Imagine that a fund house launches a new, actively managed, diversified equity scheme. Now imagine that in its presentation material it says something to the effect that if this scheme had been launched several years ago, then this is the return that that scheme would have given, and this is the alpha that the scheme would have generated. Without a doubt, that’s a really big ‘if’. My question, then, is this: is it proper for any fund house to make such an assertion?
In case you are wondering, this is not a hypothetical question. Over the last weekend, I actually came across some presentation slides for a newly launched closed-end equity scheme, in which such projections were made. And it wasn’t just the returns: the fund house had made projections regarding the scheme’s standard deviation and index correlations. According to the fund house, these numbers were derived from a “portfolio simulated with the investment process and strategy proposed for the scheme”. To tell you the truth, I don’t know if SEBI allows fund houses to make counterfactual projections and, if so, in what circumstances. Personally, though, regardless of disclaimers, I question the wisdom of doing so.
But even if I were to give this fund house the benefit of the doubt, there is still the question of how accurate are the numbers that they projected. According to the fund house, if the scheme had existed six years ago, then from 2012 to 2017, it would have generated an alpha (outperformance) of ~6% p.a. over the Nifty 50 TRI.
Right off the bat, that struck me as problematic. And it all had to with the period selected. Rather than take a single period of 6 years, a much more fairer way would have been to consider rolling periods. Given that the new scheme has a maturity of just over 3 years, it would have been appropriate to consider the rolling 3 year alpha. But probably the biggest issue I have is with the difference in market valuations. At the start of the period chosen by the fund house, the PE ratio of the Nifty 50 was 16.75. In contrast, at the current point in time, which is when the scheme is being launched, the PE ratio of the Nifty 50 is over 25 (despite the recent fall in the markets). For those who may wonder why that should make a difference, let me explain.
As most of us would know, there is a strong inverse correlation between medium-term equity scheme returns and the market valuations at the time of investing. In simple terms, at the time of investing, the higher the PE ratio of the Nifty 50 (or BSE Sensex), the lower are the chances of getting above-average returns over a period of 3-5 years. I would like to suggest that there is a similar relationship between market valuations and fund manager alpha. To put it again in simple terms, the higher the PE ratio of the Nifty 50, the lower are the chances of a fund manager generating above-average alpha over a period of 3-5 years. For those of us who like hard data, the table below might give a sense of the impact of market valuations on fund manager alpha over a 3 year period.
Rolling 3 Year Alpha Of Diversified Equity Schemes: 2000-2017
| PE Ratio at start |
% Instances Alpha > 0 |
% Instances Alpha > 6% |
Average Alpha |
|---|---|---|---|
| Below 18 | 86% | 49% | 8.1% |
| 18 - 22 | 84% | 14% | 3.1% |
| Above 22 | 45% | 3% | -0.6% |
| Overall | 80% | 28% | 4.9% |
The table considers all possible 3 year investment periods between 30 March 2000 and 29 Dec 2017 which have been grouped based on the PE ratio of the Nifty 50 at the start of each period. Alpha is calculated by subtracting the CAGR of the Nifty 50 TRI from the CAGR of the CRISIL-AMFI Diversified Equity Fund Performance Index. Total no. of observations: 3591. No. of instances of PE ratio below 18: 1602, PE ratio between 18 and 22: 1555, PE ratio above 22: 434.
Data/ Information sources: CRISIL, IISL, NSE
As you would notice in the above table, while diversified equity funds, put together, have overall generated positive alpha 80% of the time, in periods where the PE ratio of the Nifty 50 was over 22 (at the start), they have generated positive alpha only 45% of the time. Similarly, while these funds have overall generated an average outperformance of 4.9% p.a., in periods where the PE ratio of the Nifty 50 was over 22 (at the start), on an average, they have generated negative alpha.
But coming back to the fund house in question, what of its own track record of generating alpha?
As it happens, the fund house has three open-end equity schemes that are somewhat comparable with the new scheme, on account of having the same benchmark. The table below shows the rolling 3 year alpha generated by an equal-weighted buy-and-hold portfolio of these three schemes since 2007 (the time from which all three schemes have been in existence).
Rolling 3 Year Alpha Of Existing Comparable Schemes: 2007-2017
| PE Ratio at start |
% Instances Alpha > 0 |
% Instances Alpha > 6% |
Average Alpha |
|---|---|---|---|
| Below 18 | 100% | 44% | 5.9% |
| 18 - 22 | 86% | 36% | 4.4% |
| Above 22 | 48% | 5% | 0.2% |
| Overall | 82% | 32% | 4.0% |
The table considers all possible 3 year investment periods between 7 June 2007 and 29 Dec 2017 which have been grouped based on the PE ratio of the Nifty 50 at the start of each period. Alpha is calculated by subtracting the CAGR of the Nifty 50 TRI from the CAGR of an equal-weighted buy-and-hold portfolio of the three existing schemes managed by the referenced fund house that are most comparable to the new scheme being launched. Total no. of observations: 1859. No. of instances of PE ratio below 18: 501, PE ratio between 18 and 22: 984, PE ratio above 22: 374.
Data/ Information sources: NJ India Invest, IISL, NSE
As you would notice, the existing schemes of the fund house have, on an average, generated an outperformance of 6% p.a. or more, only about 32% of the time. And in periods where the PE ratio of the Nifty 50 was over 22 (at the start), they have generated such an outperformance only 5% of the time. Furthermore, the average outperformance of these schemes in periods where the PE ratio of the Nifty 50 was over 22 (at the start), was just 0.2% p.a.
In a nutshell: the fund house’s assertion of generating outperformance of ~6% p.a. over the Nifty 50 TRI is exaggerated and misleading.
