Buy companies that turn profit into real cash, and trade cheap. It was one of the biggest winners we tested, and one of the wildest.

One backtest is one story, so we built 261 portfolios from the screen, one for every month you could have started, held one, three and five years. Rank companies by how much of their profit shows up as cash and by how cheap they are, add the two ranks, buy the best thirty, rebuild each year. It made about 24% a year against the market's 13.9%, turning ₹10 lakh into roughly ₹84 lakh while a Nifty 500 fund reached ₹39 lakh. A commanding result. But before you reach for it, two things about that number deserve a hard look.

First catch: the ranking sorts, but only roughly

If the combined score truly measured quality, the best scoring companies would have earned the most, cleanly, top to bottom. They did not. Sort every stock into ten buckets and the returns climb, but raggedly: the rank correlation was about −0.67, a real signal pointing the right way, yet a noisy one. Several middle buckets landed out of order, and the very best scoring bucket was not even the top performer. So this is a useful tilt, not a precise ruler. It nudges you toward the right kind of company; it will not reliably hand you the single best one.

Second catch: look at what it actually bought

The bigger surprise is the portfolio itself. This is not a basket of famous compounders. The screen led you deep into small and micro cap companies, close to two thirds of the money in the smallest tier, heavy in materials, industrials and energy. The names it held most often were cyclical and often unloved: NCC, Oil & Natural Gas, CESC, JK Paper. That is where both the return and the danger came from. Cash rich, cheap, small companies were a wonderful place to be over this particular decade, but they are volatile, thinly traded, and easy to abandon at exactly the wrong moment.

A big number, paid out in violent years

Even a winner this size handed out wildly different results by when you began. One bar for every start month tells the honest story: mostly generous, sometimes savage. Over a single year eight in ten holds landed anywhere between −34% and +116%, the luckiest start more than doubling at +157%, the unluckiest losing −56%. It beat the index in 60 of 100 one year starts. Stretch the hold and it steadies: over five years every single start date made money and it beat the market in about two of every three runs. This is an edge that rewards patience, and punishes anyone who reaches for it near a peak and sells in the fall.

A real edge, wrapped in real risk

FCF Conversion Quality is a strong idea that worked, and it is easy to see why: honest cash and a cheap price are a sensible pair, and demanding both kept you away from expensive stories and accounting mirages. But the version that produced ₹84 lakh did it by owning deep, small, cyclical companies through falls most people cannot stomach. The return is not the number you would have kept unless you could hold a portfolio that halved and worse without flinching. Whether you can is a question about your nerve and your horizon, not about the screen, and it is exactly the thing worth learning before you commit real money to it.

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