Asset price inflation and monetary policy.

Introduction

Several recent papers have addressed the question of whether central banks should respond to changes in asset prices. Bernanke and Gertler [1999] emphasized inflation-targeting as the primary responsibility of monetary authorities whom they enjoined to respond only if asset price changes signaled changes in expected inflation. On the other hand, Cecchetti, Genberg et al. [2000] urged the authorities to react to asset price inflation as much as to goods and services price inflation in formulating monetary policy decisions. Asset price inflation in their view was a predictor of core inflation.

In two follow-up papers these authors essentially repeat their initial positions. Cecchetti, Genberg, and Wadhwani [2002] respond to their critics and find no reason to alter their recommendation to monetary authorities. Likewise, Bernanke and Gertler [2001] reiterate their judgment that monetary authorities should not respond to asset price inflation. For monetary authorities to be able to target inflation assumes that they can forecast it. A paper by Stock and Watson [2000] raises a question about their ability to do so. For G-7 countries these authors find no indicator that reliably predicts future rates of inflation, so casting doubt on the recommendation by Cecchetti et al. that reliable signals of inflation can be extracted from asset prices.

Asset price inflation, of course, can take many forms, raising prices of art objects, land, housing purchases, and equities. Cecchetti et al. conclude that housing inflation should be given a larger weight than equity prices in a measure of core inflation to which authorities should respond by adjusting the interest rate that they use as their instrument. Goodhart [2001] endorses this conclusion.

The foregoing papers restrict their consideration of asset price inflation to its relationship to achieving the target of inflation that most central banks currently regard as their mandate. Asset price inflation has other dimensions, however, that should not be neglected. It is relevant to ask whether monetary policy contributes to asset price inflation. It is also relevant to ask whether asset price inflation affects the portfolios of financial institutions in ways that spell trouble for them should the asset price inflation collapse. Monitoring financial institutions may be a responsibility of regulatory authorities rather than monetary authorities, but monetary policy decisions inevitably are influenced by a worsening condition of financial institutions. An example occurred in some financial crises of the 1990s, when monetary authorities hesitated to institute contractionary policies that they otherwise would have adopted. They feared that higher interest rates would prove catastrophic for already weaken ed banks and therefore refrained from raising them.

Monetary authorities need to be alert to policies they pursue that may promote asset price inflation. Even if they cannot be tagged with responsibility for asset price inflation, when it occurs, they should be alert to changes in portfolios of financial institutions that asset price inflation induces. In the past, as we shall see below, they have taken steps to end the asset price boom, since that was the source of the change in portfolios. An alternative response to deal with the problem would be to control the portfolio effects without directly confronting asset prices if monetary policy is free of involvement. In that event, restraint on the portfolio effects might serve indirectly to dampen or even obviate the asset price boom. If there are no portfolio effects, whether an asset price boom should be a concern of the authorities becomes a debatable question.

The motive for focusing on portfolio effects is not to rescue financial institution investors from the consequences of management's shortsighted risk-laden lending decisions. What is of concern is that taxpayer funds should not be used to ball out these institutions when their balance sheets reveal that liabilities exceed assets. If the financial institution is a bank, funds of the deposit insurance agency may be drawn upon in a rescue. The temper of the times is such that failure is regarded as politically unacceptable, so taking action that precludes a failure is held to be warranted. The ultimate effect of forbearance, however, may inflict larger costs than when a timely response by supervisors prevents institutional weakness.

There may be fewer pitfalls for the authorities in trying to determine whether asset price inflation is damaging from paying attention to changes in the composition of financial institutions' portfolios than from trying to gauge the effects of asset prices on core inflation. Assuming that monetary policy has not generated the bubble, a response by the authorities to limit the weakening of financial balance sheets as a result of asset price inflation may be a more effective solution with less collateral damage to the economy than the solution of raising interest rates to puncture a bubble. Preventing a deterioration in the quality of financial institution balance sheets has the further advantage that the need does not arise, in the aftermath of a fall in value of asset collateral backing loans, to clean up portfolios. Even if financial institutions emerged unscathed in the aftermath of an asset price boom, they might still be undermined should the authorities pursue flawed policies, as happened after 1930.

This paper investigates the condition of financial institutions when asset prices are escalating and when asset prices crash. It examines the role of monetary policy, if any, in accounting for the upswing in asset prices as well as for the downswing. Monetary policy may have had no responsibility for the upswing and yet be implicated in the unwinding of the succeeding asset price debacle.

I propose to examine the behavior of monetary and regulatory authorities with respect to the performance of financial institutions, first, during two major episodes of a sustained rise in equity prices in the twentieth century --1926-29 in the United States and 1985-89 in Japan--and, second, during the subsequent sustained fall in equity prices --1929-33 in the United States and 1989-02 in Japan. Asset prices other than those of equities also were involved during some of these episodes, and their effects on financial institutions are examined. I next review the spectacular annual growth in U.S. equity prices in 1995-2000 to check for financial institution involvement, and note the subsequent decline in asset prices. A conclusion follows. I begin with some history to learn how financial institutions fared during the upswing and then the downswing in equity prices and the reaction if any of the authorities.

Equity Price Inflation

For each episode, I review (a) the extent of the rise in asset prices; (b) the accompanying change in portfolios of financial institutions; (c) the response of authorities; (d) alternatives to that response.

The United States in the 1920s The Upswing

Prices of equities advanced through most of the months from the end of 1924 until the peak in September 1929, but the most spectacular gains occurred from 1926 on. The Dow-Jones industrial annual average in each of these years was 167, 202, 300, with a peak of 381 in September 1929. The corresponding Standard & Poor composite 500 stock price averages were 12.59, 15.34, 19.95, with a peak of 31.30 in September 1929.

Whether these price increases were justified by prospective earnings growth is still in dispute. White [1990, p. 78] believes that qualitative evidence suggests "the existence of conditions that enhanced the likelihood of a bubble," although econometric tests for a speculative bubble are inconclusive. The equity upsurge was not matched by commodity prices, which showed no tendency to rise. Wholesale prices on a 1926 base fell to 95.3 in 1929.

Financial Institution Portfolios

What is indisputable is the nature of Federal Reserve Board concern that the stock market advances aroused. Banks extended loans to investors that their security purchases collateralized. No better description is available of the Federal Reserve's revulsion with such a practice than the following extract [Willis and Chapman 1934, p. 621]:

"The banks liked the security loan because it seemed like a conveniently liquid and therefore safe asset. They did not realize that in the 1927-29 period they were thus directly aiding and abetting the common-stock boom and infecting the whole structure, that the securities were only liquid as long as speculators were willing and able to support the market for them at dizzy heights; that the enormous issues of new securities amounting to fifty billions in five years, and stock prices of 200 times earnings represented a national gambling mania; and that in thus furthering and directly stimulating industrial fluctuations and distorting the price structure, they were conducting themselves in a manner directly contrary to rudimentary banking principles."

The reference to banking principles was to the real bills doctrine, which distinguishes between productive and speculative use of credit. Credit restricted to productive uses financed additions to output, hence was non-inflationary. Credit to finance acquisition of common stocks was speculative and, since it did not increase output, was inflationary.

The absence of commodity price inflation while speculative use of credit was growing did not disturb belief in the real bills doctrine. Monetary authorities as well as regulators and examiners of banks were well informed about the change in bank portfolios as a result of the increase in loans on securities. (1)

In June 1920, loans on securities were 32 percent of the loan portfolio of weekly reporting member banks in 101 leading cities. By June 1926, the figure was 40 percent, and by June 1929, 44 percent. Loans to brokers and dealers, despite the attention the Board lavished on them, constituted only 29 percent of loans on securities in that month. White [1990, p. 74] asks how credit to buy stocks could have been easy in 1928-29 when credit in general was tight because Federal Reserve policy was contractionary. In fact, interest rates on brokers' loans increased sharply, to levels much higher than the discount rate and commercial paper rate. It was not an increase in the aggregate supply of bank credit during the boom, but a reallocation in favor of loans on securities that supported rising stock prices.