David Glasner, an economist with the Federal Trade Commission, in a new Mercatus Research Paper, offers what he calls a “deeper explanation of the financial crisis of 2008 and the subsequent recovery.” Deeper than what? Deeper than attributing the bust to the bursting of the housing price bubble.
Glasner’s new look turns on the “Fisher effect,” an old theory that holds that the real interest rate is equal to the nominal interest rate minus the expected rate of inflation, so that real interest rates will fall as inflation increases, or as expectations for inflation increase, in the absence of an offsetting increase in the nominal rate.
The other side of this is that an expected deflation will raise the expected future value of cash over that of other assets, inducing the holders of real assets to sell them. This, Glasner suggests, may have been what happened in the period 2016-18.
Nasty Feedback Loop
Indeed, if the inflation expectations of the time were endogenous to the financial system’s search for equilibrium, then a negative shock reducing these expectations could have triggered a nasty feedback loop in which deflationary expectations cause a decline in asset prices, which in turn feed a lowering of inflationary expectations, etc.
In such a situation, Glasner writes, an “exogenous commitment to stabilizing asset prices may be an essential condition for restoring asset-market equilibrium.”
The Fisher relation, on one reading, is this:
I = 0 ? r + pe .
Here the i represents nominal interest rates, r the real rate, and pe the expected rate of inflation. This way of putting the relation makes explicit the lower bound, that is, the presumption that the nominal rate will not get into the negative numbers. Further, this way of putting Fisher’s point is “unconventional,” Glasner says, in that “the brunt of an adjustment to an expected change in inflation … from the nominal rate of interest to the real rate.”
In explaining all this, Glasner clarifies that there is no lower bound of zero for the real interest rate. The bound is for nominal interest. If the “real interest rate is negative,” Glasner says in a footnote, “then the condition for avoiding a reverse Fisher effect is that the expected inflation rate exceeds the real interest rate. In other words, if the real rate is 2 percent, inflation must exceed 2 percent to avoid asset market disequilibrium and a flight from real assets into money--a crash in asset prices.”
The Empirical Study
To study such matters empirically it is necessary that there be an objective market measure of inflation expectations. Fortunately for the theory, there is. Inferences about such expectations “can be extracted from the difference between yields on TIPS and conventional Treasury securities of a corresponding duration.” Both short- and medium-term inflation expectations as so measured fell dramatically in 2008, actually turning negative, so “the interaction of inflation expectations with asset prices over time can now be observed.”
The conclusion of this study’s empirical review confirms the theoretical intuition drawn from Fisher’s work. As nominal interest approaches the zero boundary, a decline in expected inflation can trigger a decline in asset prices.
The economy fell into recession at the end of 2007. After that, the correlation between the daily change in the Standard & Poor’s 500 on the one hand, and interest rates on the other (real or nominal) became strongly positive. The strong positive correlation continued “with few exceptions” until the end of 2016. The correlation was positive when the asset price movement had headed upward, just as it was positive on the way down.
There are plenty of implications of this work for alpha hunters. Perhaps most obviously, the work will be of interest to those willing and able to make directional bets on broadly defined classes of assets, including stock price indexes. It might encourage them to bet on a crash on S&P 500 prices when a sharp drop is the arithmetically tidiest way to keep the Fisher relationship intact. Bet on arithmetic.
Making this point, Glasner explains that he is taking the Fisher equation as an equilibrium condition, not as a tautology. It is never a prediction to say that A will continue to equal A, just as one is not predicting the outcome of a football game when one says that victory will go to the team that scores the most points. Equilibrium conditions, on the other hand (which are more akin to the notion that victory generally goes to the team with a sound offensive line) can inspire informed trades.