I said a few posts back that in order to understand the assumptions behind the efficient capital markets hypothesis (ECMH), it will help us to state the case for that hypothesis with some clarity.
Here we go: when there is a way for someone to make alpha, that way comes about because there is some exploitable inefficiency in the system. For example, XYZ may be listed on stock exchanges in two cities: Philadelphia and Pittsburgh. Due to some inefficiency, XYZ trades at a lower price in Philly than in Pittsburgh. We know that’s inefficiency because the equity of XYZ is worth only what it is worth; there is only one possible complete discounting of all relevant information on that matter. If Pittsburgh and Philadelphia differ, one or both are wrong.
Here we go: when there is a way for someone to make alpha, that way comes about because there is some exploitable inefficiency in the system. For example, XYZ may be listed on stock exchanges in two cities: Philadelphia and Pittsburgh. Due to some inefficiency, XYZ trades at a lower price in Philly than in Pittsburgh. We know that’s inefficiency because the equity of XYZ is worth only what it is worth; there is only one possible complete discounting of all relevant information on that matter. If Pittsburgh and Philadelphia differ, one or both are wrong.
Since listed exchange prices are
themselves public information, arbitrageurs will almost instantly pickup on
this, and will quickly start buying XYZ at the lower price in Philadelphia and
then selling it for a risk-free profit for a higher price in Pittsburgh. [It
would be simpler than, say, buying pumpkins early in the season and re-selling
them in October.] The arb activity itself increases the demand for XYZ in
Philadelphia and sates demand in Pittsburgh. Thus, the trans-Pennsylvanian
trade eliminates the disparity, in the process discovering the intermediate price
that does discount earnings power
properly.
The ECMH, then, is that real-world
prices seldom allow for arbitrage at all, and that when they do, they create a
window that traders close quite quickly.
So: the Assumptions?
So: the Assumptions?
The most important assumptions of
this reasoning are: that markets are liquid (that is, that the arbs can easily buy and sell the assets
involved and realize their profit); that markets are transparent (prices in the
exchanges in both Philadelphia and Pittsburgh are easily obtained), and that
the property rights of participants are protected (a government authority that
purported to cancel trades after they had been settled, restoring title to
earlier parties in the chain of transmission, could wreak havoc). If one or
more of these premises fail, then the mechanisms that make the ECMH plausible
don’t work.
There may be another assumption at
work: that many – perhaps that most – of the participants in the system are
working on the basis of a rational estimation of their own interests. No one
believes it is necessary that everyone who buys or sells be just a bloodless
profit calculating robot. Heck, some people might well be buying for
sentimental reasons. Maybe the CEO of XYZ was my college roommate, and our alma
mater is in Pittsburgh. In such a case, I may sentimentally want to buy XYZ
stock at the Pittsburgh exchange, even at some sacrifice in cost, even knowing
I could get it for a better price in Philadelphia.
ECMH can survive some such “noise”
trades, so long as there aren’t so many of them as to drown the “signal” trades
on which it depends.
Cause and Effect
Cause and Effect
Let’s summarize where we are on the
issue of the Morgan Stanley downgrade of XYZ stock and the immediate drop in
its price:
1)
We have made the philosophical point
that this downgrade is a candidate for a cause of that price move in a straightforward sense of that
word: it is analogous to the action of a human hand in pulling an apple off a
tree;
2)
We have made the case that the
downgrade probably is the cause of
some price movement, inferring this from the substantial sums of money that
gamblers are willing to pay to have this information as early as possible;
3)
And we have wondered whether the
social costs of that gamblers’ den are large: how much misallocation of
resources do such downgrades and speculation about such downgrades create.
Now we can make our final point on
this subject. Though in principle there are such costs, they are small and
fleeting. Any of the markets in which XYZ would have to be listed in order to
make this matter at all are extremely transparent and quite liquid. They are
accordingly subject to lightning-speed arbitrage. Price distortions created by
the gamblers’ den are accordingly akin to a piece of meat thrown into an especially
piranha-ridden stretch of the Amazon River.
Comments
Post a Comment