Chapter 22: Day Trading

may find that the charts of very active stocks are too densely packed with information for even a day to be useful. If that is the case, you may need to utilize a shorter span for readability, perhaps half a day.

You will need to find out what the normal trading hours are for what you are analyzing; your trading desk will know. Then set those hours as the defaults for your charts. They can be changed later if something extraordinary happens-an after-hours blowup or some such-but in the meantime youll eliminate the dead space between sessions and facilitate analytical continuity from session to session. This is especially important if you are going to try to have indicators span sessions.

When it comes to chart analysis, bar charts and candlesticks rule. For the short term youll need the shortest interval that produces robust bars. The easiest way to quantify robustness for a bar is to look at the last price of each bar. If the last equals the high or the low a preponderance of the time, the bar length for that stock is too short (Figure 22.1). Lengthen the time frame for the bars slowly until you start to get properly formed bars

H i: 1 m-zt 4 14" >ww-4

b.#-IZir44liiifcri 2434*4

THE TIME PERIOD FOR THESE BARS NEEDS TO BE LONGER

Figure 22.1 Short-term bar chart, bars too short, Guilford, 10-minute bars. Too short a time frame; collapsed bars; too many closes at high or low of bar; little price development; prolonged Squeeze.

Part V: Advanced Topics

iihm nrtlllo 1 * 1 1

-THESE BARS-ARE FINE

Figure 22.2 Short-term bar chart, bars correct, Microsoft, 10-minute bars. A much better depiction of price action.

(Figure 22.2), not just lines with the last rattling back and forth between the high and the low. The goal is to get a view of the price formation mechanism at work, not just a record of the back and forth between the bid and ask price. Once you know what the best time period for your short-term bars is, you can select a couple of longer-term charts to serve in the intermediate and long-term roles.

If you can find intervals both for your bars and for your charts that make sense within the context of the market, youll be much better off. Days, weeks, months, and quarters were "gimmes" that no one had to think about, but how do you divide a day up sensibly? Start with hours-treating the first half hour as a full hour. That renders a seven-hour day on the chart. Consider including the preopening session as an hour and postclose trading as another hour. With that method, each day would have nine hours. Or try my favorite-half hours, with presession and postsession trading allocated one period each. That makes for 15 bars per day.

Youll want intervals that make psychological sense for your charts, intervals that others are looking at or are at least aware of,

Chapter 22: Day Trading

intervals that fit the natural rhythms of trading as closely as possible. Otherwise the information content will be lost. Of course, youll need to consider what suits your style and what is appropriate for your trading vehicles. But whatever you do, try to fit your charts to the reality of the marketplace as well as you can without compromising your style.1

Now, armed with a basic set of charts, it is time to consider which analytical techniques to deploy. Bollinger Bands and the related techniques discussed in this book are widely used in day trading, and much of the material discussed earlier is directly applicable. In chatting with day traders, Ive found two common themes, selling and buying the extremes and trading volatility breakouts. Both of these approaches are well suited to Bollinger Bands.

Lefs review a bit. Bruce Babcock, the late publisher of Commodity Traders Consumer Report, said his favorite approach to trading was the use of a volatility-breakout system. This kind of system has much in common with Bollinger Bands. (The use of Bollinger Bands to construct a volatility-breakout system was discussed in Chapters 15 and 16.) Typically, the average is shortened and the bands are tightened. Then when a Squeeze occurs and the upper band is exceeded, a buy signal is triggered. Generally, some stop such as a Parabolic is used for the exit. Falling back inside the bands can also be used as an exit signal. The opposite logic is true as well. A short sale signal is triggered by falling below the lower band after a Squeeze. Again a stop system such as reentering the bands or a parabolic is used. Be very careful about using breakout logic in the absence of a Squeeze. In our experience The Squeeze is a necessary element of the approach.

When it comes to transacting at the extremes, in day trading Bollinger Bands can be used as references for overbought and oversold. Rallies that carry far above the upper band can be sold at the first sign of weakness, with the immediately prior high as a stop and vice versa. And the bands themselves give signals, as in the case of the lower bands turning up after a breakaway trend to give an end-of-trend signal. Also, if the band parameters are well matched to the security being traded, then failures at the bands can be treated as entry points with an initial price target of the middle band.