By designing profitable trading algorithms on Algorier and enabling the copy feature, you can earn tokens from users' trades and generate passive income. This guide explains the quality criteria your backtest must meet before your algorithm becomes eligible for publishing.
When you publish an algorithm on Algorier, other users can copy it and run it on their own accounts. Every time a trade is executed using your algorithm, you earn a share of the tokens consumed.
At Algorier, you can earn income by enabling others to copy your algorithm. For every trade executed using your published algorithm, you receive a percentage of the tokens paid by copy traders. You always have the option to redeem your earned tokens and withdraw the equivalent amount to your bank account.
For full details on how your earnings are calculated, including examples and an interactive income estimator, please visit our Pricing Page.
To ensure that only reliable and well-performing algorithms are made available for copying, your backtest must meet the following minimum criteria.
| Metric | Threshold |
|---|---|
| Sortino Ratio | > 1 |
| Profit to Max Drawdown Ratio | > 1.5 |
| Total Number of Trades | > 200 |
| Min (Average Win, Average Loss) | > 0.25% |
| Cancellation Rate (if limit orders are used) | < 10% |
| Time Limit on Limit Orders (if applicable) | < 1,000 seconds |
Below, we explain each criterion in plain language so you know exactly what to aim for when designing your algorithm.
The Sortino Ratio measures how well your algorithm generates returns relative to the downside risk it takes. Unlike the Sharpe Ratio, which penalizes all volatility equally, the Sortino Ratio focuses only on negative fluctuations โ the kind that actually hurt your account.
The formula: Sortino Ratio = Average Return รท Downside Deviation
Why it matters: A Sortino Ratio above 1 means your algorithm produces meaningfully more return than the downside risk it creates. This protects copy traders from strategies that look profitable on paper but carry hidden drawdown dangers.
In simple terms: If two algorithms both make 10% profit, but one has wild losing streaks while the other loses small and recovers quickly, the second will have a much higher Sortino Ratio. We require this metric to ensure your algorithm earns its profits in a stable, controlled manner.
This ratio compares the total profit your algorithm generated to the largest peak-to-trough decline (maximum drawdown) it experienced during the backtest period.
Why it matters: Max Drawdown represents the worst-case scenario a copy trader would have faced โ the biggest dip from a high point to a low point. A ratio above 1.5 means your algorithm's total profit is at least 1.5 times greater than the worst drawdown it endured. This gives copy traders confidence that the strategy can recover from its worst moments and still deliver solid returns.
In simple terms: If your algorithm's worst dip was $1,000, it must have earned at least $1,500 in total profit. The higher this ratio, the more resilient your strategy is.
The total number of trades executed by your algorithm during the backtest must exceed 200. This ensures that your backtest results are statistically meaningful rather than based on a handful of lucky trades.
Why it matters: An algorithm that has only produced 20 or 30 trades might look great by chance. With 200+ trades, the performance metrics become far more reliable, giving both you and the users who copy your algorithm a realistic picture of how the strategy behaves across different market conditions.
Both the average percentage profit per winning trade and the average percentage loss per losing trade must each be at least 0.25%. This percentage is calculated purely based on the difference between the entry price and the exit price of each trade โ it does not factor in capital size, leverage, position size, or any other account-level variables.
If you enter a trade at $100 and exit at $100.30, the percentage move is ($100.30 โ $100) รท $100 = 0.30%. This qualifies because it exceeds 0.25%, regardless of how much capital or leverage was used.
Why it matters: This ensures that your take-profit and stop-loss levels are set at a sufficient distance from the entry price. If these values are too small, wide spreads โ which can occur on some exchanges during low-liquidity hours โ may eat into your profits or trigger unexpected losses.
In simple terms: Your algorithm's trades must have enough "breathing room" between entry and exit points so that normal market fluctuations and spread variations don't erode performance.
If your algorithm uses limit orders, the ratio of canceled orders to total orders must be less than 10%. Canceled orders are limit orders that were not reached by the market price and were automatically canceled after the user-defined expiration time.
Why it matters: When your algorithm is published and copied by a large number of users, every order โ including ones that ultimately get canceled โ creates load on the platform. A high cancellation rate multiplied across hundreds or thousands of copy traders places significant strain on the system's infrastructure, making it difficult to maintain reliable execution for everyone. Keeping this rate low ensures the platform can scale your strategy smoothly to many users.
Instead of placing a buy order 1% below the current price, wait until the price drops 0.9% first, then place the order just 0.1% below that level. Both approaches target the same entry, but the second method significantly reduces the number of orders that never get filled.
If your algorithm places limit orders, the maximum expiration time for each order must be less than 1,000 seconds (approximately 16.5 minutes).
Why it matters: Limit orders that remain open for extended periods are more vulnerable to changing market conditions. By keeping the expiration window short, your algorithm ensures that orders are only executed when market conditions are still aligned with the original trade logic, resulting in more predictable outcomes for copy traders.
Your algorithm does not need to use limit orders exclusively. Strategies can use market orders, limit orders, or a combination of both. However, if your algorithm does include limit orders, the Cancellation Rate and Time Limit requirements above will apply to those orders specifically.
If your backtest doesn't pass one or more criteria, here are practical steps to improve your algorithm.
Focus on reducing the frequency and size of losing trades. Tighten your stop-loss logic, add filters that skip low-confidence setups, or avoid trading during highly volatile periods. The goal is to keep your returns steady while minimizing painful drawdowns.
Either increase your total profit by refining entry/exit conditions, or reduce your maximum drawdown by implementing better risk management โ such as position sizing rules, daily loss limits, or avoiding correlated trades that can amplify losses.
If your algorithm hasn't generated 200 trades yet, consider lowering the minimum signal threshold or testing across multiple markets. The key is to reach statistical significance without sacrificing strategy quality.
If your average win or loss is below 0.25%, adjust your take-profit and stop-loss levels to give trades more room. This protects against spread-related losses during low-liquidity conditions and produces more meaningful trade outcomes.
Place limit orders closer to the current market price. Instead of setting a limit order far from the market and waiting, configure your algorithm to wait until the price approaches your target level, then place the order with a small offset.
Build your strategy in AlgoBuild, run a backtest, and check that it meets the criteria above. Once it passes, you can publish it and start earning.
