Most investors look at the point-to-point (P2P) trailing returns of the ULIP, and some tend to incorrectly focus more on the top performing funds over the past year especially. The problem with P2P returns is that they tend to get distorted sometimes due to huge outperformance in 1 or 2 years. To take an example, imagine Fund X, which is a mid-cap fund, and delivered a huge return of say 110% in year 2007. As a result, its trailing returns over 2 years and 3 years will also start looking good, due to spill-over. Hence, we may be influenced to choose the fund. But the important question is whether Fund X has been a consistent outperformer in past years as well. Given below is an illustration of the top 10 performing funds from the Morningstar India OE (Open End) Flexi-cap category over the bull market years of 2006-2007. We have chosen mutual fund flexi-cap category, as there were substantial number of funds then in 2006-07 to make this analysis legible. We have also renamed the funds as Fund A, Fund B and so on (and removed the original names), for sensitivity purposes. It can be seen that many of the top 10 performing funds from 2006-07 were among the bottom performers during the market downturn of 2008 (global financial crisis), and fell by as much as 65-75% during the year. Also, the performance ranking of the funds varied widely in the following years, and in several years, most of the top 10 performing funds (of 2006-07) failed to re-appear consistently among the top 10 ranks again in the following years, or for that matter in the top quartile (top 25% performing funds in the peer category) as well. This shows the fallacy of looking at point-to-point performance, and especially the yearly top performing funds, while investing. There is hardly any consistency in performance of the top 10 performing funds (of 2006-07) over the following years, as indicated by the large dispersion in rankings. So, then what should we look at to measure consistency in performance of a fund? One parameter to look at is the rolling returns of the fund versus benchmark index and peer category. Rolling returns helps to remove the point-to-point bias in performance, and display the returns in overlapping cycles. For example, if we take 3-year rolling returns, with monthly shift, over a 10-year period, then it will show how the fund has done in various 3-year periods, by shifting a month each time, over a 10-year tenure. So rather than looking at one discrete period of 3 years, we look at several 3-year rolling periods (in this case, 85 three-year rolling period observations), and how the fund has performed in each of them. You can analyse in how many percentage times of the rolling period observations, the fund has beaten the benchmark index as well as the peer category. You can go one step further, and analyze that within the rolling period observations how much percentage of observations feature in each of the four quartiles (within the respective peer category). For example, if 80-90% or above of a fund\u2019s rolling period observations have featured in the top two quartiles, then it indicates good consistency in performance over the fund over the long term. Therefore, along with trailing returns (like 1 year, 3 years, 5 years, 10 years etc.) of fund, one should also look at the rolling returns of the fund, to get a true and better picture of performance consistency. (By Sampath Reddy, Chief Investment Officer, Bajaj Allianz Life Insurance) Disclaimer: The opinion expressed by the author in this article is his personal opinion and readers are advised to seek independent financial advice before taking any investment decisions.