Option Investor
Trader's Corner

A Trading Tactic Revisited

Printer friendly version

A November 5, 2005 Trader's Corner article described a tactic that involved fading a first-hour breakout. For months, choppy markets have rendered many technical trading setups less useful than usual. This has been so often seen lately that I have stopped all pure directional trades and limited my plays to those that take advantage of range-bound action. Last Friday's early morning rally prompted me to revisit that trading tactic first described in November.

Those who did not read the original article, especially those who have never studied Donchian channels, might go to the Option Investor archives and read that Trader's Corner article before proceeding further with this one. The tactic involved watching for a first sixty-minute Donchian channel upside or downside breakout. That breakout would be faded when certain other conditions were met. A five-minute Keltner channel chart was used to identify those conditions.

As stunning as last Friday's upside breakout appeared to be in those first few moments after the open, the move above the top of the Donchian channels caused me to start watching for the setup that had been detailed in that previous article. The rally was so short lived that the Donchian-channel breakout was not maintained into the close of the first sixty-minute period. It did not create a Donchian channel breakout signal on the sixty-minute chart, although it did on the fifteen-minute one. That day did not set up according to the parameters outlined in the November 5 article.

Annotated 60-Minute Chart of the SPX:

As the chart illustrates, the play never set up using the parameter described in the first article. Although other play setups might have occurred and certainly did, this one did not. A second look at that chart shows a setup on March 30, however.

Annotated 60-Minute Chart of the SPX:

Having discovered an initial setup that occurred that day, revisiting the play setup from that November 5 article required turning next to a five-minute Keltner chart.

Annotated 5-Minute Chart of the SPX:

A cautious trader would have scaled out of at least part of that position when the original target was hit and would have lowered the stop on the rest of the position to stop-even or better. Then, as price moved lower, the trader could have followed prices lower with stops, choosing either to continue to do so until stopped, to exit when stochastics showed oversold conditions or when stochastics indicated bullish divergence just after one o'clock that afternoon.

Another setup, this time for a bullish play, occurred March 21.

Annotated 60-Minute Chart of the SPX:

The Keltner set-up was less optimal that day.

Annotated 5-Minute Chart of the SPX:

Another similar setup, but this time for a bearish play, occurred March 13, with the original target hit that time, too, but with the outer channel on the 5-minute Keltner channel never violated. Although these required a little more judgment as to whether to enter or not, the first condition for the setup occurred and the plays proved viable.

I wanted to revisit this first-hour setup because it has seemed that few technical setups have followed through. However, even during the last few weeks when markets seemed so difficult to judge by technical analysis, these setups appeared to work well. They did not come frequently during this last few weeks, however.

Although that November 5 article talked about the logic behind these setups, some cautionary review might be in order at the close of this article. Donchian channels were created to identify breakout plays, and some back-testing has hinted that such plays to work best, against all logic, when stochastics are already measuring overbought or oversold conditions. So, this setup runs counter to that back-testing. However, when doing that back-testing, I noticed that two exceptions to that truism occurred: when the breakout occurred during the first hour of trading and when it occurred during the last hour of trading. Those breakout plays proved so unreliable that it looked as if fading those plays was a better bet than taking them. This conclusion fit with research that noted the high frequency with which the day's highs or lows occurred within the first hour of trading, so that a breakout during those times was likely to achieve the ultimate high or low of the day, and then, logically, could be faded.

However, because fading those plays goes counter to what Donchian channels were created to do--identify times when it was a good idea to play the breakout--caution should be exercised, and something further was needed to clarify the setup. I used Keltner channels for that purpose.

Adding to the above cautions, no official back-testing of this 60-minute Donchian/5-minute Keltner setup has occurred. QCharts does not yet provide that capability, so I urge you to do testing of your own, including some paper trading, before you attempt trading this setup. This is especially true if you're going to trade it using options. Options prices tend to be inflated the first hour of trading, so that even these successful plays--in terms of the SPX hitting the target--might not have been as successful if trading such a short-term move with options. While the SPX was moving the right direction and hitting targets, some of that excess amateur-hour premium may have been leaking away, too, so that the options play might have been only minimally profitable by the time the commissions and spreads were considered. Test it for yourself, and, if you're a futures trader, test it using futures instead of options.

At the least, though, this setup and its continued reliability over the previous few weeks demonstrates how dangerous it might be to get too enthusiastic about a first hour upside or downside break. You might just be seeing the ultimate high or low of the day.

Trader's Corner Archives