The New Sensex

Sensex's 3 per cent  average performance over the last 5 years can be considered as sleeping, writes Mukul Pal.

Mr Axe and Ms Houghton were the first to build the Axe Houghton Index. Charles Dow's work came later, sometime in 1880's. Standard and Poor's (S&P) was also founded during that time. Though it's not clear who created the market capitalisation methodology, but S&P is the largest publisher of market capitalisation methodology indices.

It was about 100 years later in 1960's that MSCI took the idea ahead to global and emerging markets. And then there were broad indices from Russell's etc.

The offer just kept getting larger with Derivatives and ETF providers. A generation long financial industry built around indexing, a simple average or market capitalisation methodology. No wonder the Nifty we knew got rechristened as S&P CNX Nifty. Though the Sensex may have avoided rechristening, it also is a market capitalisation index.

150 years is a long time for an idea to survive. But if the idea is an extension of human psychology it may become eternal. Now whether this makes it a candidate for Mackay's grand illusions would be academic blasphemy, but this is what has already been voiced and proved that market capitalisation is a flawed indexing approach.

Why? First because it gives more weightage to winners, runs after performers, is predisposed to run after growth, away from value and hence forces investors to pay an undue price for owning the "good". Robert Arnott made a complete case of how market cap methodology is a flawed approach to indexing, in his work on Fundamental Indexing. Then there was Jensen's alpha, which was a proof that most fund managers don't beat the market.

We at Orpheus are on the other side, looking at markets from a Galtonian perspective. In our paper on mean reversion indexing we redefined Galton's idea on mean reversion to extreme reversion.

How does our approach challenge the market capitalisation approach? First, it illustrates that running after winners is a losing approach. Second, it suggests that divergence between sector components can be so large that the market cap benchmark's performance can be considered no short of a sleeping asset. Third, we illustrate how this polarised performance between any groups, even blue chips can be used to consistently outperform the respective universe.

Let's explain. We took the Sensex 30 components and compared their performance for 1 to 5 years. The difference between the best and worst performance ranged from 328 per cent  (1 year) to 15 per cent  (5 years). Sensex's 3 per cent  average performance over the last 5 years can be considered sleeping. There was more alpha to be made than what Sensex offered.

About winners and losers; we ranked the performance of Sensex 30 stocks and gave them a percentile score. Then we compared the performance of the best above 80 percentile with the worst below 20 percentile. On average the worst delivered 30.8 per cent  compared to the best which fell 42.6 per cent  over the last 5 years. This was not a standalone performance but worked consistently from the 3rd year onwards.

This means that if you are adopting a market cap approach i.e. running after winners, it would have worked for you for the first 2 years and then the approach would have failed you miserably compared to picking up the worst losers among Sensex 30 stocks. The losers fail to deliver in the first 2 years, but then deliver consistently ahead.

Another interesting observation was that 8 of the 30 stocks viz. BHEL, Bharti, Maruti, Reliance Industries, Sterlite, Tata Motors and TCS showed one complete swing from the best to the worst over the last 5 years. For the last two years Bharti, DLF, Reliance, Sterlite and Tata Steel are the worst in Sensex. Now if the worst performances persisted till January 2013, you know what your 'New Sensex' composition must have till December 2014.

The author is CMT, and Founder, Orpheus CAPITALS, a global alternative research firm