Focus on a green macro regime

Research · Druckenmiller macro score · Backtest

Stanley Druckenmiller has said it for decades: earnings move individual stocks, but liquidity moves the market. When money is easy and nothing is tightening, trends run further. When the dollar, rates and oil all climb together, earnings get crushed.

Breakout traders feel this without measuring it. Some months every base breaks out and keeps going. Other months the same setups fail the next day. We wanted to know if a simple macro score, built from Druckenmiller's own checklist, could tell those months apart.

We rebuilt the checklist as a monthly score, ran it on every breakout from 2008 to 2026, and tested it as a sizing dial. Here is what showed up in the data.

What the score measures

Every month starts with a score from 0 to 13, built only from data that was public before the first of the month:

Component	Points	What it checks
Financial conditions	0–5	3-month change in the dollar, the 10-year yield, oil, credit spreads (HYG vs IEF) and the S&P. Each one moving the wrong way costs points.
Excess liquidity	0–2	M2 money supply growth minus industrial production growth. Money growing faster than the economy spills into assets.
The Fed	0–3	A rate cut in the last 6 months scores 3. A recent hike scores 0. QE adds, QT subtracts.
Companies and sector cycle	3	Fixed at neutral. Reading earnings calls point-in-time wasn't possible, so they never move the score.

Two caps sit on top:

A red dashboard caps the score at 5. The dashboard is market internals (cyclicals vs defensives), the yield curve, credit and liquidity, each rated green, amber or red. One red, or three ambers, makes it red.
All three "death knells" cap it at 4. Those are the dollar, rates and oil all rising together, the setup Druckenmiller used to call the 2000 earnings collapse.

The maximum reachable score is 13. A score of 12 or 13 means easy financial conditions, liquidity growing ahead of the economy, a supportive Fed and nothing red on the dashboard. We call that a green month.

Lookahead bias is the trap in macro backtests, because economic data gets revised for years. M2 and industrial production come from ALFRED vintages: each month uses the numbers exactly as the Fed published them on that date, not as revised later. Prices and yields use closes before the first of the month.

The setup

Same momentum breakout trades as our other backtests: 1,837 trades from 2008 to 2026. Profit is measured in R, where 1R is the amount risked. Each trade gets the macro score of the month it was entered in.

Portfolio rules for every equity curve: $10,000 start in January 2008, compounding, risk as a % of closed-trade equity, and at most 10 open positions.

PnL by macro score

Macro score at entry	Trades	Avg R	Profit factor	Share of PnL
4–5 (red, capped)	648	+0.48	1.61	38%

••••••••	•••	•••	•••	•••
••••••••	•••	•••	•••	•••
••••••••	•••	•••	•••	•••

████████ ██████ ████ ███████████ ████ ██████████ ███████ ████████ ██████ ████████ ██████ ████ ███████████ ████ ██████████ ███████ ████████ ██████

████████ ████

████████ ██████ ████ ███████████ ████ ██████████ ███████ ████████ ██████ ████████ ██████ ████ ███████████ ████ ██████████ ███████ ████████ ██████ ████████ ██████ ████ ███████████ ████ ██████████ ███████ ████████ ██████

Email me members-only research and market updates. Unsubscribe at any time.