The since inception trap

A fund's ad flashed "18.9% since inception" at me last week. Right below it on the same screener sat a fifteen-year-old fund saying 13. Most people stop reading there and tap the 18.9. I did the opposite and pulled both apart on Sunday morning.

Here is what the big number was hiding.

Fund A is six years old. It happened to launch the month the market bottomed out, rode a +75% first year on the recovery, and then did a thoroughly ordinary 10% a year after that. Fund B is fifteen years old and just plodded along at 13% through two full cycles, a crash and a boom included.

So the 18.9 and the 13 are not even the same measurement. One covers six years, the other covers fifteen. You are comparing a six-year sprint to a fifteen-year marathon and declaring the sprinter faster.

Put them on the only stretch they both actually lived, the same recent five years, and the ranking flips. Fund A did 10. Fund B did 13. Rs 10 lakh over that identical window grows to 16.11 lakh in A and 18.42 lakh in B. The boring fund ends 2.32 lakh ahead. Strip out Fund A's single lucky birth year and it was a plain 10% fund the entire time.

The deeper problem is that since inception answers a question you cannot act on. What if I had invested on the founder's first day. You cannot. That date is gone. The only money you can deploy earns whatever comes next, and the since-inception figure never promised you that.

So read the standard 1, 3 and 5 year lines where the windows actually match. Ask what regime the fund was born into. A monster first year off a bottom is a birthday, not a process.

How one birth year makes 18.9

Here is the arithmetic for Fund A, so you can see how much work its first year is doing.

Rs 1,00,000 grows 75 percent in year one, to Rs 1,75,000. Then five years at 10 percent: 1.75 x 1.1 ^ 5 = 1.75 x 1.611 = 2.818. Over six years that is a CAGR of 2.818 ^ (1 / 6) minus 1 = 18.9 percent.

Now drop the first year and the same fund is a 10 percent fund. One year out of six turned a 10 into an 18.9.

Push the age up and watch the effect fade. At twelve years the same birth year lifts a 10 percent fund to about 14.3. At thirty years, about 11.7. A since inception number is most misleading exactly when a fund is young and its launch date happened to be lucky.

Same model, same end date, five answers

I ran our own Smallcap model through this test. Same curve, same end point at Jul 2026, only the start date moves.

Nothing about the rules changed between those rows. The start date did. A start in early 2023 caught the 2023 run, when the model's calendar year return was about 113 percent. A start in Sep 2024 began near a peak and sat through the fall that followed. The full record from May 2021 says 38.6 percent a year, and the file figure, measured to 27 Apr 2026, is 37.40 percent with a maximum drawdown of minus 22.34 percent (Model, back tested, as of 27 Apr 2026).

That is why our strategy pages show the whole curve with the drawdowns, not one figure from a chosen date. Any single CAGR is a choice of window, and the window can be chosen.

Three checks before trusting a since inception figure

1. When was it born?Find the launch date. A launch near a market bottom flatters every since inception number that follows.
2. How old is it?A young fund's figure leans heavily on its first year or two. An old fund's figure includes full cycles.
3. Line up the windowsCompare 1, 3 and 5 year returns ending on the same date, for every fund on your list.
4. Look for the fallDoes the window include a real drawdown? A record with no bad year in it is a sample, not a record.

On the same five years in the opening, Fund A did 10 percent and Fund B did 13. Rs 10,00,000 x 1.10 ^ 5 = about Rs 16.11 lakh. Rs 10,00,000 x 1.13 ^ 5 = about Rs 18.42 lakh. The fund with the smaller headline was ahead on the only stretch both of them actually lived.

The free Learn module on performance measurement covers CAGR, rolling returns and fair comparison windows, with quizzes and a certificate. Two related reads: total return against CAGR and the average year myth.