Prediction, Prophecy, and Gambling

November 04

I read a headline this week which summarised a 17 page report from Morgan Stanley with a single line which said that a 20% fall in prices was possible.

As misrepresentations go, the headline was fairly impressive.

Firstly, it was a comment not on the predictions that Morgan Stanley had made, but on the predictions that the IMF had made, which Morgan Stanley used as an example, and then said they did not agree with.

Secondly, the IMF report did not actually say that prices would fall either - instead it said that it is very hard to predict what house prices would do, but instead used the mathematical concept of confidence. The mathematical term confidence is easy to understand. Imagine you flipped a coin 10 times. It is impossible to predict how many heads, and how many tails you would get - but it is possible to say how likely it would be for you to get 10 heads out of 10 flips. The answer is rougly one in a thousand, which is to say that if you flipped a coin ten times before breakfast each morning, you would flip ten heads about once every three years.

Probability runs the sums forwards - you ask a question about something that might happen, and try to work out how likely it is... statistics runs the sums backwards - you say that you want to be 95% certain of something, and then ask how specific you have to be.

What the IMF report had said was that in order to make a prediction with 95% confidence, it could not be precise at all. In order to be that confident it was right, it could only make the a statement that prices would range between 20% lower and 20% higher than they are at the moment.

You would never have guessed at THAT from the headline.

To complete the story, the Morgan Stanley economists were a little more precise - their own models have a 95% confidence that prices will range between 10% lower and 10% higher.

Apart from learning not to believe headlines, what can we pick up from this as investors? Well, firstly that the best, most professional, thoughtful and rigorous economists on the planet are say that it is too hard to predict what prices will do over the next couple of years...

... but so many armchair investors believe that they KNOW what house prices will do.

I have far more respect for the IMF and Morgan Stanley economists who admit that the problem is complex than I do for the armchair investors who tell me exactly what prices are going to do over the next few years.

What we actually know is that, over the long term, house prices broadly go up in line with inflation, and over the short and medium terms, they vary wildly up and down with no detectable logic.

The moral of this story is not that you should never buy property in the hope that prices will go up over the next few years. The moral is that you should understand what you are doing - and be aware that a purchase in the hope that prices will go up over the next few years is a gamble, not an investment.

This is just the feature article from the November 04 newsletter. Subscribe for market comment, forthcoming events, and more.

---