Kiwi

Thoughts on stops: - stops almost always reduce profit simply because they take you out at the worst price (so far) of a swing - that means that if you are very well controlled the combination of a disaster stop + exit with price improvement after a condition is met may well be best - many markets test one swing but not the second so stops two swings back are recommended by some - some forex markets seem to stop for the swing, and some overrun it a bit so stops should be adjusted for that - often swing stops have the worst slippage because everyone's stops are there so sometimes a stop 1 tick inside the swing can be better ... defense of the price still exists ... but you'd have to check how this market and your data/price provider represents thing to find out if thats true.

Comments: Gabe, IMO you must have exit rules in place before the trade takes place. Preferably have some very solid guidelines for all trades. Why? Because cash and the need to be right raise emotional levels and while they distort perception before a trade they distort it far more once in the trade. Unless you are a robot you should determine your rules when you are not affected by fear, greed, or the desire to be right. Later, when one is truly confident in the method then intuition can be permitted because it won't be swamped by emotional responses. Up to you. Daedelus, Tick bars on forex are not going to be helpful. First, everyone else here is using time bars so they make comparison difficult. Second, tick on forex is almost as much rubbish as volume on forex - its just what your data provider decides is the tick rate and has no true representation of the "real" forex market. If your data provider is honest then its a representation of his microcosm of the market but you'll probably find that even the "honest" inject ticks at slow time to make the market move - and at any time the ticks are not the real market. I mean, you can choose to use them, but they are not real. And they confuse others.

FWIW. Mechanical EJ ... I must define chop zone if I'm going to automate this at some point and probably define a poke and fail to be truly sophisticated. I note in passing that the test of top and two lower highs or lower high and slight break of the first that doesn't reach the top is a classic failure sign in hsi. Sucks in some longs and then goes ... and the sucked in longs generate extra liquidity on the collapse that follows. The tough thing is to distinguish it from a resumption upwards

The thing is that I think in terms of the trend giving an advantage. With trend, the extensions are longer, the counter-extensions are shorter; timing is more forgiving. And to me the trend is read in some fuzzy way (that has to be quantified in systems) using some combination of hhs and hls, crossing of S&R areas and testing them (digging) before continuation, crossing and sloping of mas to indicate whats happening to the average of price, and new thrust (push of so many points in so few bars). So in the 930 case we had a trend up (broke support and resistance with thrust, tested on the outside lows, were above the 50ema on a 4 hour chart). In fact we still have that longer timeframe chart pointing anemically up (but oh so weak). If I was to automate this then I would be assuming longs until you had some criteria for a break down. In this case, if your system derived its information solely from the 15m timezone the three signals approach is actually surprisingly good, but could be combined with some larger breakout or an overshooting smoother like the hull ma rolling and turning down with price below. Or sufficient thrust. I'm not sure and won't know until I get a chance to test a full system (which I won't share because those little discoveries are what make the difference between profit and not and must be aligned with ones own beliefs about how the market works). To the one party who did want the material, I'll drop it to you sometime in the next day.

I now have some quite nice functionality for looking at how well it works; and Thales, it really does I included internal swings (where it does abcde break) with a lighter colour as they are also good breaks but don't show up mechanically with a simple 123break rule. I also added adjustable day and night times so you can look at whether you'd be awake and how u might react the next morning to an "open trade." I do observe that picking the trend using, higher timeframes plus standard hhhl and breaks and fakes from chopzones dramatically improves ones results. Its not so obvious in this particular picture.

Hi Thales, I refer back to your trial defining post Rules at this Post . I am coding some rules in Sierra Chart (available to anyone here who uses SC, just PM me) to help visualize the breaks so that its easy not to miss them and wondered about the potential view of a 2 overshoot and 3 overshoot. As you've specified them there is no overshoot though Even on 3 is ok. And for a "beginner" while trying the base method I'd imagine you'd stick with this but from my years on hsi I know that I'd see (and accept) slight overshoots as still "ok." So, here's the question: what overshoots would you suggest to be acceptable on forex at points 2 and 3?

There was a breakout at the point I marked 14 . It went Low at 1, high at 7, higher low 10 bars from the low and then the high at 7 was finally broken (briefly) at 14. Whoops, sorry the next one. Low above 11:45, then a high is made after one white bar and a black bar. then a higher low is made by the 1 bar pullback with the last red line touching it. Thats L ... H ... HL ... and the breakout is above 1.4664

OK. Here's another cut at what the two differences are and thus how we might qualify when to stop trading and then start again. if the image doesn't show http : // charts.dacharts. net /2009-11-05/kw8.png

LOL. You really don't want to go with the trend do you I'm going to critique it (helpfully I hope) applying Seiden's critera.

OK. It is interesting looking at what MK and Thales are doing at the same time as I am trying to fully automate. Increasingly it seems to me that Automations strength is in taking advantage of either trends or patterns. And the human's advantage is in taking advantage of Support and Resistance. Now, a smart guy by the name of Sam Seiden, who trades pullbacks into S&R uses the concept of odds enhancers to decide which ones are worth trading. One of them is strength of the initiating move (impulse), another is the shape of the S&R the move came from, another is how far to the next Very High timeframe S&R (daily here perhaps or 4 hr min). The other major is Big picture trend. I'd say that in this case the VHT S&R supported the trades but gave warning areas in the GBP case. But in both cases the trade was against the long timeframe trend. And if you look at Thales' trades with both a 15m and 4hr chart open you'll see that when there is a clear trend (move) on the 4hr then its much easier to make money out of the 15m breakouts. So here are MKs ones from yesterday (with the benefit of my perfect 20/20 hindsight). Warning Warning ... Kiwi is inclined to trade with the trend so he might just be pandering to his own biases.

Ahhh ... ok. That's where we disagree MK. To me, the reason for larger timeframes is either: - for support and resistance and any timeframe works here (daily, weekly, 4hr etc) - for trend and here I think you want the one in which the next level of trend exists. That second part means that to me, daily would be a good higher timeframe for the overarching trend on 4 hourly and maybe hourly trades (6x and 24x respectively). But on a 15m trade I think that hourly or 4 hourly would be more appropriate (4x and 16x respectively. So I'd look at (say) the hourly and ... hmmm ... I need to do pictures or this isn't adding to anyones education, mine included. I shall come back to this in an hour with a couple of pictures for the NZD and GBP trades properly annotated. Hopefully that will make my point (lol, or prove me wrong as my system trading tests have done repeatedly). Note: thanks for your comment too Thales as it appears you use the 1st reason as your main driver for a higher timeframe.

A comment and a question on the NZD and GBP trades. Comment to MidKnight: both of these trades seem to me to be counter to the larger time frame trend (HHHLs) and stopped working when the larger time frame reasserted itself. Question to Thales: do you agree and do you tend to take trades against the next timeframe up? And if you do, do you require more range to your targets (or any other multiplier?)

But the simple answer to your question is that MidKnight uses Interactive Brokers (IB). IB are one of the few who provide decent coverage of Asian markets.

Jeepers Midknight. You're beginning to look like a breakout trader. For the record, Peewee managed 3 losing trades this morning. Interesting; a fair way off its biggest losing streak though. Now working on a new scalper system called Cheeta.

Yes. Someone on another (much rougher) forum said: Man is a parading animal. So true

Clockwork, An interesting theory but its like master salespeople or master athletes. People apply discipline, control, fairness, impatience (with fools, markets etc) differently in different situations. And recent research on self-control shows that it comes from the front of your brain and (sadly) gets used up and then takes time to replenish. So the master trader applies control and perception to the markets for xx hours. Then he finishes, takes a break, goes to the pub to let it all out, or the internet and rips someone a new one. Worse though, the internet has become a place where people who wouldn't fight in person will fight. Size doesn't matter. Its often used for peacocking not communication. And you can't see that ILoveSax is starting to tear up when you point out how silly his question is so you don't feel sympathy and pull back. The only thing that saves us is the social norms of the board you work on and their ability to expel people who breach them. Expectations and Policing. Too strong and people who are smart traders get bored and leave. Too weak and you get ET (which does have some amazingly knowledgeable traders but is a little rough for ILoveSax). At least, that is my suspicion.

I would comment on Midknight's chart and the final stop at 4hr resistance. 1. Yes the resistance is there so you could have either had a bounce (likely as its the second test of a zone that has a strong thrust previously). 2. When it did test you got very strong rejection over 30 minutes. A 30 min pinbar. I think three strategies suggest themselves. 1. Preferred. On first or second approaches to strong 4 hour support just put in a limit exit like your early 15m support limits. Then if it breaks, sell the breakout of a bar that closed through??? 2. Wait for PA and exit on break of the synthesized 30m pin bar (fast rejection of price). 3. This would suit 3rd or 4th touches of the support: just play it out as you did. Option 3 has the merit of being simple and uncomplicated but I think Option 1 would be good as well given you are already exiting on S&R. You just need to make a rule so that you are not swinging in the wind when the decision time comes (zero stress). I suggest applying Seiden's criteria to the 4 hour break and if it looks strong, exiting on the first two touches and holding for 3 and 4.

Taking Thales point above into account, Snag, please don't take this as an attack on your or your intelligence. We form different opinions based on our experience and also the deletions and distortions our brains apply along the way. So (although you personally are right) one of the opinions above is IM(H:))O, wrong. I'm not saying that the rest of it is right, just dissecting one point. It mixes different things up - "If you are disorganized, you aren't disciplined." The confusion here is that organization and discipline are all that similar. They are related in the sense that we think of organized people as being disciplined but not in this context. I am very organized. I am also very disciplined about some things. I am a very organized trader. But I am not a disciplined trader in the face of risk and emotional distortions that result. Hence, although I can read the markets better than I can systematize them, my lack of really good discipline under monetary risk means that I am probably better of systematizing my ideas and auto-executing them. Similarly, my friend lives in a messy house and is bloody disorganized. He doesn't have a written trading plan and he turns up at the trading desk when he feels like it. He leaves on the same basis. He changes styles without "proper" research. But he is very disciplined under monetary risk in the sense that Mark Douglas means (operating without fear or greed). His breathing rate goes up etc but he doesn't let fear or greed change his approach - he does the same thing he'd do on paper or with twice the risk. He is very disciplined and as a result he is consistently profitable. But he is also totally xxxxing disorganized. If you see what I mean?

Very nice model (because all these discussions create mental models of the market that are true or untrue and useful or useless to varying degrees). I particularly like the resulting ideas that - what you need to imagine is where projected value will be, is going, might hold - that being small you don't have to go early because if you wait for the projection to give itself away in the move after accumulation the slippage won't kill you so you get a reduction in risk vs the participants who have to scale into possible reversals and one of my own - that signs of the movement of projection can result in an immediate move but may not, allowing larger participants (and you) time to accumulate a position with price improvement after the signs of movement.

It's not fraud to short something that is wildly and foolishly overvalued - no matter how big you are. It shows that someone there was thinking. I would like a way to short the Aussie housing market. It and NZ are about the only two left fully puffed from pre-crisis borrowing binges. Next year when the Govt Buying bounces have run out and the Chinese sit on their mineral stockpiles for the negotiations things are going to be pretty interesting down here. A leveraged short on housing would be a much more interesting bet than one on the Melbourne Cup

Unless an Indian has written one you are unlikely to find one.

FXCM probably have the out so that they can apply it to news events and/or anyone who they feel is benefiting too much from their stops.

Here's the old draft of the book. Not sure where it is currently. Its a 7zip file so u need to change the suffix from zip to 7z to open it. . Market Microstructure.zip

You might consider two elements to the grouped buying and grouped selling (trying to avoid jargon). One element that is most considered is the amount. Another that is less considered is the desperation. For example, if there is a perception that price is moving out of the intended buying zone before one could get filled, then desperation will increase pressure without increasing volume necessarily. I like to ask myself this question: - based on what I am seeing, will the volume guys think they can wait and get a better fill before this move really happens or will they perceive that it's on its way. If the first then price improvement is on the cards; if the second then I'd better just go with the move (even chase it just a tiny tiny bit). PS, There is an excellent (well?) book on market microstructure by Harris and the draft is legitimately obtainable on the internet at various places

I like Interactive Brokers because I can have both and because its an ECN model so you are not constrained by funny rules that some of the bucket shops apply. Also, you can be long futures and short cash should it suit you to be so.