, especially since the 2010s. The U.S. stock prices increased 9.9 times in 2022 compared to 1994, while Japanese stock prices rose only 1.3 times during the same period. The base year 1994was when the Japanese “bubble” had just burst, therefore, stock prices plunged for more than 20 years until finally recovering to1994 levels in 2017.
The large stock price performance gap, in addition to the difference in the accumulation pace of financial assets, would have asignificant effect on households’ attitudes toward risky investments. In the U.S., investing in stocks often resulted in increasedstock values, and such experiences create a virtuous cycle, in which an increased portion of the asset was invested in additionalstocks and other risky assets. In contrast, Japanese households cannot easily earn capital gains from owning stocks or incurcapital losses from them, resulting in a vicious cycle of reluctance to invest in risky assets.
In Japan, low-income growth and falling stock prices have persisted for decades rendering investments in risky assets unworthy.Therefore, it is economically unreasonable to ignore the differences in the basic economic and financial conditions of bothcountries and attempt to induce households to increase their investments in risky assets. Figure 2 indicates Japanesehouseholds’ rational behavior of not relocating financial assets to risky assets during stagnated economic conditions. TheJapanese government’s policy priority should be to stimulate economic growth to change households’ attitudes towardinvestments.
3.2. Portfolio balances by age-category
In addition to economic stagnation, another unique feature of the Japanese economy is its aging society and decreasing totalpopulation. In Japan, there is often criticism that financial assets are too concentrated among older people. Figure 15 shows thatin Japan, individuals over the age of 60 own 56.2 percent of total assets, while in the U.S., individuals over the age of 65 alsopossesses approximately 50 percent. Therefore, both countries show similar tendencies for older persons possessingapproximately half of total assets.
Next, Figure 16 shows that the age distribution of stock holdings, a typical risky asset households would hold, also shows a similardistribution pattern in both countries. The share of individuals over 65 years of age is higher in Japan than in the U.S. In Japan,some critics argue that “older persons hold a larger portion of financial assets, which are allocated heavily in bank deposits. Suchlarge amounts of sleeping assets are one of the main causes of long-term stagnation of the Japanese economy.” However, thedata show that aged households hold a significant share of financial assets not only in Japan, but also in the U.S.; this pattern isconsistent with the life-cycle model. The Stock holdings of both countries also show similar patterns.
Figure 17 shows the risk-asset ratios of Japanese households classified by age. The ratio is close to zero for those aged below 35years and increases to approximately six percent for the 35?44 age group. This ratio gradually increases with age[4]. This patternshows that despite some critics’ arguments, aged households are not too conservative to hold only bank deposits. Figure 15?17indicate that holding risky assets requires a certain level of total financial assets and that this condition is met only in those intheir 50s and above in Japan.
Figure 18 shows Japanese households’ net savings (gross savings minus liabilities) categorized by age. During the ages of 35?44,net savings are negative because of relatively small gross savings and a significant amount of liabilities, mostly consisting ofhousing loans. Given this financial situation, naturally the priority is to avoid investing in risky assets and secure sufficient liquidassets.
As mentioned in section 2.2, holding a substantial amount of housing loans is one of the main reasons for the low risk-asset ratioof the younger generation in Japan. According to a survey conducted by the Ministry of Land, Infrastructure, Transportation, andTourism, the purchase of single-family housing peaks at the age of 41, and the annual mortgage payment amounts toapproximately US$ 8900 to US$ 10,000, which is approximately 20 percent of their annual income. Approximately 60 percent ofthose who participated in the survey responded that paying a mortgage is a substantial burden for their family. Figure 19 comparesthe ratio of disposable income to the outstanding amount of housing loans in Japan and the U.S.
The figure shows that in Japan, households’ burden of housing loans is increasing, while that in the U.S. is decreasing. In Japan,disposable income barely grew during 2013?2022, while the value of housing loans increased. In the U.S., the value of housing loansalso increased, but the growth rate of disposable income was higher than that of housing loans, lowering the ratio. The increasedburden of housing loans for the younger Japanese generation has caused liquidity constraints and restrained investments in riskyassets.
3.3. Accumulating financial assets for life after retirement
The last section of this chapter compares the types of financial assets accumulated for life after retirement in Japan and the U.S.In the U.S., the main types of assets for such purposes are defined contribution plans such as 401(k)s and IRAs. The 401(k)commenced in 1981 and the amount outstanding is US$ 18.1 trillion. In Japan, corporate-type defined contribution plan wasestablished in 2001, 20 years after the U.S. The amount outstanding for all types of DC, including corporate type, is growing swiftlyin Japan, but the latest value remains at US$ 0.16 trillion, or one hundredth of the U.S. value. Investment value per person are US$ 34,300 for Japan’s corporate-type DC, and US$ 133,300 for the American DC[5].
Figure 20 and 21 show the purpose for holding financial assets categorized by age. Households in both countries show a similarpattern, where in their 20s and 30s, priority is placed on expenditures for housing and children’s education, while preparing forretirement gradually increases and becomes a top priority after the age of 40 and beyond.
Figure 22 depicts the results obtained from surveying Japanese households regarding the criteria for choosing their financialproducts. Profitability is the first choice for the younger generations. However, as the age category rises, their priority changes toasset safety rather than profitability. This pattern is consistent with the life-cycle theory. Younger generations prioritizingprofitability is also consistent with the accumulation of retirement assets. The contradiction between figure 17, in which youngergenerations show a low risk-asset ratio, and figure 22, in which they prefer profitable financial assets, could be explained byliquidity constraints.
4.Macroeconomic effects of shifting financial assets
This chapter discusses the third point; how the shift of financial assets from bank deposits to investment instruments, affects themacro economy. Usually, households are considered a fund-surplus sector and funds are provided to fund-shortage sectors, suchas the government or corporations, via indirect financing through banks or direct financing through capital markets. SomeJapanese critics argue that “to revive the Japanese economy, financial assets accumulated by households should be utilized.”Their argument could be interpreted as “Japanese households’ 1,000 trillion yen of bank deposits bear little interest rates and arelying dormant in banks. Such funds should be invested in growing industries, including venture companies, to revive the Japaneseeconomy.” This chapter discusses several issues related to this statement.
First, the Japanese corporate sector is not constrained by fund shortage. Figure 23 shows the lending attitudes of Japanesefinancial institutions. After the global financial crisis, lending attitudes continued to be relaxed and the corporate sector as a wholewas not facing liquidity constraints.
Second, in a normal economy, nonfinancial corporations require funds for fixed investments, such as increasing productioncapacity or constructing new buildings. If internal funds are insufficient for such investments, they must rely on bank lending orthe issuance of new stocks and/or corporate bonds. However, figure 24 shows that the Japanese corporate sector, as a whole,has been a fund-surplus sector for more than two decades. Fund surplus in the corporate sector implies either a lack ofinvestment opportunities to expand business operations and improve productivity or corporate managers are too risk-averse andreluctant to invest. Either way, it is one of the main reasons for the long-term stagnation of the Japanese economy.
Under such circumstances, a household’s direct purchase of newly issued stocks has little effect on stimulating fixedinvestments. The number of stocks newly issued by listed companies was US$ 957 million in 2022, and the five-year-average wasUS$ 4,226 million. As the total amount of nominal fixed investments in fiscal year 2022 was US$ 715 billion, even if householdspurchased all newly issued stocks, it constitutes only 0.6 percent of total investments. Meanwhile, purchasing stocks from stockmarkets has little effect on corporate funding, except that stock prices may rise with increased trading.
Third, it is often said in Japan that “to stimulate the economy, providing risk money from the household sector to growingcompanies, including venture companies, are necessary.” The amount of investments in start-up companies in Japan was only US$ 4.3 billion in 2020, whereas that in the U.S. was US$ 143 billion. However, the amount in Germany is US$ 5.6 billion, which ismarginally greater than that in Japan. This suggests that the U.S. fund-supplying system has developed outstandingly, rather thanthe amount in Japan being too small. Additionally, investing directly in venture companies or capital requires professionalknowledge that cannot be handled by households.
Fourth, how is the Investment-Savings (IS) balance affected by government policies that shift household funds to risky assets?Figure 24 indicates that the corporate sector continues to be a fund-surplus sector, whereas the government sector is running ahuge deficit financed by issuing government bonds. Currently, approximately 90 percent of government bonds are held by financialinstitutions, including the central bank.
It is widely known that IS balance holds only ex-post. As a thought experiment, if a significant amount of household bank depositswere to flow into risky assets, then as financial institutions lose deposits, they are forced to sell government bonds. This couldcause a decrease in bond prices (increase in interest rates), and if the central bank tries to stabilize interest rates, it mustconduct security purchasing market operations. This results in a shift in government bonds from private financial institutions tothe central bank. If the central bank does not intervene in the bond market, rising interest rates will exert downward pressure onbusiness activities. Additionally, as the above discussion implies, even if households purchase all newly issued stocks, thestimulating effects on productivity or economic growth rates seem to be marginal. Therefore, it seems unlikely that shiftinghousehold financial assets as a starting point for government policy will succeed in changing the IS balance toward a more growth-oriented pattern[6].
The current combination of the fund-surplus corporate sector and government deficits results from the response of individualsectors to long-term economic stagnation. Therefore, judging from the IS balance, it seems unlikely that the relocation ofhouseholds’ financial assets alone could increase economic growth. Conversely, economic theory predicts a causality of“economic growth → household asset relocation,” and not vice versa, as advocated by the government. As discussed under theChapter 3, if the Japanese economy succeeds in increasing economic or productivity growth, disposable income of householdswould also increase. Consequently, younger generations would increase the purchase of risky assets, as liquidity constraints wouldbe relaxed. This causality may be more convincing than the reverse scenario.
5.Conclusion
This study critically analyzed the Japanese government’s policy of shifting households’ financial assets from bank deposits toinvestment instruments. There were three points for discussion; first, whether to use the U.S. households’ allocation pattern as abenchmark, second, the effects of long-term stagnation of the Japanese economy on the asset accumulation of households, andthird, the macroeconomic impact of asset relocation.
Regarding the first point, the financial asset allocation of U.S. households is extremely biased toward the top 5 percent wealthiestgroup, whose risk-asset ratio is outstandingly high. Therefore, care should be taken when using the U.S. allocation pattern derivedfrom aggregate household financial assets as a benchmark for policies to relocate Japanese households’ financial assets towardrisky assets.
Regarding the second point, the growth rate of Japanese households’ financial assets is distinctly lower than that of U.S.households. This was primarily because of the long-term stagnation of the Japanese economy and sluggish stock prices.Additionally, the younger Japanese generation’s low risk-asset ratio is not caused by their risk aversion, but mainly by liquidityconstraints imposed by risky nonfinancial assets, such as real estate.
Regarding the third point, the corporate sector remains a fund-surplus sector in Japan. This implies that liquidity constraints donot hamper its growth, and that shifting households’ bank deposits to risky assets will not accelerate the growth rate of theJapanese economy. Conversely, if the economy starts to grow faster, households will spontaneously increase investments in riskyassets as rising disposable income and relaxed liquidity constraints are realized.
There are other points of discussion that are not addressed in this study. The first is the entry cost associated with startinginvestments. The entry cost for Japanese consumers includes a sound level of financial literacy, further structural refinement ofDC plans and other financial products, and improvement in the trustworthiness of financial institutions. Currently, the JapaneseFSA is planning or engaging in various policies to lower the entry cost for investments, including improving financial literacy, thestructural renewal of the Nippon Individual Savings Account, a personal-type defined contribution plan, and the establishment of anindependent financial adviser system.
Another point for discussion is that the rising inflation rates in Japan, starting from 2022, lowered the real interest rates ofbanking deposits to negative territory, as the nominal deposit rate was very close to zero percent. The significant changes in therelative profitability of major financial instruments, not seen for decades, may have caused spontaneous asset relocation byJapanese households.
(References)
BEA (2022) ”Gross Domestic Product” Bureau of Economic Analysis
BOJ (2022) “Flow of Funds Statistics of Japan, the U.S. and the Euro area” Research and Statistics Department, Bank of Japan
BOJ (2023) “Tankan, a short-term economic survey of enterprises in Japan” March 2023
CCFSI(2022) “Consumer Survey on Finance” Central Council for Financial Services Information
Eurostat (2022) “GDP”
FRB(2019) “Survey of Consumer Finances”
Hirota (2018) “Simple questions on shifting from bank deposit to investment instruments movement” Masato Hirota, Shoken-keizai-kenkyu No.101 (Japanese only)
ICI(2023)”2022 Investment Company Fact Book” Investment Company Institute
Ito, Takizuka and Fujiwara (2017) “Portfolio selection by Japanese households :Investigation using Japanese and U.S. householdquestionnaire survey” Ito Yuichiro, Takituka Yasutaka, Fujiwara Shigeaki, Research Lab No.17-E-5, Bank of Japan
Kakei-Chosa (2021) (2022) “Kakei-Chosa <Saving and Liability>” Ministry of Internal Affairs and Communication
Maeda (2017) “Challenges for campaign on shifting from bank deposit to investment instruments movements” Masataka Maeda,Gekkan-sihonn-sijo April 2017(Japanese only)
Naikaku-fu (2022) “GDP statistics”
Shouhi-Jittai-Chousa (2019) Ministry of Internal Affairs and Communication
Zenkoku-syohi-doukou-chosa (2014) Ministry of Internal Affairs and Communication