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 (Gol...