Good morning. As I write, tropical storm Henri is bearing down on me in eastern Long Island. What this looks like, in practice, is rain. If the newsletter arrives in your inbox at the normal time, it means my power has not gone out. Today’s column is even nerdier than usual, though, so non-nerds might not miss it if it never shows up.
The inelastic markets hypothesis, or, buybacks may be even more important than we thought
Regular readers will know how it pains me to acknowledge the work of rival media outlets, but here is a grudging hat-tip to The Economist, which recently highlighted a really interesting paper about what moves stock prices, by Xavier Gabaix and Ralph Koijen, of Harvard and the University of Chicago. But I think there is more to say about the paper, so here goes.
One thing we can be absolutely sure of with stock markets (because of the work of people like Robert Shiller) is that they overshoot. The fundamental factor used to make sense of the price of stocks (some version of future cash flows) goes up and down some, but stocks go up and down a whole lot more. We explain this, in general, by saying something along the lines of “people are irrational and they get overexcited or unduly pessimistic”.
And surely this is broadly true. But how does it work? Naive people say things like “people are excited, so they are buying stocks and money is flowing into the market”. At which point sophisticated people who write Wall Street newsletters have a self-satisfied laugh, because everyone knows that any time someone buys a stock someone else sells it, so the idea of “money going into stocks” is for dopes.
What Gabaix and Koijen do is make sense of the idea of new money going into the market, and then argue that these flows affect prices — a lot. This is a big deal, given that the standard picture is one in which stock prices reflect only information (or at least beliefs) about fundamental value, not supply and demand.
Gabaix and Koijen’s core insight about flows is that most buyers of stocks are very insensitive to price; that is to say, in the stock market, the price elasticity of demand is very small. The big buyers are institutions working under strict mandates governing their portfolio mix, whether it be 100 per cent equities, or 70/30 stocks and bonds, or what have you. When an investor gives such a fund a fresh $1 to invest in stocks (“money going in”), the fund must put it to work according to its mandate. Fundamentals have little to do with it.
On the classic model, this purchase might drive the price away from fundamental value but (the market being efficient) sellers quickly drive it back. But there is — to exaggerate only slightly — no such seller. There is just a bunch of other funds working under inflexible mandates.
Hedge funds, for example, just don’t own enough stocks to modulate the price impact of these flows (the authors seem to assume, quixotically, that hedge funds are value arbitrageurs, rather than just chasing momentum like everyone else). So flows just force prices up. Gabaix and Koijen argue — using both an economic model and with evidence comparing flows to prices — that a $1 flow can push the value of the market up by three to eight times that much.
The nice thing about this model is that it replaces airy explanations which wave at either “fundamentals” or “animal spirits” with something more concrete:
The mystery of apparently random movements of the stock market, hard to link to fundamentals, is replaced by the more manageable problem of understanding the determinants of flows in inelastic markets . ..
one can replace the “dark matter” of asset pricing (whereby price movements are explained by hard-to-measure latent forces) with tangible flows and the demand shocks of different investors
Of course fundamentals, at certain moments, could affect flows — but mapping out the psychology and sociology of how and when that happens seems more scientific than rattling on about “efficiency”.
One area where the Gabaix and Koijen model has very interesting implications is share buybacks, because they are by far the dominant source of inflows. Here is a (slightly hard to read, sorry) chart from Citigroup, comparing net flows from buybacks to equity mutual funds and ETFs since 2005:
It’s not an exaggeration to say that buybacks are the only flows that amount to much over time. Now, Gabaix and Koijen seem to think — for technical reasons I frankly don’t understand — that the market response to buybacks is less dramatic than for fund flows, but it is still significant.
Most people think that buybacks are good for investors because (a) they increase corporate financial leverage, boosting returns (b) they increase earnings per share growth and (c) they signal management confidence. But it may be that it’s simpler than that: they simply bid up the price of stocks.
Gabaix and Koijen make the point that the demand elasticity of individual stocks can be quite different from that of the very inelastic broad market (we know this because not all companies that do a lot of buybacks see their shares go up, though it sure seems to help). But if you think flows have a direct impact on price, then it probably matters which sectors buy back a ton of stocks. Here is a table of the composition of buybacks in the S&P 500, by sector, from the great Howard Silverblatt of S&P Dow Jones indices. Look at information technology!
Why have tech stocks led the market recently? Maybe it has less to do with growth prospects and more to do with the fact that the sector accounts for something like a third of all buybacks.
I have lots of questions for Gabaix and Koijen. I’ll try to get them on the phone and report back.
Yet one more thought on ageing and inflation
I have now been blathering for a couple of days about Goodhart and Pradhan’s demographic argument for a coming rise in interest rates and inflation. Some of you are now probably bored of the whole thing. But I received a note from my friend Andrew Smithers, economist and frequent correspondent with Unhedged, who made a point that I think is important.
Smithers argues that the biggest risk from the demographic shift that G & P describe — which is, basically, a labour supply shock as the world ages — is that it increases the probability of a very bad central bank policy mistake.
The effect of the inclusion of China and eastern Europe in the global work force thirty or forty years ago was to increase the supply of labour and lower its price per unit. With those changes came lower inflation expectations and a lower non-accelerating inflation rate of unemployment (Nairu), which is the lowest level of employment possible before inflation goes up. The ageing of the world population will have the opposite effect.
In a world with a higher Nairu, employment and participation rates have to be lower if inflation is to be avoided (a rather anti-intuitive impact of a labour supply shock, but there it is). But this is an adjustment that the economy can make without too much trauma. The trauma comes if policymakers don’t take note of what has happened:
If, as seems sadly possible, central banks fail to allow for this, inflationary expectations will pick up and the level of unemployment needed to rein them in again will be even higher than the change in demography alone would cause
The picture is, the Fed doesn’t get the memo about demographic change, so they let inflation run a bit, and inflation expectations dig in. The rate tightening then needed to get expectations back under control will be Paul Volker-like, with a key difference: the stock market was much cheaper back then, and the debt level was much lower, so it will be uglier this time.
One good read
Will Fed tapering of bond purchases drive yields up or down. Will it hurt or help stocks? James Mackintosh walks through the possibilities and uncertainties, and there are plenty, here.