Saturday, January 02, 2010

The Trading Process: Idea Generation and the Scientific Mindset

How does a trader make the transition from understanding a current market (the three elements enumerated above) to actually formulating a trading idea? That is the topic of this post.

In a word, the trading idea must integrate the information we've gained from an appreciation of market context (whether markets are trending/bracketing on longer time frame; intermarket themes that have been operative; relevant news events and economic reports); key price levels (support/resistance; average trading prices; target prices); and day structure (whether the current day appears to be in breakout/trending/range mode).

As an intraday trader, I begin the process of generating ideas by formulating hypotheses as the market is trading. Many times, the hypothesis will start by assuming that whatever has been happening at the larger time frame--what I am identifying in the market context--will carry forward to the current trading day.

Thus, on average, if I see we've been trending higher, I will hypothesize that strength will carry over to the current trading session. At this point, the hypothesis is only that: an educated hunch, not a trading idea.

As the market is trading, that hypothesis may gain or lose support. For instance, if I see that, with each bout of selling, the market holds above its volume-weighted average price (VWAP), the idea of continuing the trend higher gains support. Conversely, if we make buying efforts and volume wanes and we cannot take out prior highs, my hypothesis loses support.

If I'm thinking like a scientist, my real-time market observations are helping me to refine my hypothesis. Perhaps the stock market is not moving much in pre-opening trade, but I notice correlated markets (U.S. dollar, commodities) moving in a way that would be supportive of higher stock prices. That might not only help support my hypothesis of continuing the market trend, but might also provide clues as to which stock sectors could most benefit from such a move.

Similarly, I might see emerging market stocks leading a move higher and conclude that such speculative sentiment could carry over to the more speculative areas of U.S. stock performance. Or I might see overseas markets level off on unfavorable economic news and moderate my expectations for U.S. stocks on the open.

The point is that hypotheses are ever-evolving. New market information is helping us gain or lose confidence in our ideas.

Once I have high confidence in direction (i.e., the most recent market action is confirming my prior hypothesis), I then use my understanding of price levels and day structure to make the hypothesis more specific.

For example, if we're trending higher and I see signs of firmness in the market prior to the open, I might hypothesize that we will hold above the overnight low in Globex futures trading and trade above yesterday's high price. Alternatively, I might have lost confidence in the idea of continuing the trend higher due to tepid overnight action and failing intermarket themes and hypothesize that we will hold below the overnight high in Globex futures trading and trade back to the current day's VWAP or the prior day's pivot price level.

When I refine my hypotheses, I always make reference to key price levels and my ideas about direction must always be grounded in what I'm seeing in the most recent market behavior. If I'm hypothesizing a trend or breakout day, I'm looking for prices higher or lower to follow the most recent price action. If I'm hypothesizing a range or false breakout session, I'm looking lower or higher for prices to reverse the most recent price action.

If you were to watch me trading, you would see me jumping from screen to screen, chart to chart, constantly updating measures of sentiment (NYSE TICK, Market Delta); volume; correlated indexes, sectors, and asset classes; and behavior around price levels. Cognitively, I'm continuously weaving a narrative about what I'm seeing: assembling the emerging market data into a story line that makes sense to me. The hypotheses flow from this story line.

Many times, I find the data to be contradictory and mixed and no convincing story line emerges. At such times, I do not act. Only when the data fall into a pattern and the pattern fits into my understanding of context and day structure will I entertain a hypothesis and elaborate it with price level/target specifics.

As we will see in later posts in the series, this narrative process of hypothesis generation continues through the day and, indeed, throughout the period of holding positions. It can also occur at time frames wider than intraday, as in the case when we see stocks failing to sustain strength early in the week and hypothesize that we will trade back to last week's pivot level.

Once we understand idea generation as a kind of scientific process, we can better appreciate the role of psychological factors in trading. A good mindset will not put valuable hypotheses in the head of a scientist, but the wrong mindset can interfere with generating and refining worthwhile hypotheses. Most emotional disruptions of trading involve some kind of short-circuiting of pattern recognition and reasoning processes.