Buy whatever is already going up. It sounds dumb. It was the best performer we have tested.

One backtest is one story. We built 258 portfolios from the screen, one for every month you could have started, held one, three and five years, and asked the plain questions: how much did chasing strength actually pay, how often did it beat a plain index fund, and what did you have to survive to collect it?

One rule, no exceptions: rank every company by its return over the past twelve months, buy the thirty strongest, and rebuild the list each year. That is the entire strategy. It made about 22% a year against the market's 13.9%, turning ₹10 lakh into roughly ₹75 lakh while a Nifty 500 fund reached ₹38 lakh. No balance sheet was ever read. And yet the ranking behind it was the most reliable we have seen.

Last year's winners really did keep winning

Sort every stock into ten buckets by past year strength and the returns climb in an almost unbroken line, weakest to strongest. The rank correlation was about −0.96, near perfect, the cleanest sort of any screen we have tested. Whatever momentum lacks in logic, it made up for in consistency: the leaders, as a group, genuinely stayed ahead.

But you owned a brand new portfolio almost every year

Here is the price of admission, and it is not paid in rupees. Momentum has no loyalty. Look at the names it held most often and the striking thing is how rarely any of them stayed: not one survived even a third of the yearly rebuilds. The winners of last year become the laggards of next, so the basket is torn up and rebuilt constantly. That means real turnover, real taxes, real trading, and a portfolio of names you will never get comfortable with, because they are gone before you do.

And when momentum breaks, it breaks violently

The other bill is drawdown, and momentum's is savage. Take every start date and its worst fall over the following five years: the typical start had to sit through a drop of about 41%, and the unluckiest starts fell about 64% from their peak, against the index's 34%. Crowded winners unwind together, so momentum pays you well for years and then, without warning, takes a brutal cut back. You have to be holding through the cut to collect the years.

A big number, spread across very uneven years

Even a strategy this strong handed out wildly different results depending on when you began. One bar for every start month tells the honest story: mostly generous, occasionally brutal, and only smooth if you never looked too closely at any single year.

Over a single year, eight in ten holds landed anywhere between −23% and +79%. The luckiest start nearly tripled at +174%; the unluckiest lost −48%. It beat the index in 59 of 100 one year starts. Stretch the hold and it steadies: over five years every single start date made money, the worst still up about 4% a year, and it beat the index in 68% of runs. This is an edge that rewards holding, not timing.

The best returns here, for the least peaceful ride

Momentum is the strange case among our tests: the idea makes almost no fundamental sense, and it worked better than any of the careful, sensible screens we have tested. But the number on the chart is not the number you would have kept, because most people cannot hold a strategy that reinvents itself every year and loses two thirds of its value on the way. The edge is real and unusually consistent. Whether you can actually harvest it is a question about your patience, your tax bill and your nerve, not about the strategy. That is exactly the thing worth finding out before you commit.

So we took it to Krest, and ran it through the whole test.