A trader I know called me in March 2024, visibly shaken. He'd called the Bitcoin top within $800—impressive accuracy—then proceeded to lose more money than if he'd simply held through his entire short position. His entry was right. His sizing was suicidal.
This happens constantly in crypto. The narrative around trading focuses entirely on direction—bullish or bearish, long or short, buy or sell. But position sizing is the gravitational force that determines whether your market reads matter or merely become expensive trivia.
The Mathematics Nobody Teaches
Position sizing answers one question: how much do I allocate to this trade given my conviction, my stop loss, and my account size?
The basic formula looks simple:
Position Size = Account Risk ÷ Stop Loss Distance
If you have a $50,000 account and risk 1% per trade ($500), and your stop is 5% below entry, your position size is $10,000. That mathematics is available everywhere.
What's not available everywhere is why this formula breaks down constantly in crypto—and what to do about it.
Crypto's problem is volatility asymmetry. Bitcoin can move 8% in four hours during a liquidations cascade. It can also grind 3% over three weeks. Your stop loss distance isn't static, and treating it as such is how traders get stopped out of positions that were fundamentally correct.
Consider the current environment. Bitcoin sitting at $87,553 with bearish sentiment means wider-than-normal daily ranges and more frequent wicks through technical levels. If you're sizing based on a 5% stop because that's what worked in January when volatility was compressed, you're likely getting stopped out in February's expanded range—while the thesis remains intact.
The fix: Size positions based on current realized volatility, not historical averages. Calculate your stop distance as a function of 20-day ATR (Average True Range) rather than a fixed percentage. When ATR expands—which it does during market regime shifts—you either reduce position size or widen your stop. Most traders do neither, then wonder why they're getting chopped up.
The Kelly Criterion's Crypto Problem
Every serious trader eventually encounters the Kelly Criterion—the formula that maximizes geometric growth by sizing bets based on win rate and average win/loss ratio.
The basic formula: Kelly % = W - (1-W)/R Where W = win rate and R = win/loss ratio.
If you win 40% of trades and make $2 for every $1 lost, Kelly gives you 10% of your bankroll per trade.
Here's the problem: Kelly assumes your win rate and loss ratio are stable and your outcomes are independent. Neither holds in crypto.
Your actual win rate fluctuates based on market regime. In trending markets, you might hit 55% winners. In ranging chop, that drops to 35%. If you size using the 55% calculation and the market shifts, you're overleveraged. Kelly becomes a machine for blowing up accounts during regime transitions.
The practical approach: Use a fractional Kelly (typically 25-50% of the Kelly fraction). This reduces volatility in your equity curve while still capturing most of the growth. At current Bitcoin prices with the kind of swings we see, you're not trying to maximize growth—you're trying to survive long enough for your edge to compound.
In crypto specifically, I'd argue for even more conservative fractional Kelly than traditional markets. The regime shifts are faster, the fat-tail events are fatter, and the leverage available makes the downside of oversizing catastrophic.
The Correlation Trap
Most retail traders think about position sizing as a per-trade decision. Professionals think about it as a portfolio problem.
Consider this scenario: You have conviction on Bitcoin's next move. You size three separate positions—a Bitcoin spot ETF, MicroStrategy equity, and a Bitcoin futures contract. All three are thesis-driven bets that Bitcoin goes up. All three will move in near-perfect correlation during a risk-off event.
You've effectively created a 3x position on your thesis while believing you have three diversified bets. When the music stops—and in crypto, it stops fast—your correlation assumption gets tested and all three positions move against you simultaneously.
This is how accounts get destroyed not by bad individual trades but by correlated positions that feel independent.
The rule: Calculate your effective exposure. If three positions all have 0.9+ correlation to Bitcoin, they're not three positions—they're one position with three times the risk. Either reduce size on each or pick the most efficient expression of your thesis and consolidate there.
For crypto specifically, correlation spikes during volatility events. Assets that normally move independently (Bitcoin and Ethereum, or DeFi tokens and BTC) start moving together when leverage gets flushed. Sizing decisions that look reasonable in calm markets become reckless during regime changes.
Volatility-Adjusting Your Size
Here's a concept that separates competent traders from amateurs: position sizing should be inversely proportional to the asset's volatility.
High volatility assets deserve smaller positions. Not because you're less confident—maybe you're more confident—but because the mathematical reality of volatile assets means your stop loss will be triggered more often even if you're directionally correct.
Bitcoin's 30-day volatility currently sits around 4-5% daily moves during the bearish sentiment regime we're in. Compare that to traditional equities where 1-2% daily moves are significant. If you're treating your Bitcoin position the same as a stock position from a sizing standpoint, you're misallocating risk.
The practical formula for crypto:
Volatility-Adjusted Position Size = (Account × Risk %) / (ATR × Multiplier)
Where the multiplier accounts for asset-specific volatility. For Bitcoin, I'd use 1.5-2x. For smaller-cap altcoins with daily moves that would terrify stock traders, you're looking at 3-4x.
This means when you're trading an altcoin that's moving 15% in a day during a pump, your position size should be roughly one-third what it would be for a comparable Bitcoin trade. Not because you don't like the trade—maybe it's your highest conviction idea—but because the mathematics of volatility means you're going to get stopped out more often regardless of direction.
The One Trade That Destroys You
There's a specific position sizing mistake that accounts for more destroyed trading accounts than bad entries or poor thesis development.
It's the "I have to make it back" trade.
After a losing streak, traders feel pressure to size up. The logic feels sound: "I lost 10% of my account, so I need 11% just to get back to even, and the only way to get that is to take a bigger position."
This is mathematically backwards.
Here's why: recovering from a loss requires a larger percentage gain than the percentage lost. Lose 50%, you need 100% gain to break even. But sizing up to "make it back" doesn't just require a larger gain—it also increases your probability of another loss.
The math of trading means that after losses, your emotional state is compromised, your thesis development is likely rushed, and the market conditions that caused your loss may not have changed. Sizing up in this state doesn't recover your account—it increases the probability of further loss.
The rule: Size your positions based on current account size, not based on where you want your account to be. After losses, either reduce size or step away entirely. The pressure to "make it back" is a psychological trap that turns drawdowns into blowups.
Sizing for Different Timeframes
Your position size should vary based on your holding period, and most traders get this catastrophically wrong.
Scalping (hours to days): Use smaller positions with tight stops. You're trading noise, not signal. The probability of being correct on any individual short-term move is close to 50/50. Size accordingly—1-2% risk per trade maximum. The frequency of trades means even small position sizes compound into meaningful exposure over time.
Swing trading (days to weeks): Medium positions with wider stops. Your thesis has some fundamental or technical basis, but you're still subject to short-term noise. 2-3% risk per trade is appropriate. The holding period means you're exposed to overnight gaps and weekend moves—size for that reality.
Position trading (weeks to months): Larger positions with very wide or no stops. You've done the work, the thesis is clear, and you're willing to weather volatility. But "no stop" doesn't mean "no risk management"—it means your stop is based on fundamental thesis failure, not price noise. Size to a level where you won't sell during normal volatility.
At $87,553 Bitcoin with bearish sentiment dominating, I'm seeing retail traders either sizing way too small (paralyzed by fear) or way too large (trying to catch the bottom aggressively). The right answer depends entirely on your timeframe and conviction level.
The Leverage Multiplier Problem
In crypto, leverage is cheap and available. 10x, 20x, even 125x on some exchanges. This creates a position sizing distortion that destroys accounts faster than bad direction.
Here's what happens: A trader decides to risk 1% of their $10,000 account ($100) on a Bitcoin trade with a 1% stop. That requires a $10,000 position. They have $10,000 in their account, so they need 1x leverage to take this position. Reasonable.
Then they see an opportunity. The same $100 risk on a 5% stop requires a $2,000 position. With their $10,000 account, that's 0.2x leverage. The leverage feels too low—it feels like they're not taking the trade seriously.
So they multiply. They find a 10x leverage product that lets them turn that $2,000 position into a $20,000 position while keeping their $100 risk. Now their stop is hit and they lose $100—but if Bitcoin moves 1% against them without hitting their stop, they've lost $200 (on 10x leverage). The position feels small because of the dollar amount, but the effective risk has multiplied.
The rule: Treat leverage as a position size multiplier, not a way to make "small" positions. Every time you increase leverage by 2x, you've doubled your effective risk while keeping your dollar risk constant. The question to ask isn't "how much can I control with my $100 risk?" but "what happens to my account if this 10x position moves 2% against me?"
Practical Takeaways
Calculate position size from your stop loss, not from your conviction. Your confidence in a trade should determine your thesis, not your size. Size based on where you're wrong, not where you're right.
Adjust for current volatility, not historical averages. When Bitcoin's daily range expands (which happens during bearish sentiment regimes like now), either reduce size or widen stops. Don't do neither.
Map your effective portfolio exposure. Three correlated positions are one 3x position. Calculate your actual Bitcoin-equivalent exposure before sizing individual trades.
Use fractional Kelly (25-50%) and adjust down for crypto's fat tails. The Kelly formula is a ceiling, not a target. In crypto, that ceiling is lower and shakier than traditional markets.
Never size to recover losses. After drawdowns, reduce position size or step away. The "make it back" trade is a mathematical trap that increases your probability of further loss.
Size inversely to volatility. Higher daily moves require smaller positions. This isn't about confidence—it's about the mathematics of probability and stop loss placement.
The traders who survive long enough to compound their accounts aren't the ones with the best entry timing. They're the ones who treated position sizing as the primary skill and everything else as secondary. The math is available. Most traders simply don't apply it.