Averaging down: why it's the most reliable account-killer in retail

You buy 100 shares at $50. Stock drops to $40. You buy 100 more.

"My average cost is now $45. The stock only needs to recover to $45 to break even, not $50."

Mathematically, that's true. Strategically, it's the most reliable account-killer in retail trading.

What averaging down actually does

Averaging down doesn't change your cost basis in any meaningful sense. Your cost basis is the average you paid; your capital at risk is the total dollars in the trade.

Compare two scenarios:

Scenario A - single entry at $50, 100 shares:

• Capital at risk: $5,000
• Stop at $45 (10% below entry, 1R)
• Max loss: $500

Scenario B - averaging down. Buy 100 at $50, 100 at $40:

• Capital at risk: $9,000
• Average price: $45
• "Stop" mentally placed at... $40? $35? You're not sure anymore
• If stop hits $35, max loss: $9,000 × (10/45) = $2,000

You quadrupled your dollar risk. The "lower cost basis" is a number that exists on your statement but doesn't correspond to anything real. The market doesn't owe you a recovery to $45 just because that's where your spreadsheet says you broke even.

Why the brain loves it

Three biases stack:

1. Loss aversion: closing at $40 means realizing a loss. Adding shares means not yet realizing. Same dollar position, different emotional reality.
2. AnchoringAnchor biasOver-weighting an arbitrary reference point (entry price, recent high, the round number you remember) when evaluating new information. Why traders hold losers waiting for 'breakeven' even when the original thesis is broken. The market doesn't care what you paid.Read in glossary →: $50 is the original entry, the reference point. Everything is measured relative to it. "It just needs to come back."
3. Sunk cost fallacy: the $500 unrealized loss feels like it has to be recovered specifically through this trade. New money on the same position feels like it's for the recovery, when it's actually just new money on a worse setup.

None of those biases are about whether the trade is still a good trade. They're about whether you can avoid feeling the loss right now.

The honest test

Strip away the original entry. AskAskThe lowest price a seller is currently willing to accept. When you buy with a market order, you buy at the ask.Read in glossary →: "If I had no position right now, would I buy this stock at $40?"

If the answer is yes, you can buy 100 shares - but as a new trade, with its own stop, its own size, its own thesis. Not as an "average down" of the original.

If the answer is no - and 90% of the time it's no, because if you wanted to buy at $40 you'd have already done so - then you don't add. You exit the original at the planned stop and move on.

What this looks like on the equity curve

Two traders, same setup, 50 trades each. Both have a 50% win rate, both target +2R wins.

• Trader A sticks to single-entry, stops at -1R. Equity over 50 trades: roughly +25R, smooth curve.
• Trader B averages down on 30% of losers, doubling capital at risk on those. Stops "drift" from -1R to -2.5R on average. Equity over 50 trades: roughly -5R, smooth-then-cratering curve.

Same setup. Same probabilities. Different rule about losers. Trader B is bleeding because they let one decision (averaging down) systematically convert -1R losses into -2.5R losses. Win rate didn't change. Math did.

The fix

You don't add to losers. You add to winners. If the trade is working - past your first target, into profit, structure intact - you can scale into the position because the thesis is being validated. If it's not working, the thesis is being invalidated. Adding doesn't fix that. It just makes the loss bigger.

This is the rule professional traders summarize as: add to strength, not weakness. It's not a slogan. It's the math.