Inflation Expectations and Asset Prices

There’s been a bit of interesting discussion this week surrounding the new paper by David Glasner. (See e.g. Scott Sumner, Kevin Drum, Paul Krugman.) In it, Glasner provides solid econometric evidence of the strong correlation between inflation expectations and asset prices as measured by the S&P 500 - but only since 2008.

The heart of the paper is, as has been noted by others, a picture like this, which while not actually in Glasner’s paper nicely illustrates its central point:


Note that inflation expectations are estimated as the difference in yields between 10 year US government bonds and their inflation-adjusted counterparts.

The point is that there was a statistically significant - one might say profound - structural break in the relationship between inflation expectations and asset prices, somewhere in 2008. Before that time, inflation expectations and asset prices had little or no correlation with each other; after that time the correlation was nearly perfect.

Some commenters have cited this as further proof that the US economy has been characterized by an insufficiency of demand for the past three years. Kevin Drum asks for a better understanding of the specific reasoning behind that inference, however. Kevin is a very smart guy, and generally understand economics better than most people. He writes:

Sadly, neither Glasner, Sumner, nor Krugman explain in terms someone like me can understand why this correlation implies that aggregate demand is what's behind our economic woes. I feel a bit like a dummy, since they seem to expect this to be obvious, but hopefully someone out there in the econ blogosphere will take pity and explain this in laymen's terms.

Kevin, I think that you’re smarter than you give yourself credit for in this case. The reason it doesn’t seem obvious to you that this is evidence that the US economy is facing a shortfall of demand is because it’s not. Not by itself, anyway.

Think of it like this. Normally, when stock market investors think that inflation is rising, there are two conflicting reactions. First comes a jump for joy, because rising prices mean rising profits for the companies represented in the NYSE. But next comes a depressed reality check, as stock market investors realize that inflation rising means that the Fed is likely to raise interest rates, which tells them that maybe stocks are not going to be the best way to get a good rate of return in the future. These two reactions roughly cancel each other out (sometimes one is bigger, sometimes the other), resulting in no net correlation between inflation expectations and the stock market.

Now, however, things are different. The reason is because the Fed has already pushed interest rates (at least short term rates) as low as they can go. In fact, the Fed would probably have pushed them below zero if that was possible. In other words, the Fed’s target, or optimal interest rate is probably negative.

So in this world where we’re up against the zero lower bound for nominal interest rates, stock market investors only have the first reaction – the jump for joy – when they think that inflation is rising. They don’t have to endure the depressed reality check, because they’re not at all worried about the Fed raising interest rates any time soon. If anything, it just means that the Fed’s optimal interest rate is less negative. But the nominal interest rate, stock market investors reason, is still going to be zero. Since there’s no countervailing effect, the jump for joy rules the day and the stock market goes up with inflation expectations.

The implication is that Glasner’s evidence says something about asset price behavior and the relationship between real and nominal interest rates when we’re up against the zero lower bound on nominal interest rates (which are the conclusions he draws in his paper), but doesn’t really tell us anything more than that. In other words, the strong correlation between inflation expectations and asset prices starting in 2008 is fully explained by the fact that 2008 is when we hit the zero lower bound on nominal interest rates. No other explanation for this correlation is needed. The chart below illustrates.

It's not a coincidence that as soon as the green line (representing nominal short term interest rates, as measured by the 4 week Treasury bill rate) hits zero, the correlation between inflation expectations and asset prices is almost perfect. Starting in 2008 stock market investors no longer had to endure that depressed reality check whenever inflation expectations rose.

Of course, I know that the US economy has been facing insufficient demand for the past 3 years, as surely as I know that the earth revolves around the sun. There is a host of other evidence confirming that (very obvious) assessment of the situation. But I don’t think that this post-2008 correlation between inflation expectations and asset prices is part of that evidence.