There was a ‘flood of stories’ during the last week that highlighted the stupendous returns generated by Sensex during the past 40 years. And most of them also demonstrated it by explaining ‘how Rs 1 lakh invested in 1979 grew to Rs 3.9 crore now’.
However, no one asked this simple question — was it possible to invest in Sensex in 1979? It was not possible to invest in Sensex in 1979 because the Sensex was launched only in 1986. In other words, Sensex did not turn 40 in 2019; turned 33. Then from where did this ‘40-year story’ come? It came because at the time of launch, the index committee of BSE had selected 1979 as the ‘base year’ for Sensex.
You can see a similar situation in NSE’s Nifty as well. Though launched only in 1995, backdated Nifty values are available from 1990. And compared to its launch value of 1,000, the Nifty was available at just 279 in July 1990. Could you have invested in the Nifty at 279 in 1990Rs No, because there was no Nifty in 1990.
Though it is natural to select an early base date for index calculations, it is wrong to use that backdated base value for return computations. By doing this, you are succumbing to a serious mistake called the ‘survivor bias’. Let us explain this with the help of an example: Assume that you want to calculate the ‘average return’ of all stocks during the last 20 years. Since current prices are needed for return computation, all stocks that were delisted in the middle (i.e. mostly because of business failures) automatically get eliminated and only companies that ‘survived’ these 20 years (i.e. mostly fundamentally strong companies) get selected. And due to this survivor bias, the computed historical return will be higher than the actual return generated by investors.
Survivor bias happens with mutual fund returns analysis, too. Mutual funds have the habit of killing underperforming schemes, usually by merging them with their better performing schemes , usually by merging them with their better performing schemes. It means that most schemes that survived during the past 20 years are good performers and therefore, an average of their returns overstate the average performance of the fund universe. Mutual fund industry usually puts the survivor bias to its favour and uses these inflated returns to showcase that it generated better returns.
While the back-dated Sensex till 1986 has generated a fabulous CAGR of 28.71 per cent, it only generated a decent CAGR of 13.68 per cent from 1986. If you consider this 13.68 per cent as the long-term returns from the Sensex, an investment of Rs 1 lakh in 1979 would have grown to a decent sum of Rs 1.69 crore.
However, compared to the showcased value of Rs 3.90 crore, it is lower by a massive Rs 2.21 crore. And now you know why everyone wants to use the back-dated base value in 1979 not the Sensex’s actual starting value in 1986.