Please respond to lbo-talk at lists.panix.com
To: lbo-talk at lists.panix.com cc: (bcc: DANIEL DAVIES) bcc: DANIEL DAVIES Subject: EMH
Date: Fri, 18 Feb 2000 10:04:05 -0500 From: Enrique Diaz-Alvarez <enrique at anise.ee.cornell.edu> Organization: Cornell University X-Mailer: Mozilla 4.7 [en] (X11; U; HP-UX B.10.20 9000/735) X-Accept-Language: en MIME-Version: 1.0 To: lbo-talk at lists.panix.com Subject: EMH a hypothesis? References: <80256889.002BCCB5.00 at notesjanus.flemings.com> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit
DANIEL.DAVIES at flemings.com wrote:
>>
>>
>> Efficient market theory isn't really a "theory" in this sense -- it's
an
>> empirical claim about stock markets. There are numerous different and
>> incompatible theories of the underlying "microstructure", most of which
are
>> at least consistent with "the efficient markets hypothesis".
>>
>> dd
>>
>>
>Dumb question:
As usual, it most certainly isn't. Though it may get a dumb answer anyway because my profession is rapidly rotting my brain -- I'm forgetting all the academic finance theory I ever knew.
>EMH says that "prices in an efficient market will reflect all
>available information". But for this to be a testable hypothesis, instead
of a
>tautology, there needs to be a way to determine the "price which reflects
all
>available information" that is independent of market prices, right? Then
we
>could go ahead and *test* the theory, by looking at the prices in a
largely
>efficient market, such as the US, and checking them with our
>indepedently-determined "pricves reflecting all availabel information".
Sort of. Actually, the situation isn't quite as dire as that. You can look at big pieces of information, such as changes in the dividend, profits warnings, etc, etc, and see how quickly the price moves. If Amazon announced that one month from now it would wind itself up, and the price stayed the same, then that would *disprove* the EMH. And you can look at the difference between otherwise identical transactions when in one case they are informative and in the other case not. For example, bank stock prices react to equity issues which are opportunistic decisions by management, but not to equity issues forced on them by regulators, because in the second case, the equity issue imparts no new information to the market, while in the first case it tells the market that the management think the stock is overvalued. So it's not quite as desperate a situation as it might seem.
But for *global* confirmation or otherwise of the EMH, you'd need something like the counterfactual model you describe. So, the academics try to test weaker versions of the theory. We have:
* the "weak form" EMH. This is the theory that all past information about stock prices is reflected in today's price; that the history of the price is irrelevant to predicting its future. Also known as the "technical analysis doesn't work" version, or the "random walk hypothesis". You can test this by using the standard statistical tests of whether a given time series is a geometric Brownian motion or not (I'm trying to put this into more engineering type terms 'cos there's more hard scientists than finance guys on the list). Most of the academic evidence is that weak-form EMH is true -- prices are a random walk. On the other hand, there are no really good tests for a random walk, so these results have to be taken with a pinch of salt. This version of the EMH would have the implication that prices never "overshoot" in the short term, which isn't well supported. Robert Shiller's argument that stock market prices are too volatile to be efficient (cited on Doug's site, IIRC) would also be an argument against weak form EMH.
(disclaimer: One of my best mates is the UK's top-rated technical analyst, and has more than a slight interest in weak form EMH being false).
* the "semi-strong form" EMH. This is the theory that all information "publically available to a well-informed investor" is already reflected in the price, more colloquially known as the "Mutual Funds Always Underperform in the Long Term" thesis. There is a lot of evidence that mutual funds underperform in the long term, but there are lots of alternative explanations (mainly to do with the sociology and psychology of the fund management profession). The limited evidence on hedge fund performance, and on the trading reccomendations of analysts like me suggests that we may be able to systematically outperform.
* the "strong form" EMH, which suggests that absolutely 100% all of the information is reflected. This would, AFAICS, require the sort of counterfactual that Enrique suggests. What finance academics do is to assume that the actual realised results were the truth, and then to ask whether there was any publicly available information beforehand that would predict them. This is a *hellishly* arbitrary assumption - I've got a .pdf file by Edwin Elton about how bad it is that you can have if you like. But it's the only game in town.
Note that the strong form implies both the weaker forms, so any empirical refutation of, say, fund manager underperformance, or any proof of overshooting would be game over for market efficiency.
>Absent such a method for determining prices, isn't EMH about as meaningful
as
>saying that in a parallel universe, I am married to Tyra Banks?
heh. Yes. More exactly, it's like saying "Since you're not married to Tyra Banks, she must not like you".
cheers
dd
_______________________
----------------------------------------------------------
___________________________________________________________________________
_____
---------------------------------------------------------
This email is confidential to the ordinary user of the
e-mail address to which it was addressed. If you are not
the intended recipient, please notify the sender
immediately on (44) 171 638 5858 and delete the message
from all locations in your computer. You should not copy
this email or use it for any purpose, or disclose its
contents to any person : to do so may be unlawful.
Email is an informal method of communication and is
subject to possible data corruption, either accidentally
or on purpose. Flemings is unable to exercise control
over the content of information contained in
transmissions made via the Internet. For these reasons
it will normally be inappropriate to rely on information
contained on email without obtaining written confirmation
of it.
----------------------------------------------------------