Could Leaning Against Asset Price Bubbles Exacerbate the Business Cycle?

In my last post I argued that using monetary policy to lean against asset price bubbles would only tend to hurt the real economy in the long-run if for some reason the central banks target is constructed in an asymmetric way, i.e. it reacts to asset price deviations to the upside but ignores them to the downside. If its asset price target, however defined, is correctly chosen and designed in a similar way as its inflation and/or output targets in e.g. a classic Taylor rule, then on average there will be as much leaning against asset prices as there will be propping up asset price. But what effects would such a policy have on the business cycle in general? More precisely, would a central bank that targets asset prices tend to have a stabilizing or destabilizing effect on the overall system?

The Lucas critique as well as Goodhart’s law, of course, makes such a discussion difficult. Yet given the massive damage that asset price bubbles can have on the economy when they burst, particularly if they are credit-driven, it would seem that anything done to contain these bubbles would be good for overall economic stability. But that is not necessarily the case. How such a central bank policy influences other economic variables, such as inflation and economic growth, depends on the timing of the central banks actions. And the timing of the central banks actions in term depend on the timing of “excesses” in asset prices, both to the upside as well as to the downside. Overall standard economic theory would suggest that e.g. the stock market, but also housing prices, would tend to be fairly strongly correlated with overall economic growth – if the economy is doing good, this should be reflected in financial markets as well. For the most part this notion is confirmed by the data – there is indeed a fairly strong correlation between deviation of RGDP from trend and the growth rate of different financial asset indices, such as housing price indices or the S&P 500.

Particularly problematic cases would of course arise when the real economy is experiencing negative growth rates relative to trend yet positive growth rates in asset markets, i.e. a central bank actively also targeting the latter would be stuck between wanting to ease monetary policy to get the real economy back on track, yet would like to tighten monetary policy in order to avoid further inflating asset prices. A recent speech by Larry Summers made the point that in periods like the current one, where the concept of secular stagnation appears to become relevant again, there is indeed the danger that monetary policy trying to stimulate the economy ends up blowing ever bigger asset market bubbles. Indeed, the argument sometimes goes as far as to suggest that full employment in a secular stagnation environment inevitably comes with a more bubble-prone financial market. Whenever growth rates of anything are above real interest rates, ponzi-like schemes would seem to become more likely. Reaching for yield is also certainly an important phenomenon worth having in mind.

Yet what does the data tell us? As I mentioned, overall, general measures of different asset markets do tend to move in line with the real economy. Thus, in most cases, there is no dilemma the central bank would need to face: it would only have to “lean” against asset prices in times when the economy is doing well anyways, and try to “reflate” asset prices when the economy is not doing quite so well. Thus the effect would be to limit business cycle fluctuations while also holding bubbles in check – everyone wins. However, if we look at yearly growth rates in the National Composite Home Price Index for the United States, and compare them to the deviations from trend in RGDP for the U.S. (trend being 2.5% per quarter during 1990-2012), we find two years in which the economy is growing below trend while the housing market (measured by this index) is doing the opposite – 2001 and 2002. If we do the same for the S&P 500, we find the years 1991, 2007 and 2011 to be potentially troubling – stock prices rose while the economy fell short of trend growth.

How problematic is this? In theory, it’s a pretty big deal. It would indeed seem to support the view that low interest rates to try to revive the real economy would, particularly after economic crises (the recession in the early 1990, the dotcom bubble bursting and the 9/11 attacks, the Great Recession) could potentially be fueling bubbles. However, none of the deviations is really large – in 1991, for instance, RGDP fell by 2.6% while the S&P 500 grew by 4.9% relative to their respective trends. Big movements in asset prices are always in tune with fairly big movements in the real economy – particularly the outliers in the data tend to show a strong positive correlation. And if anything, that asset prices recover sooner than the real economy could simply be explained due to the fact that their prices are less sticky and their adjustment less sluggish. So it certainly is an issue that is worth taking into consideration, but in the majority of cases a central bank actively leaning against asset prices would tend to have a stabilizing, not destabilizing effect.

Advertisements

3 thoughts on “Could Leaning Against Asset Price Bubbles Exacerbate the Business Cycle?

  1. Interesting post.

    One question: when you are saying that the data supports that there is “a fairly strong correlation between deviation of RGDP from trend and the growth rate of different financial asset indices” did you calculate that yourself or did you take that from a paper? And if you calculated it yourself how did you do that?

    Because when I look at both series, what I see is a flat house price index until the late 1990s and then a sharp rise on the one hand, and a GDP growth rate which pretty happily fluctuates around its trend during the whole period.

    • Hmm….well I took the growth rates in the National Composite Home Price Index for the United States and plotted them against the deviation from trend of RGDP growth. Yearly growth rates since I figured, depending on causality, one might lead the other and at times this might cause some issues (e.g., if economic growth picks up, this might not feed through to the housing market until in the next quarter, or something. Causality can, of course, also run the other direction – the direction of which is of vital importance to what a central bank might want to do if it indeed does something. I might write something on that if I get a chance), but over the period of a year it should even out.

      I get a correlation of 0.66 between the two time series, and the scatter plot shows that only the three years I mentioned seem to be problematic. Even from just looking at it it would seem that one could fairly easily draw a simple linear regression line through the whole thing with a decently positive slope.

      My main takeaway is not necessarily that high economic growth always goes hand in hand with equally high growth in housing (or asset) prices (which is not always what what I find), but that it seems reasonably rare that the two growth rates (or their deviations from trend) have opposite signs, particularly in cases where one of them has a high (absolute) growth rate.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s