Why you need a scorecard at all
By this point you have four drivers, a dozen data series, and no way to combine them. Without a structure you will default to whichever driver you read about most recently, which is a reliable way to be wrong with confidence.
A scorecard does not make the drivers agree. It makes the disagreement visible, which is more valuable.
Scoring each driver
Score each driver as -1, 0, or +1 for its effect on gold:
- Real rates. Falling real yields are bullish. Use the trend over the last month, not the level — gold responds to changes.
- Dollar. A weakening dollar is bullish — but only when it is a rates story, not a haven bid.
- Positioning. This is not directional, it is a modifier. Crowded longs reduce how much you should size, they do not flip the direction.
- Official demand. A slow regime input. It sets a bias over quarters, not a signal for this week.
Handling conflict: the important part
Real rates and the dollar are usually the same story wearing two hats. When they disagree, that is information, not noise. The standard interpretations:
- Real yields falling, dollar rising. Usually a non-US risk event. The dollar is being bid as a funding currency while global growth fears crush real yields. Gold often still rallies — the rates channel dominates — but expect volatility.
- Real yields rising, dollar falling. Often a US-specific reflation or a Fed that is behind the curve while the rest of the world tightens. Gold's response is unreliable here. Reduce size.
- Everything agrees. This is the only time to size up. Congruence across independent-ish drivers is rare and worth more than any single signal.
What the scorecard must not do
It must not produce a number you obey. Its output is an input to your sizing and your conviction, and its most valuable output is the confidence column: how many drivers agree. Four drivers agreeing is a very different trade from two drivers agreeing and two abstaining.
How to check this yourself
Take a sheet of paper and score the four drivers — real rates, the dollar, crisis demand, official buying — as positive, negative, or neutral for gold right now. Then count how many agree. The net direction is the easy part; the agreement count is what should set your size. Four drivers pointing the same way and four drivers split two-two are not the same trade, even when the net score is identical.