Monday, March 26, 2007

Price-Volume Correlations: Assessing Stock Market Trends

My recent post looked at a promising stock (Manor Care; HCR) and found, among its virtues, a very positive dollar volume flow. What I noticed going back to 2004 was that, whenever there was significantly elevated volume in HCR, the stock also rose significantly. That told me that institutions--the ones who could account for such large volume--were accumulating the stock.

How can we measure such accumulation? I decided to create a version of my Power Measure, as described in my latest entry on the Trader Performance page, which consists of a 50-day moving correlation between daily price change and daily stock volume. If a stock is under accumulation, we should see positive values for this correlation; if it is under distribution, we should observe net negative values. A stock that is neither under net accumulation nor distribution (and hence probably in a trading range) should oscillate around the zero correlation level.

In the chart above, which goes back to the start of 2004 (N = 781 trading days), we can see stock price for HCR (dark blue line), the 50-day moving correlation between price and volume (pink line), and the horizontal red line at the zero correlation level. Notice that we spend far more time above the red line than below it. That tells us that HCR is a stock under accumulation. Elevations of volume are more apt to be associated with price rises than declines.

Interestingly, this price-volume correlation ends up also being a useful timing tool. Pullbacks in the correlation have corresponded to excellent times to purchase the stock. Specifically, when the correlation has been below -.10 (N = 118), the next 30 trading sessions in HCR have averaged a whopping gain of 5.65% (83 up, 35 down). When the correlation has exceeded .40 (N = 107), the next 30 trading sessions in HCR have averaged a loss of -1.86% (44 up, 63 down).

What this tells us is that it has been best for active traders to buy HCR when price-volume correlations have been negative and are now turning higher. This puts volume (and hence volatility) on the side of our trade. Similarly, once volume has already powered prices higher, it has been a good time to take profits.

In future posts, I will explore the price-volume correlations for other stocks and ETFs, including looks at different time frames. This could be a useful metric for comparing the "trendiness" of stocks: the degree to which volume/volatility is associated with directional movement.