The traditional way of creating and optimising a portfolio has been Mean Variance Optimisation based on Modern Portfolio Theory. There are several issues with MPT, as we detail here, but the most problematic one is that it relies heavily on assumptions around volatility as a driver of risk. Most retail investors do not intuitively understand volatility, but do understand drawdowns and want to minimise them - so we chose our objective as building portfolios that achieved the highest returns given a maximum drawdown constraint.
Our end goal was to build portfolios that had higher returns on average than the benchmark with lower drawdowns in bad times. One important limitation we set from the start was to only look at fully liquid products (ie. which can be liquidated or sold at any time by the investor to withdraw their money), so we avoided closed-ended mutual funds, bonds/FDs, sovereign gold bonds etc. in our research. Liquidity is important because our approach also incorporates regular rebalancing and this is not possible with illiquid investments.
Our approach was to first determine all the possible asset classes an Indian long-term investor should own. Next we constructed four risk-based portfolios combining these asset classes to maximise returns within a given drawdown constraint. Finally, we identified the best funds within each chosen asset class to build the actual portfolios for the end user. To analyse asset classes, we chose the nearest possible indices or funds as proxies.
The main limitation we had in our work was length of data available - in the Indian markets, the longest we could go was up to 1995, which gave us 27 years of data. To over come this, we looked at various holding periods & time horizons and only used results that were consistent across these. We have detailed this here.
We started by closely reviewing the historical performance of four domestic equity indices NIFTY 50, NIFTY Next 50, NIFTY Midcap 150 and S&P BSE Smallcap 250. Their definitions are here. We wanted to use only broad-based indices and not narrower indices based on any sector or style.
The NIFTY 50 is the most commonly owned index by both domestic and international investors because it has exposure to the largest Indian companies and tends to be more stable than the other indices. It also has the longest history among market indices, other than BSE Sensex which we did not consider as it is less diversified than the NIFTY 50. Hence, we automatically included NIFTY 50 as one of the assets we would own for our ideal portfolios.
We compared the returns of the other three indices we had chosen across various holding periods and starting periods, concluding that NIFTY Next 50 is clearly a superior performer once the holding periods are 5 years or longer. It also has the advantage of longer data availability which would improve the quality of our results.
The above chart represents 5 Year horizon returns of various investors having different starting point of investment in 3 of the indices i.e one from 1-4-2003 to 1-4-2008, 1-4-2004 to 1-4-2009 and so on….We can observe from the chart that Nifty Next 50 has outperformed both Mid-cap and Small-cap indices in 10/15 of the 5 Year periods while underperforming by huge margin only in the two early periods
We also looked at drawdowns experienced by these three indices from 2003 to 2022 and found that prior to 2010, the Small-cap Index had lower drawdowns than NIFTY Next 50 and NIFTY Midcap 150, but since then NIFTY Next 50 has clearly outperformed the other two indices.