I guess the debate is on! But let's not detract from jjrvat's thread. I think jrrvat deserves our respect. If you want to continue this, then let's start a new thread.
TRO, A couple of points Limit orders are all well and good but you will always be filled if you are on the wrong side of the market (as price moves through). You will not necessarily be filled if you are on the correct side. I searched for information about order precedent and matching on interbank networks and whether you can submit things like iceberg orders. I could find nothing. Some participants leave resting limit orders all the way up the book so they are first in the queue and pull them if they decide they don't want to trade at that price. In short Limit orders can be a double edged sword. Having said that I use them 90% of the time but tend to hit the ask with a limit buy. Markets do 'change' but I guess it depends on how you define 'change'. I like to think of them as having different 'personalities' depending the 'mood' they are in. Volaitility changes, pace changes, persistence changes, the amount of backing and filling changes. Having said that when a market is in a particular 'mood' it will exhibit similar characteristics as when it was in that mood before. Final thought it is exits that determine profits, and stops that determine losses. The entry is probably the least important of the 3! Having said that this approach has great value as it will tend to keep your entry on the right side of the price action for the timeframe you are trading. With methods like this knowing when to scratch when to take one off at a couple of pips or when to press for a bit more is what determines success. This is where novices (and experts sometimes) struggle. This depends partly on the character of the market at that time. For scalping keeping an eye on a lower time frame chart can help.
Although I am just a beginner and not very happy with my exits yet, I'd like to humbly mention, in the spirit of sharing, a few things my limited experience have tought me so far: (the discussion below assumes candlestick bars) - do not exit as long as the candlesticks are green even if the price hits a resistance area. - do not exit if the red candlestick body is engulfed in the body of the previous green candlestick. - set a tentative target 0.5 to 1 point close to and before the next major/minor S/R area. That'll take you out with a profit before the price turns around and goes against you without hitting the S/R line. This is because sometimes the price acts this way and because sometimes the S/R line drawn is not accurate. (and it does not have to be. S/R lines should thought as 1 to 1.5 point thick horizontal zones, not just thin, straight lines) - if the price hits the S/R, exit the half position and move the rest to break even, if the psychological comfort of scaling out is more important for you than a worsened R/R - if, after a rally, a red candlestick with a strong body (a body which is not so small and is outside the preceding green bar) shows up, wait until that red candlestick bar closes (unless it starts going against your position fast, you should then exit quickly) and exit on the next bar after it goes against you by 1 tick below the red bar before (the red bar with the strong body). This is kind of "magic tick on the exit". - consider tightening your stop as the price approaches a major/minor S/R area As can be seen above I am trying to keep my exit rules simple and according to the price action. I am sure good traders do not need my tips. However, the tips might give some ideas to those who struggle with exits. Comments and criticisms are welcome..
umitd Thanks for taking the time to post. Exiting is definitely the area that gives me the most trouble. I'll definitely look at your "rules". Cheers
First i wanted to thank jjrvat for his help in developing/modifying my trading plan. I made three trades yesterday, all were profitable. First trade, half was filled at predetermined target (cover fees etc...), remainder breakeven. Second trade, half was filled at predetermined target, remainder at a trailing stop. Third trade, half hit target, remainder exit on first Heiken-ashi red. I am still testing out better entry's/exits, but this seems to be working for the last couple days. I am a bit agressive on my entries, but still trying to break some habits. Advice/critiques welcome. Thanks again for this thread, and the invaluable information within.
ps. What are your views on time charts? I am use to the 5min chart, and is why i use it for my entries. I see that you all use tick/volume charts, which makes sense to me. I'm watching a 4000 tick chart, secondary confirmation i guess..
I've noticed that a lot of times all profits will be wiped out in one bar (the slope changing bar). However I won't deny that in trendy markets this system results in epic win (if you can hold on through the >10 tick drawdown at the beginning): And I am still having tons of difficulty with this situation (perfect setup yet the price reverses as soon as I enter = loss): (With reference to the situation in the above pic, people have suggested that this is not a good entry point because the new downtrend is starting too close to the previous low. In my experience this is not a statistically significant point. In other words, there are plenty of massive trends that begin/resume near/at/beyond the previous high/low, and also plenty of failures that begin a good way below/above the previous high/low. And as I've mentioned before, in my backtesting, capturing these massive trends is necessary in order to have positive expectancy) Additionally, I'm still not sure what constitutes a "trend." How many successive HH and HL are required? I made a post with many example pics pages back but I never got a reply. I also don't like bars with giant range because they are causing huge drawdowns and loss of profit. I've changed from 500 volume bars to 233 tick and 144 tick (I'm not on the Fibonacci bandwagon I swear... I'm sure 232 and 234 would work equally as well as 233). But every bar interval has these as they are relative to the interval selected. So I thought maybe I would use range bars, so I loaded some 5 range bars on the YM and they trended very nicely but also new bars formed every 4 seconds and I can't trade that fast. So I'm stuck with big tall candles stealing my profits on volume/tick/time charts, and charts updating way too fast on range charts. I've also been experimenting with different ways to tell how choppy the market is/is going to be so I can just sit out during that time. I fully accept that reducing your losses is just as valuable as increasing your winners. However, every signal/method/indicator I've tried/coded/invented for this purpose doesn't work, and almost inevitably result in one of two outcomes: - By the time it says "hey, we're trending" the trend is over. (so I adjust it to make it react quicker and get...) - Too many false positives: "Hey, we're trending... [enters order] JUST KIDDING! [exit for loss]" Comments/critiques/corrections welcome.
reading through the entire thread, and seeing all of your issues/questions ironfist, it all stems down to your entry/exit. In my opinion, no one can spell out where you should enter/exit as each person's risk aversion is different. Your first chart, you state all your profits wiped out in one bar. Well what is your 'target'? You think if you have this issue all the time, maybe you should employ a trailing stop, or a tighter goal? Or accept that you wont catch everything? Your second chart, a 10 tick drawdown is peanuts if you are intending to catch all 100 ticks... it is all about perspective in my opinion. Third chart, valid trade. But if selling into a previous low, maybe tighter stops in order? These are only my opinions, what works for me may not be your cup of tea. The way i interpret jjrvat's ideas is simple. He presented a framework outlining when to trade with highest probability of success. That is it. Never said take every trade, enter at specific rules, exit with other rules. That is the beauty of it, in my opinion. He outlined a principal many preach, trade with the trend... All i can suggest, is experiment with different entrances/exits. Seems you are trying to automate it too much, and relying entirely too much on your MA...
ddoonie, Thanks for the reply and welcome to ET I'd say it's my exits I have a problem with. And as you read on you'll find out why... My target is as much as the market gives me. I have backtested using various fixed point targets and, while it increased my % of winners, it caused my testing to end with a net loss. That is where I came to the conclusion that the huge runs are required in order to have a positive expectancy over time. So I can't exit with +2 or 3 or 5 or 8 or 10 or 15 ticks because this trend might end up being one of the +100 runs which statistically speaking is required in order to have a positive expectancy. Re: a trailing stop - I haven't done this because I have a lot of trades that go +/- 20 ticks before deciding that they want to continue with the trend. Trailing stops would cause me to be stopped out (even at break even) which could prevent the big winners which results in net loss. Re: a tighter goal - I explained above why I don't do this. Although I am kind of OCD about efficiency, I have accepted the fact that I won't catch everything. However, again, I don't not exit early because I need to catch everything from a psychological standpoint; I don't exit early because I need the big winners to have a positive expectancy over time from a math standpoint. I know. But sometimes 10 tick drawdown turns into 30 before the slope changes back so I have to hold on regardless of how much price moves against me. I should also add that this has made me excellent at following rules and having no emotion. Yeah it sucks to be down 30 ticks, but it sucks worse seeing that go +100 ticks in your favor after you exit. Perhaps. But I don't like stops for the reasons above. And I don't want to re-enter the same trade again if I get stopped out because it goes the other way before going with the trend I originally predicted. I know. That's the part I'm trying to figure out on my own (and failing). I believe any strategy can be automated (and therefore back and forward tested). Discretionary strategies can be automated. The user is basing his "discretionary" decisions on facts which can be automated (altho it might take a while and be extremely complicated). A "feeling" about what the market is going to do appears to be discretionary at first but is actually the result of subconscious calculations based on previous experiences (ie. mechanical rules). The only way a system could not be automated would be if it was based on totally random entry like flipping a coin "heads = enter, tails = don't enter." Even systems that give automated signals and then the user chooses whether or not to take those signals based on discretionary decision could still be automated. The user is seeing something in the market that causes him to think it's a good idea to not enter even tho his system tells him to. But this is an entirely different topic. It would be nice if jjrvat gave specific rules for entry and exit but I don't think he's going to At any rate, I have still learned a lot from this thread. Honestly, I never even considered HH, HL, LL, LH before reading this thread. Unfortunately, price doesn't behave in pretty little waves or we'd all be millionaires.