Bell Curves, Black Swans And Sell Discipline

by: Bob Wilson


Investors soon learn that corrections can be sudden and large, whereas the market climbs its wall of worry more slowly. I'll call that asymmetry No. 1: trading can be panicky on the downside whereas euphoria is slow to develop.

Economists say that people generally act to minimize their expense and maximize their income. Behavioral economists say that this old mantra about the "invisible hand" has lots of exceptions because people's emotions or values interfere with "rational" choices. Let's call that asymmetry No. 2: our choices can be financially "irrational".

Then we have the random walk model, where there is no overall pattern. One assumption is that buyers and sellers all have access to the same public information, so this contributes to the randomness. But we know that some people do better research or have a smart, rich uncle, etc. So call this asymmetry No. 3: not all of us, in fact, have the same information.

Now along comes a Wall St. quant. He produces a frequency histogram of market returns (bell curve) and says the risk is defined as one standard deviation because that's how he calculates volatility. OK, fine. But a bell curve shows how often various returns have occurred, not the market behavior that produced them. That's what those asymmetries allude to: actual market behavior. And therein lies the bite of the poisoned apple the bell curve invites you to take.

It's all too easy to convince ourselves that past results suggest at least similar future results despite the risks we are accepting. We know that the usual mutual fund disclaimer is true, for the downside as well as the upside. The odds of avoiding serious damage to our portfolio are no better just because of our past success in avoiding that damage.

I'm not overly concerned about a nice symmetrical bell curve and the volatility ("risk") it reflects. To me, my risk is the probability that my portfolio will get its butt kicked. And if I'm honest with myself, I'll admit that my portfolio's greatest risk is me. A Black Swan is highly unlikely but could have disastrous consequences. I'm the one who thinks I can defend against this.

In the fall of 2008 we saw lots of volatility started by a meltdown in subprime mortgage securitizations. Gretchen Morgenson, Pulitzer Prize winning columnist, wrote that, ".....traders have been blindsided by a correlation between bonds and stocks that they never expected would occur." That correlation wasn't a tail event. It wasn't even on their radar screen because their algorithms didn't account for that possibility, "bell curve blind" we might say.

In that article she quoted Matthew S. Rothman, writing in a Lehman Bros. research report: ".....the models (ours included) are behaving in the opposite way we would predict and have tested for over...45 plus years...our risk models are miscalibrated for the current market....This appears to be an event with little precedent....". It was a costly miscalibration. We know what happened to Lehman Bros. in 2008.

Long Term Capital Management LP, a hedge fund, was initially funded with $7 billion in 1994. The partners included two Nobel laureates and several successful big money bond traders. There were two dozen Ph.D.s on the staff, primarily in economics and mathematics. Lots of brains, only one problem.

In 1995 LTCM made a gross profit of 42.8% and then 40.8% in 1996. In 1998 Russia defaulted on its sovereign debt, touching off plunges in various derivatives tied to that market. Because LTCM was being guided "impartially" by econometric models that had proven their value in previous "normal" market climates, management's strategies did not allow for catastrophic drops in prices. LTCM was trading on unregulated margin as high as 50:1. The fund imploded and was bailed out by several large investment banks, in consultation with the Fed, closing its doors in 2009.

John Meriwether, one of LTCM's founders, after losing $150 million of his own money, launched a new hedge fund in 1999 and later said, "Our whole focus is on the extremes now--what's the worse that can happen to you....because we never want to go through that again." (see The (Mis)behavior Of Markets, B. Mandelbrot and R. Hudson, Basic Books, 2004, p.107. Ibid. pp.105-107 is my source for the information about LTCM).

One of the most succinct summaries on this topic I've found comes from the legendary investor Irving Kahn, now nearly 108: "Value investing will almost always be right....The economy, the market, they don't care about your feelings. Buy the out-of-favor, the unpopular. Nobody can predict the market....Think value. Think downside risk. Think total return with dividends....(With today's) extraordinarily low rates, be careful with fixed income....You hope you're in a recovery, but you don't know....Protect yourself."

Protect yourself, indeed.

OK, here's how to protect yourself.

PART II -- Sell Discipline

"Buy and Hold" has been a market mantra for some time. But between the tech meltdown in 2001-02 and the subprime meltdown of 2008-09 many investors still haven't recovered losses. It's usually preferable to hold most positions through market corrections of 10% or so. But if your portfolio starts showing larger losses you might need to take action and this means having a sell discipline.

Until those two bubbles, a policy of buy and hold had been smart for a long time. It still might be now. After the market dropped 22.4% on October 19, 1987 it didn't take long to recover. But the collapses of those two bubbles were far worse. There were tech stocks going from 80 to 8 in a few weeks or less. Then the entire market plunged starting in the fall of 2008, with huge swings.

There are two purposes for a sell discipline. One is to prevent an ordinary loss from becoming a disaster. The other is when you can at least take some profits while you still have them. In the first case your sell discipline preserves most of your capital. In the second it helps preserve profits.

The simplest form of sell discipline is based on percentage declines. Rather than try to interpret various macro indicators, I've often felt it's better to let market behavior guide your decisions. Here's a guideline that I've used and suggested to others. It can be modified to suit your holdings, your risk tolerance, etc.

% Down % Of Position Liqidated
13%-17% 15%-25%
23%-27% 35%-50%
33%-37% 60%-75%
>40% you're out

OK, there are some glitches with this scheme.

First, volatility varies with the stock. Where the chart shows low volatility you can hold those stocks to stricter loss limits. If a stock is more volatile you can give it more slack, just not a lot more.

Second, your faith in a business might give you more confidence than the stock's falling price reflects. If so, don't hesitate to hold on through rough weather if you feel sure the company will solve its problems. But you'll really need to be right about this.

Third, sometimes, just after you sell, the darn stock will start back up. But this is the risk you agree with yourself to take in exchange for knowing that if the stock continues to fall you could be out with more than half your capital left. Selling doesn't rule out buying back later on if the stock looks good to you again.

Just because a company seems to be a pillar of the market doesn't necessarily mean you just sit on it. In 1999 IBM went from $140 to $90 and in 2002 it went from $125 to $55. Investors considered "Big Blue" to be a core holding to be kept through thick and thin. But that's a lot of thin.

The idea is to be raising more cash the longer a stock or the market falls. You'll never know where the bottom is, but a sell discipline will provide dry powder to buy into cheaper prices, bottom or not.

Have you seen those slow motion commercials where a car with a dummy behind the wheel collides with a wall? Sell discipline advises that as soon as you see your bumper crumple, get ready because it might not be a fender bender. If a stock hits your downside price tolerance, hit the eject button.

A sell discipline has limited applicability to Black Swans. With stocks you can defend against a Black Swan with GTC sell-stops ("stop losses"). But a stock that's in free fall can dive right through your sell-stop price to give you an execution a lot lower than you expected, so this is not foolproof.

Another choice would be to buy puts. This is gutsy because your time premium on any long option's price will time-erode. You can renew sell-stops for free, of course.

With mutual funds, if the market looks bad at breakfast and you punch in a sell order, you could sit there all day while your fund's NAV continues to dive into a large loss. This is a good reason to have at least some of your equity portfolio in stocks and/or ETFs.

There are other forms of sell discipline based on metrics, charts and other factors, but I'd recommend using percentage loss if you haven't used one before. Another nice feature is that sell discipline tends to take emotion out of your trading. When your money is running out the door in a bear market that's a pretty good defense to have.

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.