Five weeks ago, I posted the following to a journal: It should be remembered that there are five phases here: observation, backtesting, forwardtesting, simtrading, real trading. First, one must observe in order to find those "tells" which show the market's hand. In broad categories, these tells will manifest themselves in reversals, breakouts, retracements. This is analogous to meditation in that if one finds himself thinking about entries rather than focusing on price behavior, he should start over from the beginning. Eventually he will get tired of doing this and focus on the behavior in order to get through it. Second, he then finds instances of these tells in old charts (i.e., anything before now) to see if he's onto something or if what he thought was a tell was just his imagination. And since he'll be working right to left, this can be done fairly quickly. Third, he tests these suppositions via replay, "reading" the chart from left to right. If the suppositions don't pan out, then he goes back to step one. If they do pan out, he moves on to simtrading. Fourth, he simtrades his suppositions to see if the data matches what he came up with during earlier steps. If it doesn't, then he has to back up until he finds where things went wrong. Part of what may be wrong is that in simtrading he cannot pause or stop or replay what he's just done. In replay, the trader is in control of just about everything. In simtrading, he has no control over anything other than his entries, management, and exits. The market doesn't pause because the trader is confused. There are no do-overs. The shock of moving from replay to simtrading is a reality check. It is here that the trader begins to realize that he didn't understand this stuff as well as he thought he did. Therefore, to get to the point, finally, marking up charts and reviewing the day and even going through the day again in replay isn't going to accomplish much unless one has gone through the first two steps thoroughly. This is why those who read chat logs and thus review the day, with or without charts, are pretty much wasting their time. A great many traders just cannot reconcile the facts that each moment is unique and the outcome of any given trade is unknowable with the idea of determining the probabilities of success of a particular "set-up", or cluster of behaviors. One cannot trade successfully by staring into the rearview mirror. He must look forward, but while doing so he must also know what he's looking for and what to do with it if he finds it. Otherwise he can literally spend (waste) years examining charts and dwelling on couldawouldashoulda. I go into this at this length because this whole subject is so murky to so many people, and too often they persuade themselves that they can become successful traders if only they study enough old charts. This is of course wildly unrealistic. A week ago, I started the Observation Journal to address the difficulties of observing price without concerning oneself with "trades": entries, exits, management, risk. Doing so is next to impossible for the average trader but it is also essential to mental clarity in real-time trading. I don't know whether that thread will ever achieve -- much less maintain -- that clarity or not. But I know that posting anything there that has to do with trades would only defeat the goal of the thread to keep trading out of it. I said to someone yesterday that I would not be beginning a journal on backtesting because the SLA had already been backtested and backtesting of course includes a wide range of approaches that need not have anything whatsoever to do with the SLA, such as backtesting having to do with indicators. Yesterday's chart, however, has been niggling away at me ever since. I missed something, and geeky though it may be, that kind of thing drives me crazy until I come up with at least a half-assed answer. The process of figuring this out, however, has to do with a missed trading opportunity. So if I want to get this down, much less seek the opinions of others, I pretty much have to start this journal, even though it can't in all fairness be limited to the SLA. So. Backtesting. Second, he then finds instances of these tells in old charts (i.e., anything before now) to see if he's onto something or if what he thought was a tell was just his imagination. And since he'll be working right to left, this can be done fairly quickly. I'm probably bumping up against the post length limitation, so the charts will follow in the next post.
I've examined this chart backwards and forwards and up and down and drawn trendlines and means in every location I could conceive of and could not come up with anything that made foresight sense in terms of trend. However, there is this hinge. Although someone may have pointed out this hinge in real time yesterday, it went right by me. Looking back on it, I must admit that I would not likely have taken the failure at 48 anyway since I don't like to trade inside hinges, which are essentially a pretty form of chop. Note that the line drawn across the last two swing highs could have been drawn prior to the open and extended forward so that when price hit it, the rejection could easily have been noted. This might constitute a "tell":
Zooming in a little, we get this. The price break clearly is at 10, but the break of the hinge is a little lower:
What I find as much interesting about the above chart as the hinge is the regularity of the movement, characteristic of AMT. Clearly the trading is intentful. Although this is of course hindsight, a trader who stuck with this from an AMT perspective would have noted by 1300 that he had a nice little supply line going for himself. And even if he missed or decided not to take the price break at 10 or even the break of the hinge, the retracement to the bottom of the hinge, coinciding as it did with the test of the supply line, would have been more than sufficient justification for taking the short, worth in this case over 30pts. Of course the trader may not have been there to take the short. This did, after all, take place late in the day. But that's the difference between CouldaWouldaShoulda and backtesting. This is a definite tell and it's worth filing away and testing in order to see how often it comes up and whether or not it's worth trading. CWS is more often a boo-hoo that's soon forgotten with no follow-up. But there it is, the best I can come up with. There may be other perspectives, which is why I went ahead and started this journal. In any case, this appears to be a worthy addition to the AMT/SLA methodology.
I usually discard trading inside big hinges because of your advice, and because of the overlap that I have observed (casually, not statistically significant) within these areas. Yesterday for example, it was a downtrend, it was easy to call it by 12:00 . But for those who were there trying to get in was not easy, as we were rapidly scratched and rejected. Now that can be because of the places we used to enter and the entry method (other methods perhaps might do the trick) and is very interesting that there is now a place to discuss and to go further into R&D. For now I have to focus on my journal but I will be back later in the day. Great that you started this thread. Thanks.
Which is a big part of what pisses me off about yesterday. It isn't as if I just started doing this, and for that in utero hinge to not even ping my radar . . . Maybe next time. But that's why this is done: next time.
I'm confused by this statement. You had previous to the open posted a down sloping channel within the daily chart. If you bring this channel forward to an hourly chart you get what I've posted. Do you need more here than to notice the rejections on the median of the channel?
I've re-read your posts a few times and I'm still confused about what tell it is we are backtesting? You mention that a trendline drawn across the higher lows and the rejection there may be a tell (assuming this is a side note for those interested). Is the tell we are interested in the break of the bottom of the hinge followed by the retest at 15:00?