MACD Crossover

What it is

Trading the signal-line cross on the MACD indicator.

MACD subtracts a 26-period exponential moving average from a 12-period average and compares the result with a 9-period signal line. These historical inputs create lag. A crossover can persist or reverse; the calculation does not establish a win rate, profitability or how many false starts a rule filters.

Market conditions that matter

A sustained historical move and a sideways sample can produce different numbers of crossovers. That observation does not prove a universally favorable asset class or timeframe. A filter based on a longer average is another rule to test with the same data and costs, rather than an established improvement. For a concrete comparison, label a five-bar sequence of hypothetical MACD values and identify exactly which bar produces a crossover rather than drawing a line after seeing the outcome. Record whether the signal was calculated at bar close or while the bar was still forming; that difference can alter whether a crossing appears at all. Because the indicator is derived from closing prices, it contains no information about bid-ask spreads, execution priority or transaction fees.

Study context

Studying differences between moving averages, signal-line crossings and alternative exit assumptions.

Limitations

Lag and repeated crossings can create unfavorable hypothetical outcomes; no market regime guarantees performance.

The steps

• Specify the price series, interval and EMA calculation; 12, 26 and 9 are conventional settings, not validated defaults.
• Record each selected signal-line crossing without discarding later unfavorable cases.
• Choose fictional entry and exit assumptions before calculating results, including a case where price moves before entry.
• Compare costs and alternative rules in a worksheet. TradeHQ does not provide a MACD overlay or genuine short selling.

Worked example

Hypothetical worksheet, not an observed trade: NVDA MACD crosses up at $130 with stop $124 — held for 6 weeks to $165. Stop and target levels are conceptual assumptions, not resting TradeHQ orders; gaps and fees can change the result.

The numbers behind it

Hypothetical arithmetic: fixed realized 43% wins at 2.5R and 57% losses at 1R give 0.505R gross expectancy before costs. A planned reward-to-risk ratio is not the realized payoff. Losing-run probabilities depend on sample length, independence and the assumed win rate; a six-to-eight-loss run is not a universally expected outcome.

How it fails

• Reporting only favorable crossings while excluding costly reversals.
• Treating a planned reward-to-risk ratio as a realized payoff.
• Assuming histogram divergence establishes a reversal or a profitable trade.

Scope and source limitations

MACD results vary by timeframe, market and exit rule. Track payoff distribution and drawdown instead of assuming a fixed win-rate range.

Educational simulation only — not financial advice.