Raisin Capital Solutions is a curated marketplace. Submit your algorithm, pass our evaluation, and reach investors — while your intellectual property stays yours, protected from the first line of code.
Every requirement is laid out below. One button copies the full checklist as a ready-to-use prompt — paste it into your favourite AI assistant with your code to bring your algorithm up to standard before you submit.
Drawdown may run up to 30%, but only if the Calmar ratio holds above 1.0 — annual return must at least match the worst drawdown. Recovery factor of 2.0 or more: total profit at least double the deepest loss. We care more about surviving bad periods than peak returns.
At least a 2-year backtest, 100+ trades total, with at least 2 every month — no long dead stretches. Profit should build steadily across the whole period, not come from one lucky window. The edge must hold across every market condition — which is only possible with adaptive sizing and levels, never fixed pip values (see Honest Engineering).
Our advice: run your test over a period, then leave room for a 6-month out-of-sample (OOS) backtest — and be honest about whether it still holds. OOS is where a strategy proves it wasn’t curve-fit.
No lookahead: the strategy can’t use price data it wouldn’t have known at decision time. No repainting indicators that redraw past signals. Stops may only tighten, never widen. Maximum risk known and capped at all times. No martingale or loss-doubling — ever.
No fixed pip values. Hard-coded pip or point distances for stops, targets or sizing are prohibited — they silently break risk across instruments, brokers and market regimes. Levels and sizing must be derived from adaptive measures: ATR, % of range, realised volatility, or trend/momentum tools built to hold up in any condition. Build the edge around these robust measures, not around a brittle fixed number.
Risk stays constant across instruments — no accidental 10× exposure from price decimals or contract size. Handles broker symbol variants (XAUUSD, XAUUSD.m, XAUUSD_i). Respects each broker’s min/max lot, lot step, and stop distances. Spread and slippage guards skip trading when conditions are poor.
Key settings — risk, sessions, filters — exposed as adjustable inputs, with sensible defaults so it runs well out of the box. Input names in clear, plain English, not cryptic abbreviations.
You can add, for the user’s preferred approach, an input to skip months where the strategy tends to underperform or lose (e.g. September–October).
Clean, readable code: no dead code, redundancy, or logic flaws. Fully self-contained — no external DLLs or internet calls. No hard-coded account numbers. A clearly marked slot for our licence-key check. The work must be your own original code.
We’ll ask whether out-of-sample testing has been done. Live OOS is best, but backtest OOS counts. If no OOS exists, we require a 30- to 90-day live test — length depends on the strategy’s overall quality. The goal is to prove execution: correct sizing, controlled drawdown (it won’t crash), and expected trade frequency — not to prove performance. The weaker the Calmar ratio, the longer and more careful we’ll be. If you submit during a month the algo historically struggles (Sep/Oct), we want to watch it handle that live. Performance data refreshed every 6 months — stale listings may be paused.
What to send us (MetaTrader 5). Two exports from the same backtest: the full Strategy Tester report (right-click the Results tab → Report) for the trade list and summary statistics, and the tester graph as CSV (right-click the graph → Export as CSV) for the equity curve. Together they let us rebuild your equity curve, statistics and Monte Carlo on exactly your data. Backtests must use 100% real-tick modelling.
Reserved for the most consistent, robust strategies. We judge it on the linear-regression fit of the equity curve and its standard error — the straighter and tighter the return line, the stronger the strategy. Premium algos produce steady returns across every market condition and every year, not just in favourable regimes.
You set your subscription price; we handle billing, licensing and distribution.
Revenue split: 70% developer / 30% Raisin. Your share is paid directly at the point of sale — no waiting, no clawbacks. Our 30% funds what keeps your algorithm selling and protected: marketing and distribution, VPS-ready hosting, ongoing code hardening and QA, and active defence of your intellectual property. You keep the majority and full ownership; we handle everything around making it earn.
Before you send anything, we sign a Non-Disclosure Agreement in your favour. Your algorithm stays your intellectual property. We evaluate it only to decide on commercialisation — and if we don't list it, we destroy every copy on our end.
The NDA is a plain-language template. It is not legal advice; have it reviewed by a qualified lawyer before relying on it.
Signed the NDA and ready to be evaluated? Send us your algorithm and reports. We'll review it against the checklist and get back to you.
Submit an algorithmSends to contact@raisincapitalsolutions.com
