NOONOO TRADINGStart in the bot

Portfolio Heat: Why an Account Can Fall Sharply Despite a 2% Risk Limit per Trade

Every trade has a stop, and the risk on each is set to 2% of the account. Yet the account falls more than 10% in one day. The rule was not broken; it was applied only to individual trades. What the account actually bears is the sum of everything currently open, not the risk of one trade.

Heat is the sum of the losses currently open

Portfolio heat is the amount the account would lose if every open position hit its stop. Rules that limit each trade to a percentage of the account are widely used, but that rule alone protects the account only when there is one position.

The calculation is simple. For each position, calculate (entry price − stop price) × quantity, add the amounts, and divide by the account balance.

$10,000 account · 2% risk-per-trade rule

Position A: −$200 to its stop
Position B: −$200 to its stop
Position C: −$200 to its stop
Position D: −$200 to its stop
Position E: −$200 to its stop

Heat = 1,000 ÷ 10,000 = 10%

No individual trade broke the rule,
but the account's simultaneous exposure is 10%, not 2%.

The objection that all five positions are unlikely to stop out together often fails. Crypto assets tend to move together during declines, so stop-outs tend to cluster. This is the issue discussed in the diversification trap.

How correlation increases effective risk

If five positions moved independently, the 10% above would represent the worst case, with the average much lower. The problem arises when they are effectively the same bet, such as five altcoin longs.

With complete independence, combined risk scales with the square root of the sum of squares rather than the simple sum. When correlation is close to 1, it approaches the simple sum.

Five positions, each at 2%

Fully independent, correlation 0 → √(5 × 2²) = about 4.5%
Fully synchronized, correlation 1 → 2 × 5 = 10%

Five altcoin longs are usually closer to the latter.
They look like five positions but resemble one 2% position enlarged fivefold.

Heat should therefore be grouped by direction and asset category, rather than simply by position count. Treating a BTC long, an ETH long, and an altcoin long as three separate bets understates their actual risk. The correlation article explains how assets come to move together.

Setting a numerical heat limit

A total heat limit below 5–6% of the account is often suggested, but it is better to work backward from the drawdown you can tolerate than to adopt someone else's number unchanged.

There are two inputs: your tolerable maximum drawdown and a losing-streak length within the normal range.

Working backward

Tolerable drawdown: 20%
Expected longest losing streak, with a 55% win rate over 200 trades: about 8 losses

Allowed risk per trade = 20% ÷ 8 = 2.5%

With 3 simultaneous positions, heat for one group
= 2.5% × 3 = 7.5% → number of full stop-out groups within a 20% drawdown = 20 ÷ 7.5 ≈ 2.7

At 7.5% heat, 3 consecutive complete stop-outs reach the limit.

The losing-streak length that counts as normal comes from calculation, not intuition. Looking at losing-streak probability together with risk of ruin makes the limit more concrete. The position size calculator is convenient for calculating individual position quantities.

Heat can be reduced rather than remaining fixed

One useful property of heat is that moving a stop can reduce it immediately. Even if heat is 6% at entry, once one position moves into profit and its stop is raised to breakeven, that position's contribution becomes zero.

$10,000 account · 3 positions at 2% each

Heat immediately after entry = 6% ($600)

A moves into profit → move its stop to breakeven
A's contribution changes from $200 to $0
Heat = 400 ÷ 10,000 = 4%

Remaining capacity = 6% limit − current 4% = 2%
→ Room becomes available for 1 new position.

This calculation becomes a baseline for new entries. Instead of deciding intuitively whether to take an attractive setup, check whether unused heat capacity remains. Without capacity, skip even a good setup. Setups return; an account may not.

Four ways heat is measured incorrectly

1. Counting a position without a stop as zero risk. Without a stop, its risk is not 2% but the entire amount down to liquidation. A stop that exists only in your mind cannot be included in the calculation.

2. Using the liquidation price instead of the stop price. Under cross margin, liquidation prices move with other positions' unrealized profit and loss. If one position deteriorates, the others' liquidation prices move closer too. This is why heat can grow particularly quickly with cross margin.

3. Ignoring gaps and slippage. Heat is calculated assuming execution exactly at the stop price, but a sharp decline can jump past that price. Actual losses then exceed the calculation. Leave headroom rather than using the entire limit.

4. Omitting holding costs. A group of positions concentrated in one direction also pays funding in the same direction. The account can shrink gradually even without reaching the stops.

Five practical checks

1. Calculate current heat before entering a new trade. Counting after entry is already too late.

2. Group positions in the same direction and asset category together. Five altcoin longs resemble one large position more than five independent ones.

3. Skip new setups if they would exceed the limit. Reducing size to squeeze in another position still increases heat.

4. A position without a stop cannot be included in a meaningful heat calculation. In that situation, the limit itself loses meaning.

5. Record days when you broke the limit, regardless of the outcome. Making money that day can be more dangerous: rewarding a broken rule encourages breaking it again. Record this in a trading journal.

Three key points

1. The account bears the sum of open stop-loss amounts—heat, not just the risk of one trade. Five positions at 2% make 10%.
2. Positions in the same direction are not diversified in proportion to their count. Several altcoin longs are effectively one large bet.
3. Heat is an amount that can decrease when stops move. Base new entries on remaining heat capacity, not merely the attractiveness of a setup.

Caution

The account sizes, risk percentages, losing-streak lengths, and correlation assumptions in this article are examples illustrating calculation methods, not measured results from a particular strategy. The square-root-of-squares calculation assuming zero correlation is closer to a theoretical lower bound. In real markets, correlations rise together during declines, so using the simple sum as the reference is safer. Leveraged trading can lose all principal. Investment decisions and responsibility remain yours.

NOONOO TRADING invites you to follow live trading in our free chat.

Start in the bot

📈 OKX trading fee discount for new registrations

Register for the OKX Fee Discount →