Is quantitative investing fatally flawed?

NEW -> Contingent Buyer Assistance Program
profette_IHB

profette_IHB

New member
"Quantitative investing, a widely-used strategy that employs complex computer models to make rapid trading decisions, has taken a blow to its reputation in the fall-out from the credit squeeze. The models failed to cope with sudden and unusual market conditions and banks took heavy losses.

Riccardo Rebonato, global head of market risk and of quantitative research and analysis at the Royal Bank of Scotland, warns of the dangers of taking human decision making out of risk management in his new book, Plight of the Fortune Tellers.

He argues that the current risk management rests on conceptually shaky foundations and that we should restore genuine decision making to our financial planning, using a proper understanding of probability, experimental psychology and decision theory." Here is an interesting Q&A with Rebonato
 
Human beings are herd animals, in part.

They often do not act logically. They are irrationally influenced to follow the herd. I think this is biology, and it was successful biology in the past, or it wouldn't be there in the present.

Human beings are also status seekers. They will sell their futures down the river for a little extra present status. This was also at one time a sucessful biological strategy, especially for males, who may more successfully reproduce perhaps. If he disappears later, and the female he chose to reproduce with succeeds in raising the kid, he has won the biological jackpot. (Not the kid and mom, of course.)

Combine these unrecognized forces (at least unrecognized by the participants) with the stock market, and the ability to buy houses you can't afford and huge gas-eating monsters of cars, without the normal controls which impede this foolishness (underwriters, scoffing mothers-in-law), and you have a true disaster.

I still think we are in for a depression, and my hub still thinks I'm nuts.

Can you mathematically represent these status seeking and herd instincts? I dunno, but what I do know is even if you could, while an event like this is going on, the participants in the event, even perhaps the mathemeticians that devised the model, will pay no attention to the prediction. I think we are victims of our biology, and can stand against it only so long and only so hard.
 
Lawyerliz, you asked: Can you mathematically represent these status seeking and herd instincts?

In a word, yes. Here is a citation for a paper that might be of interest:

Conversation, Information, and Herd Behavior

Robert J. Shiller

The American Economic Review, Vol. 85, No. 2, Papers and Proceedings of the Hundredth and Seventh Annual Meeting of the American Economic Association Washington, DC, January 6-8,1995. (May, 1995), pp. 181-185.
 
From Shiller's "Workshop in Behavioral Finance":

"There are many other psychological theories that have been applied successfully to understanding phenomena in financial markets. Extensive psychological research has documented that people tend to be overconfident in their judgments. People tend to show a wishful thinking bias, believing what they want to believe. People show problems of self control, and know that they may be unable to control themselves in the future. People often make judgments using the representativeness heuristic, that is when judging the probability that a model is true, they base their estimate on the degree to which the data resemble the model, rather than do appropriate probability calculations. People tend to exhibit belief perseverance, hanging onto past beliefs long after they should have abandoned them. People tend to make the error of anchoring, that is, when making difficult quantitative judgments they tend to start from some arbitrary initial estimate, often suggested to them by something in their immediate environment, and then allow that initial estimate to influence their judgments. People tend to show an availability bias, overweighting evidence that comes easily to mind, thereby allowing their decisions to be over-influenced by evidence that is more salient and attention-grabbing".

How true.
 
The same thing happened in '98 during the Asain meltdown.

The same thing happened in '87 with the computerized sell orders.

The problem is market liquidity. When everybody freaks out and stops trading all the models in the world don't help you.

Anybody remember LTCM (Long Term Capital Managment)? They got busto (in great part) to liquidity.
 
Just about every model used is based on the assumption of standard distribution--that returns will follow the classic bell-shaped curve. Everyone from Markowitz to Fama and French build off this assumption. But we know the actual distributions have fat tails. There are, all in all, quite a few more "odd" outliers than what the models allow. They can be good or bad. But frequent outliers increase overall risk.

The manager who accounts for this outlying risk would be more conservative than his colleagues. And in most years he would underperform them. How long would he have a job?

It's very interesting to me that current asset management and valuation methods have not been around all that long. In most hard sciences, the guys that came up with the foundational equations have been dead for a long time. In finance, they're still alive and in the game.
 
Ok, I thought, I asked, I will try to read Shiller. But I found you have to pay to get the paper. Not very much, but some.

Hey Prof Shiller, how's about posting your paper here?

I repeat however, that in the thros of nutso behavior, NOBODY'S immune. Anthropologists at a voodoo ceremony have been known to start participating in the services, because the stimuli evoke a response.
 
Well, I read the first article, and except for a brief mention of eigen values and vectors, it was comprehensible. When I have had time to digest the reading, I will ask the hub what they are.

I have long thought that insurance, which is really what he is talking about, might not be such a good idea, in the long run.

It looks to me that Shiller, being a good person, does not have a deep appreciation of just how evil people can be. (He should practice law for a while.) And how easily the good idea of insurance can turn into a con job, even with the very biggest companies.

Also, most people, for most of history, have existed on the margin of starvation; any loss of resources under those circumstances could mean death. Therefore, it may be that we have evolved to not let go of any resources. Which could be the reason that people don't want to sell for a loss, even tho they get some equity back, and would rather take the risk of keeping the whole thing vs. losing the whole thing. Hmmmm.
 
I think all investing methods are fatally flawed. I am not kidding. I think that as soon as you find success using any particular investing method, model, or whatever, you had better get humble quick and start getting ready to change your method, model, or whatever.
 
More on herd behavior, for Liz:

Abstract:

“We study the relationship between asset prices and herd behavior, which occurs when traders follow the trend in past trades. When traders have private information

on only a single dimension of uncertainty (the effect of a shock to the asset value), price adjustments prevent herd behavior. Herding arises when there are two dimensions

of uncertainty (the existence and effect of a shock), but it need not distort prices because the market discounts the informativeness of trades during herding. With a

third dimension of uncertainty (the quality of traders' information), herd behavior can lend to a significant, short-run mis-pricing.”

Multidimensional Uncertainty and Herd Behavior in Financial Markets

Christopher Avery; Peter Zemsky

The American Economic Review, Vol. 88, No. 4. (Sep., 1998), pp. 724-748.
 
Gosh, Profette, what do you do for a living? Are you Tanta under an alias?

Ok, the more uncertain you are about more things or aspects of things, the more the pricing is likely to be off? The quality of trader's info is never certain. (By private information, is inside info meant? Or, info you honestly ferreted out, or got all insightful about?) Nobody's ever sure of the existence of a shock (the time to buy is now; we are at the bottom!!) At least until the dust settles and everything is over with. And nobody is ever sure of the effect of a shock either. (The housing crisis is contained!) So herding is always with us.

So pricing is always off. Or, really, it's never off; the value of something is what a willing buyer is willing to pay. There are very few trades in housing now; few willing buyers, and fewer qualified buyers. Just because somebody jumped in at a time that it would be hard to make a profit, does that mean that at that time the pricing was wrong? (Absent frauds and cons of course.)

Following a trend might be just following the herd, it could be random, or it could be that the trader followed unique reasoning and came to the same conclusions as everybody else. Stock mkt traders might be different, I take that back, are almost certainly different from people who are buying and selling houses; even speculators are a tad different, I think, tho less different. People who intend to live in the houses that they buy are certainly different. Every time I've bought a house, I've felt that I was jumping aboard a moving freight train going full steam ahead, and that's over a period of from 1970 and until, in Fla, say 2003. The poor suckers who bought after that and just wanted to live in their houses had no experiential basis for thinking they were going to lose their shirts; logic would say that at some point this nonsense has to stop, but most people are not particularly smart (average IQ 100, remember?), and even the smart can be illogical.

Nobody thinks stuff is mispriced if it goes up a lot after you buy, but isn't that being mispriced too, at the time you buy?

So I was, in a sense part of the herd. But I would never have agreed to borrow using the toxic stuff on offer.

Generally I've noticed that it takes about a year and a half, after I conclude that someone's business is hopeless, for said client or acquaintence who is going down the drain to actually go out of business. My broker buddies are hanging on by their fingernails, and plan to keep the ofc open til Febuary, a bit short of my rule of thumb, but business got horrendous so fast that I guess it is foreshortened.

As to the stock mkt, there is some reality there, lots of techies and whatnot making money, but I fail to see how it can last. There is a thread about the Wiley Coyote economy, and that's what I think is happening. So I think the big bust will be any time between now and a year from now, since I started thinking the mkt was totally ridiculous about 6 months ago.

Pricing like everthing else is an ever-changing dance.
 
Ok, the Shiller article I read had a example of human nitwittedness, which I think is something an ubernerd would come up with and sound and is very logical, but may not actually be under the circumstances that most of us evolved in. To wil:

The choice--20% chance of getting $4000. 50% chance of getting $3000. People pick the former. Baaaad math.

But, suppose your clan or tribe needed assets of $4000 (or cowrie shells, or mammoth meat) to survive the winter. With $3000, you don't survive, or if you let a bunch starve, you know darn well the tribe over the hill, which is in much better shape is going to wipe you out come summer. So, what you really have is a 20% chance of surviving, vs a zero chance of surviving.

This, I think, is hard-wired into us. We can think with the math part of our brains, but the hard wiring is still there warning us we'd better max out our assets, even at a very high risk of total failure. The hundreds of generations of this type of hard wired calculation totally trumps the accurate baby math of the choice.

Hence the reluctance to take a loss, and keep 50 grand, while losing 100 grand vs losing the whole house, trying to preserve everything. Status seeking is in there too.
 
You must log in or register to reply here.
Back
Top