Creative uses of efficient markets

The “efficient market hypothesis” argues that markets quickly and correctly incorporate all publicly available information into prices. Under the strong version of this theory, the only reason prices of assets like stocks move is because new information becomes available. (The ideas underyling efficient markets are largely associated with the University of Chicago in the 1960s and 1970s.)

Most economists these days would say that there is a lot of truth to the arguments that markets are good at building information into prices, but that there are also other forces that affect prices, like “noise” trading, bubbles, etc.

If markets are efficient, then you can reverse the logic: any change in prices must be due to some new information. This is what an economist would call an “event study.” You look at some unexpected event, measure how much stock prices changed, and assuming that event was the only important piece of information that emerged in a short time window, the change in the share prices tells you how important the event was. For intance, recently there have been some nice academic papers looking at how the stock prices of companies with ties to political officials fluctuate in response to unexpected political events (e.g. Senator Jeffords switching parties, Representative Livingston resigning, Suharto from Indonesia getting sick).

There is an interesting article in the Economist showing how this logic landed a guy named Jamie Olis in jail.

Like all economic tools, however, one needs to be careful in how one applies and interprets “event studies.” After the Sept 11 terror attack five years ago, I was discussing with a prominent economist friend of mine about how important the attacks were. He argued (pretty convincingly) that the attacks were hugely important for the economy — the S&P 500 fell about 10% after the attacks, wiping out about $600 billion in market value. Under the logic of the efficient market hypothesis, that meant the events of Sept 11 had been an economic disaster.

The problem with the argument: three months later the S&P 500 was 4% higher than before the attacks!

There are two possible explanations for this. Maybe lots of good news hit the markets in those three months offsetting the bad news from 9-11. More likely, the price response after 9-11 was an enormous overreaction. The markets got it wrong, and slowly people caught on to this and realized that, at least in economic terms, 9-11 was not that big an event.


Chris Mealy

My take is that on Sept 12 there was more information: not another attack. And on the 13th, 14th, etc. I'm not denying that there was a panic. Or maybe a month latter there was irrational exuberance? And a year later the market was even further down?

Attributing share prices to anything is pure folly. I wish NPR would stop saying stuff like, "The Dow was up 12 points on news that it's Thursday," on the hour. They have to know it's nonsense. They wouldn't report sports like that. "The Giants won on news that it was raining."

striker

Both the 10% drop and the 4% increase in S&P reflected the market's expectation at the corresponding moment. You can't say market's inefficient just because realization deviates from expectation. Realization deviates from expectation exactly because market is efficient and incorporates new information.

sophistry

Another interesting article looking at political assassinations and stock market prices in Israel.

econ.tau.ac.il/summer_workshop/zussman.pdf

Joe Miller

I like to think of the DOW as a middle-aged guy like me--"The DOW rose 120 points today based on a particularly good cup of Starbuck's coffee and a really good night's sleep."

semivoid

Read 'The Economist' this week - the EMH has made it in. Of course why it made the cut has to do with how it is being used; it is apparently the only economic hypothesis that is used in federal sentencing.

Yes, the Duke Energy employee that was sentenced to >20 years in prison had his sentence calculated based on what the EMH would 'show' to have been the markets reaction to news about 'Project Alpha'. Funny thing that his convinction was overturned on that point but it will be tested again - the prosecution has hired an economist (from a consultancy of course not academia) to go for the max. The defense has a former SEC Comissioner working pro bono who notes that on the day Project Alpha's existence was revealed there wasn't a price move. That happened in the next few days - california energy rev clawback proposals, accounting issues, etc occurred then.

Anyway Project Alpha basically smoothed income so that even under EMH there shouldn't be much issue; the future earnings were there.

Anyway it's sad to read about the EMH being used to obstensibly but a 'loss' value used for sentencing that one's actions caused another - especially in stock movements.

Read more...

semivoid

^^^
I am paraphrasing The Economist's article. I am not certain whether it is one of the 'subscriber only' articles that it makes available online.

If it is one of their 'free' articles then read it - I was a little appalled. Otherwise buy a copy or read my rather inadequate review from memory (read it a few days ago).

nphebel

In the benefit of 20/20 hindsight, aren't the markets always "getting it wrong?" Any price move destroys the validity of the previous minute's version of the "truth." I agree with Striker.

dolcevita1972

Or how about this? In Sept of 2001 the Fed floods the market with liquidity to prevent a collapse. That money winds up somewhere, like the Pig (SNP400) Just look at any chart http://www.zealllc.com/c2003/Zeal110703A.gif

jreifler

There is the old Wall Street adage that "the market" abhors uncertainty. It is hard to imagine an event in recent memory that caused more forward-looking uncertainty than 9/11. How does one price the future value of something when the future is "more" unknown than before? If there is a distribution of future possible outcomes, then the immediate reaction to 9/11 would understandly be thinking the distribution is much more skewed to the bad side than before. Hence lower valuations.

It may be that bubbles (like the Internet stocks bubble) is an example of a more uncertain future that makes people see the distribution as skewed to the positive side. Hence higher valuations.

I think the first poster has it right--markets got significant additional information every additional day we didn't suffer another attack. Once again, the future became less uncertain.

John Mueller, in the current Foreign Affairs, asks a very important, and dare I say Freakonomics inspired, question about our expectations when it comes to terrorist attacks--why don't we see more of them? If we are so vulnerable at so many points, and there are so many dedicated terrorists, shouldn't we be experiencing more attacks?

This reminds me of the failed DARPA attempt to create a terrorist attack prediction futures market.

Read more...

AdamRuby

The S&P 500 and the DJIA dropped about %15 and %20 respectively from the middle of May 01 through Sept 10, 2001. This may be the markings of an efficient market in that either A) many people knew about the coming attacks, B) many people knew about the corporate fraud being committed. How can we know what the market is predicting?

tlianza

I have to agree with nphebel and striker. It doesn't make logical sense to say the market "got it wrong" on September 11th by virtue of the fact that the price was different 3 months later. You're using a second value, also determined by the market, to define what "right" is.

The S&P closed .62 points higher today - did the market get it wrong yesterday? Or, more to the point - I bet the S&P will close at a different value tomorrow than it did today... so it's probably wrong today... is it ever "right"?

What is "right"?

jroane

As the posters have stated, the "right price" is only right at the moment the buyer and seller agree. That can change from moment to moment as new information is processed. By definition stocks are "opinions" of the future. One's opinion of the future of the U.S. was much different at 9:30 am on 9/11 then it was at 8:30.

ensembio

Also, note that you can't claim the 9/11 price response was an overreaction just because the S&P was 4% higher a few months later. The market could have gone up for completely different reasons (new labor reports, interest rates, home prices). It's possible that the S&P500 in December would have been even higher if it hadn't been for 9/11.

Bruce Hayden

For an interesting look at how bad information can affect an inefficient market, TCS Daily has an article: Investment Spam Scams Work; What Should We Do About Them?. My suggestion was to sell short shortly after receiving such spam.

Chris Mealy

My take is that on Sept 12 there was more information: not another attack. And on the 13th, 14th, etc. I'm not denying that there was a panic. Or maybe a month latter there was irrational exuberance? And a year later the market was even further down?

Attributing share prices to anything is pure folly. I wish NPR would stop saying stuff like, "The Dow was up 12 points on news that it's Thursday," on the hour. They have to know it's nonsense. They wouldn't report sports like that. "The Giants won on news that it was raining."

striker

Both the 10% drop and the 4% increase in S&P reflected the market's expectation at the corresponding moment. You can't say market's inefficient just because realization deviates from expectation. Realization deviates from expectation exactly because market is efficient and incorporates new information.

sophistry

Another interesting article looking at political assassinations and stock market prices in Israel.

econ.tau.ac.il/summer_workshop/zussman.pdf

Joe Miller

I like to think of the DOW as a middle-aged guy like me--"The DOW rose 120 points today based on a particularly good cup of Starbuck's coffee and a really good night's sleep."

semivoid

Read 'The Economist' this week - the EMH has made it in. Of course why it made the cut has to do with how it is being used; it is apparently the only economic hypothesis that is used in federal sentencing.

Yes, the Duke Energy employee that was sentenced to >20 years in prison had his sentence calculated based on what the EMH would 'show' to have been the markets reaction to news about 'Project Alpha'. Funny thing that his convinction was overturned on that point but it will be tested again - the prosecution has hired an economist (from a consultancy of course not academia) to go for the max. The defense has a former SEC Comissioner working pro bono who notes that on the day Project Alpha's existence was revealed there wasn't a price move. That happened in the next few days - california energy rev clawback proposals, accounting issues, etc occurred then.

Anyway Project Alpha basically smoothed income so that even under EMH there shouldn't be much issue; the future earnings were there.

Anyway it's sad to read about the EMH being used to obstensibly but a 'loss' value used for sentencing that one's actions caused another - especially in stock movements.

Read more...

semivoid

^^^
I am paraphrasing The Economist's article. I am not certain whether it is one of the 'subscriber only' articles that it makes available online.

If it is one of their 'free' articles then read it - I was a little appalled. Otherwise buy a copy or read my rather inadequate review from memory (read it a few days ago).