The D-Score is Darwinex Zero's 0–100 verdict on the quality of a DARWIN, not its raw profit. It is built from a set of investable attributes that reward experience, restraint and consistency over headline returns. The attributes below are described conceptually — the exact formula and weightings are proprietary and change, so confirm the current model with Darwinex.
Why quality, not return
Because VaR normalization already standardizes every DARWIN to the same risk band, raw return is a weak differentiator — magnitude is equalized. What remains is the shape and reliability of the return stream, and that is exactly what the D-Score is designed to measure. Two DARWINs can post the same annual figure while one earned it in a smooth, controlled line and the other in a jagged, lucky-looking scramble. The D-Score is meant to rank the first one higher, because it is the more investable asset.
The D-Score answers an investor's real question: not "how much did it make?" but "how likely is this to keep behaving the way it has?"
The investable attributes, conceptually
The score is a composite of several attributes. Think of them as different lenses on the same track record:
- Experience. A longer, richer track record with more independent trades carries more statistical weight than a short, lucky sample.
- Loss aversion. How well losses are contained relative to gains — controlled drawdowns and the absence of catastrophic single positions.
- Market and correlation. How the returns relate to the underlying markets and whether the edge looks like genuine skill rather than a passive beta bet.
- Return consistency. How positive returns diverge from negative ones — a track record whose winning periods dominate its losing ones in a stable, repeatable way.
- Duration consistency. Whether trade durations and holding behaviour are stable, rather than lurching between scalps and multi-week holds unpredictably.
- Performance and capacity. The strength of the risk-adjusted return and how much capital the strategy could realistically absorb without decaying.
No single attribute makes a DARWIN. A high return with erratic duration and thin experience can score poorly; a modest return with long experience, tight loss control and clean consistency can score well. Treat the D-Score as a weighted verdict, never a formula you can game one input at a time.
Return divergence, made concrete
Return consistency is often the attribute traders misjudge, so make it concrete. Imagine two DARWINs that both average roughly +0.8% per month:
- DARWIN A: months of +1.2%, +0.8%, +1.1%, +0.9%, −0.3%. Positive periods cluster tightly; the one negative month is shallow. Positive-versus-negative return divergence is strong and stable.
- DARWIN B: months of +6%, −4%, +5%, −4%, +1%. Same average, wildly different character. The winners and losers are large and interleaved; the divergence looks like noise, not edge.
Under VaR normalization both are pinned to the same risk band, so their averages converge — but A's return consistency and loss aversion score far higher. A is the DARWIN that attracts and keeps allocation. B looks like a coin flip that happened to land up.
Common mistakes
- Chasing raw return. Optimizing for a big month usually raises variance and wrecks the consistency and loss-aversion attributes, netting a lower score.
- Trading a thin sample. A short track record caps the experience attribute no matter how good the numbers look; there is no substitute for time and trade count.
- Erratic holding behaviour. Randomly switching between very short and very long trades dents duration consistency.
- One catastrophic position. A single outsized loss can dominate the loss-aversion attribute for a long time, even after many clean months.
- Treating the score as a formula. You cannot reverse-engineer exact weightings; build genuine quality instead of fitting to a guessed input.
Inspect the same qualities in your own curve
The good news is that the qualities the D-Score rewards — consistency, controlled drawdown, positive return divergence — are visible in your own equity curve long before a DARWIN exists. The Backtest Report Analyzer lets you turn a strategy report into exactly those metrics: paste your trades or an MT5 report and it returns profit factor, expectancy, payoff ratio, max drawdown and a return-versus-drawdown verdict. A high profit factor with a shallow drawdown and steady expectancy per trade is the same profile the D-Score is looking for. Use it to test whether your edge is consistent and well-controlled before you commit a track record to it.
Takeaway
The D-Score rewards the quality of a return stream, not its size: experience, loss aversion, market behaviour, return and duration consistency, and performance with capacity. Build a long, steady, well-controlled track record where positive returns reliably outweigh negative ones, and stop trying to reverse-engineer exact weightings. Confirm the current attribute model with Darwinex, and use your own equity metrics to check you are building the profile the score rewards.
