The fee you never see: what 13 years of fund data say about Regular plans

Every Indian mutual fund comes in two versions: Direct, which you buy yourself, and Regular, which you buy through an agent who is paid a commission out of the fund's returns. Same fund, same holdings. The only difference is about 1% a year.

One percent sounds small, so I built Fund Reality to see what it adds up to with real prices instead of illustrations.

What the data show

• ₹10 lakh invested through an agent in January 2013 ended up ₹9.9 lakh smaller on average than the same money invested directly.
• In Nippon India Small Cap the gap was ₹21.2 lakh.
• The Nifty 50 index fund returned 5.8% a year over the last five years, but the typical five-year period since 2013 earned 13.3%. The latest period started near a market high.
• No kind of equity fund escaped the 2020 crash: average worst falls ranged from −35% for mid caps to −43% for small caps.

Why one return number misleads

Fund factsheets show a single trailing return. That number depends heavily on the start date. So the site computes every possible five-year window, 459 of them, and shows the whole spread. A first-time investor sees the range they might actually get, not the best-looking slice.

The catch in my own list

The 21 funds were chosen because they are well known today, which means they are survivors. Funds that closed or were merged are missing. That is why these large-cap funds beat the index in 71% of five-year periods, while S&P's SPIVA scorecard found 84% of all Indian large-cap funds lagged their benchmark over five years.

I put that caveat on the site itself. If a data story only works when you hide its weaknesses, it isn't a good story.

How it's built

Python and pandas download and clean the NAV history once, compute rolling returns, drawdowns and XIRR, and write static JSON. The Next.js site reads that JSON, so public pages need no database. Supabase handles sign-in for the members-only portfolio cost check.