Saturday, September 04, 2010

Index investing and quantitative strategies

I was reading the buttonwood blog in economist - regarding bond indices. A side argument there pointed me to an interesting thought. Index investing has picked up in a big way (at least in terms of mindspace) in the developed economies. It is finding some favor with indians as well. This was no doubt a reaction to the exhorbitant asset management fees charged by managers who ultimately ended up underperforming the index. The thought pioneered by Vanguard and the likes was that lower cost model of index investing is likely to be remunerative in the medium term since no one can sustainably outperform the index anyway.

Leaving aside that reaction, if one were to explore the world of quantitative rules driven investing as against index investing, one reaches the same conclusion sooner or later - index investing is an oversimplified quant strategy! The rules are deceptively simple - in fact so simple that they do not appear to be rules at all! Invest in stocks in the same proportion they form in the index and stick to it. Invest/divest when index constitution changes. Period. A quant model can be extremely complex or relatively simple. The simplest of them can start to look very similar to index investing. And then one realizes that index investing is also a quant strategy - the difference is one of degree and not of nature.

If that be so, can we think of outperforming the index using quantitative strategies? The above-mentioned article in economist spoke of synthetic bond indices which do not over-weight more indebted borrowers unlike the commonly used ones. Thus they end up performing better than the conventional index. Likewise a 'remixed Nifty' ETF launched by one of the new AMCs in India is an attempt to redefine the allocation within Nifty stocks using factors other than market value. Both approaches merit attention. Leaving aside the special status accorded to the index, these discussions are akin to evaluation of two alternate quant strategies. Clearly one with market cap as the sole basis of weightage (i.e. the index) seems less sound than one with some other parameters that incorporate attractiveness of stocks/bonds.

What factors? That remains the holy grail of quantitative strategists! More on that later.

No comments: