**The price columns (open, high, low, close) are almost perfectly correlated with one another. Volume is essentially unrelated to any of them.** These are Pearson correlations across all 422,729 rows, pooled over every company and date. **Price columns (r ≈ 0.9998–0.9999)** - The strongest pairs are high–close (0.99989), low–open (0.99989) and open–high (0.99987). - Even the weakest pair, open–close, is 0.99976. - This is expected. A stock's open, high, low and close on one day are all close to its current price level. - Pooling 149 companies with very different price levels inflates these figures. The correlation mostly reflects differences in price level between companies, not day-to-day co-movement. The four price columns carry nearly the same information, so using them together in a model would be redundant (multicollinearity). **Volume vs. price (r ≈ −0.008)** - Volume's correlation with open, high, low and close is about −0.0081 to −0.0080, which is effectively zero. - In this pooled data, higher-priced stocks do not systematically trade more shares, and the reverse is also true. - The relationship would likely look different within a single company, or if you used returns or range instead of price levels. **Time and row order (weak)** - Date has a weak positive link with close (r ≈ 0.088) and almost none with volume (r ≈ 0.003). Prices drift upward over 2012–2025 in the pooled data, but time explains very little of the variation across companies. - A row-index-style column correlates at about −0.17 with close and −0.22 with volume. That probably reflects how the file is ordered, not a real relationship. **In short:** the price columns are one tightly linked group, and volume is an independent variable. For more useful signals, compute correlations within each company, or use derived measures such as daily returns, the high–low range relative to price, and volume spikes.