How Many Crypto Signals Should You Take at Once? Portfolio Heat and Correlated Risk

You size one trade perfectly. Stop loss placed, risk set to 1% of the account. Then a second signal fires, then a third. Each one is "only 1%," so surely three of them is fine? On a calm day, yes. On a bad day, those three positions can behave like one large bet, and that is exactly when accounts blow up. The number that matters is not risk per trade. It is total portfolio heat: how much of your account is exposed at the same time, adjusted for the fact that your positions are not independent.

What "heat" actually means

Portfolio heat is the sum of what you would lose right now if every open position hit its stop loss simultaneously. If you have three trades each risking 1% to their stops, your raw heat is 3%. That is the honest ceiling on a single moment of pain, and it is far more useful than staring at unrealized PnL.

Getting per-trade risk right is step one, and it is a topic on its own: see how to calculate position size with the risk-to-stop method. Heat is simply step two, done across every open trade at once.

Why BTC, ETH and SOL are not three separate bets

Here is the trap. In crypto, the majors move together. When BTC dumps 4% on a macro headline, ETH and SOL rarely sit still, and they usually fall further because they are higher-beta. So three "diversified" long signals on BTC, ETH and SOL are mostly one directional bet on crypto, cloned three times.

Correlation is not a fixed number, but on stressed days the majors routinely move in the same direction. That means your real worst-case is not softened by diversification the way it would be with genuinely unrelated assets. If anything, three same-direction alt longs can hit their stops together on the exact candle you feared.

A quick worked example. Say each position risks 1% to its stop:

• Best case (uncorrelated): losses partly offset, effective heat feels like roughly 1.7% (the square-root-of-3 intuition for independent bets).
• Reality on a bad day (highly correlated, same direction): the stops cascade together and you take the full 3% in one move, sometimes worse if slippage and a fast market push fills past your stop.

You do not get to choose which day you get. So plan for the correlated case, because that is the day that actually threatens the account.

A simple heat cap you can run in your head

You do not need a covariance matrix. Use two rules:

1. Total heat cap. Never let combined risk-to-stop across all open positions exceed a fixed number. A common, survivable choice is 2% to 3% of the account. If new signals would push you past it, you skip them or size down.
2. Correlation grouping. Treat highly correlated, same-direction positions as one "cluster" and cap the cluster. BTC-long + ETH-long + SOL-long is one cluster; a BTC-long and an ETH-short partially hedge and are not.

Worked version. Account of $10,000, heat cap of 3% ($300):

• Signal 1: BTC long, stop implies $150 loss. Heat used: $150. Fine.
• Signal 2: ETH long, stop implies $150 loss. Heat now $300. You are at the cap.
• Signal 3: SOL long fires. Taking it at full size pushes heat to $450 (4.5%) in a single correlated cluster. Skip it, or halve all three so the cluster fits inside $300.

That one decision is the difference between a manageable 3% down-day and a 4.5%+ gut-punch that tempts you into revenge trading.

Leverage does not change your heat, but it changes your survival

A subtle point that trips people up: your heat is defined by your stop distance and position size, not your leverage. Risking $150 to a stop is $150 whether you use 3x or 20x. What leverage changes is your liquidation buffer. Crank leverage high enough and the exchange can force-close you on a wick before your stop ever fills, blowing your neat heat math apart.

Before you stack correlated positions, sanity-check that none of them sit close to liquidation. The liquidation calculator shows how much room you have, and the risk/reward calculator confirms each stop is where you think it is.

How many signals is the right answer, then?

There is no magic count. The right number is however many positions fit inside your total heat cap once you have grouped correlated trades. For most retail accounts running a 2-3% cap at 1% risk per trade, that lands at two to three concurrent positions in the same direction during correlated conditions, and a few more only when trades genuinely offset (different directions, different drivers).

Two more honest guardrails:

• Watch shared catalysts. Funding flips, a CPI print, or a large liquidation event hits every major at once. If you already hold a cluster into a known event, that is not the moment to add. The funding-rate tracker helps you see when the whole market is leaning the same way.
• Measure it after the fact. Over a month, are your correlated clusters actually helping? Feed your closed trades into the profit-factor calculator and check whether stacking signals improves your edge or just amplifies variance. Often it is the latter.

The point of a cap

A heat cap is not about maximizing returns. It is about guaranteeing that no single day, no matter how ugly, can take you out of the game. Signals are only useful if you are still trading next month to act on them. We publish every signal, winners and losers, on our public track record precisely because the losing days are real and you need to size for them.

Take the signals you want. Just make sure that when BTC, ETH and SOL all turn against you on the same candle, the total damage is a number you chose in advance, not one the market chose for you.