Debugging Strategies

When a backtest produces unexpected results — no trades, wrong signals, or poor metrics — use these patterns to diagnose the issue.

No Trades Generated

The most common problem. Your strategy runs but places zero orders.

Check 1: Indicators returning NaN

If your warm-up is too short, indicators may be NaN for the entire run:

function onBar(ctx, i) {
  // Add temporary logging
  if (i < 5) {
    console.log('Bar', i, 'EMA:', ctx.ind.ema[i], 'RSI:', ctx.ind.rsi[i]);
  }
}

Fix: Increase warmupBars to at least max(all indicator periods) + 1.

Check 2: Conditions never met simultaneously

Multiple filters may be individually reasonable but never true at the same time:

// These might never all be true at once
if (rsi < 30 && close > ema200 && volume > avgVol * 3 && atr < threshold) {
  // Too many conditions
}

Fix: Test each condition independently. Remove filters one at a time to see which one blocks all entries.

Check 3: Position check prevents re-entry

// If you close and re-enter on the same bar, the position
// might not be 0 when you expect it to be
if (pos.qty === 0 && buySignal) {
  ctx.order.market('ASSET', 1, { signal: 'buy' });
}

Fix: Remember that orders fill on the next bar's open. Position state reflects fills, not pending orders.

Unexpected Entry/Exit Timing

Order Fill Timing

All market orders fill at the next bar's open, not the current bar's close:

Bar 100: Signal detected (close = $50,000)
Bar 101: Order fills (open = $50,150)  <-- actual entry price

This is by design — it prevents look-ahead bias. If you see entries at "wrong" prices, check that you're comparing against the correct bar.

Crossover Fires Once

q.crossOver() returns true only on the bar where the crossing happens, not while one series remains above the other:

// This fires once (correct)
if (q.crossOver(fastEma, slowEma, i)) { ... }

// This fires every bar while fast > slow (different behavior)
if (fastEma[i] > slowEma[i]) { ... }

Excessive Trading

If your strategy generates too many trades with high turnover:

Choppy Signal Filter

Add a minimum bars-between-trades cooldown:

function onBar(ctx, i) {
  const pos = ctx.position('ASSET');

  // Simple cooldown: don't re-enter within 5 bars of an exit
  if (pos.qty === 0 && pos.lastExitBar && i - pos.lastExitBar < 5) {
    return;
  }
}

Trend Filter

Only trade in the direction of the larger trend:

function init(ctx) {
  ctx.addIndicator('fastEma', 'EMA', { period: 10 });
  ctx.addIndicator('slowEma', 'EMA', { period: 50 });
  ctx.addIndicator('trendEma', 'EMA', { period: 200 });
}

function onBar(ctx, i) {
  const inUptrend = ctx.series.close[i] > ctx.ind.trendEma[i];

  // Only take long entries in an uptrend
  if (inUptrend && q.crossOver(ctx.ind.fastEma, ctx.ind.slowEma, i)) {
    ctx.order.market('ASSET', 1, { signal: 'buy' });
  }
}

Poor Performance Metrics

High Drawdown

Position sizing too aggressive — reduce quantity from 1 to 0.5 or less
No stop loss — add ATR-based or percentage-based stops
Trading against the trend — add a trend filter

Low Win Rate

Entry signals may be noisy — add confirmation filters
Stops too tight — widen stop distance (e.g. 2x ATR instead of 1x)
Wrong timeframe — some strategies work better on higher timeframes

High Fee Drag

Check total fees in the performance metrics
Default fee is 0.1% per side (0.2% round trip)
Tight stop-losses in choppy markets generate many small losing trades where fees dominate

Strategy Checklist

Before running a backtest, verify:

All indicators declared in init() with correct parameters
warmupBars >= max indicator period + 1
q.isNaN() guards on indicator values
Position check (pos.qty === 0) before entries
Both entry AND exit logic defined
Signal names are descriptive
Quantity is reasonable (0.1 to 1.0)

Warm-up Periods — Indicator initialization
Order Fill Timing — How orders execute
Performance Metrics — Understanding results