Tim is mocking Will Hutton and Richard Murphy for having a selective attitude towards the ability of stock markets to predict the economy.
This raises an interesting and often overlooked possibility - that it’s possible for stock markets to be rational and yet very often wrong.
To see how, think of a share not as the discounted present value of future dividends, but instead as a state-contingent security which pays out a lot in a boom and very little in a slump. Its price is therefore the probability-weighted average of different scenarios.
If we think of equities in this sense, then it is quite possible that Will is wrong to believe that the stock markets sharp rise is a portent of a strong economic recovery.
This is because, early last year, the market was pricing in a high probability of a disaster - the possibility that a large chunk of the market (banks) would become worthless, and that a long depression would destroy other stocks. Shares have risen since then because the probability of this disaster is thought to have receded.
This need not tell us that a strong recovery is on the way. It might instead be merely that we are heading for a weak upturn in the economy, but that the market believes that the downside risk to this is smaller than it thought - in particular, that the risk of catastrophe has receded.
You should always think of the future - if you must do so at all - as a probability distribution, not as a single point.
Thinking of equities as state-contingent securities gives us (at least) three implications:
1. The stock market will look like a lousy forecaster. Imagine the probability of a depression were to increase from (say) 5% to 10%. Shares would fall sharply. But it would still be very likely that the economy would continue to grow steadily. With hindsight, it’ll then look as if the market has predicted nine of the last five recessions.
2. Shares will be more volatile, without the market necessarily being irrational. With hindsight, it looks as if share prices were too low in the 1930s and 1970s, and too high in 2000. They look much more volatile than the underlying path of dividends, as Shiller famously pointed out (pdf).
But this needn’t mean the market irrationally over-reacted. It could be that in the 30s and 70s they attached a small but reasonable weight to scenarios (revolution, depression, boom in the 00s) that might have materialized but in fact didn’t. As these weights were revised shares rose in the 40s and 80s and fell in the 00s. Such moves are evidence that the market was wrong - but it doesn’t necessarily mean it was irrational.
3. Share prices will, to a very large extent, be unpredictable. Their moves depend not (just) upon actual economic conditions, but upon the changing weights attached to possible futures which are not in fact subsequently observed. Such changes cannot be anticipated in advance.
This raises an interesting and often overlooked possibility - that it’s possible for stock markets to be rational and yet very often wrong.
To see how, think of a share not as the discounted present value of future dividends, but instead as a state-contingent security which pays out a lot in a boom and very little in a slump. Its price is therefore the probability-weighted average of different scenarios.
If we think of equities in this sense, then it is quite possible that Will is wrong to believe that the stock markets sharp rise is a portent of a strong economic recovery.
This is because, early last year, the market was pricing in a high probability of a disaster - the possibility that a large chunk of the market (banks) would become worthless, and that a long depression would destroy other stocks. Shares have risen since then because the probability of this disaster is thought to have receded.
This need not tell us that a strong recovery is on the way. It might instead be merely that we are heading for a weak upturn in the economy, but that the market believes that the downside risk to this is smaller than it thought - in particular, that the risk of catastrophe has receded.
You should always think of the future - if you must do so at all - as a probability distribution, not as a single point.
Thinking of equities as state-contingent securities gives us (at least) three implications:
1. The stock market will look like a lousy forecaster. Imagine the probability of a depression were to increase from (say) 5% to 10%. Shares would fall sharply. But it would still be very likely that the economy would continue to grow steadily. With hindsight, it’ll then look as if the market has predicted nine of the last five recessions.
2. Shares will be more volatile, without the market necessarily being irrational. With hindsight, it looks as if share prices were too low in the 1930s and 1970s, and too high in 2000. They look much more volatile than the underlying path of dividends, as Shiller famously pointed out (pdf).
But this needn’t mean the market irrationally over-reacted. It could be that in the 30s and 70s they attached a small but reasonable weight to scenarios (revolution, depression, boom in the 00s) that might have materialized but in fact didn’t. As these weights were revised shares rose in the 40s and 80s and fell in the 00s. Such moves are evidence that the market was wrong - but it doesn’t necessarily mean it was irrational.
3. Share prices will, to a very large extent, be unpredictable. Their moves depend not (just) upon actual economic conditions, but upon the changing weights attached to possible futures which are not in fact subsequently observed. Such changes cannot be anticipated in advance.
"Shares have risen since then because the probability of this disaster is thought to have receded."
Hasn't QE had something to do with asset price rises ?
Posted by: Laban | January 04, 2010 at 12:35 PM
Possibly - though I'm sceptical of how far fund managers would use cash raised by selling a safe asset to buy very risky ones.
But even if QE has contributed to rising share prices, it is consistent with my story.
In helping to unfreeze capital markets, QE has helped reduce the risk of disaster.
Posted by: chris | January 04, 2010 at 04:09 PM
I couldn't agree more. I generally dislike "aren't people stupid!" arguments, but it's amazing how often journalists, for instance, fail to understand the basics of probability and what it means for a prediction to be the right prediction. I really think journalists would benefit greatly from a basic grounding in probability and statistics, hence I disagree strongly with Bryan, here:
http://www.bryanappleyard.com/blog/2009/10/smith-gladwell-humphrys-and-empty.php#links
Posted by: Luis Enrique | January 05, 2010 at 12:44 PM
"This raises an interesting and often overlooked possibility - that it’s possible for stock markets to be rational and yet very often wrong."
This does seem somewhat counter-intuitive. What do we mean by "rational" if most predictions are wrong or if you can't get any concrete predictions out of them at all? Does "rational" then just mean, "follows some rules of its own, even if these don't tell us anything about our world"?
Posted by: phil jones | January 09, 2010 at 10:56 AM
Never frown, when you are sad, because you never know who is falling in love with your smile.
Posted by: Ugg london | January 12, 2010 at 12:39 AM
CHI flat iron by Farouk system. Direct from the manufacturer, this genuine Chi ceramic iron comes with valid, one year warranty!
Posted by: chi flat iron | January 18, 2010 at 09:44 AM
Interesting article. I work with this company and often visit their education center: http://education.fisherinvestments.com/ It has a lot of useful information and links to a number of investing books and articles.
Posted by: FI | March 29, 2010 at 07:25 PM
HHH Yes, the design of national policy is important, how our economic development plans for the next five years, how the implementation, how to make our economy even faster. Are designed to advance our focus to invest money in what ways it should be carefully arranged.
Posted by: christian louboutin schaussures | September 29, 2011 at 11:58 PM