Jeremy Grantham's quarterly letter, published yesterday, is a classic. It's also a two-parter, so we'll get the pleasure of another half in two weeks. In the meantime, there's plenty to chew on in this half. We'll carve it into a few posts this weekend.
First: As we've noted before, the three great stock-market bubbles of the 20th Century--US in 1929, US in 1965, and Japan in 1989--were all followed by price troughs that were 50% below fair value. (See charts below right). The bubble that peaked last fall was every bit as spectacular as these earlier bubbles, so it seems reasonable to expect that stock prices might trough 50% below fair value this time, too.
Jeremy puts fair value on the S&P 500 at about 975 (vs. Friday's close of 940). A trough of 50% below fair value, therefore, would be about 500, or some 45%+ below today's levels.
The good news: US stocks are finally below fair value for the first time in two decades, so Jeremy finally feels comfortable buying them. Also, three previous bubbles do not, in Jeremy's opinion, constitute enough of a data sample to bank on. So he's hoping, for the sake of all of us, that it's different this time and the market won't fall 50% below fair value (while keeping in mind that this is a distinct possibility):
We have had some confidence in saying that by October 10th global equities were cheap on an absolute basis and cheaper than at any time in 20 years. Full disclosure requires that we add that, in our opinion, this is not as brilliant as it sounds, for markets have been more or less permanently overpriced since 1994 and have not been very cheap since 1982-83 and perhaps a few weeks in 1987.
There is also a terrible caveat (isn’t there always?), and that is presented in Exhibit 3, which shows the three most important equity bubbles of the 20th Century: 1929, 1965, and Japan in 1989. You will notice that all three overcorrected around their price trends by more than 50%!
In the interest of general happiness, we do not trot out these exhibits often and, until recently, they would have been seen as totally irrelevant and perhaps indecent. But, after all, it’s just
history. Being optimistic like most humans, we draw the line at believing something so dire will happen this time.
We can hide behind the fact that there are only three data points, and therefore no self-respecting statistician can give them much weight. We can convince ourselves that things are different this time since the background to each of the four events, including this one, is different.
One of them had high inflation; three, including the current situation, did not. Japan and 1929 were characterized by complete incompetence, while this time we had only – shall we say – very widespread incompetence. This time we have thrown ourselves more quickly into battle, although not so quickly as some would have liked.
Not all of the differences are favorable: we have a more global, interlocking, and complicated system, including non-bank players like hedge funds. We also have the “financial weapons of mass destruction” – asset-backed securities that are tiered and sliced and repackaged – and, perhaps most destabilizing of all, totally unregulated credit default swaps. Did we have even more greed and short-term orientation this time than they did? Well, we certainly didn’t have less!
Still, a 50% overrun seems unacceptable. Probably governments would feel that the consequences of such a loss in asset value would simply be too awful and would do anything to prevent it. And perhaps, just perhaps, their “anything” would work. But a reasonably conservative investor looking at the data would want to allow for at least a 20% overrun to, say, 800 on the S&P 500, and have a tiny portion of their brain loaded with the notion that it just might be quite a bit worse.
So what the hell to do? Start buying, cautiously, saving some powder in case we get the same complete collapse that has followed other bubbles of this magnitude.
We at GMO have a strong value bias, and our curse, therefore, like all value managers, is being too early. In 1998 we saw horribly overpriced stocks that at 21 times earnings equaled the two previous great bubbles of 1929 and 1965. Seeing this new “peak,” we were sellers far, far too early, only to watch it go to 35 times earnings! And as it went up, so many of our clients went with it, reminding us that career risk is really the only other thing that matters.
The other side of the coin is that only sleepy value managers buy brilliantly cheap stocks: industrious, wide-awake value managers buy them when they are merely very nicely cheap, and suffer badly when they become – as they sometimes do – spectacularly cheap.
I said as far back as 1999, while suffering from selling too soon, that my next big mistake would be buying too soon. This probably sounded ridiculous for someone who was regarded as a perma bear, but I meant it. With 14 years of an overpriced S&P, one feels like a perma bear
just as I felt like a perma bull at the end of 13 years of underpriced markets from 1973-86. But that was long ago.
Well, surprisingly, here we are again. Finally! On October 10th we can say that, with the S&P at 900, stocks are cheap in the U.S. and cheaper still overseas. We will therefore be steady buyers at these prices. Not necessarily rapid buyers, in fact probably not, but steady buyers. But
we have no illusions. Timing is diffi cult and is apparently not usually our skill set, although we got desperately and atypically lucky moving rapidly to underweight in emerging equities three months ago.
That aside, we play the numbers. And we recognize the real possibilities of severe and typical overruns. We also recognize that the current crisis comes with possibly unique dangers of a 0.3
global meltdown. We recognize, in short, that we are very probably buying too soon. Caveat emptor.
