Research ·
How small can a copy account be? Binance minimum orders vs our trades
With a $100 copy, 68% of the bot's orders would be below Binance's minimum order size. At $500 it's 17%, at $1,000 7%, and from $2,500 almost none. Below about $1,000, your copy isn't really a copy.
Copy trading looks like it scales down to any amount: set a ratio and your account does what ours does, smaller. It doesn't quite work that way, because every exchange has a minimum order size, and a trend bot sends a lot of small orders.
Binance's minimums
On Binance USDⓈ-M futures, an order has to be worth at least a certain amount. On 29 September 2026 the minimums for our ten coins were:
| Coin | Minimum order value |
|---|---|
| BTC | 50 USDT |
| ETH, LINK | 20 USDT |
| BNB, XRP, SOL, DOGE, ADA, AVAX, DOT | 5 USDT |
Taken from Binance's public exchange information; Binance changes these occasionally.
How small the bot's orders are
A new position is typically about 10% of the account. But most orders aren't new positions. They're adjustments: making a position a little bigger or smaller as prices and volatility move. In the backtest of the live rules (3,944 orders from 2021 to 2026), the typical add was 2.7% of the account and the typical trim 3.7%. On a $100 account, that's $2.70: below every minimum on the list.
What a small copy can't do
We took every order from the backtest and checked, for accounts of different sizes, whether it would clear Binance's minimum for that coin:
| Copy account | All orders below minimum | New positions below minimum |
|---|---|---|
| $100 | 68% | 42% |
| $250 | 37% | 17% |
| $500 | 17% | 6% |
| $1,000 | 7% | 1% |
| $2,500 | 1% | 0% |
| $5,000 | 0% | 0% |
An order below the minimum can't be placed. Your copy skips it, and from then on your position is a different size from ours. A skipped trim leaves you with more risk than we have; a skipped add, with less. Skip enough of them and your results stop tracking ours in ways neither of us can predict. With $100, even two in five new positions would be skipped altogether.
Stop-losses aren't affected: they close the whole position, whatever its size.
So how much?
- Under $500: don't. Too many orders get skipped for the copy to mean anything.
- $1,000: the practical minimum. About one order in fourteen is skipped, almost all of them small adds and trims.
- $2,500 or more: close to a true copy.
That's the mechanical answer. The other answer is on the copy page: only copy with money you can watch fall 20% without needing it back. If $1,000 is more than that for you, the right amount is zero, and following the journal costs nothing.
The code is research/min_orders.py in the bot's repository.