Does U.S. monetary policy react to asset prices? Implications of stock market bubbles, volatility and productivity

Indian Journal of Economics and Business, Dec, 2007 by Rajeev Sooreea

The striking feature in Figure 1 is the huge bubble that occurred in the late 1990s: real stock prices are clearly above their fundamental values. That stocks were overvalued in the 1990s is well supported by this estimation. The figure also shows the (recessionary) years around the 1980s to be a period where stocks were under-valued. It appears that historically fundamental values and asset prices tend to have a strong correlation moving closely together until the early 1990s where they diverge. However, because of frequent revisions in expectations, asset prices tend to have a volatile component as well. In the 1990s asset prices have departed drastically from fundamental values as Figure 1 shows. The implication of such a misalignment or stock market bubble is that if fundamental values tend to have the same growth path, then it means that asset prices should revert back to it in the coming years, all else constant. The bursting of the bubble is indeed a cause for concern for the Fed as these certainly have implications for price stability.

[FIGURE 1 OMITTED]

Stock Market Volatility: Leverage Effects and the E-GARCH Model

While volatility in part reflects the nature of asset prices (driven primarily by revisions in expectations of future returns), large movements have important implications as they pose a threat to price stability which is the overriding goal of monetary policy. Large and persistent volatilities in stock prices can result in higher market uncertainty which could create macroeconomic imbalances and adversely affect the real economy. To the extent that stock prices signal potential inflationary pressures and investor sentiment or the health of the economy at large, such vulnerabilities may become an issue for monetary policymakers (Chadha, Sarno and Valente, 2004). However, although many studies have recognized that volatility in asset prices is a cause for concern, none of them to our knowledge has integrated a systematic measure of volatility into a monetary policy rule. This section addresses the issue of how to gauge stock market volatility in a systematic way taking into account the asymmetries involved in stock returns variance.

In their study on monetary policy and asset price volatility, Bernanke and Gertler (1999) model volatility as the once-lagged log level of the stock price relative to its steady-state value [log([S.sub.t-1]/S)]. In their empirical model they use the log-differenced change in stock prices to capture this. While this is traditionally used as a proxy for volatility, a more appropriate method of measuring volatility would be the use of Nelson's (1991) Exponential Generalized Autoregressive Conditional Heteroskedastic (E-GARCH) model. The E-GARCH model has the property that it can capture the asymmetries and leverage effects which are typically observed in stock returns volatility.

The E-GARCH model is an extension of the ARCH model introduced by Engle (1982) and generalized as GARCH (Generalized ARCH) by Bollerslev (1986). In the standard GARCH (1,1) specification the mean equation [y.sub.t] given in (12) is written as a function of exogenous variables, [x.sub.t], with an error term, [[epsilon].sub.t], which has conditional mean zero and variance [[sigma].sup.2.sub.t] given by (13)