AIN'T MISBEHAVING
by Bill Bonner
The financial world was undressed recently. A new book by Benoit Mandelbrot and Richard Hudson, The (Mis)behavior of
Markets: A Fractal View of Risk, Ruin and Reward, revealed
the naked and revolting truth: Modern investment theories
are "nonsense."
Mandelbrot is a mathematician at Yale who has popularized
the idea of fractals - elaborate and unpredictable natural patterns, such as frost on a windowpane.
Fractal patterns have been thought to describe patterns of market prices. As far as we know, no one has succeeded in developing a successful trading model based on fractal mathematics (of course, if we had, we might not tell anyone, either.)
Several Nobel laureates owe their prizes, their prestige
and their incomes to what is known as the Efficient Market Hypothesis (EMH), or simply the Random Walk. No idea about finance that was ever flushed out of the frontal lobe ever enjoyed greater acceptance. None ever received more recognition. And none ever sucked so much money out of the lumpen. But the Efficient Market Hypothesis is really nothing more than an elegant subterfuge, based on propositions everyone knows are false.
The basic idea is that the market is smarter than any
investor. No matter how hard you try, you can't beat the market... because at any given moment all that is known about a share price is reflected in the share price itself.
The price - no matter how absurd - is thought to be
"perfect." No better price can be imagined. Yet a different
price tomorrow is almost a certainty. Each day, prices
move. One day gives you a price that is perfect. The next
day may give you a price only half as much, but that too is
perfect. Mr. Market is always right; he just changes his
mind.
Since an investor cannot have a better idea of what a share
is worth than the market itself, he cannot do better than
buy at the market price, whatever it is. The actual
performance will be random. He might as well throw a dart
at pages of The Wall Street Journal. Or buy the index. He
will get the same return. Anything else would call into question the whole idea - unless it were purely luck, purely random... in no way connected to conscious effort on
the part of the analyst or stock-picker.
A corollary of the EMH was the idea that prices and rates
of return were distributed evenly, according to what is
known as the bell curve. Another way of saying this is that
prices will usually be normal... or average... and that
they will only be abnormal in a normal way... that is, abnormally high or low such as might be predicted by a standard bell curve distribution.
This was the idea that led Long-Term Capital Management,
for instance, to invest billions of dollars on the
proposition that if prices fell in a range outside of the
norm, you could calculate the odds of how likely prices
were to come back. LTCM had 25 Ph.D.s on the payroll. John Meriwether put them to work calculating standard deviations
on various debt instruments - among them, Russian bonds. In
1995, the fund made 42% on its money. In 1996, the rate of return was 40%. But in 1998, Russia defaulted on its bonds.
But the bell curve, or the standard distribution, applied
to finance is "nonsense," says Mandelbrot. It may be
nonsense applied to many other naturally occurring
phenomena, too. The seas in Holland, for example, were
thought to rise 3.83 meters above "normal" only once every 10,000 years. Yet they rose that high in 1953. Records showed that they hit that level back in 1570 too. There was
something wrong with the calculation. Standard deviation measures didn't seem to work. The world was a riskier place
than people thought.
It came as a shock to Myron Scholes and Robert Merton. The
two Nobel Prize winners at LTCM had staked their careers, reputations and money on the perfectly logical idea that bond prices followed a bell curve pattern... and that if they reached the kind of extremes LTCM was seeing in 1998, it was a not a defect in their logic, but a huge buying opportunity. The odds were vanishingly small that prices would go further out of whack, they thought. Finance was a science, after all, and science gives predictable results. Every time.
And yet bond prices went further out of whack. Traders at
other companies knew what was happening. They had studied
the same theories. They guessed that LTCM had loaded up
with risky bonds... they also guessed that the company
tottered on a high wall of debt. They did just what you'd
expect: They gave it a push. Prices fell even further.
The old-timers knew the Random Walk theory was nonsense.
There is always a lot of random noise in markets - as in
life itself. But there are patterns, too. The trouble is,
the patterns - like fractals - are variable and largely unpredictable. Markets go up and down. The movements are not only mostly unforeseeable, but perverse; they move in tandem with the broad strokes of human sentiment. As people
become more comfortable, more sure of themselves and the future... more confident and expansive... they tend to bid up prices. A house that was worth only $100,000 when they expected nothing but shelter from it rises to twice as much
when they expect it to finance their retirement.
But investors only expect such a result after a long
experience of pleasure returns. That is, they expect asset prices to rise after they have already risen, not before. The mood of the crowd reflects what has just happened, not what will happen next.
House prices, like everything else, cannot go up forever.
Who would be able to afford them? Instead, they go up and down... and perversely, go down when they are most expected
to go up... and up when a bull market is least expected.
This is, of course, not an insight based on Efficient
Market Hypothesis, but one that springs from experience...
and intuition. No one ever won a Nobel Prize for saying so,
but it never made anyone go broke.
Mandelbrot points out that when you actually look at market
prices, you find patterns that differ greatly from those predicted by mainstream financial theories. "Outlier events" - those that are supposed to happen only once in a blue moon - actually happen all the time.
Citigroup looked at daily changes in the yen/dollar
exchange rate, for example. It found that moves of five standard deviations were not uncommon, even though they should only happen, according to the theory, about once every 100 years. They even found one "heart-stopping" move,
10.7 standard deviations from the norm. What were the odds
of that? Even if Citigroup had been trading dollars and yen
every day for the last 15 billion years - from the Big Bang
to the day after yesterday - such an event shouldn't have happened a single time. But it did.
And look at what happened to the Dow Jones on Oct. 19,
1987. It fell 29.2% - an event that registered 22 on the standard deviation scale.
Not only do the markets produce events that should never happen, they also act in other ways that intuition might predict, but existing theories cannot fathom. According to EMH, prices - like dice - have no memory. One day's perfect
price is thought to be completely independent of
yesterday's perfect price. This is what mathematicians call
"independence" and contrarian investors call "more
nonsense."
It is nonsense because human beings use prices as points of
reference. A man who owns a stock worth $50 one day is reluctant to believe it is worth only $5 the next. He has no idea of what it is worth except what the market tells him; yesterday, it told him it was worth $50... and it's a shock to discover that it is worth only a tenth as much.
This, or something we can't explain, causes the phenomenon described by the old saying, When it rains, it pours. Periods of intense volatility cluster together; bad news piles up in what is known as "positive short-term serial correlation." Markets don't really just bounce around randomly, in other words, but there are periods when things
go well and periods when they don't. There are bull markets
and bear markets, as any old-timer's intuition could have
told you.
Investing isn't rocket science. In fact, it isn't any kind
of science. Instead, it is a human study, like poetry or
prize fighting. And as in all the human studies, the
problems one confronts are not bounded engineering problems
but unbounded, infinitely complex ironies. If you do the calculations correctly, you can fire off an artillery shell
and it goes where you expect. But market prices often go in
the exact opposite direction, no matter how well you do the
numbers. It is easier to send a spaceship to the moon, in
other words, than it is to win an argument with your spouse
or figure out which way stocks are headed. And if you are
going to follow advice, an old-timer's intuition is
probably a better guide than modern portfolio theory.
The markets ain't misbehaving, in other words: They're just
doing what they always do.
Regards,
Bill Bonner
The Daily Reckoning |
