The short version: comparing returns without comparing the ride is meaningless. The tools that adjust for this are genuinely useful and genuinely misleading on an asset like Bitcoin — and the most honest comparison isn't a ratio at all, it's whether you'd have held on.

There were two hunters in the colony last year. The tabby brought something back roughly twice a week, reliably, year round. The grey one brought back far more in total — and vanished for three or four days at a time, came home thin, and twice came home injured.

If you count catches, the grey one won comfortably. If you count catches, you have also failed to describe anything useful about the difference between them, and if you were choosing which one to be, the number would actively mislead you.

Almost every comparison of Bitcoin to other assets is a count of catches.

1. Why raw returns don't compare

"Bitcoin returned more than the S&P over period X" is true for many values of X and tells you very little, because it omits everything about how that return arrived:

The size of the drops along the way. An asset that gets you there with a 15% maximum decline and one that gets you there through an 80% decline are not comparable experiences, even with identical endpoints.

The duration underwater. How long between a peak and recovering it? Years of being down is not the same as a bad month, and it's the dimension people are least prepared for.

Whether you could have stayed. The decisive one, and the one no chart shows. A return you didn't capture because you sold at the bottom is not your return. It belonged to whoever bought from you.

Where you started measuring. Shift the start date by six months and most long-run comparisons in this asset change dramatically. Anyone who picked their start date after seeing the data is showing you a conclusion, not an analysis.

2. The standard tools, and what they do

The finance industry's answer is to divide return by variability.

Standard deviation measures how much returns bounce around. Higher means a rougher ride.

The Sharpe ratio takes return above a risk-free rate and divides by that variability. Loosely: how much return per unit of bumpiness. Higher is better. It allows two very different assets to be put on one axis.

Maximum drawdown is the worst peak-to-trough fall over the period. Blunter, and in my view more useful to an actual human than any ratio, because it answers a question you can feel: what's the worst this has put people through?

Sortino is Sharpe but counting only downside variability, on the reasonable grounds that nobody complains about upside volatility.

These are real tools and they're a large improvement on counting catches. They are also where it gets complicated.

3. Why they fit Bitcoin badly

They assume a distribution Bitcoin doesn't have. Standard deviation describes well-behaved, bell-curve-ish returns. Bitcoin's returns have fat tails: extreme moves occur far more often than the model expects. A ratio built on standard deviation systematically understates the chance of the events that actually matter.

They punish upside. Standard deviation counts a violent rally as risk. For an asset whose entire historical return has come from a handful of explosive periods, a metric that treats those periods as a penalty is measuring something other than what you care about.

They're wildly period-dependent. Bitcoin's Sharpe ratio over its history, over the last five years, and over the last two years are different numbers telling different stories. Whichever one appears in the chart you're being shown is the one that supported the point.

Past volatility isn't future volatility. Every one of these measures is backward-looking and is routinely presented as though it were a property of the asset.

4. A more honest comparison

Four questions that beat any ratio:

What's the worst drawdown, and how long did recovery take? Then ask yourself, specifically and without flattery, whether you'd have held through it. Most people say yes about the past and sell during the present.

What would a fixed monthly purchase have produced, across several different start dates? Running the same strategy from a dozen starting points — rather than the flattering one — shows you the spread of outcomes instead of one path. This is the single most useful exercise in this article and almost nobody does it.

What's the position size at which I'd behave badly? Size, not asset, determines most real outcomes. The right allocation is the one where a 70% decline is unpleasant rather than life-altering.

And what am I comparing against, honestly? Against an index, a savings account, your mortgage rate, inflation in your own currency? These give completely different answers and people switch between them mid-argument without noticing.

5. The uncomfortable half

Volatility isn't risk, and that slogan is also overused. The point is fair: price movement is not the same as permanent loss, and an asset can be volatile and sound. But it's become a way to wave away the subject entirely, and that's wrong too. Volatility is the mechanism by which real risk — being forced to sell at a bad moment, or choosing to — actually reaches you. The people who lost most in this market lost to a sequence of price moves plus their own response. Calling that "not risk" is a technicality that has cost people a lot of money.

Risk-adjusted metrics are also a marketing surface. Funds and products advertise the ratio and the period that flatter them, and a good Sharpe over a chosen window is one of the easiest numbers to produce in all of finance. Treat any ratio presented without its period, its start date and its comparison set as advertising.

I've argued for drawdown as the honest measure, and it has its own flaw: it's a single historical worst case, and the future is under no obligation to respect it. "The worst was 80%" is a fact about the past, not a floor.

And the deepest problem is that the hardest variable is you. Every metric here describes the asset. None describes the holder, and the holder is where almost all the variance in real outcomes lives. Two people with identical positions over identical periods routinely end up with wildly different results, and the difference wasn't the asset.

One cat brought back more and came home injured. Count the catches if you like, but the question was never how many — it was whether you could live the way that cat lives. 🐾⚡