In my ultimate pipe dream world, I'd be able to trade a variety of markets intraday, meaning I'd always be out by the close.  There would be no overnight nightmares, the likes of which I endured during the aftermath of 9-11.  Opportunities might come from a slight bias over a heavy volume of signals (as in the types of either-or daily indicators I've been outlining), or something less frequent but with a higher degree of accuracy.  For the latter, if you find something reliable but scarce, you could still potentially have something valid given enough markets.  (For one thing, you could maybe apply individual stocks to your trading universe, thereby enjoying continuous action even from systems signaling once a week, once a month or less.  For another, if you sufficiently trusted your signal, you could up your trading size).

Like any mechanical technician, I can imagine an almost infinite array of such indicators.  How many times have you been looking to be a buyer, for example, in a market that has been down many days in a row before finally giving you a higher close?  At that point, don't you get a sense of "about time!"  Maybe your instincts told you that at the very least, the bearishness was overdone and you can expect a little follow-through in the way of dead cat bounce. 

Obviously in today's computer environment, you can do more than guess.  I'm wondering as I write this.  Be back in a minute...

Ok, here are the results of buying after the first up close following five down closes in a row, and selling a down close after five higher in a row (open to close).  

They suggest several things.  First, you don't have a lot of trials.  Several people in my recent book, Market Beaters, claim they won't be fortified before seeing several hundred or several thousand instances.  But they also said they would need less in the way of corroboration if the results conformed to their theories about the market drivers â€" what they believed to be true about the underlying forces. 

In this case, we can certainly argue that results conform to theory.  A market, however one-way prone, is always going to back and fill.  The more closes you have in a given direction, the closer you are to a reverse close.  That's pretty much true by definition.  Once you get the reverse direction, you can say that a trend has been broken for however short a period.  (If it's for one day, obviously you're going to be booking a loser in trying to go with it). 

As it turns out, five days, or a week's worth of one way action followed by a final break, will produce a bias.  That's not a crazy left field number.  Furthermore, it works, however faintly, in the nine markets we've already determined to be our focus.  We didn't just include the Swiss franc because it happens to conform well while throwing out others we've been using throughout this column's history. 

Eye-popping summaries aren't logistically possible as we just don't have enough instances of the signal.  Still, the research would inspire me to heed future signals.  I'd even argue it's reasonable to include it in a future combo-signal amalgam â€" the kind we'll hopefully be presenting soon. Stay tuned.

Art Collins is the author of Market Beaters, a collection of interviews with renowned mechanical traders. He is currently working on a second volume. E-mail Art at artcollins@ameritech.net.