Is the model wrong, or is reality wrong?

September 3, 2009

From Carl Richards, as seen on

“Maybe, just maybe something is wrong with the model. Maybe there is nothing standard at all about risk. Maybe there is nothing normal about returns. Maybe risk is not like a little dial we can control on clients’ portfolios like we tune in a radio…Or maybe, just maybe, reality is wrong.”

This is a snippet from Mr. Richards’ short piece talking about what he’s learned since Lehman collapsed. (That last line is sarcasm, if you didn’t catch that.)

I wonder how many asset managers really think of modern portfolio theory as immutable. Most have to know it’s flawed. I suspect many adhere to it (and other imperfect models) because they have no choice. They have assets to invest and MPT is a convenient guidepost, however imperfect it may be.

As violent as the September thru February period was, at this point, many have moved past it. It was a big downturn, sure, but when the system was rescued, so were the troublesome theories and operating methods that led to the crisis in the first place. The Great Depression was so long and so deep, that folks couldn’t help but take certain lessons to heart … for instance, that debt is dangerous.

For much of Wall St., it’s back to business as usual.


We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see

I found the post interesting.

Posted by Kirsten | Report as abusive

All the fundamentals changed before, during and after 9/11. That moment in time showed up the treachery in and amongst the financial, arms, energy, commodity, drug and slave trades. Property investment to me a secure one, interest bearing obviously a joke, shares easy to predict ex strike price to free cash flow multiple ratios, options a good one, but I think the derivative collapse scared people off from the latter.

‘It’s back to business as usual’, small investor versus large portfolio algorithms.

Posted by Casper Lab | Report as abusive

I was reading another blog (don’t worry, your my favorite) that had a great discussion about ?Walter? Bagehot and the banking system. There was a line from Bagehot about how people take it for granted that you put money in a bank and get 5% (yeah, for the youngsters, you used to get 5% from “passbook” saving at a bank) as if it was equivalent to the sun rising in the east. We’re going on a century of money being merely paper – and I actually think that has served us well…up to now. Just as conquering scarcity of food has now led to obesity, the generation of ever more credit may have reached its limit. Interesting times.

Posted by fresno dan | Report as abusive

People **need** random inputs to operate properly, but Monte Carlo methods don’t work real good if the casino is rigged. That’s why we got so wrapped around the axle about high-frequency trading (HFT) this summer.

I’ve been revisiting our transcript of AEI’s amazing March 28, 2007 seminar on the subprime crisis. Alex Pollock has a couple of great quotes from Bagehot’s (pronounced “badge-it”) classic 1873 book “Lombard Street”, and there’s an etext of the whole thing here. adm10.txt

That modern portfolio theory (MPT) model must be depending on a theory of risk where black swans and other nasty animals show up in a truly random way. Just looking at the gyres in yesterday’s NYSE indexes makes it obvious that there are non-random hidden hands (another acronym, the PPT, comes to mind) at work there.

Posted by John McLeod | Report as abusive

The casino is only rigged if you allow it to be rigged.

In line with the present economic ‘weirdness’, follow this link for an article that popped up in the South African Press this morning: ntent.aspx?id=80483

Now that’s weird !

Posted by Casper Lab | Report as abusive

I think it was Albert Einstein that first clued me in to the idea that it takes a generation (20 years) for a major new idea to be fully absorbed into common usage and understanding. Roughly ten years since the Long Term CM crash, so it hasn’t fully sunk in that implied volatility might have some empirical problems. If Black-S has conceptual flaws, what do people really have to work with?

MPT (and EMH) were true advances, they were just not final answers, and anyone who ever studied intellectual history finds that as no surprise. The financial industry that exists today, could not have worked with the pre-Markowitz/Sharpe methods. The world is, on the whole, a better place with some moderately accurate method of valuing options. But there is no return without risk. Even the RiskFreeRate, as we may soon learn, is not really a RFR. And if there are no free lunches, there are also no bets certain on the future.

Posted by ARJTurgot | Report as abusive