Mitigating Bubble Risk

We’re working on a book called ‘The Right Answers to the Wrong Questions’, but the book has been put aside because the research behind it  took us in an unexpected direction and that’s one that allowed us, as it turned out, to actually see early warning signs of bubbles developing. There are some entertaining aspects to this but the serious implication is for investment profit.

We’ve gone through at least three bubbles in the past decade, including the real estate bubble, but policy makers say they can’t spot a bubble.

That’s the interesting thing. The reason this research got pushed ahead a little bit is because we’re talking with central banks, among them the Bank of Canada. They’re now being pretty much forced into a role of trying to do something in anticipation of bubbles, whether they like it or not. Actually, I think even (former U.S. Federal Reserve Board Chairman Alan) Greenspan admitted eventually that they did know when bubbles were developing — and many people agree.

Our initial research was of no use whatsoever to central bankers, even though it’s of tremendous use to asset managers and traders because you could see the problems in bank stocks and financial indices worldwide and also in equity indices — before the current crash — easily in late 2006 and early 2007. Now of course, the regulators wanted to get in ahead of Northern Rock collapsing. They wanted to know that well in advance of 2007, even though the markets didn’t go over the cliff  until late 2008.

What’s an early warning depends on the context. If you’re a portfolio manager, it’s longer than if you’re a trader and if you’re  a central banker, it’s longer than if you’re a portfolio manager.

What metrics do you use to determine a bubble forming?

Many economists have put a lot of effort into trying to understand modelling of, among other things, bubbles. There was a lot of work put into it after the tech trouble, but if you take the simplest and most widely accepted economic model, bubbles don’t even exist. So that’s one of the issues.In a sense it’s not entirely fair to the accuse the economists of not having done this because over an awful lot of the last century, it was extremely difficult to use the data that was available. People had to transcribe from the newspaper clippings what the prices were if they wanted an historical record to go back at all. So now we have enormous amounts of data, easily available, not all of these are as clean as you might like them to be, but we’re now in a position to do a lot of the observational work that any scientific approach to the subject would have started with and in our case we have some mathematical tools that are extremely good at analyzing statistical extremes. That’s exactly the sort of thing  that strikes you as markets collapse.

What early bubble warning signs do you look for?

At the end game of the bubble, everyone knows after the fact that it was the buyers that were taking the most risk. For an extended period, even though the market was going up, it was riskier to be long than to be short. That’s sort of the opposite of what you ought to see. If things really are all different this time and everything’s going to go up nicely forever, it’s the short-sellers and the profit-takers who are leaving money on the table. They’re the people who are more exposed to risk than the people who are buying and holding. But that’s not what the statistics told us. That statistics told us that over long periods the relative risks are saying the market should be correcting, not continuing to go up.

How do markets correct?

You don’t need to be able to predict when a wave is going to form. You just need to know you’re on it before it’s too late. You don’t need to be able to anticipate when prices are going up. But if you want to buy low you need to notice things going up and see signals that it’s going to continue to go up and you need to be able to get out before it crashes down on the other side. That’s a much lower bar to get over than coming up with a model of prices that would lead economists to explain why bubbles occur and tell you when they are going to happen and when they’re going to burst. All of that would be a wonderful thing but it’s quite unnecessary if all you need is an early warning system.

Do your early warning signals give any indication of the likely duration of a bubble?

The big ones turn out to be bubbles and they will collapse catastrophically, like the ones we all know about. What we’ve identified is a misalignment of risk. We call it an unstable expansion. The market is going up but in spite of that it’s riskier to be a buyer than a seller. Typically they go for at least six months, but then they very often correct, without a market collapse. They simply go back to where they were.

If we say a potential bubble is forming here, say in October 2009 in the TSX Composite Index, what we do is take the level that that index was at and imagine if we had invested our money at the current bank rate at that point. So now the measure is the market above or below that. The intuition, the rationale behind that is whether the buy and hold investor ends up worse off in the markets than if he just put his money into the bank. Anybody that buys above this potential level when a bubble is opening up can compare where they are to where they would have been with a bank.

Unfortunately there’s no way of saying how long it can continue. It’s something we’ve observed. In the most liquid equity market in the world, the U.S., these things don’t persist for very long – anymore. They did back in 1929.

To learn more about the Risk Management Conference, please visit the events section of the website. If you are interested in attending the event, please email Garth Thomas to be considered, as limited space available.