Price Dispersion as a Smart Money Indicator

Before I get into the topic at hand, let me say I have not seen the following stock price data interpreted or studied like I am about to show you. As far as I am aware my approach is unique in that it is not overly complicated, can be generalized across a large cross section of asset class ETFs, and makes intuitive sense regarding market structure. 

Before I introduce the chart it is important that I clarify some definitions. 

What is Price Dispersion?

I'm sure this may have many meanings among market participants but for our purposes the Blackarbs definition of price dispersion is as follows:

Price Dispersion is an alternative measure of a security’s volatility. Specifically, it is used to track market participant’s agreement regarding a security’s value. Major value disagreements show up as spikes in the level of price dispersion. Equation: (bar.high - bar.low) / (bar.close)

In this study price dispersion is the stock's daily price range expressed as a percentage of the Adjusted Close price.

How Does Price dispersion work as a smart money indicator?

The interpretation is based on the following assumptions.

  • Assumption (1):  Smart money is big money. These are the major players, the whales if you will, who move markets when they make trading decisions. These players tend to hold medium to long term views on positions and as such are primarily concerned with value. They tend to be buyers when others are selling and sellers when others are buying. 
  • Assumption (2):  Daily range is inherently a measure of market participants agreement or disagreement regarding the price(value) of a security. If the daily range is increasing or relatively large there is value disagreement. If the daily range is decreasing or relatively small there is value agreement.

There are a few more assumptions that will make more sense after viewing the chart. I must warn you the chart may look complicated on the surface but I assure you, the interpretation is relatively simple after I explain the details. 

XLK - Technology Select Sector SPDR ETF L/63 Days

What is this chart? what is contained in the two subplots?

Above is a chart of XLK. I have numbered the two subplots respectively.  The plot marked (1) shows an exponentially weighted cumulative return over the last 63 days. On the secondary axis I have plotted the daily adjusted close price. The black horizontal line is the 0% return value. 

The plot marked (2) contains a barplot of the daily price dispersion. The black line is a threshold value calculated as the top quintile (top 20%) of dispersion values over the period studied. On the right hand axis or secondary axis, is an exponentially weighted moving average of the daily dispersion. The red dotted line is also a threshold value defining the top quintile (top 20%) of all EMA values over the period. 

The blue vertical lines represent the bars where price dispersion exceeded the threshold value. The blue verticals are plotted on both subplots and correspond to the same dates. Note: due to formatting issues, at times the blue vertical lines are not aligned perfectly however, they still represent the same dates as the dispersion subplot. 

What are the final Three assumptions used to interpret this plot?

  1. Assumption (3): Smart money traders create the largest value disagreements therefore spikes in price dispersion indicate areas of trading opportunity and/or significant support and resistance levels.
  2. Assumption (4): Single interspersed vertical lines are more often associated with position closing events (liquidation/profit taking/short covering). Clustered or consecutive vertical lines (>=2) are indicators of buying and/or an interim bottom, however this is not always true. More importantly, as the size of the vertical cluster grows the more likely a sustainable trend change is occurring.
  3. Assumption (5): Rising dispersion as measured by the 21 Day EMA indicates increased risk of declining(negative) returns. Declining dispersion is associated with increased probability of increasing(positive) returns. 

put it all together, what is the chart saying about xlk?

Examining the plot we can see there has been much disagreement over the ETF value during the 63 day period. Resistance is ~$43.50 which coincided with a clustered dispersion spike in late May.  Price trended negatively over the period until disagreement in the ~$41.50 range indicating an interim bottom formed during late June/early July. However the outlook moving forward is mixed with a negative bias. Cumulative returns over the period are slightly negative and the dispersion EMA trend is clearly elevated above the threshold value. 

Let's look at another one.


This is the same ETF over the last 126 days or roughly 6 trading months. We can see the similarities in structure which reinforces some of the interpretations made previously. I've poorly circle the clustered areas. Notice how they coincide with high conviction trend changes.

The first occurred late March and happened to form a significant bottom ~$41.25. This value was not retested until July. The second cluster occurred  in early May and also formed an interim bottom around $42. From there price advanced until the next cluster which formed a significant top ~$43.50 in late May. 

To reiterate the outlook for XLK is mixed. The most recent cluster triggered during August 11/12. Generally this is a bullish sign, however with price dispersion clearly elevated on two timeframes and cumulative rolling returns below zero I would have a bearish to neutral bias.

how is this intuitive to understand??

Big money moves markets. Big money is opportunistic and likely to get involved at advantageous prices. By measuring the disagreement over a security's value, as measured by price dispersion, we can identify significant areas of perceived value.  Everything else provided in the chart is simply used to help identify and contextualize these areas. 

Composite Sector ETF Valuation Report [6.15.2015]

Check out the updated IPython Notebook where I take a look at changes and trends in ETF valuations using the Implied Cost of Capital model. To learn more about the model and the methodology used see here and here

For reference here is a Table of Contents, but due to some technical issues the TOC is not working properly on the page. I'll keep working to fix it for the next issue.


Check out my updated IPython Notebook where I take a look at changes and trends in ETF valuations using the Implied Cost of Capital model. To learn more about the model and the methodology used see here and here

Composite Sector ETF Valuation updated [5.24.2015]

Composite Sector ETF Valuation updated [5.10.2015]

Check out my updated IPython Notebook where I take a look at changes and trends in ETF valuations using the Implied Cost of Capital model. To learn more about the model and the methodology used see here and here

Composite Sector ETF Valuation updated [5.10.2015]