One day, a calf needed to cross a forest in order to return to its pasture. No one before this had ever ventured in to the forest. Without any rational, it forged out a long and difficult path full of bends, uneven ground and steep climbs. The next day, a dog took the same path following the calf’s footprints and a flock of sheep followed. As the path started taking some sort of visible shape, men started using the same and gradually, it became a well defined, accepted and the only way to cross the forest.

After many years, the trail became the main road to the village. Everyone complained about the traffic, cursed the long distance and treacherous turns, up and down hills but never thought of a better alternative. The old and wise forest smiled at how men tend to easily accept the way already open, without ever questioning whether it’s really the best choice.

In the same way, if you look at, the mutual funds’ investors. They seem trapped in a similar kind of concept called as – The Past Performance.

Past Performance as selection Criteria

Since quite some time, Past Performance has become a major criterion if not the Holy Grail of the mutual fund selection system. In fact, one of the leading business magazines in association with one of the leading rating agencies went ahead and mentioned “We take into account a much longer period for mutual fund evaluation as that can serve as a serious guide to future performance”… (Their long term means 3 years in this particular case and they used return and risk adjusted  numbers for analysis)

Further, rather than challenging the concept, there has been a continuous debate if investors should look at 1-3 year performance or a longer period like 5 years to make mutual fund selections. The treacherous path to the village is already created.

Proponents of the long-history case argue that a long term analysis ensures that the performance is analyzed over various market cycles and if the fund has done well across the long term horizon, it stands a good chance to do so in future.

Sounds logical. Is it really?

I wanted to examine if it really works. For me and for others like me who would like to know the truth and may be many more whose investments have been in red, thanks to these ratings. I gathered historical data of equity mutual fund schemes and worked out the numbers. (3rd chapter of my first book also has same kind of data)

The results were startling !

The core of the finding is “Past performance hardly relates to future returns”… and here we are, pumping our hard earned money into mutual funds, depending  on these ratings that rely heavily on the past returns generated.

Analysis Details

• The top 5 schemes by their 1yr, 2 yr, 3yr, 4 yr and 5 yr returns were selected. Thus 5 portfolios consisting of Top 5 schemes were created.
• The performance of these 5 portfolios was observed over the next 1 year (e.g. say for Dec-09 analysis, the return for 2010 was observed).
• Steps 1 & 2 were repeated every quarter for the past 3 years. The objective was to establish if the relationship with past performance that exists consistently over the period of time.

Findings & Explanation

How to read this Chart ?

The graph shows 5 bars, each bar represents the average next 1 year return generated by the portfolio created on the basis of historical returns. The left most bar shows how much average return was generated by portfolio created on the basis of scheme’s past 1 year return; the second bar is the return of the portfolio created by past 2 year return and so on.

So say the analysis is done on Dec-09, the past 1 year refer to 2009 and the returns are calculated for 2010.The above graph shows aggregate returns of the analysis done every quarter.

The graph ‘suggests’ that the ranking by past 2 years is of greatest significance while the ranking by 5 years is least significant. Please note that this is just an observation and not a conclusion. Statistics is a sensitive subject and any data tortured, throws some outputs. It’s important to delve deeper and see if the outputs can be supported with reason and logic.

To validate if the inference really holds true and if it can be used for decision making, I dissected the data of all equity schemes into two parts: The Large Cap Schemes and The Mid/Small Cap schemes. The result of analysis conducted over Large and Midcap funds is presented in the charts below.

Take that. While the large cap funds ‘seem-to-be’ driven by their past 4 year returns, the mid caps ‘seem-to-be’ to be driven by their past 2 year returns. I have consciously quoted the word ‘seem-to-be’ as I can’t find a suitable reason to defend even these findings. There aren’t actually. For anything to be considered as a general rule, it should be consistently true. I couldn’t find that when I looked at individual analysis done quarter on quarter.

If I look at each analysis done across quarters, the 2 year return is not the significant driver of future returns always. A considerable number of times the other ones (1, 3, 4, 5 yr returns) gain importance. Just like “past performance” parameter, there are many other mistakes which an investor does in his financial life, and we have decided to talk on some other aspects like those in our upcoming workshop in Mumbai on 10th March. Incase you are in Mumbai, dont miss that event.

Dont use Past Performance to Predict Future

It is quite evident that the past returns cannot be a torch bearer for investment decisions. Following the past returns as a guide or using ratings that rely heavily on past returns is like shooting at a dart board in dark. For any doubts that remain, consider this

Reliance Equity Fund was the top performer in the Large Cap category in 2012, with a return of 41%. Did you know it was the worst performer amongst Large Cap funds by historical return? i.e. if you were in Dec-11 and would have picked up this fund’s historical analysis, it was the worst performer by 1y, 2y, 3y, 4y and 5y return.

Another best performer, SBI Bluechip Fund (2012 return: 38%), never beat more than 33% of its peers ranked by past 1y, 2y, 3y, 4y and 5y return as of Dec-11. Not surprisingly, a leading mutual fund star rating agencies top picks of 2011, underperformed the index in 2012. The agency boasts of using a good mix of longer period historical return and risk adjusted performance.

In my next blog, I shall discuss what can be the reasons of looking at past performance, where did the whole thought possibly evolve and where do  things  go wrong. The views expressed above are personal and for a change, I own them. All arguments and points welcome. I really value justified facts and accept only my wife’s opinions…!!!

PS: While I finished writing this, there is a news regarding S&P being sued for damages worth USD 5 Billion for misrepresenting the credit worthiness/rating of an issuer due to conflict of interest. (https://on.mktw.net/YRehB0)

Does it sound any bells?

About the Author

Sharad Singh, a serial-entrepreneur, has spent more than 14 years in the analytics domain and has done extensive big data work in finance. He runs Valuefy, an investment portfolio analytics firm that provides portfolio management solutions to BFSI clients. Valuefy has recently launched www.theFundoo.com, a niche portal that aims to make investment decisions easy and effective for individual investors. Sharad is an engineering graduate with PGDM from IIM, Ahmedabad.

Do you use past performance as one of the criteria when you select mutual funds ?