HOW TO REDUCE THE LUCK FACTOR IN INVESTING (Sunday Times May 2005)
In his book “A Mathematician Plays the Stockmarket”, author John Paulos demonstrates how the law of large numbers can produce some excellent results on the stock exchange. He does this by showing how to set up a hypothetical scam. First you get the names of 64000 potential investors. Post them all a letter explaining that you are a JSE expert and have a foolproof system to predict individual stock movements. Tell them that they don’t have to take your word for it – over the coming weeks you will prove it to them. Then you take a share, any share, let’s say Sasol. To half the 64 000 potential clients you write that on close of business on Tuesday in three weeks time Sasol will definitely be trading above its present level. To the other 32 000 you write that the price will be lower. Be sure to keep a record of to whom you sent the “ups” and to whom the “downs”.
In three weeks time, Tuesday arrives, and let’s say Sasol ended the day up. Discard the 32 000 clients to whom you predicted the price would be down. Take the remaining 32 000 and divide them into two. Send 16 000 a second letter predicting that in three week’s time Investec (for example) will be up, and 16 000 a prediction that it will be down. Repeat this process four more times in a row, always halving the remaining clients and splitting your prediction. Once completed, you will have 2000 clients left that you have been spot on with your predictions six times in a row! Now send them a letter saying that you think they will agree that your prediction strategies are excellent and in order to carry on receiving your predictions they will need to pay you R 2000. Doubtless they will be impressed with your track record. If they all take up your offer – you have made yourself 4 million Rand!
Although this is a hypothetical scam, it harbours valuable real life lessons. When looking at investment performance commentary, there is always a lot of technical and complex jargon – but never any mention of luck. However, the above algebra can be applied to investment analysts too. Start with 64 analysts and presume that on average half of them make bullish investments and half make bearish investments. Let’s say that they invest every two months. After a year, two of the original 64 analysts will have got every call right simply by the mathematics of chance. Naturally these two analysts won’t attribute their success to chance, but to superior investment strategies. And their investment portfolio will have such a good 12 month track record that investors will be piling in. But, if their strategy was as random as this example suggests, the next 12 months are unlikely to deliver such good returns, as, eventually everyone’s luck runs out.
So how do you avoid this trap? First you do not base your investment decisions on past performance only. You look at the caliber of the people making the decisions, and the philosophy of their investment strategies. When you do look at past performance figures, look for the investment managers that have consistently delivered good results over the long term. There are undoubtedly some very capable portfolio managers out there, and in the long term they will succeed, no matter what luck they are dealt. Above all – do not part with your hard earned cash to anybody who claims to have a foolproof system for making money (no matter how good their first six predictions look). comment