Using the Correlation Heatmap of Variables to Analyse Silver Price

 The above image shows two heatmaps (one with a 104 week rolling window  to capture the correlation and the other with a 26-week rolling window) of economic and financial  market variables that are known to be correlated with the price of Silver. By looking at the change in the degree of correlation of each variable in the 26-week heatmap from the 104-week heatmap, we can get a sense of the probable directon of each variable in the next few weeks.

Data & Method

Frequency: Weekly (W-FRI).

Transforms used before correlation:

• Silver, S&P 500, Gold, Brent, USD proxy: weekly log returns

• M2: weekly log change

• 10Y Treasury yield: weekly first difference (percentage-point change)

Heatmap coloring: Fisher z-transform z = atanh(r) applied to the correlation matrix for improved visual scaling near high |r|. Cell annotations show the raw Pearson correlation r for interpretability.

Tickers / Series: Yahoo Finance: SI=F (Silver), ^GSPC (S&P 500), GC=F (Gold), BZ=F (Brent), DX-Y.NYB (DXY).

FRED: M2SL (M2), DGS10 (10Y), DTWEXBGS (Broad Dollar Index; used only if DXY is unavailable).

Dashboard: Two windows are shown: 104-week (regime) and 26-week (recent regime).

Analysis

Look first at left 104-week panel then at right 26 week panel.

In general, currently the only significant correlation (R) with Silver is Gold price and R is expected to increase 0.71-->0.78

Although R is insignificant (<0.5) for the rest of the variables; we can see Brent will have R changing in direction either from positive to negative. 

Brent: 0.04--> -0.30 (change in direction)

SP500: 0.29 to 0.31

USD: -0.20 to -0.28

10yr yields: 0.13 to -0.22

M2: -0.19 to -0.26

Conclusion

Currently, the performance of Gold is crucial to the performance of Silver.