How Veksi Evaluates Strategies
Understand the data inputs, scoring model, and review workflow behind every strategy surfaced on Veksi. Use this guide to interpret the score and decide how to incorporate candidates into your trading process.
What the Veksi score represents
Each strategy receives a composite score (0-100) that summarises current performance quality, stability, and data confidence. The score is recalculated daily.
- PF & profitabilityProfit factor, cumulative PnL, and daily profit carry the largest weight to reflect edge persistence.
- ConsistencyWin ratio, streak behaviour, and variance bands adjust the score when outcomes are volatile.
- Data depthTrade count, timeframe coverage, and data recency ensure forward-looking confidence. Strategies with limited history are capped.
- Quality gatesStrategies failing minimum thresholds (for example trade count or negative profit factor) are excluded or assigned review status.
Weights and gates may evolve as the model improves. Check the changelog for updates when score behaviour changes.
Data pipeline and versions
- V2 (simulated): Primary dataset for research. Each trade is generated by our backtesting engine with exchange-quality prices.
- V1 (realised): Historical live-executed trades. Useful for validating behaviour in production environments.
- Refresh cadence: Analytics tables update hourly with the latest closed trades. Score rollups recalc after each refresh.
- Validation: Automated checks detect missing candles, outlier PnL, and latency. Failing checks create a banner on the strategy detail page.
Collect raw trades from exchanges and backtests.
Screen for missing candles, outliers, latency, and data drift.
Calculate PF, PnL, streaks, factor tags for each strategy.
Update composite score and publish to Analytics.
Workflow: evaluate a candidate
- Filter for fit — Start on Analytics or the Strategy Index. Filter by pair, interval, and minimum trades that match your mandate.
- Check the score context — Read the score, daily profit, and profit factor together. Investigate dips or spikes in the timeline view.
- Open the detail view — Inspect recent trades, factor tags, and streak information for behavioural clues.
- Apply risk guardrails — Benchmark drawdowns against your limits. See the risk guide for position sizing worksheets.
- Journal the decision — Record why the strategy passed or failed your checklist and review weekly.
Set pair, interval, and minimum trades.
Review score, profit factor, and daily profit.
Inspect trades and factor tags.
Check exposure versus guardrails.
Document rationale and follow-up tasks.
Key metrics reference
- Score: 0-100 composite for fast sorting. Use alongside the full metrics to avoid overreliance.
- Profit factor: Gross profits divided by gross losses; above 1.3 often signals edge when sample size is sufficient.
- Daily profit: Average PnL per day in quote currency. Normalise against your capital before comparing strategies.
- Trades: Closed trades in the selected timeframe. Low counts require caution—consider extending the window.
- First/Second type: Factor tags describing the core idea (for example “Trend / EMA”). Use them to diversify playbooks.
See the glossary for full definitions and formulas behind each metric.
0-100 snapshot combining all factors.
Values above 1.3 suggest edge; confirm sample size.
Translate the average to your active capital allocation.
Fewer trades mean more variance; extend the timeframe.
Mix factor tags to avoid concentration risk.
Limitations and disclaimers
- Simulated trades do not account for slippage, fees, or liquidity constraints unless noted.
- Data refresh delays can occur during exchange maintenance windows. Use timestamps to confirm freshness.
- Veksi provides research only. You remain responsible for execution, compliance, and capital allocation decisions.
- Strategy code can evolve. Monitor changelog entries and revalidate after significant updates.