Best Arbitrage Fund? Returns Compared by Holding Period

The Indicator model sometimes allocates a monthly contribution to arbitrage. That raises a practical research question: does the historical leader change when we extend the holding period?

One holding period cannot answer that. We need to move the starting date, extend the holding period and see whether the same name keeps coming up.

Let's run that comparison with actual fund NAV histories.

Put every fund on the same starting line

We screened Direct Growth arbitrage schemes from MFAPI. The final group needed current NAV data at the time of the study, a usable ten-year rolling return and at least 180 shared calendar start dates for comparison.

That left 12 funds. For a given holding period, a start date counted only if every fund in the group could be measured from it.

These are lump-sum annualised return comparisons, not monthly SIP returns. A fund wins a period when it produces the highest return from that common start.

Watch what happens as the holding period changes

All names in this table refer to the Direct Growth series in the public evidence.

The one-year leader was Nippon India. At three years, it was Invesco. So the fund that led most often at one holding period did not necessarily lead at another.

At five, seven and ten years, Edelweiss led. Its share of first-place finishes increased across those longer tests, reaching all 759 shared ten-year starts.

What does a first-place finish leave out?

Imagine two funds that differ by only a few hundredths of a percentage point. One can finish first repeatedly without producing a large difference in rupees.

So the win frequency is one view of the result. It does not show the size of each lead, short-term NAV fluctuations or the consequences of an early exit.

There is also a selection effect. Requiring ten years of history removes newer funds from the group, including from the one-year comparison. The short-period result therefore describes these 12 established funds, not every arbitrage fund available.

How does this connect to the Indicator research?

The Indicator model uses arbitrage as an allocation that may be held while the equity reading is expensive. Edelweiss's longer-period record makes it a historical candidate for further research in that role.

But the model may move that allocation into equity before five or ten years. Long-horizon leadership does not, by itself, establish the best result for those shorter and variable holding periods.

The Indicator strategy backtest also uses a fixed arbitrage return assumption. This NAV comparison is a separate study. It does not silently replace that assumption with Edelweiss's realised returns.

Sources and the limits of this study

Source: public Direct Growth arbitrage-fund evidence, generated from MFAPI NAV histories on 16 September 2026. It contains the fund universe, common-date method and detailed results.

NAV already reflects ongoing fund expenses. This comparison does not separately model investor taxes, exit loads or the operational cost of switching. It also does not assess future changes to a scheme's process.

Calendar start dates overlap, so these are not independent experiments. The figures describe this snapshot and selection process. Arbitrage funds have investment risks and do not provide guaranteed returns.