Copy trading risks, and what limits each one

1. The strategy loses

The obvious one. Every strategy has drawdowns; the question is whether they are bounded. Look for a documented worst drawdown and the controls that produced it. Ours: −18.8% in backtest, from volatility-targeted sizing, a 25% cap per asset and an exchange-side stop on every position. Guide 04 lists the controls.

2. The lead changes behaviour

A lead with a good record and a growing follower base has every incentive to raise leverage for a bigger profit share. A discretionary trader can do that overnight. A systematic strategy cannot without a code change, and a public one leaves a trail: the exposure figure on the track record, and a journal entry for any rule change. Prefer leads whose exposure is published and whose rules are fixed.

3. Copy drift

Orders below the exchange minimum are skipped, and your position sizes diverge from the lead's. At $100 on Binance, 68% of our book's orders could not be placed; at $1,000, 7%. Control: copy with enough capital, and use fixed ratio rather than fixed amount.

4. A copy stop-loss set too tight

Binance lets you cap the loss on the whole copy. Set at 10% on a strategy whose normal drawdowns reach 15–19%, it fires in the middle of a drawdown, closes every position at the low and ends the copy. Set it wider than the documented worst drawdown, or not at all, and rely on the lead's own stops.

5. Fixed-amount sizing

Fixed amount puts the same sum into every trade, overriding the lead's sizing. A position the lead sized at 3% of equity because the asset is volatile becomes as large as one sized at 12%. On a volatility-targeted book this multiplies risk in exactly the wrong places.

6. Funding and fees the record does not show

Your copy pays funding on every position and fees on every order, at your account's rates, whatever the leaderboard shows. For a leveraged long-only lead, funding alone ran about 8.8% a year on average since 2021 and far more during 2021 itself. Ask whether the lead's return is net of funding; ours is, and the cost is published (Note 01).

7. Platform and counterparty

On Binance your capital sits with the exchange; on Hyperliquid, in a smart-contract vault. Exchanges have failed, frozen withdrawals and changed rules. A vault leader cannot withdraw depositors' funds, but the protocol can have bugs. Control: size the allocation as money that could be inaccessible for a period, and do not hold everything at one venue.

8. Edited records

Leaderboards show ROI over short windows, sometimes computed on margin rather than equity, and a lead can retire a losing account and start a fresh one. Controls: a record whose start date has never moved, visible losses, and, once live, a link to the exchange's own page for the account. Guide 06 has the full list of red flags.

9. The copier

Stopping in a drawdown and restarting after a recovery is the single most reliable way to lose money in a profitable strategy. In our backtest, 52% of weeks were negative in a book compounding at 41.8% a year. Control: size the allocation to a 20% drawdown you will not react to, decide once, and judge at twelve months, the shortest window over which the record has been consistently positive.

What a copier controls

Of the nine, the copier chooses six: the lead (1, 2, 8), the amount and mode (3, 5), the stop (4), and their own behaviour (9). The platform accounts for one (7), and the costs (6) are known in advance. Copy trading is not passive; it is one decision made carefully, then not revisited.