Monday, September 17, 2007

The Tradeoff between Output Price Stability and Asset Price Stability

One area of interest to me is the distinction between malign and benign deflation. As I have noted before, malign deflation is the result of negative shocks to aggregate demand and is what most observers invoke when discussing deflation. Benign deflation, on the other hand, is less well known and arises from positive shocks to aggregate supply (AS) not accommodated by monetary policy. Now positive AS shocks, such as those coming from rapid productivity growth, generally mean a higher neutral interest rate and faster trend growth for the economy. If the monetary authority, however, wishes to stabilize the price level in the face of such positive aggregate supply shocks it will have to lower its policy rate below the neutral interest rate. In turn, this response by monetary policy lowers the cost of capital relative to the return on capital. More leverage now makes sense and sets the stage for a potential credit boom. Throw in some "irrational exuberance" to this mix and you have an asset price boom party.

Now had monetary authorities allowed the benign deflation to emerge, the policy rate and the neural rate would be better aligned and there would be no credit boom. Asset prices, in turn, would better track fundamentals.

This story I just told implies that there is a trade off between stabilizing output prices and asset prices when AS is growing. Is there any evidence for this claim? Below are several scatterplots that plot real S&P stock price growth rate against the CPI growth rate for the period 1871 to the present. The data are in monthly frequency and come from Robert Shiller's database. There are several graphs based on different growth rate measures for each variable: a 2-year growth rate, a 3-year growth rate, and a 5-year growth rate. Here is the scatterplot using 2-year growth rates:


Not much going on here, so now take a look at the scatterplot using 3-year growth rates:



The results are getting more interesting as there is almost a Phillips curve-like form emerging amidst the noise. To make better sense of this graph, recall that the output price-asset price trade off is conditional on there being positive economic growth. The above graph, however, includes all data points, including those that occurred during during economic downturns. To clean up this potential noise, I created another graph where I excluded all points where output prices and stock prices are both negative (i.e. eliminate points from quadrant 3). The assumption here is that if both variables are negative there must be economic weakness. After this adjustment, the trade off curve is even more apparent.

Now, take a look at the scatterplot using 5-year growth rates (without any adjustments):

Again, an output price-asset price trade off is suggested by the figure. To whiten some of the noise in the scatterplot, I once again remove points from quadrant 3. Now the scatterplot looks as follows:

These figures all indicate there is a potential trade off between stabilizing output prices and asset prices. While further refinements are needed in these figures, they add perspective to what I have been arguing in my blog: the Fed in its attempt to prevent benign deflationary pressures from materializing over the past few years has been fueling financial imbalances, particularly asset prices. In short, the Fed has been exploiting the trade offs implicit in the figures above. Now, I need to coin a clever name for the trade off curves above. Any suggestions?

Update

The figures above are illuminating, but I do not want to oversell them. While I am still convinced the Fed was trading asset price stability for output price stability over the past few years, I am eager to find more solid evidence than that presented above.

Below is another figure I created using the aggregate asset price indices from the BIS. The BIS aggregate asset price indices consist of real estate and stocks so they are a broader measure than a stock index alone. In this figure I plot the yoy % change in the real asset prices against the yoy % change in the price level for the years 1971-2005. Like the graphs above, this figure is suggestive too of a trade off between output price stability and asset price stability.

No comments:

Post a Comment