Shantam Mutual Fund's low volatility index fund backtested at 15.1% a year against 13.2% for the parent index, with lower volatility, but in three live years it lags by 1.8% a year and holds 38% in consumer staples. Is the factor broken, or is this what the strategy looks like?

The factor is not broken: the live lag is what the backtest itself predicts for strong markets. In the backtest's four years when the parent rose 20% or more, the index lagged by 3.8 points a year on average; the three live years were all such years and the lag was 1.8. The backtest's whole edge came from two down years. The real question is the 38% staples weight, which the index rules should cap.

A cautious driver reaches the destination later on an empty motorway and earlier in a storm. Over many trips the average speed can still be higher, because the storms cost the fast driver far more than the clear days cost the cautious one. A low volatility index wins over a cycle by losing less in bad years, and it pays for that by lagging in strong ones. Shantam's backtest says exactly this: the index was ahead in only 5 of the ten years. In the two years the parent fell, it beat the parent by 10.5 points on average; in the four years the parent rose 20% or more, it lagged by 3.8 points on average. Its downside captureThe average return of a fund in the years its benchmark fell, divided by the benchmark’s average return in those years. Below 100 per cent means the fund fell less. was about 9%: in the parent's two down years it fell about a tenth as far.

Put the three live years into the backtest's own pattern. The parent rose 20.3% a year, which is stronger than any three-year stretch in the backtest. The backtest's strong years lagged by 3.8 points; the live years lagged by 1.8. The fund has done slightly better in strong markets than its own backtest, so the live record is consistent with the strategy, not evidence against it. The longest run of years without beating the parent in the backtest was 2; three is longer, but three strong years in a row is also rarer than the backtest contained. Nothing in the live data tests the claim that matters, which is behaviour in a fall, because there has not been one.

The 38% in consumer staples. A volatility screen with no sector limit crowds into whichever sector has been calm lately, and calm is not the same as safe: staples that trade at high multiples can fall hard when rates rise or growth slows, and that fall would not be in the backtest if the sector was never this expensive during it. The factor is fine; the construction has let one sector become two fifths of the fund, and that is a bet on staples, not on low volatility. Check the index methodology for a sector cap and a single-stock cap; many factor indices cap sectors somewhere between a quarter and a third, but confirm this index's rules. If there is no cap, propose one at the next index review, and show investors the fund measured against a sector-neutral low volatility version so they can see how much of the return is the factor and how much is the sector.

Two honest warnings belong in the answer. First, a backtest of Rs 100 growing to Rs 393 against Rs 315 over ten years rests on two down years; remove them and the strategy simply lags. Second, no live record of three up years can confirm the strategy, and none can refute it; the test arrives with the next fall. Neither merging nor changing the index is justified by a lag the backtest itself predicts. What investors need is a factsheet that says plainly: this fund is expected to lag in strong markets and is judged over a full cycle.