Why the fifth trade is the one that kills you

Every position sizing guide I've read treats trading like you're a sniper with one shot. Take the trade, set the stop, manage the risk, move on.

That works for someone taking one setup a week. The moment you're holding four positions and a fifth alert fires, the math changes. Now you're not sizing one trade. You're sizing it against the heat already sitting in your book.

That's where most retail accounts die. Not on the bad trade. On the fifth trade that looked fine in isolation but pushed total portfolio risk past survivable.

The formula that actually matters

Position size comes down to one equation:

Position Size = Account Risk Dollars ÷ Per-Unit Risk

Where:

  • Account Risk = Account Size × Risk Per Trade (%)
  • Per-Unit Risk = Entry Price − Stop Price

That's the whole game. Everything else — leverage, scaling, correlation — modifies the application, not the formula.

Worked example. You have a $50,000 account, you're risking 1% per trade, BTC is trading at $83,945, and you've identified a stop at $82,500:

  • Account Risk = $50,000 × 1% = $500
  • Per-Unit Risk = $83,945 − $82,500 = $1,445
  • Position Size = $500 ÷ $1,445 = 0.346 BTC

Notional value = 0.346 × $83,945 ≈ $29,045. That's the size. Not "how much can I afford," not "what feels right," not "what 3Commas suggests by default." The size that risks exactly $500 if the stop hits.

Fixed percentage vs fixed dollar

Two camps, both wrong sometimes.

Fixed percentage means every trade risks the same % of current account equity. After a win streak, your position sizes grow. After a drawdown, they shrink. The advantage: mathematically tied to your current state, so you naturally bet more when ahead and less when behind. The disadvantage: a deep drawdown makes your sizing almost un-tradeable, which can cause traders to override the formula.

Fixed dollar means every trade risks the same dollar amount regardless of account size. Useful in two scenarios: (1) you've hit a personal drawdown threshold and want to lock position sizes down while you reset; (2) you're scaling up a sub-account and don't want one bad week to wipe a meaningful chunk of capital.

Most professionals use fixed percentage as default and switch to fixed dollar when they're already bleeding and need to prevent tilt-sizing.

Calculating from stop distance

The formula above already does this, but the practical lesson is what most traders miss: the stop sets the size, not the other way around.

If you want to risk $500 on BTC and your stop is $1,000 away, you can take 0.5 BTC. If your stop is $200 away, you can take 2.5 BTC. Wider stop, smaller position. Tighter stop, bigger position.

This is why low-timeframe entries with tight stops can be sized up, and weekly swing trades with multi-thousand-dollar stops must be sized down. If you're placing 0.1 BTC on every trade regardless of where your stop sits, that's why your risk is wildly inconsistent across setups.

Worked example, two BTC entries on the same $50K account at 1% risk:

  • Trade A: entry $83,945, stop $83,500 (tight scalp). Per-unit risk $445. Size = $500 ÷ $445 ≈ 1.12 BTC.
  • Trade B: entry $83,945, stop $80,000 (wide swing). Per-unit risk $3,945. Size = $500 ÷ $3,945 ≈ 0.127 BTC.

Same account, same risk %, wildly different position sizes. That's the math working correctly. The stop dictates the size, every time.

What leverage actually does to sizing

Leverage doesn't change your risk. It changes your margin requirement.

Going back to the BTC example: 0.346 BTC at $83,945 has a notional of about $29,045. Spot that and you need $29,045 of capital. On 5x leverage, you need about $5,809 of margin. The $500 risk to your stop is identical in both cases.

Where leverage kills people is the inverse mistake. They decide "I'm risking 1%" and then choose a tight stop because they're using leverage, not realizing the liquidation risk changes. On 5x, a 20% adverse move is a stop-out. On 20x, a 5% adverse move is liquidation, regardless of where your technical stop sits.

Practical rule: your stop distance should be wider than your liquidation distance at the leverage you're using. If it isn't, the platform liquidates you before your stop ever fires.

Take that 0.346 BTC position at 10x leverage. You'd post about $2,905 margin. A 10% adverse move — about $8,400 — wipes that margin and you're rekt before your $82,500 stop even registers. The trade had a plan. Leverage made sure you never got to execute it.

Scaling in and out

Scaling solves two problems: it reduces entry regret and it lets winners run.

Scaling in. Standard pattern: 1/3 size at initial entry, 1/3 at confirmation (breakout retest holds, structure confirms), 1/3 at extension. The first tranche tests your thesis. The second confirms it. The third rides the momentum.

You keep the same total risk budget ($500 in our example), split across three entries. If the first tranche gets stopped, you've lost $167 and you have dry powder for the next setup. If all three fill and the trade runs, you've added size into strength.

The mistake: scaling into losers to "average down." That's not scaling, that's martingale. Different game, almost always ends in liquidation.

Scaling out. Standard pattern: take 50% off at 1R (where R equals your initial per-unit risk), 25% at 2R, let 25% run with a trailing stop. This locks in gains while preserving exposure to continuation.

Worked example, your 0.346 BTC at 1% risk:

  • Entry $83,945, stop $82,500 (R = $1,445 per BTC)
  • 1R target = $85,390. Close 0.173 BTC, lock $250 of the original $500 risk as profit.
  • 2R target = $86,835. Close another 0.087 BTC.
  • Trail the final 0.086 BTC with stop moved up to $85,390.

You've banked a guaranteed profit on 75% of the position while keeping a runner. The mental game matters as much as the math: most traders close winners too early because the unrealized P&L feels too good to lose. Scaled exits make letting winners run mechanical, not willpower-dependent.

Portfolio allocation — the missing chapter

This is where most sizing guides stop, and where most accounts actually die.

You can run perfect 1% risk on every single trade. If you take five trades simultaneously at 2% each, your heat is 10%. If four of them go against you in the same session, you're down 8% before any one trade hits its stop. That's the point where most people start cutting winners to cover losers, or worse, adding to losers.

Portfolio heat = sum of (per-trade risk %) across all open positions.

The rule professionals use: total heat stays under 6-8% in normal conditions, under 4% in high-correlation environments.

Two BTC-correlated longs are not two trades. They're effectively one trade at 2x size. If you're long BTC and long ETH and both are beta-correlated, your heat calculation should treat them as a single position with combined risk. Same logic applies if you're running the same direction on every L1.

This is exactly where current conditions get tricky. The bullish lean across the majors — BTC pressing the $84,000-$84,400 resistance cluster with a 67/100 bullish confluence, ETH flashing a perfect 100/100 with a confirmed higher-high at $2,720.50 per BullSpot's market report — tempts you to load up. But BTC and ETH are correlated. ETH's clean chart isn't a free lunch if it just doubles your BTC beta.

Where to size differently right now: SOL. SuperTrend is bearish, structure is bearish with the $120.70 swing high still holding, and the long side is already crowded at 61.8% longs vs 38.2% shorts on OKX, per the same report. If you're tempted to chase the next leg up, your stop has to be tight and your size has to be small, because the squeeze risk is real. A crowded long with no fresh catalyst is a vulnerability, not a setup.

PAXG is the mirror image. Bearish across all timeframes, 4H RSI at 24.96, but that doesn't make it a buy. The trend is down. Bounces in bear trends have terrible R:R because stops have to be tight against momentum and the winners don't run. The discipline is to skip it or size it down to almost nothing.

The sizing checklist

Before any position, ask five questions:

  1. What's my account risk % for this trade? (Default 1%, lower if heat is high.)
  2. Where's my stop, in absolute distance from entry?
  3. What position size does the math give me?
  4. What's my total portfolio heat if this trade loses?
  5. Is this trade correlated with what I already hold?

If the answer to question 4 puts heat above 8%, size down or skip. If the answer to question 5 is "yes, very," treat the combined positions as one trade and risk accordingly. Five BTC-correlated longs at 1% each is not five independent trades. It's one trade at 5x size.

What a real book looks like

Concrete portfolio, $50K account, current conditions:

  • Position 1: BTC long, $84,000 entry, $82,500 stop. Risk 1% ($500). Notional ~$28,000. Heat: 1%.
  • Position 2: ETH long, $2,700 entry, $2,640 stop. Risk 1% ($500). Notional ~$22,500. Heat: 1%.
  • Position 3: SOL long or skip? Crowded longs, bearish structure. If taken, 0.5% risk ($250). With a $115 stop, size comes to about 2.17 SOL. Heat: 0.5%.

Total heat: 2.5% across three name-different but beta-correlated trades. That's the realistic lower bound because BTC and ETH share directional exposure. If a BNB or altcoin breakout fires as a fourth setup, you don't get to take it at full 1%. You get 0.5% or less, because the book's already holding correlated risk.

That's the discipline most traders skip. They treat every entry as a fresh slate. The math doesn't care. The math knows how many positions you already hold.

Takeaways

  • The formula is Account Risk $ ÷ Per-Unit Risk = Position Size. Memorize it.
  • The stop sets the size, not the other way around. Same risk %, different stops = different sizes.
  • Leverage changes margin, not risk. Pick leverage that doesn't liquidate you before your stop fires.
  • Scale in to reduce entry regret. Scale out to bank profits mechanically. Don't average into losers.
  • Portfolio heat is the missing variable. Total open-position risk under 6-8% in normal conditions, lower when correlated.
  • BTC and ETH are correlated. Stacking full-size on both is an effective 2x BTC bet, not two separate positions.

The fifth trade problem isn't about finding the fifth trade. It's about having the discipline to size it correctly when your book's already warm.

Source context: BullSpot report from 2026-09-29T07:58:14.092Z (Fresh report: generated this cycle).