Is the Popular Way to Measure Fund Performance Wrong?
MAS Team | 18 May 2013

Open any newspaper, magazine and TV channel and website that reports on fund returns and you will find every analysis on mutual fund performance calculates returns in just one manner: 1-year, 3-year and 5-year periods, starting on a fixed date and ending on a fixed date. Does it make sense? Investors enter and exit at different periods. Therefore, if an investor entered the market exactly five years back, his portfolio may have underperformed the market. If he had entered a year earlier, it might have outperformed the market. How can you measure something that is highly variable by putting into a fixed start and fixed end dates? If you shift the start dates and end dates you will get a different figure of performance. Sometimes the figures will be dramatically different.

Between January 2012 and January 2103, the Sensex has gone up by 23%. Shift the dates a bit you get a completely different picture.  
12-Jan-12       16037.51          21-Feb 12    18428.61
11-Jan-13       19663.64          11- Jan13     19663.64
Return             23%                  Return          7%
The same would be the story with different start and end dates over three years and five years periods. Quite simply, returns vary hugely, depending on the start and end date. This is natural because equity funds are not fixed income products. Fixed-period returns are impractical in another sense. No investor jumps into a scheme with all his money at one go on a specified date. He is consistently advised to invest systematically over different periods, over the ups and downs of the market. It is obvious that to come to any conclusion about returns, one will have to take into account different periods that cover the bulk of market trends. 
What should be done? Funds should apply rolling returns as it captures performance across different time periods, including upturns and downturns of market cycles. This is the method employed by Moneylife to analyse fund performance (Read about it here), although we were also guilty of point-to-point comparison in the past. 
There are other ways fund returns can be misleading. Sometimes, ranking categorisation depending on investment styles can be misleading as well. Morningstar has something known as ‘Style Box’ where nine categories of investment styles are delineated. It is pertinent to note that mutual fund schemes are not structured this way. Even if they are, fund managers, more often than not, sidestep a scheme’s objective or mandate. For instance, you would probably see Reliance Industries find a place in both a ‘value fund’ (risk averse) as well as a ‘growth fund’ (higher risk). Besides, there are simply too many schemes having nearly identical portfolios simply masquerading under different fancy names and investment styles.
Statistics also play a crucial role in performance measurement. For instance, assume that a fund is up 40% in one year and 40% down the next year. What are the average returns? Zero would be a common answer. Did you know that if you had invested in this fund, you would have actually made a loss of 16% over two years? This in the first year, you made a 40% profit and the fund value becomes Rs140. The next year you make a 40% loss but now this loss is on Rs140, not on Rs100. The loss is Rs56 and so your fund value is now Rs84, a 16% loss. The way to calculate returns in such cases is to us geometric mean, not an arithmetic mean.