The Coffee Can portfolio crushed the market. It has also never seen a bad one.

Ten lakh rupees run through the screen since 2020 grew to about ₹43.7 lakh. The same money in a plain Nifty 500 index fund grew to about ₹29.9 lakh. Over its run the strategy compounded at roughly 34 percent a year against the market's 24.5 percent, an enormous margin, and it did it with a shallower fall along the way. On every count an investor cares about, it won.

It is the kind of chart that sells a strategy. So before you are sold, read the one thing that chart is quietly telling you: the dates on the bottom axis.

Now look at the dates

That chart begins in June 2020, at the very bottom of the COVID crash, and ends in 2025. Five years, and not one of them a real bear market. The Coffee Can, as tested here, was born at the perfect moment and has lived its entire life inside the strongest bull run in a generation. It has never once been asked the only question that ultimately matters: what happens when the market falls and stays down.

The edge was really two years

Break the run into calendar years and that spectacular average dissolves into a couple of extraordinary ones. In 2020 and 2021, riding the rebound, it returned 62 and then 115 percent. Take those two years out and what remains is ordinary: it lagged the market in 2022, again in 2024, and again in 2025. The ten point a year edge was not a steady drumbeat of skilful picking. It was two moonshots, and then a strategy that mostly kept pace or fell behind.

A gentle fall, in a market that never truly fell

Its worst drop was about 32 percent, milder than the screens that lived through 2018 and 2020 in full. That reads like safety. It is not. A shallow drawdown inside an uninterrupted bull market is not evidence of resilience; it is the absence of a test. We simply do not know how this screen behaves in a crash, because it has never been in one.

The range, in a very short sample

Even across this brief history the one year outcome swung widely, from a 15 percent loss to nearly doubling your money, depending on the month you began. But notice how few independent windows there are: a handful of overlapping stretches, all inside the same boom. Treat every number on this page as drawn from a sample of one very good market.

What it actually owned

The screen leaned on a familiar cast of quality names, holding some for most of its short life, and tilted toward materials, technology and industrials. A reasonable, recognisable portfolio, which is part of the point: it looks sensible, it performed brilliantly, and none of that tells you how it survives a market that finally turns.

The most dangerous backtest is a short, winning one

None of this means the Coffee Can is a bad idea. Marcellus may well be right, and quality may well compound. But a five year record set entirely inside a bull market is the single most seductive and least reliable kind of evidence there is. It shows a strategy at its best and hides the only thing you truly need to know, which is how it behaves when things go wrong. The number to distrust is the amazing one. The only way to earn real conviction is to test an idea across history that includes the bad years, not just the good.

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