Thursday, May 18, 2006

Random ramble on forecasting foolishness

Call it a bloodbath , call it a carnage, call it whatever Udayan Bose wishes to term it on CNBC . It’s a 826 point fall. Actually the absolute numbers are scary but in percentage terms around 7% for a market at 12000. I am no expert on the markets, in fact I doubt if anyone is.

One of the prime reasons given was that FIIs (in grey suits that dignify their greed) are selling because they are about to be taxed. Just note how the market always goes up due to fundamental factors like the booming GDP growth, great middle class consumerism and always falls due to sentimental factors like rumours of taxation, FDI restrictions and left govt in Bengal. Ever wondered why markets do not fall on rational expectations. Because it means all these FIIs, fund managers and a lot of us were smoking dope or in some specific cases speculators and hedge funds had run amok.

During my break between jobs last month, I read a book titled ‘ Fooled by randomness’ by Nicholas Taleib. In a refreshingly simple but intellectually sophisticated way, the writer manages to educate us about how randomness and not superior wisdom or intelligence is a significant and decisive component in markets, business and hence careers. From the first question he asks in the book “If you re so rich why aren’t you so smart”, you are immediately drawn to his writing style and you start reminiscing various pub and coffee table discussions like “ arre yaar yeh CEO kaise ban gaya, boss that banker is chhaaping a 20 lakh bonus this year and so on”.
Through interesting examples like “if you keep sending a report to a sample size of 200,000 where 50% get an up report and 50% get a down report , at the end of eight quarters there would still be 12500 people who would rate you as a genius for predicting markets so accurately”. Now extrapolate that to 2 million and you know where the fat bonuses of many fund managers are coming from.

The most wonderful insight is on how humans cannot think expected value. Imagine someone told you there is a stock that has a 95% probability of gaining Rs.100 in a week and 5% probability of losing Rs.2000( assuming it is within the circuit filter). The expected value is actually – Rs. 5. He explains how we associate a low probability of an event to determine the risk of an investment rather than the extent of loss or expected value. Most of us would actually buy such a stock since it seems ‘less risky’.

There are other good fundas like if you decrease the frequency of observations there is lower volatility because of the reversion to the mean. So basically do not watch Udayan Bose and CNBC everyday and get mini heart attacks but watch him (or your stocks) once a month and your health would improve.

The author also clarifies that his observations should not be used as a recourse to cynicism and laziness. He just urges to be more wary of stock analysts, traders, consultants, poll forecasters and news anchors and others who take themselves too and the quality of their knowledge seriously & those who don’t have the guts to sometimes say: I don’t know....". By the way I think the markets will bounce back today since FIIs have been net buyers of F&O yesterday , so there will be some squaring up.


Post a Comment

<< Home