No. 15 — April 14, 2000

Feature Article

WHERE WILL JAPAN'S MATURING POSTAL SAVINGS GO?

Arthur J. Alexander

Issue Index aaaa 2000 Archive Index aaaa Subscriber Area aaaa Home


Summary

Measured by deposits, Japan's postal savings system is the world's largest financial institution. Its almost ¥260 trillion ($2.4 trillion at ¥110=$1.00) on hand at the end of January 2000 represented one-third of all Japanese bank deposits. Driving the current interest in the postal savings system, though, is the fact that some ¥100 trillion ($909.1 billion) in high-yielding deposits that were guaranteed for up to 10 years — so-called teigaku accounts — will be reaching maturity in the 24 months beginning in April 2000. In that month alone, the peak period for maturing accounts, an estimated ¥5.9 trillion ($53.6 billion) will be up for grabs. Clearly, the flow of money into and out of the postal savings system is important to Japan and, given its scale, to financial markets everywhere.

Financing for many of the government's off-budget operations could be in jeopardy if the money in the maturing accounts leaves the postal savings system. Such a massive outflow is the exact hope of other financial institutions, including foreign investment firms with Tokyo offices. Some financial analysts have speculated about a possible surge in Japanese money seeking higher returns in U.S. markets, an outcome with implications for exchange rates, interest rates and American stock yields.

This report assesses the likelihood of a large volume of money leaving the postal savings system. It finds that depositors have few reasons to transfer maturing funds to other assets since the postal savings system not only offers comparable returns but also greater safety than the alternatives. Moreover, Japanese savers have shown little inclination to put their money into riskier, higher-return investments in numbers large enough to make a difference. Considerable movement into and out of various accounts is likely, but when the dust settles, little net change will have occurred. A huge amount of money still will be socked away in postal savings accounts.

Top


Scale Of The Issue

In April 1990, when their accounts reached the end of a 10-year period of guaranteed rates, Japanese savers withdrew ¥22.2 trillion ($201.9 billion at ¥110=$1.00) in deposits from the postal savings system. One of the main forces behind this massive withdrawal was that interest rates on alternative instruments had risen sharply above postal rates. In the fiscal year that began that April, some ¥76.7 trillion ($697.3 billion) was pulled out of so-called teigaku accounts. Almost immediately, however, the bulk of this money was redeposited back into the postal savings system and, within a year or so, most of it had made its way back into teigaku accounts.

All told, ¥165.9 trillion ($1.5 trillion) surged out of the postal savings system in FY 1990. However, that figure was more than matched by the fairly quick return of most of those funds plus the inflow of new money (see Table 1). The echo of those decade-old events will be heard in FY 2000 and in FY 2001 as the current crop of teigaku accounts reaches maturity.
e--aaaaa- --aaaaa-

Table 1: Japan's Postal Savings, FY 1970-FY 1998

(in billions of yen)

Fiscal Year

Postal Savings Balance*

Postal Savings Withdrawals

Postal Savings Receipts**

Teigaku Balance*

Teigaku Withdrawals

Teigaku Receipts**

1970

¥7,744

¥4,857

¥6,284

¥5,431

¥1,353

¥2,540

1971

9,654

n.a.

n.a.

7,042

n.a.

n.a.

1972

12,293

6,898

9,351

9,278

2,274

4,508

1973

15,377

8,261

11,344

11,776

2,742

5,238

1974

19,431

11,154

15,209

15,012

3,702

6,939

1975

24,566

12,041

17,176

19,649

4,303

8,939

1976

30,525

13,340

19,298

25,300

4,637

10,289

1977

37,726

14,014

21,216

31,932

4,971

11,603

1978

44,996

15,389

22,658

38,272

5,324

11,664

1979

51,912

24,506

31,422

44,248

12,022

17,999

1980

61,954

42,585

52,628

54,470

27,125

37,347

1981

69,568

23,301

30,914

61,529

8,363

15,422

1982

78,103

25,447

33,982

69,499

8,453

16,423

1983

86,298

26,223

34,419

77,338

9,574

17,414

1984

94,042

30,709

38,452

84,051

12,128

18,841

1985

102,998

33,066

42,022

92,856

12,701

21,505

1986

110,395

32,605

40,002

99,626

10,670

17,441

1987

117,391

36,001

42,996

106,260

11,876

18,510

1988

125,869

37,146

45,625

114,880

11,250

19,870

1989

134,572

48,990

57,693

118,561

18,455

22,136

1990

136,280

165,879

167,587

109,567

76,707

67,714

1991

155,601

97,986

117,306

133,737

26,499

50,669

1992

170,091

107,870

122,360

147,446

30,299

44,007

1993

183,535

93,183

106,627

159,958

20,443

32,955

1994

197,590

111,925

125,981

172,289

28,587

40,918

1995

213,437

105,798

121,645

184,774

23,204

35,689

1996

224,887

109,687

121,137

193,859

19,461

28,546

1997

240,546

124,666

140,324

202,128

17,113

25,382

1998

252,549

119,322

140,010

214,170

16,772

21,112

*As of the March 31 end of the fiscal year.
**Including interest payments on existing deposits.

Source: Ministry of Posts and Telecommunications, Postal Savings Bureau

An outflow of maturing deposits on an unprecedented scale is possible. The Postal Savings Bureau, part of the Ministry of Posts and Telecommunications, estimates that about ¥106 trillion ($963.6 billion) will be available for withdrawal over the next 24 months (see Table 2). Since taxes on interest payments are deducted automatically when teigaku accounts mature, Ministry of Finance budget planners are looking forward to windfall revenues of almost ¥7 trillion ($63.6 billion) in FY 2000 and FY 2001.
e--aaaaa- --aaaaa-

Table 2: Teigaku Accounts Reaching Maturity in FY 2000-FY 2001

(in trillions of yen)

Fiscal Year

Original Deposits

Deposits Remaining

Accumulated Interest

Total

Tax on Interest

Free Funds

2000

¥62.1

¥40.0

¥18.0

¥58.0

¥3.6

¥54.4

2001

44.4

32.0

16.0

48.0

3.2

44.8

Sources: Ministry of Posts and Telecommunications, Postal Savings Bureau, and JEI estimates

The Finance Ministry is not the only contender lining up for a piece of the action when all this money comes into play. Domestic and foreign banks as well as other financial services providers are promoting the advantages of transferring the U.S. freed-up funds into an assortment of investment vehicles that were not available in 1990. In particular, managers of mutual funds, known in Japan as investment trusts, have been developing products as fast as the Big Bang financial reforms permit.1 Many mutual funds have grown rapidly over the past two years, stimulated in part by rises in the Tokyo Stock Exchange, which finally seems to have shaken the post-"bubble economy" lethargy that marked the 1990s.

The Postal Savings Bureau is trying to keep its hands on the money it now controls. Since these deposits are used by the government to fund public corporations through the Fiscal Investment and Loan Program,2 the second capital budget, a huge outflow of funds would require Tokyo to drastically shift its financing strategy. The Bank of Japan has established contingency plans to provide at least near-term coverage in case massive withdrawals from the postal savings system cause revenue shortages.

Economic analysis and econometric estimates suggest, however, that most of the money will remain in the postal savings system. This conclusion will be discussed after the postal savings system is described.

Top


Japan's Postal Savings System

Japan was one of the first countries to create a postal savings system. The government used the United Kingdom's system, established in 1861, as a model for its 1875 creation. New Zealand, Canada and Belgium followed in the next few years. The stated goals of the Meiji planners were to encourage thrift and to gather small savings to finance industrial development. Although those early bureaucrats took credit for teaching thrift to their fellow citizens, Columbia University's Hugh Patrick has suggested that the remarkable growth of postal savings in the early years attested to an already widespread habit of savings.3 The post office began offering money order services in 1875, giro (direct transfer) services in 1906 and life insurance in 1916. Approximately 10,000 post offices provided savings vehicles between 1900 and 1940.

The United States established a postal savings system in 1910; it was officially discontinued in 1967. However, the U.S. system never reached the size of its Japanese counterpart. Nor did it offer the same range of services. For example, in contrast with Japan and many European systems, the American plan did not provide direct transfers. The rationale for the establishment of the U.S. system followed that of the other countries with postal savings systems. Its main customers, though, were not people in rural areas. Rather, they were urban immigrants who did not trust banks and who had participated in postal savings systems in their native countries.4 U.S. postal savings peaked in the mid-1940s, when the interest rates on those accounts exceeded the ones offered by banks. At that time, post office accounts represented 5 percent of deposits in all deposit-taking institutions.

Funds placed in the American system were used to purchase U.S. government bonds and played an important role in both world wars. However, competition with the commercial banking industry in the 1930s and the 1940s led to closer examination of the system's rationale and costs. The Hoover Report5 and subsequent studies called for its abolition. President Lyndon Johnson's legislative proposal to close the institution was supported by the postmaster general and passed with little congressional debate in 1966. The only strong opposition came from the postal workers' union.

As in the United States, the Japanese government sold bonds to the postal savings system; in addition, starting in 1898, deposits were used to make loans to companies through private-sector banks. By the 1930s, postal savings deposits had been placed under the jurisdiction of the Finance Ministry's Trust Fund Bureau and were being lent to government-affiliated institutions. According to Columbia University's Patricia Kuwayama, "the convenience of using postal savings as a huge slush fund for various policy lending purposes seems to have won out easily over time."6

The post office offered convenience and safety to depositors, although rates usually were somewhat less than what was available from commercial banks. As Japan's financial sector was gradually liberalized in the 1980s, the range of products offered by the postal savings system expanded to remain competitive. Indeed, these investment options must have been more than competitive because, despite the government's implicit guarantee from the 1950s on that no bank would fail, postal savings continued to grow.

As of January 31, 2000, almost ¥260 trillion ($2.4 trillion) was deposited in postal savings accounts. They now are 52 percent as large as commercial bank deposits and equal about one-third of total private and public deposits.

These shares have risen in the past two decades despite the April 1988 elimination of the exemption, up to a prescribed limit, of interest income from taxation. This exemption applied to the earnings on all deposits, but its abolition particularly affected post office customers, who often had multiple accounts to avoid the ceiling on individual deposits. The Postal Savings Bureau appeared to be especially lax in policing this practice, perhaps, as it argued, because it was insufficiently computerized. In any case, the ratio of postal savings deposits to those of commercial banks, which had been 39 percent in 1986, fell to 29 percent in 1990 after the tax advantage was discontinued. Since then, though, the figure has climbed back to the previously mentioned 52 percent.

Top


Teigaku
Accounts

Most postal savings are in teigaku chokin (postal savings certificates). Since 1980, their share of total postal deposits has been in the 85 percent to 90 percent range. Introduced in 1941, teigaku chokin paid 1 percentage point less than the interest rate on the longest-term time deposit available at commercial banks. Minor exceptions arose when official rates were changed; then, the postal rate lagged for a few months. When compounding was included, the yields on teigaku chokin approached what could be made at banks.

This 1-point differential formula was applied until the liberalization of Japan's financial markets started in earnest in the mid-1980s. In the 1990s, commercial banks were allowed to offer market rates on deposits. In 1994, MOF and MPT agreed from October of that year to liberalize interest rates on liquid deposits held by both private financial institutions and the postal savings system. Furthermore, the rates on the two types of deposits would be linked.7 In practice, this policy change meant that postal rates would follow market-determined deposit rates.

In fact, the differential has moved in a small range around zero. The authorities agreed that the teigaku three-year rate would be 5 percent (not percentage points) below the bank three-year rate, or 10 percent lower when interest rates were high.8 However, the spread has varied as banking operations have been deregulated and as interest rates themselves have responded to market forces. The gap, though, has tended to narrow with the recent fall in interest rates. The one major anomalous period occurred in 1989, when banks introduced money market certificates and their rates soared above those offered by teigaku accounts.

The outstanding characteristic of teigaku accounts is their guaranteed interest rate for up to 10 years. After the first year, the rate rises every six months for three years according to a schedule fixed at the time of the deposit; it then is constant until the end of the 10-year period. No-penalty withdrawals, however, can be made at any time after the first six months.

These features are shared by American-style put options, which give the purchaser the right to sell the asset at any time before maturity at a predetermined price. In contrast, European options convey the right to sell the asset only at maturity. Valuation of a teigaku account, therefore, must consider the option premium implied in the terms of the deposit.9 The most direct comparison is the five-year deposit offered by commercial banks. Until late 1993, though, banks had only a three-year instrument.

Banks and other financial institutions have argued that the cost of teigaku accounts to the government has been greater than simple interest rate calculations would suggest because of the option premium granted to savers. A comparison of the value of the implied teigaku premium with the deposit instruments available from banks indicates that the difference can be worth as much as 1.5 points. This spread helps to explain the growth of postal savings in the face of nominal interest rates that are below those offered by banks.

Higher nominal interest rates are about the only advantage for individual depositors that banks have over the postal savings system. Otherwise, the postal system offers the better investment option for risk-averse savers. Moreover, postal savings are implicitly guaranteed by the government, whereas the safety of bank deposits has become a concern despite the existence of unlimited government deposit insurance (see JEI Report No. 2B, January 14, 2000) because of the failure of banks both large and small and the collapse of other financial institutions.10 Postal savings accounts also provide greater convenience since there are 24,000 post offices versus 16,600 bank branches.

Top


Predicting Teigaku Trends

The first issue that must be addressed is that teigaku deposits and withdrawals move together. In fact, within the same month, they are almost the same. The correlation of 0.80 implies a high incidence of short-term churning of accounts. Funds are redeposited in the same month that they are withdrawn or vice versa. Most of the action seems to flow from withdrawal to deposit, though. Funds in accounts that mature are automatically deposited in regular postal savings accounts. Some of the money is withdrawn immediately, a portion is reinvested in teigaku accounts in short order and the rest remains in the other accounts for longer periods. Part of this balance later migrates back into teigaku accounts.

A statistical regression analysis of monthly new deposits (subtracting interest payments) indicates that some 50 percent of withdrawn money is redeposited within 30 days and another 25 percent within 12 months. Of course, some of the action occurring within the same month could represent deposits that are quickly pulled out.

Although possible, this behavior seems to be less likely for two reasons. First, withdrawals from teigaku accounts within the first six months incur a penalty; short-term funds could be parked more profitably in other types of accounts. Second, the statistical evidence suggests that lagged withdrawals show up in later deposits, but that lagged deposits are not reflected in withdrawals, at least within the same 12-month period.

If savers were using their teigaku accounts as short-term refuges for their loose change, deposits likely would be withdrawn before they were one or two months old. This has not happened. In fact, deposits lagged one month have a negative effect on withdrawals. Put another way, the greater last month's deposits, the smaller will be this month's withdrawals — an indication that money is deposited for longer-term periods.

Figure 1 shows teigaku deposits of new funds (not including interest) and withdrawals for 1990. That year was the most volatile in the history of Japan's postal savings system. Yet even in those wildest of times, inflows closely tracked outflows except for the month of April.

To deal with the problem of correlated deposits and withdrawals, it is useful to focus on net changes. Although several different choices are possible, monthly changes in teigaku balances and changes in balances minus interest payments are examined here. (The reason for ignoring earnings is that these payments are automatic and do not reflect current customer decisions.) The problem with these choices is that the two data series are published many months after the close of the fiscal year. Consequently, the most recent information extends only through March 1999.

The Postal Savings Bureau does publish a series on teigaku receipts and withdrawals on a current basis. However, these data are incomplete. They do not include automatic payments from postal savings accounts for such purposes as insurance or retirement income; nor do they include direct payroll deposits into the accounts. Therefore, it is impossible to calculate teigaku balances directly from this information. The fact remains, though, that these statistics represent the current choices of consumers and may be a more sensitive index of contemporary decisionmaking processes.

The same basic equations are used in all the estimates. The dependent variables are the differences between seasonally adjusted deposits and withdrawals. The independent variables include the one-month lagged dependent variable and the following lagged independent variables: interest rates on teigaku accounts, the spread between one-year teigaku rates and one-year time deposit rates for amounts greater than ¥10 million ($90,900), the difference between one-year teigaku rates and the return on the benchmark government bond, and seasonally adjusted disposable income. Also used are the one-month differences in the three interest rate variables and disposable income. U.S. One additional feature is a dummy variable that equals 1 if the interest rate on government bonds has risen over the past six months and 0 otherwise. The rationale for this is that declining interest rates could motivate savers to put their funds into guaranteed teigaku accounts to lock in high rates, while rising rates might encourage people to cash out their accounts and open new ones in the postal savings system or elsewhere, depending on spreads. As expected, the dummy variable always has a large and statistically significant negative coefficient.11

The results from the equation estimating the difference between teigaku receipts (minus interest payments) and withdrawals are shown in Figure 2, which compares actual teigaku balances with those produced using the equation. The computation was a full dynamic simulation. The estimated value of the lagged dependent variable rather than the actual amount was used to estimate the current value, along with the other independent variables. The only actual teigaku balance was the first observation in April 1988. All subsequent quantities were calculated by adding the estimated change to the previous month's estimated balance. Therefore, except for that initial month, the entire predicted curve shown in Figure 2 was derived from the equation (with interest payments added back in). Such a procedure was employed to avoid exaggerating the predictive worth of the equation by biasing the results.12 Even though the available teigaku data only go to March 1999, the other series run through December 1999. Therefore, the equation can be used to predict up to that endpoint.

Two findings stand out. The equation projected no decline in teigaku balances in FY 1999, and it anticipated the FY 1990 outflow quite well. In order to examine the predictive power of the equation, Figure 3 blows up the curves right before and just after FY 1990. It shows that the sharp fall in teigaku balances in April 1990 and in subsequent months was forecast well in advance. Throughout FY 1989, the equation was indicating a decline that accelerated in March, April and May 1990. The actual collapse in April of that year occurred when a substantial number of accounts matured and depositors were compelled to make choices.

The actual contraction in teigaku balances, though, overshot the prediction by ¥3.1 trillion ($28.2 billion) or about 2.9 percent of the total. Still, the equation accounted for most of the actual shrinkage and predicted it well in advance. Relatively stable teigaku interest rates and increasingly negative spreads between them and what was available from competing financial institutions explain the forecast decline. For example, rates on time deposits at banks had climbed from 4.8 percent in December 1988 to 7.6 percent a year later. In contrast, postal rates went up by only 1.4 points in the same period.

Finally, the teigaku data through December 1999 were examined. As noted, these statistics do not include various automatic transactions, but they may be more sensitive indicators of current investor thinking. Figure 4 plots the difference between actual teigaku receipts and withdrawals and those estimated dynamically from the equation. While the reported balance fell below zero for most of 1999, the predicted value remained stable and in positive territory. The most recent data series (admittedly incomplete) are the only indication of possible coming trouble in attracting teigaku savings, although the equation fails to detect such a trend.

Top


Trends In Postal Savings

What might happen to teigaku accounts this spring is only part of the story. It also is important to ascertain whether funds leave the postal savings system when they are withdrawn from these term deposits. To illustrate the necessity of looking beyond teigaku accounts, Table 3 presents the monthly changes in teigaku and total postal savings in 1990. Although almost ¥11.8 trillion ($107.3 billion) left teigaku accounts in April, ¥9.8 trillion ($89.1 billion) was parked in other post office accounts. Total postal savings, therefore, fell by only ¥1.9 trillion ($17.3 billion). Much of this outflow seems to have gone into time deposits at commercial banks, which surged that month. Over the longer term, a simple statistical trend suggests that time deposits at banks will pick up about two-thirds of the net outflow from the postal savings system.
e--aaaaa- --aaaaa-

Table 3: Monthly Changes in Postal Savings Accounts, 1990

(in billions of yen)

Month

Teigaku Accounts

Other Postal Accounts

Total Postal Savings

Bank Time Deposits

January

¥297

¥119

¥416

¥1,166

February

173

656

830

2,082

March

382

19

401

1,074

April

-11,764

9,836

-1,928

3,702

May

-961

377

-584

1,989

June

-650

1,060

410

1,994

July

164

-40

123

2,464

August

-9

326

317

1,377

September

179

-469

-290

1,638

October

-110

-156

-266

1,463

November

-970

415

-556

2,381

December

1,410

489

1,900

-32

Source: Ministry of Posts and Telecommunications, Postal Savings Bureau

In order to predict the effects of the basic forces acting on the volume of postal savings, the same type of equation employed to estimate teigaku accounts was used. Predicted seasonally adjusted postal savings were so close to the actual total that it is not very revealing to reproduce the data depicting the two. Significantly, however, there was no hint of a decline in postal savings amounts in 1999 that might presage sharp withdrawals in April of this year or in subsequent months.

As was true for teigaku accounts alone, it may be more revealing to look at estimated net additions to total postal savings balances minus interest payments. Actual and estimated values of this perhaps more sensitive measure are shown for recent years in Figure 5. Again, since 1998, a slight general downturn in actual net additions to postal savings accounts is apparent — a result not predicted by the equation, which indicates continued inflows. In fact, Figure 5 tells a story similar to that in Figure 4. In any event, total net withdrawals in all of 1999 added up to just ¥1.1 trillion ($10 billion) or 0.4 percent of total postal savings, hardly a sign of a major outflow of funds in coming months.

Top


Alternative Portfolio Choices

The theory underlying the equations used to predict behavior in this report is based on a simple portfolio-choice model that allows people to allocate their assets among postal savings, bank deposits and such other safe instruments as government bonds. The equations signal no decline in postal savings primarily because the financial reasons to move that money into other assets are lacking. Put another way, interest rates on teigaku deposits are at historic lows but so are the rates on bank deposits and government bonds. In fact, rate spreads have been remarkably constant since 1996. In contrast, when savers took their money to post office branches 10 years ago, teigaku earnings finally were equivalent to those offered by banks since long-term rates on teigaku accounts rose to more than 6 percent from April 1990 onward.

The failures over the last two-plus years of several major banks and a number of their smaller competitors as well as the widely publicized weaknesses of Japan's hundreds of shinkin (credit associations), shinkumi (credit cooperatives) and agricultural banks raise legitimate concerns about the safety of alternatives to the postal savings system for individuals' assets. Studies in both the United States and Japan show that bank failures are a powerful determinant of the attractiveness of postal savings.13

Another consideration that favors postal savings is the likelihood that interest rates eventually will rise above the floor on which they now rest. Investments in such fixed-rate assets as bonds would suffer capital losses if market rates rose. Also, the commitment of funds to fixed-term instruments like certificates of deposit could lock funds up in low-return investments if rates do increase. The flexibility and the safety of postal savings guard against these possibilities.

Mutual funds are a possible portfolio choice. In fact, they are attracting a great deal of attention from industry analysts and the media because of the huge amount of money tied up in low-return deposits at the same time that these investment alternatives have become much easier to buy and far more numerous and diversified.

One advantage of stock-based mutual funds is their expected gain in value if a resumption of economic growth in Japan causes interest rates to rise. Moreover, the 20 percent to 40 percent annualized jumps in various Tokyo Stock Exchange price indexes over the past six months or so (see JEI Report No. 14B, April 7, 2000) are obviously in a different league than the near-zero rates on bank or postal savings deposits. Similar or even bigger increases in American asset values in recent years have drawn the attention of Japanese investors to foreign funds as well.

Mutual funds, however, have negative attributes. Just as prices can go up, they also can go down, as American investors have been reminded recently. Whether Japanese savers, noted for their aversion to risk, will find the returns on mutual funds high enough to overcome their latent fears remains to be seen. Moreover, even though stock funds may increase in value if interest rates rise because of faster growth, bond funds will fall for the same reason. And, today's high returns on foreign funds incorporate the added risk of exchange rate volatility.

Table 4 shows the amount of money invested in Japanese-style domestic mutual funds as well as assets denominated in foreign currencies. In order to estimate the rate of new investment in these financial alternatives, the value of stock funds has to be adjusted to reflect changes in stock prices since these contribute to their market value.
e--aaaaa- --aaaaa-

Table 4: Value of Investments in Japanese Investment Trusts, by Type, 1997-2000*

(in trillions of yen)

End of Period

Stocks (Unadjusted)

Stocks (Adjusted)**

Domestic Bonds

Other

Total (Adjusted)

Foreign Assets

1997

¥5.4

¥5.4

¥23.0

¥12.3

¥40.7

¥3.5

1998

5.1

5.5

21.7

15.9

43.1

5.2

1999

10.4

7.1

22.3

18.7

48.1

3.6

2000*

11.9

8.2

26.4

19.9

54.5

3.6

*As of February 29, 2000.
**Value of assets in stocks adjusted to yearend 1997 base by Tokyo Stock Exchange price index.

Source: The Investment Trusts Association

Recently, bonds have represented close to half of all domestic mutual funds assets. Stocks, though, have gone up in absolute value and as a share of the total, mainly because of the rise in share prices. After adjusting for such increases, the volume of new money invested in domestic mutual funds grew by ¥5 trillion ($45.5 billion) in 1999 and another ¥6.4 trillion ($58.2 billion) in just the first two months of 2000. Much of the recent gain has been in bond funds rather than in more volatile stock funds. At the same time, the yen value of money invested in mutual funds denominated in foreign currencies has changed little since 1997, another reflection of the risk-averse character of Japanese investors. Given that the amount of postal savings available for reinvestment in FY 2000 is almost the same as the value of the assets in all of Japan's domestically oriented mutual funds, the securities industry will do well to attract even a small share of these funds.

Recent surveys of Japanese savers indicate a continued flight to safety.14 A rising proportion of respondents rank this investment consideration before liquidity or profitability. Although the earnings objective rose slightly in importance during the asset-price bubble of the late 1980s, only 15 percent of the participants in a July 1999 survey chose it as their determining investment factor, whereas 56 percent opted for safety and another 25 percent picked liquidity. Unless these preferences shift markedly in the next few months, mutual funds are not likely to be the main choice of savers looking to invest their maturing teigaku monies.

Top


Conclusion

As FY 1999 was coming to a close, various organizations made predictions about where all the maturing postal savings money would go. A survey by Japan's leading business daily, Nihon Keizai Shimbun, of Tokyo-area households suggested that some ¥3.3 trillion ($30 billion) would be invested in mutual funds over the next two fiscal years. A Sanwa Research Institute study forecast a total outflow of ¥16 trillion to ¥19 trillion ($145.5 billion to $172.7 billion). According to its estimate, ¥2 trillion to ¥4 trillion ($18.2 billion to $36.4 billion) would end up in individual stocks or mutual funds, ¥13 trillion ($118.2 billion) would be put into savings accounts at commercial banks and the remaining ¥1 trillion to ¥2 trillion ($9.1 billion to $18.2 billion) would be invested in accounts denominated in foreign currencies.15

Last summer, the Postal Savings Bureau predicted that deposits under its control would fall by ¥16 trillion ($145.5 billion) during FY 2000 and by another ¥15 trillion ($136.4 billion) over the following fiscal year. These numbers reflect the estimated net impacts of withdrawals, redeposits and new deposits as well as the outflow of maturing teigaku account balances worth ¥27 trillion ($245.5 billion) — about half of the maturing funds. MPT has charged local post offices with retaining 70 percent of the assets that remain after the deduction of taxes and the mandatory withdrawal of amounts over the ¥10 million ($90,900) limit on individual accounts.16

The Japan Economic Institute's statistically based projections are sharply at odds with other forecasts. Although large outflows from teigaku accounts were recorded in FY 1990, more than 80 percent of those funds stayed in the postal savings system. The regression estimates indicate that nothing like the FY 1990 experience will occur. Even if large withdrawals are recorded in the first few months of FY 2000, later redeposits will result in a small net impact on total postal savings. Equally probable is a high incidence of short-term churning as funds are moved from one type of account to another. In other words, the system may experience temporary withdrawals as savers consider how to reinvest their formerly high-yielding funds. However, without drastic changes in interest rates — especially the spread between those on teigaku and commercial bank accounts — or major policy shifts that affect the role of postal savings, big transfers of money lasting more than a few months are not likely.

As it happened, the rate of reinvestment of maturing deposits in the first week of April was higher than what MPT had projected.17 Certainly, the experience of a single week is an inadequate basis on which to make a judgment about the entire fiscal year, but the continued popularity of postal savings accounts could fool a lot of analysts.

Kanako Yamada provided research assistance.

The views expressed in this report are those of the author
and do not necessarily represent those of the Japan Economic Institute

Top aaaa Issue Index aaaa 2000 Archive Index aaaa Subscriber Area aaaa Home


Notes
aaa

1aa For an overview of the first phase of Big Bang reforms, see Jon Choy, "Japan's Financial Market Big Bang: The First Shock Waves," JEI Report No. 22A, June 12, 1998. For an update on securities industry reform, see Jon Choy, "Japan's Securities Industry: From Big Bang To E-boom," JEI Report No. 22A, June 11, 1999. Return to Text

2aa For a discussion of the postal savings system's financing of the FILP and FILP reforms, see Douglas Ostrom, "Government Deficits And Debt: Tokyo's Dilemma," JEI Report No. 2A, January 14, 2000, pp. 9-11. Return to Text

3aa Cited in Patricia Hagan Kuwayama, Postal Banking in the United States and Japan, A Comparative Analysis (Discussion Paper No. 99-E-18) (Tokyo: Bank of Japan, Institute for Monetary and Economic Studies, June 1999), p. 27. Return to Text

4aa Ibid., p. 2. Return to Text

5aa The 1949 report of the Commission on the Organization of the Executive Branch of the Government, which was created by Congress in 1947 to recommend ways to improve the efficiency and the economy of government operations and was chaired by former President Herbert Hoover. Return to Text

6aa Kuwayama, op. cit., p. 29. Return to Text

7aa Bank of Japan, "About the Bank, Chronology of Events in Fiscal 1994." Available at http://www.boj.or.jp/en/about/nenpyo94.htm. Return to Text

8aa Ministry of Posts and Telecommunications, Postal Savings Bureau, Postal Savings in Japan: 1998 (Tokyo: 1999), p. 15. Return to Text

9aa Koichiro Kamada, "The Real Value of Postal Savings Certificates," BOJ Monetary and Economic Studies, XI, No. 2, November 1993, p. 60. Return to Text

10aa See Jon Choy, "Japan's Banking Industry: The 'Convoy' Disperses In Stormy Seas," JEI Report No. 10A, March 12, 1999. Return to Text

11aa A similar variable was used effectively in Thomas Cargill and Naoyuki Yoshino, "Japan's Postal Savings System: Financial Liberalization, Dilemmas, and Solutions." (Unpublished paper, April 1999.) Return to Text

12aa Since the lagged dependent variable appears on the right side of the equation and since there is a modest degree of autocorrelation in the data, using the actual lagged dependent variable would force the estimation to track the actual data and make the equation look like a better predictor than it may be. Return to Text

13aa Kuwayama, op. cit., p. 31. Return to Text

14aa Central Council for Savings Information. Available at http://www.saveinfo.or.jp. Return to Text

15aa "Maturing Of Postal Savings To Deeply Impact Economy," Nikkei Net Interactive, March 13, 2000. Available at http://www.nni.nikkei.co.jp. Return to Text

16aa "Postal Savings Braces For Outflow," The Nikkei Weekly, March 20, 2000. Return to Text

17aa "Ministry Says ¥1.9 Trillion Reinvested In Postal Deposits," Nikkei Net Interactive, April 5, 2000. Available at http://www.nni.nikkei.co.jp. Return to Text

The views expressed in this report are those of the author
and do not necessarily represent those of the Japan Economic Institute

Top aaaa Issue Index aaaa 2000 Archive Index aaaa Subscriber Area aaaa Home