Crypto bot trading generates continuous data — every trade produces an entry price, exit price, result, and context. Most automated traders either ignore this data entirely (checking only account balance) or obsessively over-monitor (reacting to every individual losing trade). Both extremes lead to poor decisions. A structured monthly review bridges these extremes: reviewing monthly allows enough trades to accumulate for statistical patterns to emerge, without leaving serious performance issues unaddressed for too long. This guide provides a systematic monthly review framework covering trade-level analysis, equity curve health, metric tracking over time, and the key decision rules for when to investigate vs when to let the strategy continue through a normal drawdown. See also our trade journal guide for ongoing record-keeping practices.
Related: bot monitoring guide, backtesting guide, risk management.
Monthly Review Checklist (30 Minutes or Less)
- Export trade history: Download the current month's trades from DennTech's trade log
- Calculate core metrics: Win rate, average win/average loss, Profit Factor for the month
- Compare to backtest baseline: Are current month metrics within normal variation of the backtest's expected ranges?
- Equity curve check: Is the current drawdown within the backtest's historical Maximum Drawdown?
- Signal frequency check: Did the strategy generate the expected number of signals this month?
- Trade review: Spot-check 3–5 trades — did they execute correctly per strategy rules?
- Market regime assessment: Was this month trending or ranging? Does current performance match strategy's known behavior in that regime?
- Decision: Continue / Investigate / Pause (use decision rules below)
Key Metrics to Track Monthly
| Metric | What It Tells You | Warning Sign |
|---|---|---|
| Profit Factor (monthly) | Per-trade profitability this month | Below 0.7 for 2+ consecutive months |
| Win Rate | Percentage of winning trades | Drops 15%+ below backtest average |
| Avg Win / Avg Loss ratio | Risk-reward realized | Ratio degrades below 1.0 for 2 months |
| Signal frequency | Strategy generating expected entries | Zero signals or 3× expected signals (indicator misconfiguration) |
| Current Drawdown | Unrealized losses from peak | Exceeds backtest's Maximum Drawdown |
Distinguishing Temporary Drawdown from Strategy Failure
The hardest judgment call in automated trading: is this losing period temporary (stay the course) or a genuine strategy failure requiring intervention? Apply these decision rules:
- Continue the strategy: Current drawdown is within the backtest's Maximum Drawdown range AND Profit Factor for the month is above 0.7 AND signal frequency is normal AND you can identify a market regime reason for the current losses (e.g., choppy BTC month performing in line with the strategy's known weakness in choppy conditions)
- Investigate the strategy: Profit Factor below 0.7 for 2 consecutive months OR current drawdown exceeds the backtest's Maximum Drawdown by 20%+ OR signal frequency is dramatically off (0 signals or 5× expected signals)
- Pause and reset: Current drawdown exceeds the backtest's Maximum Drawdown by 50%+ (something has changed significantly) OR execution errors detected in trade log (wrong prices, unexpected order types)
Market Regime Context Assessment
Always assess whether current month performance is consistent with the strategy's known behavior in the current market regime:
- Trend-following strategies (EMA, MACD, Ichimoku) underperform in ranging, choppy markets — this is expected, not a failure signal
- Mean-reversion strategies (RSI, Stochastic) underperform during strong trending periods — expected
- Grid strategies underperform during sharp unidirectional breakouts — expected
Before declaring a strategy failure, confirm: is this type of market (trending/ranging/volatile) the type your strategy is expected to struggle in? See our timeframe guide.
Parameter Adjustment Rules
Do NOT adjust strategy parameters monthly based on recent results — this is reactive optimization ("curve fitting to the last month") that almost always degrades future performance. Parameter adjustments should only be made based on: (1) a genuine structural change in market microstructure (e.g., exchange fee change), (2) a confirmed indicator calculation error, or (3) a 6-month performance review showing statistically significant underperformance across multiple market regimes. Monthly reviews should only drive Continue / Investigate / Pause decisions, not parameter changes. All strategies available at strategies page. Documentation at DennTech docs. Compare editions at pricing page.
Frequently Asked Questions
- How many months of live data should I collect before deciding a strategy is underperforming?
- For strategies generating 5–15 trades per month (typical for Daily indicator strategies), you need at minimum 3–6 months of live data (15–90 trades) before making any parameter modification decisions — fewer trades have too high statistical variance to draw reliable conclusions. For strategies generating 20+ trades per month (4H strategies, Grid), 2–3 months may be sufficient. A practical rule: never modify a strategy based on fewer than 30 live trades. The monthly review's role for the first 3 months is to verify the strategy is executing correctly and staying within expected parameters — not yet to judge long-term performance. See our backtesting guide for expected ranges by strategy type. Compare editions at the pricing page.
- What should I do if my DennTech strategy's current drawdown exceeds the backtest's maximum drawdown?
- Exceeding the backtest's Maximum Drawdown requires immediate investigation — but not automatic panic-stopping. First, verify the backtest was conducted over a sufficiently long period including bear markets (a backtest using only a 6-month bull period will have an artificially low Maximum Drawdown that live trading will likely exceed). If the backtest was comprehensive, a drawdown exceeding the historical MDD suggests the current market condition is more adverse than anything in the backtest period. Actions: temporarily reduce position size to 50% to limit further drawdown accumulation, review recent trades for execution errors, assess whether the current market regime (high volatility bear market, flash crash, etc.) is genuinely outside the strategy's tested conditions. Only permanently pause after ruling out execution errors and confirming the excess drawdown is not a statistical outlier. See our Maximum Drawdown guide. Start at the pricing page.
- Should I monitor my crypto bot every day or is monthly review sufficient?
- A tiered monitoring approach balances oversight with avoiding over-reaction. Daily: verify the bot is running (systemd service active, no error alerts), confirm no unexpected positions are open, check that account balance is within expected range. No decision-making at daily level — just confirming operational health. Weekly: quick scan of the week's trades and current P&L — is it roughly in line with expectations? Flag anything unusual for investigation. Monthly: full structured review per this guide — all metrics, equity curve, regime assessment, Continue / Investigate / Pause decision. This tiered approach keeps you informed without the overreaction risk of daily performance evaluation. See our monitoring guide. Explore the live demo.
Management guides: Monthly Review (this guide), Monitoring, Trade Journal. Start at the pricing page.
A key input to your monthly review is the balance report — read the guide to reading bot balance reports.