Shiller Chicanery

Here’s why I’m not too worried about this. The link will take you to a chart (I’m speaking here about figure 1.1 – there are three, selectable through tabs on the bottom) on Robert Shiller‘s homepage. It purports to show a disturbing widening in the gap in the Price-to-Earnings-Ratio of stocks and the actual earnings of the companies they represent.

I think this data is spurious. The scary bits in the chart for me are regions like that in the 1960s and 1970s, where the P/E ratio grows in contrast to a relatively steady state for corporate incomes. Periods like those are a pretty stark demonstration that stock prices were doing a bad job tracking corporate incomes. By contrast, periods like the one in the 2000s don’t bother me so much. True, the P/E ratio is sky-high compared to corporate earnings – dramatically so. But it’s exactly this kind of thing that a statistical charlatan exploits to sell books. When you consider what the P/E ratio represents, it’s pretty easy to convince yourself that what it actually is matters not a whit, so long as it generally trends in the same direction as corporate profits.

The P/E ratio is simplicity itself. All you do is take the stock price (P) and divide it by the company’s earnings (E). The P is easy to calculate – you simply look up the stock’s value on the NYSE index (or whatever market it’s traded on). E is a bit harder. For that, you take the dollar amount of the company’s earnings and divide it by the number of shares (a “share” is, after all, just what it sounds like: your own private piece of the company). But no matter – if you own the stock, it’s usually easy to get that number too, and P/E ratios typically posted somewhere public to save you the trouble.

Now, since this is a simple fraction, it’s easy to puzzle out what it would mean if company earnings drop, but the P/E ratio for a stock rises. Since the P – the price of the stock – is in the numerator, and the E – the company earnings – is in the denominator, then for the P/E ratio to rise while E falls it would have to be that at best corporate earnings are falling faster than the price of the stock, but more likely that the price of the stock either holds flat or even rises as company earnings fall. Same thing for the situation where corporate earnings rise but the P/E ratio falls. For that to happen, it means that E is growing faster than P – which is to say, corporate profits are either (a) growing faster than the price of the stock, (b) growing while the price of the stock holds steady or (c) growing while the price of the stock falls. In either case, it means that the price of the stock is probably not tracking corporate earnings – which is to say, probably not reflecting economic reality. THAT’s the sort of thing that we should worry about – when the stock market is tracking essentially nothing but the machinations of traders.

By contrast – consider the situation where P/E is growing faster than E, or falling faster than E, but in the same direction. That simply means that stock prices are more dramatic than corporate earnings (they rise or fall faster than earnings) maybe, but they are still basically tracking them. It’s impossible for the P/E ratio to rise while corporate earnings rise if P is falling, and likewise impossible for the P/E ratio to fall while E is falling if P is rising.

I won’t say this is exactly what we want. In the utopian world where the Efficient Market Hypothesis were transparently true without qualification, then I suppose the P/E ratio would be less volatile than E itself, and that P/E would stay close to 1. But that’s a world of total transparency in both reporting and prediction. I really do think the future holds something like that, but probably not in my lifetime. Right now, the world economy remains uncertain enough that what tomorrow holds is still guesswork, to a large degree. That means that P/E is both volatile and a long way from unity. But that’s OK with me so long as it generally tracks the real economy.

What Shiller’s graph says to me, actually, is that the stock market is doing a better job than ever tracking the real economy. Because even though it’s true that the P/E ratio has a wider gap with E than ever before, notice how closely it tracks E. Starting in the late 90s, it’s actually just a REALLY EXAGGERATED proxy for E. It rises when E rises, falls when E falls, which is to say that E is doing the real heavy lifting of explaining where stocks prices go. P rises faster than E and falls faster than E, but always in the same direction. In other words, the market is working, and the direction stock prices take in general gives us a good idea where the market is headed. Better still – P/E isn’t just tracking E, it’s correctly predicting it. P/E seems to rise just before E does, and fall just before E falls. It’s a good demonstration of the wisdom of crowds. Prices are volatile, so that means that individual traders have no clue which direction the market is headed. However, traders as a whole seem to be getting it more right than ever. Which is to say, as a group they’re doing their jobs.

Why is the gap between P/E and E so stark all of a sudden, then? No clue – but here’s my stab at it. I would say that it coincides with a massive expansion in stock speculation. That is, back in the 70s there weren’t so many traders; these days, nearly everyone with a job is a trader. The amount of fodder traders have to play with is much larger, so of course there’s some drama. It’s sort of like the difference between shooting someone with a .45 and shooting them with an AK. Dead is dead either way, but it’s a messier kill the more firepower you have. Traders are playing with larger volumes of money than ever (what with the existence of 401Ks and such), but if Shiller’s chart is to be believed, this has only made them more precise in their predictions recently.

Now – one fear that probably still needs calming is the idea that with stock prices in general being higher than ever, isn’t a lot of that money “not real?” Well, no – all it means is that there’s a higher barrier to entry into the market than before. A good way to think of it is to imagine a poker game. You can’t play without first paying for your chips. Likewise – in the stock market, you can’t play without first paying for your stocks. Once you start playing, you’re mostly trading chips for chips (stocks for stocks), but the total amount of money in the stock market can’t grow without someone first playing in. Granted that there’s a caveat here – some of these purchases may be made with borrowed money, and if these traders go bust and can’t repay their loans then yes, some of the volatility they caused was based on money that turns out not to be real. Dispelling this worry takes a longer post – but the general idea is that so long as they never become the majority of traders we’re probably OK, the noise comes out in the wash (some of them, after all, make back what they borrowed and pay back their creditors). That caveat aside, it’s just like a poker game. When you take your chips out, all you’re claiming is money that’s already been deposited. And if there’s more money total on the table than there was when you started, then that’s because others have been buying in with more (real) money to stave off their losses. Either that, or more players have joined the game since you did. (In fact, I think this analogy probably works for the widening of the gap in P/E and E too. The more total money there is at the table in a poker game, the bigger the bets – and especially pots – tend to be. The underlying reality of how you win and lose hasn’t changed, but the stakes have is all.)

So I’m not too impressed with Shiller’s scary graph. I think as long as P/E tracks E, then E is doing the real explaining, and that means the market is working. P/E holding relatively flat would be even better. What’s REALLY scary is when P/E goes in the opposite direction as E. Then the stock market is only tracking its own wankery and has effectively ceased to function as an economic indicator. It shouldn’t surprise anyone that this happened at the height of Socialism in this country in the late 1960s/early 1970s. “We” (meaning the Johnson and Nixon Administrations) adopted an anti-market policy, and the market therefore stopped telling us much that was useful, just as Ludwig von Mises warned it would.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>