A note to readers: While Dr. Tharp’s content is timeless, this article is from our newsletter archive and may contain outdated information, missing links or images.
Most of you probably know people who thought they had a trading strategy buying technology stocks in the late 1990s. By the time 1999 came around, they considered themselves market geniuses. In 2000, however, they probably continued the strategy even as they watched the market type change. The following chart shows the market type for the QQQ (which is where most people where investing during that time period).
Notice that the market type fell from Strong Bull in late March into Neutral by mid-April. Then, the market stayed either Neutral or Bear for many months and it finished the year in Strong Bear territory.
The market gave another strong sign that the type changed in 2000 — volatility. The next chart shows the volatility of the QQQ during 2000:
Notice that the market was literally nearly off the chart in terms of volatility in the early months of 2000. That’s not a sign of safety, that’s a sign of danger. The market volatility decreased as it became more bearish but it stayed in the very volatile category all year long. This is a great period for day traders who thrive on volatility but it’s not a time for a buy and hold strategy in tech growth stocks. Most of you probably remember than many tech shares dropped 90% or more.
And how long did the Bear phase last? Well, the market type was bearish for most of 2001, with just a short time in neutral territory as shown by the next chart.
In 2002, there was a brief period where it looked like the market type might change, but volatility remained high throughout the year and kept the market in Bear, Strong Bear, or Neutral for nearly the entire twelve months.
As you can see in the next chart, the market type didn’t turn around until 2003 when the market was either neutral or bullish and the last 2/3rd of the year was bullish.
So What’s My Point?
Systems that worked well in bull markets would have been dangerous to trade by the end of 1999 because of the extreme volatility — which usually signals an upcoming bear. And throughout most of the period from 2000 through 2002, bullish trading systems would not have worked at all. Generally, you could have determined that by the market type.
Market type was even clearer at the end of 2007. Our Market SQN® score for the 100 day period changed to a bear market in late 2007 and it remained in bear or strong bear mode for nearly a year and a half. And, as is common in extreme bear markets, volatility was off the charts. After the market bottomed in March, 2009 the Market SQN moved out of bear and into neutral in April 2009 so you had a chance to catch a really nice bull market. You can see this in our Monthly Update on the market by looking at back issues from late 2007 through early 2009.
Rather than acknowledge the importance of market types, I tend to see people look at systems first and try to find one that works. And what they typically mean by ‘work’ is that it works in all markets. To me this is foolhardy. If you understand market type, then you can understand that it is quite easy to develop a fairly good system that will work in any one market type. But the insanity is in expecting a system to work in all market types.
My initial attempt to measure market type was based upon simple logic. Markets go up, down, and sideways. They are either quiet or volatile. That makes for six market types. And my initial definition of market type was to simply look at a six month chart. If it was going up it was bullish. If it was going down, it was bearish. And if I couldn’t really tell, then it was sideways. If the movement was very choppy, then it was volatile and if the movement was fairly smooth then it was quiet.
The next chart shows the S&P 500 ETF SPY from June, 2013 through January, 2014 right before the recent dip. I generally use a six month window to determine market type for US stocks. Based upon my old way of determining market type by looking at the last six months, what is it?
Well, it’s pretty obvious that the market was going up, so it has been a bull market. And at least in the last three months on the chart, there wasn’t a lot of chop so I’d call it a bull quiet market. It’s that easy.
The drawback to my original method was that we couldn’t really objectively classify it and know when it changed. As a result, I decided to look at the daily percent change in the SPY over the last 100 days and then look that the Market SQN score of that data. I made it a little more complicated by having five directions: strong bull, bull, neutral, bear, and strong bear.
In addition, I looked at the ATR % change of the last 20 days and compared it with the historical mean and standard deviation of the last 60 years. I developed four classifications of volatility: very volatile, volatile, normal, and quiet. And most of you know that I give you the market type on the first Wednesday of each month based upon that classification.
You probably don’t need more than the six original market types — but what you do need to know is the conditions in which your strategy works. And when conditions change, you either stop trading or you move to a different strategy that works well in the new conditions. It’s not rocket science, but it is so different from the way most people trade.
Trading is a statistical game. And under different market types you are going to get different conditions. Pollsters know that not all people are alike so they only ask questions of the type of people they are interested in learning about. Traders should know that not all markets are alike and they should only trade systems under the conditions in which those systems are likely to work.