Gerald Appel
The MACD crossover: momentum turning through its own average
A moving-average crossover waits for price to overtake an average. MACD asks an earlier question: is the gap between a fast average and a slow one growing or shrinking? When that gap turns up through its own smoothing, momentum has shifted before the averages themselves cross. Gerald Appel's original method is exactly this one rule — the histogram and the divergence readings that now crowd every charting package were added later, by others.
- Style
- Momentum
- Approach
- Mechanical
- Difficulty
- Beginner
- Horizon
- Swing (days to weeks)
- Holding period
- Days to weeks
- Time needed
- 10 minutes a day
- Markets
- Stocks · Index futures · FX · Crypto
- Source
- The Moving Average Convergence-Divergence Trading Method (1979) — Gerald Appel
The rule set
- Compute MACD as the 12-period EMA minus the 26-period EMA of the close
- The signal line is the 9-period EMA of that difference
- Buy when the MACD line crosses above the signal line
- Sell when the MACD line crosses back below it
- Hold a 2×ATR stop underneath, so a single bad swing cannot run
What makes it distinctive
- Faster than a raw moving-average cross, because it reads the distance between two EMAs rather than price itself
- One indicator and one rule, available in every charting package ever written
- Works on any timeframe — which is also precisely how it gets misused
When it works
Markets with clean momentum swings lasting several weeks — trending stocks and index futures, where the gap between the fast and slow average widens and narrows in readable waves.
When it fails
Sideways markets produce cross after cross, each one a small loss. And because the whole construction is averages of averages, tops and bottoms are always signalled late — the speed advantage is relative to other crossovers, not to the market.
How a decision moves through it
Input
Daily bars
About a year of daily closes for the instrument. Appel developed the method on daily charts, and daily is where the 12/26/9 defaults were tuned.
Measure
MACD(12/26/9)
Two EMAs of the close, 12-period and 26-period; their difference is the MACD line. A 9-period EMA of that difference is the signal line. Alongside it, a 14-period ATR is computed for the stop.
Decide
MACD crosses above its signal line
The momentum measure turns up through its own average. This is the entire entry condition — there is no volume test, no pattern, no second timeframe in the original method.
Decide
MACD crosses back below the signal line
The same rule mirrored is the planned exit. Symmetry is a feature: the trade ends the moment the condition that justified it stops being true.
Size & protect
A stop at 2×ATR
The crossover exit can be slow when a trade goes wrong quickly. The ATR stop is the backstop for that case — it caps the single bad swing the signal line would ride too far.
What the crossing actually measures
MACD is not a mysterious oscillator. It is a subtraction: the 12-period average of price minus the 26-period average. When price accelerates upward, the fast average pulls away from the slow one and the difference grows; when the move tires, the gap narrows before either average has turned. MACD is that gap, plotted as its own line.
The signal-line cross asks whether that gap is now above or below its own recent average — whether momentum is expanding or contracting relative to its recent self. That question resolves earlier than 'has price crossed an average', which is the whole speed advantage.
Appel's method is one rule, not a dashboard
The indicator now ships with a histogram, divergence lore, and zero-line studies, and it is easy to assume the method uses all of them. It does not. Appel's original use is the signal-line crossover alone: cross up, buy; cross down, sell. The additions came later and belong to other methods.
That restraint matters for testability. A rule set with one condition can be checked against history and either survives or does not. A dashboard of five MACD readings, consulted selectively, can justify almost any trade after the fact — which is why the extras make the indicator more popular and the method less falsifiable.
The cost is paid in ranges
A rule this simple has no way to know what kind of market it is in. In a range, the gap between the two averages oscillates around zero and crosses its smoothing constantly — each cross a signal, each signal a small loss. The system's failure mode is not dramatic; it is a long file of minor debits.
Well under half the trades win. The system's arithmetic depends on the few multi-week swings paying for the many range losses — which makes the failure page and the sizing page the two that decide whether anyone can actually run it.
Five ways into this system
- The construction and the crossing: 12/26/9, and the one rule read from itTwo subtractions and a smoothing, then one rule applied in both directions. The parameters are historical conventions, and there is a reason to leave them alone.7 min read
- Sizing a system where well under half the trades winThe crossover tells you when; it says nothing about how much. The stop distance decides the size, and the win-rate arithmetic decides why that discipline is not optional.6 min read
- Ten minutes after the close, on markets that actually swingThe system needs momentum swings that last weeks, a daily check after the close, and the restraint not to run it on a five-minute chart just because the software allows it.5 min read
- The range, the lag, and why filters do not rescue itOne structural failure it announces in advance, one it cannot announce by construction, and an honest account of the standard patch.7 min read
- MACD for beginners: building 12/26/9 from scratchNo jargon assumed. Start with what an average of prices is, build the indicator one subtraction at a time, and end with the single rule this system trades.7 min read
The ideas behind it
This system assumes you already know these. Each one is explained from scratch in Investing 101.
Compare with
- SMA 5/10 Adaptive CrossoverBuy when the 5-day average crosses above the 10-day and sell when it crosses back — but after a choppy stretch, make the buy signal prove itself for a few bars first.
- ADX + DI Directional MovementGo long when the positive directional line crosses above the negative one — but only while ADX, a separate reading, says the market is trending at all.
These are documented methods described for study. Nothing here is investment advice, a recommendation, or a claim about future returns — every system on this page has losing periods, and the pages say where.
Reading about a system is not having one.
Plutux is where you write your own rules down, test them against real data, and keep the record your memory would otherwise rewrite. Join the waitlist for early access.