The first step I took was to run a 12 month bar-by-bar walk-forward test of SLA on its own. I would step through each 90 minute morning session and draw in the SL/DLs, and place the trade without seeing what came next, using nothing but the basic rules - I believe this was coined "surfing" at some point. That turned out to be profitable on its own. I of course went back and tried adding various entry filters, trade management rules, etc., and some worked in some situations, but a few days/weeks later, I would find situations where those new rules were really just becoming a hindrance. This is when I then went back a third time and ran a 4HR and 30M chart alongside the 1M and I realized the primary "filter", if you want to call it that, for moving beyond "just profitable", was location.
Wow.. impressive that you did this for a whole year. I did this in real time for 2 weeks, as in my journal, and had great results. The location filter, was it trades at the extremes or means according to AMT? Care to discuss your findings some more... but no obligation of course since each person should be doing this work themselves.
It didn't take me a year to do the walk-forward testing, it was just a year's worth of data. All the time spent testing took several months. Application of the principles was/is a whole other game. Its amazing how hard it was to simply do what I was supposed to, what I had planned to, do. To be honest, AMT in the diagonal sense is something I've only more recently been incorporating and it has certainly helped add to the roadmap. As I mentioned in the Ghost thread, DBs discussions of boxes (ranges) is what originally clicked with me - lateral extremes. To answer the question, I have mostly found frustration when my entries are too close to the mean. I find that price is most undecided and I have a harder time deciphering what traders want to do. And I think in many cases, I'm not the only one. The mean often is a point where I see price bounce up and down, and I really have trouble determining which way it will break. At the extremes, be it lateral (box) or diagonal (trend channel - as was the case this morning), is the location that I am referencing. If you watched a DOM, or time and sales window, or even just your chart this morning, during the testing of that extreme, price appeared to be almost rabid, if traders, from both sides were so eager to do something they were foaming at their mouths. Niko started a conversation on TL concerning fear, following something DB mentioned in his PDF. You can see that in the tape at these extremes - breakout shorts fearing they'll miss the boat, early longs fearing they were too early, long time longs that think their position is now in trouble, etc. - you can feel it. That is being emotionless and reading the emotions of everyone else and capitalizing on it. So by location, I mean extremes - this is where price risk is minimal, and information risk is at its highest. But is it really? When traders are at the mean, they are undecided which direction to go. When traders are at the extremes of a range, their most likely course of action, having tested that extreme, is to test the other side. Of course the market cares nothing of your lines on your chart, and could easily fall through whatever you marked as an extreme. But you know quickly, and often with tighter risk that you were wrong. The mean is filled with scratches trying to find a direction - the extreme is right or wrong, and quickly discovered. [That is just my perspective].
Fantastic stuff Fluke. Learning a lot from your posts. I too had/have a difficult time reconciling lateral extremes with diagonal. One of my early misconceptions was confusing Support/Resistance with mean reversion. While S/R can cause mean reversion, mean reversion need not be caused by S/R. One reverts because of the number of trades that took place, while the other reverts because of a lack of trades. Diagonal Extremes still cause me to distrust them - but after seeing time and again (in the Ghost Thread) how PA evolves at these locations, I have started to come around - today being an example.
And a wonderful perspective it is. I can't argue with anything you say. (except that I figured you meant a year's worth of data, not that it took you a year to backtest! )
Thank you Game - I will never forget the post of 40 Draws, I believe it either in the original SLA thread or the next one, where he posted his 2 hour chart, showing price at the lower extreme of a down trend channel. That morning's open was violent - it dropped suddenly below the trend channel, and immediately reversed (eerily similar to this morning). By the time the first SLA entry from the 1M chart was triggered, the next RET following the entry would have made (and did make) most SLA followers exit due to its depth. However, his entry was much lower, if I remember correctly, or perhaps he just had the conviction necessary, due to his pre-market analysis and was able to hold through the first major RET (which I believe tested the 50% of the move up from the trend channel lows). If I recall correctly, he ended up holding part of that position through the day session and overnight - that is still where I struggle. I don't necessarily aspire to hold overnight, but when you're in at the right location, you should milk it for "as long as the trade continues to make a profit".
Within the world of short- term trading, if we consider scalping as one end of the spectrum and multi day swing trades as the other, it seems that the most powerful combination is to enter trades and manage losses like a scalper but to develop the perspective of a position trader when managing winners. Here is a relevant quote from Oldtrader: To me the market is a lot like shooting pool. You constantly play for position. Let's say you're in a position that has moved as you expected, and is now near your target area. If you sell, what's your next move? In other words, if you get out, and the move continues will you have a way to know that, do you have a way back in? It's one thing to achieve a high profit factor in longer term trading, but quite another to do the same in a high turn, short term trading environment. The way to match Db's or 40' PF of 10+ is by playing for position, without taking the larger risks taken by swing traders. And by always rooting each decision to the fundamentals, i.e Behavior. Another quote from Acrary that you may find interesting: Monee, it sounds like "you want to eat your cake and have it too". Profit targets improve the % wins and keep drawdown periods small. However, you never get the benefit of the non-normal distribution of gains in the market (fat tail trades). Trailing stops, allow participation in the big moves, but because of their lagging behavior, you're stopped out more often with a loss. Many traders want to use targets for consistency, but when a huge move happens and they only have a fraction of the profit, they get frustrated. You need to decide which goal you want to pursue (home run champ - trailing stops, or batting ave. champ - profit targets). Because of the non-normal distribution of profits in the markets, the home run champ is the better of the two ways to go. I'd guess that most amatuers go for the high percent wins and may have 60 or 70% win records. However, the pro's know it's the big trades that make the big money.
Apr 14 - Pre-market On the hourly, we can clearly see that we have a bit of a rally happening here after reaching an oversold position based on your new "red" channel. The mean of this channel is 3515 as a possible target, but 3510 would also we the level of the lower line of our previous channel. We could also turn around at open, in which case going down to the bottom of the channel would bring us to 3439. We also have a level of 3414 marked in, first noted last week as a level at which price turned around twice, so this could halt a drop in price. Given the bullish nature of the overnight activity, I would expect a continued rise to most likely the mean of the channel. Taking this overnight action into account, we turned around to go back up at 3426 and 3428, and we also were in a tight trading range which had a mean 3439 roughly. We are well past all that now, sitting at 3465 and continuing to climb.
Good day, but tough day. A - Was very much expecting price to take off. Had I gotten up a few hours sooner it would have been nice to ride the wave up. But right at open, we quickly hit a high of 72 and went nowhere. B - This seems like a good entry for short, but it really took a few minutes to figure out that price wasn't taking off. Plus, I would have been stopped out on the bar at C. My entry here would have been about 67.50, but it went as high as 70.25, and I don't think I want more than a 2 point loss. C - Shorting again here at 65 works for a few minutes, but SL breaks and would have been out for hopefully BE. This clearly isn't easy, even though it looks like we aren't going up by now, so shorting is the way to go. D - We have a SL break... some sideways action, and we resume our drop so draw in a new SL. E - Here I take an actual trade and it goes pretty good for 2 mins, but soon a SL break and I'm out for only a loss of one tick, -$9 with commissions. Naturally after this, I'm not in autopilot anymore so lets just watch the action to regroup a bit. F - Considered a long here, but our DL breaks. G - We have a strong rejection here, and on the next bar, the long would trigger technically. I would have taken the trade off by now though had I been looking for a live trade. But I have seen that these strong one minute rejection bars generally tend to hold for price moving in the opposite direction for at least a few minutes... perhaps this could be a good for a quick scalp?? H - A hinge starts forming here, and although we initially break out of the top, we come bar down. Look at these bars though, so much overlap! If we are to get into a trade like a short anywhere, we would be stopped out for sure. I - This ends up being the low, which I see now is even a higher low from the earlier low that ended up being the low of the day, but with so much choppy action, I'm just not looking for any trades. J - A long above this bar is a completely legitimate trade. Our SL is broken (not drawn), and we have a clear RET on the one minute chart. This would have also been a zero risk trade as price moves instantly up and doesn't come back. But I just didn't even mark it in because I was perhaps thinking the high just before H might form a double top. K - So after that was missed, we are approaching the opening high of 72 so lets wait to see what happens there. We have a nice thrust up to it, so a possible long at K could be taken. Would be filled on the next bar. The trouble is that price would have moved against you over 2 points three bars later, so I would be stopped out for a 2 point loss. L - Here of course we have another possible entry for a long. Now that I think about it though, we can't fan the DL until we clear the highs of the bars before, and this entry is below those highs, so in real time, I doubt I would see it, so this entry for a long just doesn't exist in real time, and once its obvious, the trade is gone. So today was quite tough, but putting money on the line is an excellent teacher I think. The only really good trades I see now are roughly around J. Getting here was tough though, so its hard to be eager to take trades here as if you're fresh. On the bright side, staying out of trouble is important, and other days, the price will just take off so as long as losses are kept to a minimum on tough days, it can't be considered a bad day. I am clearly not taking all the entries that I should, even if I follow SLA, so I will have to work on that this week. I am also looking for RETs within the bars because I want to get in sooner and at a better price, but this might lead me to over trading, and of course if I starting picking and choosing randomly, that isn't good for discipline.