Thursday, March 12, 2009

Trading Mentors: Learning by Seeing, Doing, Teaching

A good trading mentor, like a good teacher of physicians, will both model skills and observe the skills of students, providing timely feedback as part of a review process. This supervision, based on the "see one, do one, teach one" model of training, emphasizes learning by doing, not by classroom teaching. Much of that learning entails pattern recognition: seeing so many cases that you begin to appreciate similarities and differences.

One of the first sets of skills aspiring physicians learn is diagnosis. You can't effectively treat an illness until you clearly identify it first. Similarly, you can't effectively trade a market unless you acquire an understanding of the market you're trading.

In the doctoring world, physicians run various tests on patients, from blood workups to checks on vital signs to imaging studies. These tests are their "indicators" of a patient's condition. From the results of these tests, physicians learn to rule out certain problems and entertain the possibility of other ones.

The informed trader is not so different. Indicators emphasized in this blog, such as relative volume and NYSE TICK, help a trader rule out certain possibilities (an upward trending market) and rule in other ones (a range market). From this "diagnosis" of the market, the trader can make inferences regarding the likelihood of hitting certain price levels, such as the prior day's average price (pivot) or the current day's volume-weighted average price.

If a physician takes a sick patient's blood pressure and finds that it's normal, the physician doesn't complain that the "indicator"--blood pressure--"doesn't work". Rather, that information is used to infer that perhaps the illness is not heart-related. Ruling out possibilities is just as important to diagnosis as ruling ones in. If a market is moving higher, but NYSE TICK is mixed, I don't conclude that TICK is unreliable. I know from past market studies that low TICK rises are especially vulnerable to reversal, and I start to question the likelihood that the current rise will be a sustainable upward trend.

Notice one of my early Twitter tweets from yesterday:

8:34 AM CT - Overnite low represents initial downside target; watching for signs of range day, so seeing how we trade rel to open & vwap.

Just several minutes into the trading day, I'm seeing signs of mixed strength in the market, so I'm entertaining the hypothesis of a range day. At this early juncture, my hypothesis is just like a physician's tentative diagnosis that a patient might have an ear infection based on the pattern of presenting symptoms. That tentative diagnosis leads to further tests (such as examining the ear for signs of inflammation), just as a tentative hypothesis about the structure of the market day leads me to further examine certain indicators (sector behavior for mixed vs uniform strength; relative volume increasing or decreasing) for signs of trending or reversal.

The challenge for the trader is not so much one of making market predictions as refining trading hypotheses as market data unfold. The effective trading mentor models this reasoning process, but eventually needs to let junior traders try out their own reasoning, with prompt review of solid and flawed reasoning. That is why sound mentoring has to occur during the trading process, just as effective training of physicians has to occur at the bedside or in the clinic.

The intraday Twitter posts (free subscription) are my way of modeling my thinking about markets as trading unfolds. Imagine, however, an interactive messaging environment, in which students can work as teams to diagnose markets and propose winning trades--just as medical students rotate through different specialties as teams and learn from one another. The challenge for mentoring is to turn one-way mechanisms of delivering information into effective, interactive platforms for mutual seeing, doing, and teaching.