Evaluating the New MACD and 9/21 EMA Confluence Trading Signals

Before anyone treats this as a signal worth chasing, the math behind the filter deserves the same scrutiny you'd give any backtest — and the headline number tells only part of the story.

What the rule actually does

The entry requires two things on the same daily close: the 9-day EMA crossing above the 21-day EMA, and the MACD line crossing above its signal line. The exit fires the moment either indicator flips bearish. The universe is narrow — NASDAQ, NYSE, NYSE ARCA, and AMEX common stocks with market caps above $10 billion and daily volume above 500,000 shares — with five years of daily OHLC as the lookback window.

That universe filter matters more than the confluence itself. Screening a $10B-plus, high-liquidity universe removes the slippage and gap risk that distort most retail backtests. What you're looking at is closer to an executable signal than a theoretical one — the kind of setup a systematic book can actually hold without bleeding on fills.

The payoff ratio, honest version

5.41% average run-up against 3.02% average drawdown is roughly 1.8 to 1, a defined-exit trend-following profile with positive expectancy on paper. That's a working signal, not a magic bullet. The edge lives in the exit logic, not the entry — the moment one of the two indicators turns, you're flat, and that's what caps the drawdown at 3.02% instead of letting a losing trend breathe.

What this release doesn't show is the standalone performance of MACD alone or the 9/21 EMA alone over the same universe and lookback. Without that comparison, the incremental value of stacking the two is an open question. I'd want the delta before sizing into CNC or any peer that matched today. Sample size on a single ticker firing one signal is noise, not evidence.

The tape these signals are firing into

Nasdaq Dorsey Wright's August 31 update noted the Russell 2000 pulling back toward the middle of its 10-week trading band after a decline of more than 1%, while the Nasdaq Composite reversed back into Xs and maintained its buy signal. That's a mean-reverting tape with an intact index-level long bias — not a runaway melt-up, not a breakdown. In that kind of environment, confluence signals on liquid large-caps have historically had more room to run than they would in a vertical trend, because volatility provides the swings the system is built to capture.

If this goes on a watchlist, size against the 3.02% drawdown, not the 5.41% run-up. The drawdown is the honest risk budget before the exit logic fires. The run-up is what you hope for, not what you plan around.