Averaging down makes your losses look half as big, but you’re actually losing twice the principal — confirmed by 965 daily candles
“Buy more when it drops to lower your average cost” is one of the most popular lifelines in crypto. I tested it against 965 daily BTC candles (2024-01-01 to 2026-08-22): within 20 days of buying, if the price fell 10% below your entry price, buy another unit to average down. Compared with “buy and hold” and “cut your losses at -10%,” the results are counterintuitive.
First, the percentages make averaging down look good. Of 945 rolling entry points, 197 (20.8%) fell more than 10% within 20 days. Among those 197 cases, the median return on day 20 for buy-and-hold was -11.16%, with a win rate of just 3.0%. Buying another unit after a 10% drop to average down brought the median loss to -5.30% and raised the win rate to 15.2%. It looks like averaging down “halved” the losses and lifted the win rate from 3% to 15%.
But that’s an accounting illusion. The denominator changed: averaging down means doubling your principal, using twice the money to dilute the loss shown on paper. In actual dollars lost, the median loss from averaging down was -10.60%, essentially on par with -11.16% without averaging down. You doubled your principal for just half a percentage point of improvement on paper. The worst-case results are even harsher: without averaging down, the worst loss was -34.34% on a single unit of principal; with averaging down, it was -30.37% — but that was -30.37% on double the principal, equivalent to -60.7% of a single unit. At its worst, you lose 1.77 times as much as if you’d simply held.
So the conclusion is clear: averaging down doesn’t reduce your losses; it turns a “single-principal loss” into a “double-principal loss” that merely looks better as a percentage on paper. The only time it’s genuinely useful is when you already planned to buy that amount of this asset, and a 10% drop puts it at the price you wanted. That’s scaling into a position, not averaging down. The difference is whether your position size was planned before entering, or whether you’re putting in more money on the fly to rescue a losing trade.
Leverage makes this even more lethal: my data uses no leverage, adds only once, and has a fixed 20-day window — already the mildest possible version of averaging down. Average down on a 20x leveraged contract, and one move in the wrong direction can liquidate you before you even make it to day 20. What’s the most you’ve ever put into averaging down on a trade — did you break even, or did you just dig yourself in deeper?
$BTC
#Crypto
“Buy more when it drops to lower your average cost” is one of the most popular lifelines in crypto. I tested it against 965 daily BTC candles (2024-01-01 to 2026-08-22): within 20 days of buying, if the price fell 10% below your entry price, buy another unit to average down. Compared with “buy and hold” and “cut your losses at -10%,” the results are counterintuitive.
First, the percentages make averaging down look good. Of 945 rolling entry points, 197 (20.8%) fell more than 10% within 20 days. Among those 197 cases, the median return on day 20 for buy-and-hold was -11.16%, with a win rate of just 3.0%. Buying another unit after a 10% drop to average down brought the median loss to -5.30% and raised the win rate to 15.2%. It looks like averaging down “halved” the losses and lifted the win rate from 3% to 15%.
But that’s an accounting illusion. The denominator changed: averaging down means doubling your principal, using twice the money to dilute the loss shown on paper. In actual dollars lost, the median loss from averaging down was -10.60%, essentially on par with -11.16% without averaging down. You doubled your principal for just half a percentage point of improvement on paper. The worst-case results are even harsher: without averaging down, the worst loss was -34.34% on a single unit of principal; with averaging down, it was -30.37% — but that was -30.37% on double the principal, equivalent to -60.7% of a single unit. At its worst, you lose 1.77 times as much as if you’d simply held.
So the conclusion is clear: averaging down doesn’t reduce your losses; it turns a “single-principal loss” into a “double-principal loss” that merely looks better as a percentage on paper. The only time it’s genuinely useful is when you already planned to buy that amount of this asset, and a 10% drop puts it at the price you wanted. That’s scaling into a position, not averaging down. The difference is whether your position size was planned before entering, or whether you’re putting in more money on the fly to rescue a losing trade.
Leverage makes this even more lethal: my data uses no leverage, adds only once, and has a fixed 20-day window — already the mildest possible version of averaging down. Average down on a 20x leveraged contract, and one move in the wrong direction can liquidate you before you even make it to day 20. What’s the most you’ve ever put into averaging down on a trade — did you break even, or did you just dig yourself in deeper?
$BTC
#Crypto