Another crash not far off?
A Crash Course in Market Dislocations
Now I would like to take some time to discuss probabilities/guesstimates about the likelihood of a stock-market crash. When I say "crash," people usually think of the October 29, 1929 or October 19, 1987 stock market crashes, i.e., a one-day event, but I don't necessarily mean just that. My working hypothesis has always been that a market dislocation needn't necessarily happen in one day. It could happen over a series of days. I am referring to a dislocation or, to borrow a term from the commodity markets, a "market re-pricing." I prefer the term "dislocation," since it's less likely to have people immediately think of the 1929 or 1987 crashes.
If you wanted to try to estimate how rare a dislocation ought to be, you can. If you looked at the S&P 500 monthly data since 1973 (I only use monthly data here to make the calculations a little simpler), you'd find that a month where the market went down, say, 14%, like it did in August of 1998, was approximately a 3 standard deviation move. A month where the market dropped as it did in October 1987 would be about a 4.5 standard deviation move.
To put those numbers in perspective, a 3 standard deviation move is roughly 1 out of 740 observations, a 4 standard deviation is 1 in 31,600, while 4.5 standard deviation is a staggering 1 out of 294,000. The minutiae of the math that could get us to different numbers totally misses the main point that crash-like moves are very rare.
Peril, on the Perimeter of a Punchbowl
I bring this up because I believe that the risks are now far higher than "normal." This is the point I tried to make last weekend on "Cavuto on Business." However, to say that the chance of a crash-like event is far higher than normal doesn't really mean much if you don't know how likely it is in the first place.
When pressed, I told Neil I thought the chance might be as high as 1 in 5. Well, compared to the statistics I've given you, 1 in 5 is a gargantuan number. Now I don't know if 1 in 5 is the right number. One in 10 would still be a giant number, as would 1 in 20, relative to "normal." I have no idea what the true probabilities are -- because they are not knowable. Everyone who looks at the reasons for a dislocation, which I'll get to in a moment, will arrive at different conclusions.
However, what I think is not debatable is that the possibilities are far higher than most people would think (if they ever even thought about it). Folks need to factor the possibilities into their investment strategies, or think through the possibilities and dismiss them, should they come to that conclusion. With that preface aside, let me move on to what I think are the ingredients necessary to cause a market dislocation.
Speculation: Yeast for a Crash Cake
First of all, you need to have a fair amount of speculation, in order to extend a market in a large enough way to make it "crash-able." We obviously had that in 1987. Once you say that we had enough speculation in 1987, and you compare 1987 to the current period, it's clear to me that we have enough speculation now, and then some. If we just look at the P/E of the S&P 500, it's about the same now as then.
However, the earnings quality is arguably far lower and less stable (i.e., earnings are potentially overstated today, thanks to compensation in the form of options not being expensed). In addition, I could list a whole host of speculative ideas that today sport multi-billion-dollar market caps, whereas back then, by my observation, speculative ideas were still measured in the multi-millions.
Structures Conducive to Dislocation
Then, more important than speculation, and the instability caused by it, you need a structure that is crash-able. That was really the key in 1987. We had a structure called portfolio insurance that caused people who might ordinarily be inclined to sell not to sell. As long as the market didn't drop by a certain pre-established amount, they all just sat on their hands. When the market started to drop by the pre-described amount, say 3% to 8%, they all sold (via their money manager) at the same time, and that's how you had everybody pounding the exits all at once. That was the key feature of 1987. We experienced speculation, but the structure was more important than the speculation.
We can't know if today's structure is more crash-able than 1987's or not. We know we've got 8,000 hedge funds, plus or minus, most of which have hair-trigger fingers. We know we've got 7,000 or 8,000 mutual funds, most of which will allow redemptions on a phone call -- which to me makes the structure more crash-able. Back in 1987, we did not have as much mutual-fund money in the market. A lot of money was run by dinosaur types like I used to be, managing individual accounts, versus today, where so much money is being run as OPM (other people's money), which causes higher levels of risk-taking.
Importantly, I also think that due to the market declines of the last couple years (before the recent 15-month rally), we've established a predisposition for people to cut their losses if the market heads down again -- something that wasn't present during the 2000-2003 market decline. People now know that stocks can go down, and go down a long way, so the precondition potentially exists for them to exercise their right to the 1-800-GET-ME-OUT phone call.
Derivatives in the Dislocation Equation
Lastly, in the structure department, we have trillions of dollars of derivatives, about which we have no knowledge concerning how they might work in a market meltdown. As a subset to that, we also have dynamic hedging on Wall Street, which might (again, unknowable) make the market more crash-able. So, I would say we have more speculation now, but we had a structure in 1987 that was almost guaranteed to precipitate a crash. Whether the structure now is powerful enough to trigger a dislocation or not is perhaps debatable.
Finally, you need a catalyst. Here I think the catalyst is far more powerful now than it was in 1987. I was running money in 1987, and due to the climate I have described, I believed a crash was possible, which was a pretty radical thought back then. At that time, we had an inflation problem building, a belief that the Fed was behind the curve, and the dollar that was under pressure. The fact that the dollar appeared to be breaking in an uncontrolled manner to the downside in October 1987 was the catalyst that allowed the speculation and the structure to create the implosion we saw.
Next Time Down, Minus the Mighty Fed
Today, we not only have a Fed that is behind the curve, but far worse, the Fed is trapped and unable to ease to rescue the market. It's a variation of the theme I talked about recently in my New York speech -- that the next time down in the economy or the stock market, folks will realize that the Fed can't save them. If folks realize that the Fed can't save the day, that the stock market and economy are "on their own," and potentially heading south, that could easily foment panic.
The other potential catalyst now, as in 1987: The dollar is (potentially) coming unstuck, and foreigners could pressure the dollar, or in other ways get folks sensitized to the macro problems that exist, such that they'd want to sell stocks, at roughly the same time. Most people do not realize that the decline in the dollar over the last two years has been bigger than the drop leading up to October 19, 1987.
My gut feeling -- though there is no way for me to quantify it -- is that the chances of a crash at some point in the next six months to a year are far higher now than in 1987. One subjective reason is that I just don't think it's possible for all the thousands of hedge funds and hundreds of thousands or millions of people who think they're talented enough to outwit the stock market -- and who believe they can play this game of speculating in an overvalued, dangerous stock market -- to get out whole.
My belief in the perversity of markets leads me to conclude that after 10 years-plus of folks being able to get away with whatever they wanted, Mr. Market might just be ready to take some folks' money. What comes to mind are put sellers, who -- euphemistically -- win 99 times in a row, but on the hundredth time, they get wiped out. That's sort of my vision of the stock market eventually taking back a lot of the paper wealth that's been created.
Dismissing Dislocation, at One's Own Risk
In summary, I would like to say that just because the risks may be higher than "normal" does not mean that I think a crash-like event will happen. To believe that something would happen, you'd have to believe that the probabilities were at least over 50%, and probably bordering on 75% or 80%. And obviously, just because the probabilities of an event may be higher than normal also does not mean that an event will happen. As we go forward and get more data, each of us can decide whether the probabilities are getting higher or lower. But for people to conclude that it's business as usual, i.e., a dislocation is an extraordinarily low-probability event, is dangerous.
Bill Fleckenstein |
