How statistics makes you better at investing? An excerpt from Analytics Stories

If you invest in mutual funds, you might have come across a question, follow past returns or follow ratings by Value Research/Morningstar/Groww?

Ratings are decisive factor for most people when choosing mutual funds, because they are simple to interpret. Rather than looking at a dozen factors like CAGR, rolling returns, fund manager, alpha, beta, etc. you get look at a single & easily interpretable number for your decision

But the question is, do ratings work? If you choose are 5-star rated fund, how confident can you be about its returns?

The book Analytics Stories has a chapter devoted to this question: - 'Does a mutual funds past performance predict future performance?'

A one word (obvious answer) is NO

Lets look at the data:

The data is a great illustration of 'Reversion to Mean', essentially meaning the numbers converge back to average over a sufficiently long period of time. The 5-stars get to 3.1, which is just slightly higher than average. 4-stars get to 2.9, which is a bit lower than the average. Similarly, 1 & 2 stars move towards the average of 3.

A more detailed view is available in the following image

The way to look at the image is from left to right, ex: the G5 cell shows, out of all the funds having a 5-stars in the past, 25% reduced to 4-stars. In the row 5, the highest number is 29% i.e., 29% funds moved from 5-stars to 3-stars i.e., the ratings reverted to mean.

Having said that, if you invested in a 5-star funds, there is 68% chance that the fund would be at least a 3-star fund in 3 years. That's still a high probability in the world of finance.

To conclude, statistically speaking, past performance and ratings are no definite indicators of future performance. But investing as per past performance and ratings increases your chances of getting average returns as the returns revert to mean in the long term.

PS: Finance is not pure numbers, it's also about whether the fund manager is ethically handling the funds and a lot of other qualitative factors. That's why despite being a numbers game, finance is more of an art than a science.