Imagine measuring distance in kilograms, or speed in litres. Makes sense? Not at all! This brings out the importance of measuring anything the right way. The financial world is no different.
When it comes to mutual funds, there are many ways to measure their performance and a combination of a few or all might be required to gauge the performance, depending on your investing horizon, style and risk appetite. Rolling return is one such important and appropriate tool. Here is a low down on this.
Why not trailing return
Let us try to understand the concept in terms of numbers and metrics with an example to drive home the concept better.
Sensex stood at 29,468 points as on March 31, 2020, and at 27,459 as on March 31, 2015. On an absolute basis, the index seems to have given paltry returns of 1.4 per cent annually over the five-year period. But the above dataset, measuring returns between two static points, is skewed owing to the nosedive that occurred exactly on Mar 23, 2020.
More often than not the process of investing is continuous and dispersed across time-scapes. Hence, using absolute or trailing returns, entails the error of according higher weights to the beginning and end points, while ignoring the intermittent market volatility inherent in this class of asset.
Why rolling returns
Three-year rolling return, calculated with monthly intervals, over the same five-year period considered above, stood at 11.5 per cent. This is interpreted as – investors who continued investing every month and held the investment for 3 years from the date of investment enjoyed an average return of 11.5 per cent during the period. How and why this world of difference?
Rolling return considers a series of point-to-point returns within the overall time frame taken for analysis. In the above case, 25 datasets are arrived at by calculating the returns starting from March 2015 to March 2018, April 2015 and April 2018, and so on until March 2017 to March 2020, and the three-year rolling return is the average of the 25 point-to-point returns within the five-year period.
Moreover, analysing the individual data sets within the rolling returns throws out more data like – one, the maximum return over a three-year period stood at 15.8 per cent per annum and the worst, at negative 17 per cent, during the same five-year period; and two, investing across a three-year time horizon in Sensex during this period also ensured a 96 per cent chance of greater-than-zero returns. When viewed in this prism, for a systematic investor, risk in equity investing reduces looking at the data.
Pros and cons
Stretched over long time periods, rolling returns stand a better chance of weeding out volatility and giving a representative summary of the historical returns than trailing figures, for systematic investors. In general, data for at least two complete market cycles comprising a bull and a bear phase, or at the least one complete cycle helps rolling return analysis be more effective. In Indian stocks, a typical bull-bear cycle lasts around 12 years. In bond markets, your analysis needs to stretch to at least 10 years.
And the three-year rolling returns considered in the example quoted should be adjusted per the investment timeframe of each investor. Similarly, the monthly intervals could be adjusted for daily, weekly or yearly based on the frequency of investments.
As is the case with all historical data, rolling returns are again not the indicators of future returns, but do help you clearly gauge the riskiness of an asset class more lucidly.
