Are Exchange Trading Bots Halal? Grid, DCA and Copy

Open the trading bot tab on any large exchange and you get a menu: spot grid, futures grid, DCA, rebalancing, arbitrage, copy trading, martingale. They sit next to each other in the same interface with the same styling, which strongly implies they are variations of one thing.

They are not. Some of them are a scheduling tool. Some of them are leveraged derivatives with a friendly name.

Here is each one, sorted by verdict. The underlying doctrine is wakalah, agency, and I have set that out properly in the piece on AI trading agents. This article is about the specific products.

The clearly permissible ones

DCA bots. You are buying a fixed amount of a screened asset on a schedule. There is no leverage, no borrowing, no counterparty being harmed, and no uncertainty about what you are buying. It is a standing instruction to make a permissible purchase, which is exactly what a wakil is for.

Worth noting that this is not just our reading. Bybit's Shariah-certified Islamic Account, which blocks margin and perpetual funding at the account level, explicitly permits a DCA bot. That account carries certification from the Shariyah Review Bureau, whose position we cite in our scholar dossiers. When an advisory firm that certifies exchanges signs off on a product category, that is a stronger precedent than my opinion.

Spot grid bots. A grid bot places a ladder of buy orders below the current price and sell orders above it, and works the range as price oscillates. On spot, with no borrowed funds, every one of those is an ordinary purchase or sale of an asset you own. You are providing liquidity into a range and earning the spread, which is trade, not interest.

Bybit's Islamic Account permits the Spot Grid bot too. Note the word spot in the product name, because it is doing all the work.

The clearly impermissible ones

Futures grid bots. Identical logic to the spot version, applied to perpetual contracts. Now every rung of the ladder is a leveraged derivative position paying or receiving funding, which is riba, and the position can be liquidated. The bot did not make this permissible and it did not make it worse. The instrument was already the problem. Everything in our reasoning on Hyperliquid applies.

Martingale bots. These double the position size after each loss on the theory that a win recovers everything. Set aside the fiqh for a second: this is a strategy with unbounded downside and a mathematical guarantee of ruin given enough attempts, and it is usually sold with leverage. The structure is a wager on the next outcome, with stake escalation, which is the shape of maysir rather than of commerce. There is more on that line in the piece on maysir.

Arbitrage bots that extract from other traders' orders. Cross-exchange arbitrage on spot, buying cheap on one venue and selling higher on another, is legitimate trade and has been for as long as merchants have existed. What is not legitimate is the on-chain variety that watches pending transactions, buys in front of them, and sells into the resulting slippage. The profit there is precisely the amount by which someone else's trade was made worse, without their consent. That is appropriating a Muslim's wealth by deception, and the twelve-millisecond execution time does not soften it.

The two that depend on a setting

Copy trading. The mechanism is a valid wakalah: you appoint someone to trade on your behalf. The problem is that you have appointed them to do whatever they do, and you cannot see it in advance.

Two conditions of a valid agency are that the mandate must be defined and that you cannot authorise what you may not do yourself. A copy trading arrangement where the lead trader uses leverage, shorts, or buys unscreened tokens is a mandate you were not entitled to give. Most leaderboard traders rank highly precisely because they use leverage, since that is what produces the returns that get them onto a leaderboard.

If you want to use it: filter for spot-only leads, check their actual position history rather than their percentage, and confirm the assets they trade against our screening list. If the platform does not let you see the instruments, the mandate is undefined and you should not appoint them. There is also a fee question, and the answer is that a flat subscription is ujrah while a profit share sits closer to mudarabah, with the condition that the manager cannot be guaranteed a return.

Rebalancing bots. Rebalancing a portfolio of screened spot assets back to target weights is clean, useful, and arguably a discipline the tradition would approve of. The failure mode is the asset list. A rebalancing bot pointed at a preset index basket will buy whatever is in the basket, which routinely includes Aave, Ondo, or a tokenized Treasury product. Build the basket yourself from assets you have screened, or do not use it. Our portfolio piece goes into how to construct the target weights.

The setting that governs all of it

One configuration decision matters more than every verdict above.

Scope the API key to spot trading, with margin and futures disabled at the account level and withdrawal permission off.

Do that and the entire impermissible half of this list becomes unreachable. The bot cannot open a futures grid because the account cannot open futures. You are not relying on your own restraint, on the bot's configuration, or on a vendor's promise. You have removed the authority rather than declined to use it, and in agency terms that is the difference between a defined mandate and a hope.

While you are in there, check whether idle balance is being swept into an Earn product between trades. Many bot frameworks do this by default because leaving capital uninvested looks wasteful, and it quietly generates interest on your account. That is the single most common way a compliant strategy produces non-compliant income, and if it has happened, purification is the remedy.

Two things I am not certain about

Grid trading on spot in an extremely thin market is the case I keep turning over. In a liquid book you are providing genuine liquidity and earning a spread for it. In a book with almost no natural volume, a grid bot can end up trading substantially against itself, generating volume that looks like real interest in an asset and is not. That edges toward the prohibition on najsh, creating a misleading price signal. I do not think an ordinary retail grid on a major pair is anywhere near this. I do think someone running grids on a token with no organic volume should think about what their volume is telling other people.

The second is high-frequency spot trading generally. Nothing in fiqh sets a minimum holding period, and rapid buying and selling of a thing you actually own and take possession of is trade. But there is a point where the holding period is short enough that the asset is a token in a game rather than property being exchanged, and I cannot tell you where that point is. Nobody can, currently, which is worth saying rather than papering over.

The products in the first two sections are not close calls. Start there.