Bank of Japan Purchases of Risky Assets: Lessons from Colonial America


Mark M. Spiegel

FRBSF Economic Letter 1999-17 | May 21, 1999

In its recent efforts to assist Japan’s troubled banking sector, the central bank has purchased large amounts of assets other than Japanese government bonds, including private sector commercial paper. The holdings of the central bank, the Bank of Japan (BOJ), were not trivial.

In its recent efforts to assist Japan’s troubled banking sector, the central bank has purchased large amounts of assets other than Japanese government bonds, including private sector commercial paper. The holdings of the central bank, the Bank of Japan (BOJ), were not trivial. As of November 1998, it held 6 trillion yen in commercial paper, about 43% of the entire market and 7% of its total assets. While these numbers recently have come down somewhat, they still represent significant components of the BOJ’s balance sheet. Private commercial paper differs from government bonds in that it bears some degree of default risk that is directly linked to the risk of insolvency in the firm backing the paper. It is unlikely that the prices paid for the paper adequately compensated for this default risk.

This Economic Letter explores the macroeconomic implications of central bank purchases of risky assets at non-market prices. It is worth noting first that such purchases do not raise concerns about the solvency of the BOJ. The BOJ issues “fiat” currency, i.e., currency which cannot be redeemed for some other asset, such as gold. Since an issue of fiat money carries with it no promise of future redemption, the solvency of the BOJ is never at issue. Unlike a commercial bank, the BOJ can always create additional assets in the form of additional currency.

Instead, the differences resulting from the purchase of risky commercial paper will depend on how the accumulation of this alternative asset alters expectations about future monetary policy. In this Economic Letter, we trace out the differential impact of expansionary monetary policy of this type. We argue that purchases of risky paper rather than government bonds can affect the public’s perception of implied future monetary growth and thereby affect current price levels.

We begin by examining a monetary regime in which money is formally backed by promises of future redemption in gold—the case of the American colonies. In this regime however, the public was uncertain as to whether these promised redemptions would take place. Thus, the inflationary impact of a monetary expansion was dependent on the public’s perception of the credibility and value of the promised future redemption.

We then turn to Japan. Despite the fact that Japan operates a fiat money regime, we demonstrate that the inflationary impact of a monetary expansion in Japan will also depend on the public’s perception of the quality of assets held by the central bank. In Japan’s fiat money case, this works through the impact of perceived central bank asset quality on expected future monetary growth.

Evidence from the American colonies

To motivate the argument that the expansionary impact of an increase in the money supply in Japan can depend on the assets purchased during that expansion, it is useful first to consider the effects of a monetary expansion in a simpler regime. One historical example which fits well is that of colonial America.

The colonies were free to issue their own notes, which carried with them the promise of future redemption; that is, the notes would be repurchased for assets of value, usually precious metals. When issuing these notes, colonial governments typically also legislated future tax levies to finance their redemption.

However, the degree of credibility of these promises of future redemption varied across colonies. Holders of these notes sometimes were uncertain whether these future tax levies would generate adequate funds for the redemption of the new notes issued, or whether the future revenues would be allocated towards note redemption at all.

Smith (1985), studies the cases of North and South Carolina and Maryland. In the case of the Carolinas, the credibility that future tax revenues would be used to retire most issues of new notes varied with changes in either the fortunes of the Carolina colonies or perceptions about the policy preferences of the colony governments.

When the public doubted the credibility of promises of future redemptions, they treated new note issues as permanent increases in the money supply. As a result, these issues led to decreases in the value of existing Carolina notes relative to other currencies, such as the British pound.

In contrast, during episodes when the governments moved to back their note issues by providing adequate financing to redeem outstanding notes and assuring that redemptions would actually take place, the public placed a high probability on future note redemption. Expansions of notes during these episodes were not seen as affecting the permanent money stock and, hence, had little impact on the value of Carolina notes.

The experience in Maryland was similar. Maryland issued paper currency in 1733 backed by promises of future redemptions to be financed by newly designated tax levies. The proceeds of these tax levies were to be invested in Bank of England stock and redeemed at fixed dates which were announced at the time of note issue. Because these Maryland notes were to be completely redeemed by the sale of the assets in the fund of Bank of England stock, there was no systematic relationship between new money issues and the value of existing Maryland notes. Instead, the impact of an expansion in the stock of Maryland notes depended on the value of assets in the fund set aside for future redemptions and the public’s confidence that these assets would be used for their announced purpose.

The experience in the colonies teaches us that impact on the general price level of an expansion in the money stock is likely to depend on the public’s expectations about the credibility of assets backing the future monetary redemption. If the public doubts the credibility of future redemption, they will perceive the increase in the money stock as permanent. The public’s perceptions about the future money stock affects their expectations about the future price level.

Increases in expected future prices, holding all else equal, will lead to increases in the current price level. Inflation acts as a tax on holders of cash balances. Investors perceiving an increase in this tax on holdings of cash balances would be unwilling to hold the entire outstanding currency stock at current prices. Current prices would therefore need to rise to equilibrate the money market.

Japan’s fiat money regime

Like most modern economies, Japan has a fiat money regime whose currency is not backed by a promise of future redemption. Instead, these notes have value due to their legal tender status, which ensures by law that they must be generally accepted in exchange for goods and services at prevailing price levels. As in the case of the colonies, where investors doubted the credibility of the backing for notes, one would expect a positive relationship between the rate of inflation and the expected future growth rate of the money stock in a fiat money regime.

However, the composition of assets purchased during a monetary injection can have an impact on the consolidated government’s balance sheet, i.e., the sum of assets and liabilities of the Bank of Japan and the Japanese Ministry of Finance. In doing so, it can influence the public’s expectations about the magnitudes of future monetary injections.

For example, consider the difference between the BOJ’s “open market” purchase of one yen’s worth of commercial paper bearing default risk and one yen’s worth of Japanese government bills, which presumably carry no default risk. Assuming that these assets are perfectly substitutable except for default risk, they would yield the same increase in the monetary base. Any difference in impact of these two transactions must then depend on their implications for the asset side of the BOJ’s balance sheet.

However, the manner in which open market purchases affect the expected future growth in the money stock differs in the case of a fiat regime such as Japan’s, as compared with a specie regime. While redemption of existing currency is not promised in the fiat regime, the assets acquired during the monetary expansion can have implications for future consolidated government deficits. Since the expected present value of a given nominal amount of paper bearing default risk purchased at non-market prices would be lower than the expected present value of the same nominal amount of government bills, the expected future revenue stream from the purchased paper also would be lower. This implies a reduction in central bank revenue. Since central bank revenue is handed over to the government, a reduction in central bank revenues implies a reduction in government revenues.

Holding government expenditures constant, this revenue reduction must be offset in the future by one of three responses: (i) increased government borrowing, (ii) increased taxes, or (iii) further expansion of the monetary base. The manner in which the type of assets acquired by the BOJ affects the impact of a monetary expansion on the Japanese price level cannot be assessed without knowing the public’s expectations about how a possible future shortfall in revenues would be financed.

If fiscal policy is used, either in the form of a tax increase or an increase in government borrowing, the net impact of a purchase of paper that ended up in default by the BOJ is likely to be less expansionary than if the given monetary injection were used to purchase government bonds. Either form of fiscal policy would reduce aggregate demand: a tax increase would directly reduce aggregate demand, while increased government borrowing would push up the real interest rate. On the other hand if the future shortfall is not expected to be financed through fiscal policy, the shortfall in revenues implies the need for an additional future monetary expansion in order for the government to satisfy its budget constraint. As we described above, this increase in the expected future money stock will lead to increased inflation today.

While fiat monetary injections are not formally backed, then, the budgetary implications of the assets purchased through a monetary injection can affect their impact. Expected future revenue from assets purchased through monetary expansions can change the expected future expansion of the money supply. Expectations concerning future monetary policy will influence expectations about future interest and inflation rates, and therefore influence long-term nominal rates today.

Monetary or fiscal policy?

Once a central bank purchases assets other than government debt, it is no longer pursuing pure monetary policy. Indeed, some describe such activities as implicit fiscal policy. This is precisely why, as we have demonstrated in this Economic Letter, the assets that are purchased can affect the influence of a monetary injection. Assets bearing default risk purchased at non-market prices have implications for the future consolidated government balance sheet.

In addition, the purchases of risky commercial paper at non-market prices may have distributive implications. If the BOJ purchases assets bearing default risk at non-market prices at face value, it is making a net transfer to the private parties who are selling the paper. In general, these distributive effects do not affect the macroeconomic implications of BOJ purchases, though they do provide additional grounds for evaluating the appropriateness of the policy.

As such, we do not attempt here to label the purchase of risky paper by the BOJ as representing either monetary or fiscal policy, or to characterize the merits of such a policy. We are merely interested in tracing the possible impacts of such a transaction. In doing so, we find that, as in the colonial case, the quality of assets “backing” a monetary expansion can affect its impact. In particular, when the public places a positive probability on monetization of the expected future revenue shortfall associated with the purchase of assets bearing default risk, the expected shortfall should have some additional inflationary impact.

Mark M. Spiegel
Research Officer


Smith, Bruce D. 1985. “Some Colonial Evidence on Two Theories of Money: Maryland and the Carolinas.” Journal of Political Economy, Vol. 93 (6) pp. 1,178–1,211.

Opinions expressed in FRBSF Economic Letter do not necessarily reflect the views of the management of the Federal Reserve Bank of San Francisco or of the Board of Governors of the Federal Reserve System. This publication is edited by Anita Todd and Karen Barnes. Permission to reprint portions of articles or whole articles must be obtained in writing. Please send editorial comments and requests for reprint permission to