Friday, March 16, 2007

Learning How To React To Changing Markets


Back in the late 1990s, I decided to take trading very seriously. I began to archive market data in Excel, starting a collection of one-minute price and indicator data that I have maintained faithfully since 1998. I also printed out charts such as the one above, in which I viewed each major indicator vs. stock index price and identified patterns that occurred at important turning points. To this day, I have a stack of printed charts that fills quite a few filing cabinets.

When you notice a pattern and then notice it again, and again, and again in different variations, eventually you become sensitive to what is pattern and what is noise. The process is very similar to the training of a radiologist, who learns to read X-rays or scans by seeing very many normal and abnormal ones, or a pathologist, who--after seeing many normal and abnormal tissue samples--becomes very sensitive to what is disease and what is expectable variation in health.

Most important, seeing patterns again and again enables a trader to recognize those in real time as they are forming. As I stress in my book, I have found simulated trading--particularly simulators that enable you to play and replay market days--to be especially powerful as learning devices. Seeing patterns unfold again and again significantly reduces one's learning curve. Once you've seen a trading pattern unfold dozens and dozens of times, you develop a feel for its future unfolding. That's the process of implicit learning I've described in past posts.

And that is how you learn to react to changing markets: By experiencing many, many market changes.

Wednesday was a particularly valuable day for observing patterns, because so many of them occurred within the span of a single trading session. I have labeled the above chart with four letters to indicate some of the more noteworthy patterns:

Point A) Notice how, in the opening minutes of trade, the market is much more reactive to the drops in the NYSE TICK than to the bounces at the very start of the session. When we jumped to positive TICK readings, these were followed by immediate selling that took us back into negative TICK territory. By the time we saw buying at point A, even the strong TICK of +1000 couldn't vault us above the opening highs in the S&P index futures. I refer to this pattern as inefficiency: Quite simply, if we measure the amount of upward price movement we are getting per unit of NYSE TICK, it is quite low. It's like a runner who is huffing and puffing, but getting nowhere. Such inefficiency is common at market turning points: sentiment is no longer able to sustain a short-term trend. Before we see a surplus of selling, we'll see buying sentiment (positive TICK) that no longer can move price higher. I generally sell those inefficient TICK bounces, particularly when they're occurring near tops of ranges.

Point B) Commencing at point B, we see that we get a pattern of lower highs and lower lows in the NYSE TICK. That is the classic pattern of a short-term market decline. As long as that pattern holds, you keep selling the TICK bounces. The thrusts to negative TICK, especially on elevated volume, are opportunities for the active trader to take some short-term profits off the table, even as he or she might ride a core position toward a designated target (such as the pivots mentioned in the Weblog). Recall that the NYSE TICK is telling us the relative balance of stocks trading at their offer price (upticks) vs. bid price (downticks). When we see a distribution of TICK values that is tilting downward (lower highs, lower lows), it means that bearish sentiment is trending. Those are markets we can ride lower.

Market Low) The market low for the day occurred on a very negative thrust in the TICK. A pattern that doesn't show up on the above chart, but that I had mentioned in my blog post that day was that, while the ES contract was making new lows for the correction, other stock indices (such as NQ and ER2) were not. If a majority of issues are not participating in a move, there's an enhanced likelihood that the move will not be sustained. Notice, however, that someone *only* looking at intraday data (such as the above chart) would have missed this vital longer-timeframe pattern. The important point here is that patterns are generally nested within larger patterns, and it's those larger patterns that create trends. Even very active traders need to train themselves to see patterns present in daily and even weekly data. Someone with their nose in the one-minute data on Wednesday would have been completely run over at the market low if they were short, oblivious to why the market had suddenly turned.

Point C) Notice the shift that occurs following the market low. We quickly go from a downthrust toward -1000 TICK to an upthrust above +1000 TICK. This creates a breakout in TICK, revealing a sudden surge in bullish sentiment. Even more important, we're no longer seeing the inefficiency associated with Point A. Rather, price is quite responsive to the broad upticking among stocks. By the time you get another thrust down below -500 TICK, price is well off its lows. What that tells us is that, unlike early in the day, high TICK readings are being followed by further high TICK readings: the bullish sentiment is sustained from minute to minute. This tells us that longer time frame participants have identified the market lows as value and are jumping into the market, lifting offer, to take advantage of bargain prices. This transitional pattern--from TICK extreme to opposite TICK extreme, with price responsive--does not occur very often, but can be seen at major market turns and at times when the market is roiled by unexpected news and economic reports.

Point D) At Point D we have the pattern from Point B, but in reverse. Now we're seeing higher TICK highs and higher TICK lows. Bullish sentiment is now uptrending. That is a great indication that we can afford to ride long positions and even add to them on TICK pullbacks. In an uptrend, successive TICK lows will occur at higher price lows; in a downtrend, successive TICK highs will occur at lower price highs. By the time we hit Point D, bouts of selling can no longer move the market lower. They merely serve as further encouragement for the bargain hunters.

Seeing patterns on a chart is only the first step in learning how to react to changing markets. The chart review tells you what to look for. The hard part, then, is to make the transition from chart review after the fact to recognition of patterns in real time. That's a frustrating time in the learning process, as you'll first make progress by recognizing patterns a bit too late to truly profit from them. But it is better late than never. Eventually the patterns become such second nature that you can see them well in advance, just as I noted the emerging divergences minutes before the market low.

No amount of positive thinking or empty-headed glorification of "you are what you believe" can substitute for knowing what patterns to look for and then immersing yourself in those patterns so thoroughly that they become part of your perceptual apparatus. That is true whether you're a scalper noting patterns in a depth-of-market screen; an active trader trading off the one-minute NYSE TICK; or a swing trader playing off Woodie's CCI patterns: learning to react to changing markets means that you shift your perception before the market makes its shift.

6 comments:

bananajoe said...

Excellent post! Thank you so much for sharing your knowledge and priceless experience, you're real "Trading Guru" for me as I learned so much reading your blog! I just want to say that it's amazing to see people like you, sharing their great ideas in exchange for nothing. Hope I'll also get there one day :)

Brett Steenbarger, Ph.D. said...

Thanks much, Banana Joe; I feel I've been richly rewarded for my work on the blog with all the great contacts I've made with traders and trading firms. I appreciate the feedback--

Brett

MikeH said...

"The hard part, then, is to make the transition from chart review after the fact to recognition of patterns in real time"

I think this is one of the most pertinent lines to trading (or any performance activity). It's exceeding to armchair quarterback a static picture of what has happened after the fact. You can't underestimate the experience factor when it comes to realtime performance.

If you haven't turned on the TV in the last 3 years, there are a lot of televised poker tournaments. And you'll hear the announcer talk about the "kid" at the table who is only 25 against these veterans. What they don't mention is that the veterans got their experience playing live in casinos, where you get about 30 hands an hour. The "kids" get experience playing online at 60+ hands an hour, at 4 to 6 tables simultaneously, at any time of the day.

FYI I have a post on the screen recorder I use for review that performs quite well.

Brett Steenbarger, Ph.D. said...

Hi Mike,

Thanks for the comment and thanks for that link to the excellent post. I will also link from my personal site tonite. Your use of the screen capture for review is a great performance application; best of luck with it--

Brett

Yaser Anwar said...

Awesome postttttttttttttt!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Brett Steenbarger, Ph.D. said...

Thanks, Yaser; I will be posting much more "how-to" content in the near future, especially with the integration of video--

Brett