The Magic Formula beat the Indian market. That's the least interesting thing about it.

One backtest is one story. We built 258 portfolios from the formula, one for every month you could have started, held one, three and five years, and asked the only question that matters: how often did the famous recipe actually beat a plain index fund?

Over one year: 53 of 100 starts beat the index. Over three years: 45 of 100. Over five: 47. Read those again. The longer you gave the world's most famous stock formula, the less likely it became that it was beating the market at all. Every average ever quoted for it sits on top of that coin.

One number is quoted. The market paid a range.

And what a range. Over a single year, half of all starts landed anywhere between −21% and +62%. The luckiest start earned about +106% in a year; the unluckiest lost −52%. Nothing about the companies changed between those two outcomes. Only the month you happened to begin. When a strategy's single year fortunes can differ by a hundred and fifty points on timing alone, the average is not telling you what to expect. It is hiding what to expect.

The famous win, at actual size

Run straight through the decade, the top thirty compounded at about 14.79% a year against the index's 13.9%. ₹10 lakh became roughly ₹40 lakh; the index alone reached about ₹37 lakh. That is the entire celebrated victory: under one point a year, a margin thinner than a single bad week, earned over ten years of effort. The formula did beat the market. It just barely mattered.

What you had to survive to collect it

That sliver of a margin was not free. Take every start date and its worst fall over the following five years: the typical start had to sit through a drop of about 55%, and the unluckiest starts fell 68% from their peak, against the index's 34%. Two thirds of your money, gone on paper, defending a lead of less than one point a year. Patience is usually a footnote in how strategies are sold. Here it was the whole price.

What the formula actually bought

Not thirty famous blue chips. The high return, low price combination led straight down the size ladder into small commodity and industrial names, more than four fifths of the basket in small and micro caps, nearly a third of it in materials alone. Hindustan Zinc appeared in 95 of every 100 rebuilds. These are real companies, but they are far smaller and far swingier than the formula's reputation suggests, and those swings are where the 68% falls come from.

Range, not a number

None of this makes the underlying idea foolish. Good companies at good prices is a fine principle, and the formula did, on the thinnest of margins, beat the Indian market. What the decade actually tested is the promise of the single number, and that promise did not survive. Behind the one famous figure sits a coin toss that never improved, a range of outcomes wide enough to hold triumph and ruin at once, and a ride that shook out anyone who believed the average was a guarantee. The formula is not magic. It is a range, and every strategy you will ever be sold is a range. The only way to know its shape is to test it, before you trust it.

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