**Van's Market Type Classification System**

Van's Market Type Classification System | |||||||||||||||||||||||||||||||||

In 2008 and 2009, I found that the way I was defining market types wasn’t working for me so well anymore. I’d based my definition on the idea that if you put a quarterly chart on the market, a bull and bear market should be obvious, and that if you couldn’t tell, the market was sideways. The problem was that this definition didn’t provide me with what I wanted—a way of determining market type on a weekly and mechanical basis. Consequently, I had to revise my definition. Guiding me in this effort were three important observations:
With those ideas in mind, I created a new way to define market types using the same two components I’ve used in the past:
Direction (or trend) can be measured in a number of ways, and numerous indicators help traders do so. I discovered a way that lets me use my SQN to help measure the quality of market movements. A trading system’s SQN score tells you how easy it will be to use position sizing strategies to reach your objectives. To calculate it, you need to have a sample of R-multiple results for a particular trading system.
Applying the methodology to measuring the market, however, is not intuitive; you cannot use R-multiples or calculate the expectancy of the market. Rather, I find the daily percentage change for the S&P 500 from close to close, which then becomes the input for the SQN calculation that considers the direction of the move, the degree of the move and the efficiency (smoothness) of the move. I watch the Market SQN scores for 25-, 50-, 100-, and 200-day periods, but I use the 100-day score to define market type.
Here are the ranges for the Market SQN scores that I came up with for market direction:
To come up with these ranges, I looked at the index charts and I also calculated frequency distributions for four SQNs, for the average percent change for days within each category, and for the average percent change for the following day.
For every Market SQN period grouping, the average percent change got smaller (or more negative) as we went from strong bull to strong bear. Clearly, we did a fairly good job of distinguishing the categories. The percent change for the days after each market direction day also looked good. Except for the 50-day SQN, the average percent change the day after progressively decreased (or went negative) as we moved from strongly bullish to bearish.
I also found another interesting phenomenon: when the market became strongly bearish, we had the largest percent changes the day after. In fact, except for the 25-day Market SQN, the largest day-after average daily percent change seemed to occur when the market type was strongly bear—amounting to over 0.5%.
Now, let’s look at volatility.
We measure volatility by the ATR% against data from today back to the mid-1960s. To calculate this, I take the 20-day ATR and divide that by the closing price for the index each day. This keeps the comparison for the volatility measurement relative, whether the index is at 600 or 1400.
Again, I used data going back about 50 years and found out that the mean ATR % doesn’t change that much from decade to decade. The overall mean ATR % is 1.3, with a standard deviation of 0.72.
After looking at the data, I decided to sort the market types into four volatility categories based on the statistics for the ATR%: - Normal: The average ATR% plus or minus 0.5 standard deviations.
- Quiet: Anything less than a 0.5 standard deviation less than the mean.
- Volatile: between 0.5 standard deviations and 3 standard deviations above the mean.
- Very Volatile: Anything more than three standard deviations above the mean.
Here are the ranges for each of those categories:
Very volatile markets are rare, occurring only about 1% of the time; volatile markets occur about 25% of the time; normal markets occur almost 40% of the time; and quiet markets occur about 35% of the time. These volatility categories seem to work well.
The primary objective of my market type system is to tell me what the market is doing now and to help me determine which trading systems to use. Trading systems developed for a particular market type perform best compared to trading systems that try to work in several or many market types.
Remember that this market type classification system is the set of all my beliefs about markets, and it works well for me right now. You may or may not find it useful. If you don’t find it useful, look at your own beliefs about the market and come up with your way of defining market types. A day trader’s market type system will be different from a swing trader’s or a trend follower’s.
What way of looking at the market will help you choose which trading systems to use? |
|||||||||||||||||||||||||||||||||

Click here for archived issues of Van's free weekly newsletter, Tharp's Thoughts! |
|||||||||||||||||||||||||||||||||

Back to the Top |