My alpha version of this measure disaggregates moment to moment upticks and downticks across the broad list of NYSE stocks. Unlike the standard NYSE TICK, which monitors upticks minus downticks across stocks, the new indicator measures upticking and downticking separately on a high-frequency basis and then assembles them into measures of buying and selling interest. From 2012 to the present, the resulting measure of Buying Power correlates with Selling Power by -.22. In other words, less than 5% of Buying Power can be explained by Selling Power and vice versa.
This is important because it suggests that buyers and sellers are different entities, behaving relatively independently. We can have days in which buyers and sellers are both active; days in which buyers are active and sellers are not; days in which sellers are active and buyers are not; and days in which both buyers and sellers are inactive. This differentiation opens the door to unique analyses, in which direction and volatility can be tracked as a function of the interplay between buyers and sellers.
For example, the market might move to new highs because of a surplus of Buying Power or because of an unusually low level of Selling Power. Those scenarios have led to different price paths over the near term. When we made a price high in the latter part of July, Selling Power was very low, but Buying Power was also well off its peak. The market held up, not because buyers were aggressive, but because sellers were on the sidelines.
Here's another interesting observation: Since 2012, when Selling Power has been in the top half of its distribution (high levels of selling pressure) over a four-day period, the next four days in SPY have averaged a gain of .45%. When Selling Power has been in the bottom half of its distribution (low levels of sellers) over a four-day period, the next four days in SPY have averaged a gain of only .04%.
Also since 2012, when Buying Power has been in the top quartile of its distribution (high levels of buying pressure) over a four-day period, the next four days in SPY have averaged a gain of .43%. All other occasions averaged a gain of only .18%.
Together, these findings invite the hypothesis that stock market strength can be attributed to two factors: momentum effects from high levels of buying and value/mean reversion effects from high levels of selling (capitulation). By independently parsing Buying and Selling Power, it may be possible to better anticipate momentum and reversal in the market.
Of course, much more investigation remains, especially over longer periods of market history. There are also interesting questions that remain to be addressed, such as Buying and Selling Pressure measures for other indexes and for individual stocks and ETFs.
My initial work in this area suggests that it is fertile ground for the development of new and valuable market measures. For instance, during the recent market decline, Selling Pressure hit a peak several days before we hit the ultimate low price in SPY; that is also when Buying Pressure bottomed. Price continued to make new lows, but the dynamics were already shifting away from Selling and toward Buying. With the market bounce on August 8th, Buying Power absolutely swamped Selling Power and the conditions were set for momentum continuation.
I will be updating this research--and its application--in future posts. It's a great example of how one of the best strategies for adapting to changing markets is to adapt quicker.
Further Reading: When is There Significant Buying and Selling in Stocks?
.
Posts
