USD Impact Score

Weekly cross-asset dollar regime reading
Latest reading
-0.72
Regime
Soft dollar regime
Week-over-week change
-0.12
Data as of
2026-08-21
USD Impact Score — Eleven-Year Record

Automated Regime Commentary — Week of August 21, 2026

USD Impact Score: −0.72 | Soft dollar regime

This commentary is generated automatically from the same weekly score data used by the dashboard. It adds no external market-event claims and makes no forecast.

What the score is saying

The score is −0.72, with a week-over-week change of −0.12 and a four-week change of −0.28. The regime remains Soft dollar regime. Across the eight inputs, 3 contribute toward a firmer-dollar reading and 5 toward a softer-dollar reading.

What is driving the reading

- gold: z-score +2.99; score contribution −0.374, a softer-dollar contribution. - S&P 500: z-score +2.53; score contribution −0.316, a softer-dollar contribution. - U.S. 10-year Treasury yield: z-score +1.66; score contribution +0.208, a firmer-dollar contribution.

These are the three largest absolute contributions in the current calculation. Each component is standardized and receives the same fixed transmission weight every week.

What the score is not saying

The framework does not predict next week’s return for the dollar or any component asset. A positive reading describes a cross-asset configuration associated with firmer-dollar transmission; a negative reading describes one associated with softer-dollar transmission. It does not establish timing, causality or a trade.

What to watch over the coming week

Watch whether the dominant components retain their current standardized positions and whether the reading broadens across more inputs or narrows around one outlier. The nearest regime boundary is −1.00. The dashboard label changes only after a completed weekly observation moves through a boundary.

Methodology reminder

The USD Impact Score combines DXY, WTI, S&P 500, VIX, Bitcoin, gold and the U.S. 2-year and 10-year Treasury yields. Inputs are z-scored against the full sample, clipped at ±3.5 standard deviations and combined with fixed transmission-logic weights. The same calculation is used every week.