Gaining a Feel for Markets by Immersion in Historical Trading Patterns
A central part of my preparation for the trading day consists of running historical investigations of current market patterns. For example, if I notice that an index has been down for two consecutive sessions, culminating in a 10-day closing price low, I'll consult market history to see what has typically occurred thereafter. Sometimes, the investigations will reveal a potential edge: a directional bias following the current pattern.
I don't trade that bias mechanically; rather, I treat it as a hypothesis for the coming days' action. If real time market behavior supports the hypothesis, the historical background provides me with a little greater conviction in the trade idea. In that sense, the historical studies might be viewed more as hypothesis generating (qualitative) research than hypothesis testing (quantitative) research.
All of my studies are conducted with a database of historical data that I've assembled, with analyses and charts in Excel. A chapter in the Daily Trading Coach book explains in detail how I created the database and conduct the analyses.
Much of the art behind the investigations, however, consists of knowing what to investigate. Anyone can conduct 20 studies and find the one result that is significant at a .05 chance level! The most promising studies are those that make solid conceptual sense; that capture how traders actually think and behave.
For example, how traders view the shares of emerging markets speaks volumes regarding their attitudes toward risk assets in general. If traders are bullish on emerging markets, they are probably also bullish toward a host of assets, from commodities to emerging market debt. Such sentiment factors are likely to lead to trade-worthy directional tendencies.
As a simple example, I went back to 2006 and examined what happens when the five-day change in emerging market shares (EEM) is up by 2% or more and when it is down by 2% or more. Interestingly, when EEM has been strong on a five-day basis, the next five days in SPY have been weak, averaging a loss of -.44% (130 up, 162 down). When EEM has been weak on a five-day basis, the next five days in SPY have averaged a gain of .08% (128 up, 97 down).
But when we look 20 days out following five-day strength and weakness in EEM, we see a different pattern. Twenty days after a strong five-day EEM, SPY has averaged a loss of -.51% (162 up, 130 down), not significantly different from its five-day prospective return. Twenty days after a weak five-day EEM, SPY has shown further weakness, with an average decline of -1.21% (119 up, 106 down).
In other words, on a short time frame, we have seen countertrend behavior in SPY following strength/weakness in EEM, but over an intermediate time frame that has dissipated. Of course, this is but a starting point for further investigation to see if other factors (larger market trends; behavior of other assets) might affect the relationship between EEM strength/weakness and prospective price change in SPY. Many good trading ideas come from looking deeper into apparent patterns and asking why they might occur.
It's surprising how conducting these studies day after day gives traders a feel for how market move and how different markets are connected. There is no contradiction between being immersed in data and being discretionary in trading: over time, the data are a powerful source of pattern recognition.
.
I don't trade that bias mechanically; rather, I treat it as a hypothesis for the coming days' action. If real time market behavior supports the hypothesis, the historical background provides me with a little greater conviction in the trade idea. In that sense, the historical studies might be viewed more as hypothesis generating (qualitative) research than hypothesis testing (quantitative) research.
All of my studies are conducted with a database of historical data that I've assembled, with analyses and charts in Excel. A chapter in the Daily Trading Coach book explains in detail how I created the database and conduct the analyses.
Much of the art behind the investigations, however, consists of knowing what to investigate. Anyone can conduct 20 studies and find the one result that is significant at a .05 chance level! The most promising studies are those that make solid conceptual sense; that capture how traders actually think and behave.
For example, how traders view the shares of emerging markets speaks volumes regarding their attitudes toward risk assets in general. If traders are bullish on emerging markets, they are probably also bullish toward a host of assets, from commodities to emerging market debt. Such sentiment factors are likely to lead to trade-worthy directional tendencies.
As a simple example, I went back to 2006 and examined what happens when the five-day change in emerging market shares (EEM) is up by 2% or more and when it is down by 2% or more. Interestingly, when EEM has been strong on a five-day basis, the next five days in SPY have been weak, averaging a loss of -.44% (130 up, 162 down). When EEM has been weak on a five-day basis, the next five days in SPY have averaged a gain of .08% (128 up, 97 down).
But when we look 20 days out following five-day strength and weakness in EEM, we see a different pattern. Twenty days after a strong five-day EEM, SPY has averaged a loss of -.51% (162 up, 130 down), not significantly different from its five-day prospective return. Twenty days after a weak five-day EEM, SPY has shown further weakness, with an average decline of -1.21% (119 up, 106 down).
In other words, on a short time frame, we have seen countertrend behavior in SPY following strength/weakness in EEM, but over an intermediate time frame that has dissipated. Of course, this is but a starting point for further investigation to see if other factors (larger market trends; behavior of other assets) might affect the relationship between EEM strength/weakness and prospective price change in SPY. Many good trading ideas come from looking deeper into apparent patterns and asking why they might occur.
It's surprising how conducting these studies day after day gives traders a feel for how market move and how different markets are connected. There is no contradiction between being immersed in data and being discretionary in trading: over time, the data are a powerful source of pattern recognition.
.
Comments
Post a Comment