- This topic has 74 replies, 11 voices, and was last updated 11 years, 2 months ago by
.
- Posts
Maybe his random entries mean making random entries in the direction of very strong trend momentum, because if there is a very strong trend, price seldom retraces and it doesn’t really matter where u enter, I just enter whenever there is free margin when price goes to my direction. I once grew a 1k initial capital to 9k paper profits by aggressively adding positions using free margin available (leverage 1:100) but price also retraces very fast, therefore I got margin call because my margin from 1k becomes 6k or something like that, needless to say it doesn’t really work unless your SL placement is good meaning every position should be able to cover itself. Or u can close your positions fast enough. Sorry for babbling.
I only know I know nothing
Skype: learnalways@outlook.comHi, It is nice that you are going that way. While managing existing position (bad position)……only two option exist… 1. Recovery 2. Somebody called it “biting the bullet”…in the other words closing down… 1a. Bad position – wrong timing or wrong direction (against the flow/trend) And personally I think that this is the ONLY way to go…so absolutely adore the idea, only problem is that no robot can substitute for human being in vastly changing conditions, however other part of me knows that it is possible as nothing is impossible. regards D. ps. there is one guy who was talking about it (first point above) and he was retired after 2x years as pro trader for major tier 1 bank (dealing with spot fx on top senior position) as far as I remember he said that you have to understand: – martingale – Fibonacci progression – log arithmetic spirals …and that all secrets lies in the pi and its skew…he was applying those things to SIZE..not price..think about it.. and most importante… it is about what market is doing…like gg said it, methods cannot be rigid as they have to adapt to market conditions…and market is not rigid place
There is another way:
Instead of StopLoss – use hedging.
That’s what I’m using – but it’s not a simple method and quite difficult and requires experience to get out of such hedging.
As to pi & its skew – related to lot SIZE and trade scaling – that’s what I mentioned in the first post (scaling position size). A feature that is ignored by most traders. HUGE mistake.
G.
I spent about a year trying to develop such a system. It wasnt a random entry system, it was in fact the ‘back end’ or ‘recovery’ component of a normal signal based system. It all started from a train of thought of running a system without a stop loss. Things go really good until the inevitable margin call. As stated above there are two options when trades go bad: 1) cut your losses (ie stop loss) – but this is not in spirit of the thread… 2) dig a deeper hole (ie enter a recovery trade at a better price point) I would always end up crashing the account due to the problem of: At what point do you enter a recovery trade and for what (lot) size. The bigger the recovery trade lot size the quicker it would sink the ship as price moves further against you. However if you are in a counter trend trade the recovery trade lot size must be big enough to make enough money in the small amount of pips that is the pullback. I went in circles for a long time trying to overcome this issue.
Every position that I take have an alternative StopLoss in the form of hedging.
Invest in a method of getting out of an hedging position with minimal profit or BE. It works for me.
When price hits SL – It’s a definite LOSS.
When you are hedging – your money is still in the “game” and you can recover with at least BE.
G.
Don’t you think that there is an issue (contradiction) in what you have just wrote?
Yes, absolutely! gg53 constructed an extreme goal in his 1st post: profitability at random entry. I hardly can imagine anyone around would consider testing that with real money. So I’m taking it as some kind of a mental experiment, set up to direct our focus more on money and trade management than on entry signals. So I consider it inevitable that our discussion will lead to contradictions from time to time. I believe that’s part of the process, as well as resolving contradictions and trying to gain some insight into possible ways to achieve the goal to define a ‘Perfect Trader’ as gg lined out in his 1st post.
What’s random in particular?
See the question I threw in for discussion in my 1st reply: what is it that’s random? I think that randomness is limited and has to be well defined. I think that if ‘Perfect Trader’ doesn’t act randomly, the strategy that enters our trades can’t either. There has to be some kind of synchronization between them, and that particular requirement will make any random entry strategy non-random instantly. A lack of synchronization between any two EAs acting on the same account and on the same trades will certainly lead to chaotic behaviour of the system as a whole. For me, the basic question is: Can Perfect Trader be designed in a way to manage trades profitably after random entries? If we should manage to set up a prototype working properly to a certain extent, acting as a proof of concept, the secondary question arises: Which modifications are required for that prototype to adapt to non-random entries? Now considering a more or less complex and time-consuming process of design and development, a meta-question is showing up: If our discussion leads to the conclusion that considerable modifications would have to be applied to PT-Random in order to create PT-Nonrandom, does it make sense trying to create a completely random version first?
“Perfect Trader” – a trading EA that can statistically win and be profitable, even with random entry.
So, in a way, gg’s approach is carrying that contradiction you pointed out as an intrinsic property. We have to be aware of that and consider what the implications for a process of design and development are.
Wrong on first sentence.
I didn’t suggest “an extreme goal in his 1st post: profitability at random entry”.
I suggest a “perfect trader” that will trade even a “bad entry” to a MINIMAL, logical loss – after managing the trade the best way an experienced human can.
G.
Let me be clear on the subject:
I’m talking about an adaptive “trade management” and not a static system that trade random entries.
“trade management” is NOT just setting a lot size and SL+TP. It’s dynamic management as the trade develops – either toward SL or toward profit.
Meaning moving and adjusting SL, dynamic trailing, adding to a position, monitoring total equity, monitoring related correlated pairs, etc. Even exiting a trade BEFORE StopLoss in some conditions.
From my experience – those are much more important than “perfect entries” – and that’s the target of this thread.
G.
Maybe his random entries mean making random entries in the direction of very strong trend momentum, because if there is a very strong trend, price seldom retraces and it doesn’t really matter where u enter, I just enter whenever there is free margin when price goes to my direction. I once grew a 1k initial capital to 9k paper profits by aggressively adding positions using free margin available (leverage 1:100) but price also retraces very fast, therefore I got margin call because my margin from 1k becomes 6k or something like that, needless to say it doesn’t really work unless your SL placement is good meaning every position should be able to cover itself. Or u can close your positions fast enough. Sorry for babbling.
Please re-read my posts.
I’m not talking about “random entries” system. I’m talking about dynamic, adaptive trade management that makes logical decisions on current active trade.
G.
I didn’t suggest “an extreme goal in his 1st post: profitability at random entry”. I suggest a “perfect trader” that will trade even a “bad entry” to a MINIMAL, logical loss – after managing the trade the best way an experienced human can.
That sounds more specific to me. Maybe I didn’t get your 1st post right:
It’s about time to invest in a “Perfect Trader” – a trading EA that can statistically win and be profitable, even with random entry.
And maybe my wording “an extreme goal in his 1st post: profitability at random entry” wasn’t really perfect. Hope I didn’t annoy you.
What I’m really interested in at the moment is your opinion about my previously stated argument that entry strategy and Perfect Trader need to be synchronized to a certain extent, at least to avoid whipsawing IN – OUT – IN – OUT – …
A good trader is a realist who wants to grab a chunk from the body of a trend, leaving top- and bottom-fishing to people on an ego trip. (Dr. Alexander Elder)
Maybe a simple example will steer this discussion away from “random entries” strategy, and more to the point (just one example of “adaptive trade management”:
When you should exit a trade even before StopLoss is hit?
G.
I didn’t suggest “an extreme goal in his 1st post: profitability at random entry”. I suggest a “perfect trader” that will trade even a “bad entry” to a MINIMAL, logical loss – after managing the trade the best way an experienced human can.
That sounds more specific to me. Maybe I didn’t get your 1st post right:
It’s about time to invest in a “Perfect Trader” – a trading EA that can statistically win and be profitable, even with random entry.
And maybe my wording “an extreme goal in his 1st post: profitability at random entry” wasn’t really perfect. Hope I didn’t annoy you. What I’m really interested in at the moment is your opinion about my previously stated argument that entry strategy and Perfect Trader need to be synchronized to a certain extent, at least to avoid whipsawing IN – OUT – IN – OUT – …
Entry and my idea of trade management are two separate issues in my trading.
Entry is decided based upon current known market data. Now you decide to enter.
As time goes by, market is developing and changing – and at this point the “perfect trade manager” should adapt to changes – even if the trade is going “your way” or against it.
Leaving it to entry fixed setting at initial entry time, that was some time ago, is wrong IMHO, and not the way I manged “on going” open trades.
G.
Entry and my idea of trade management are two separate issues in my trading. Entry is decided based upon current known market data. Now you decide to enter. As time goes by, market is developing and changing – and at this point the “perfect trade manager” should adapt to changes – even if the trade is going “your way” or against it. Leaving it to entry fixed setting at initial entry time, that was some time ago, is wrong IMHO, and not the way I manged “on going” open trades.
Yes, I think I got that. I’m absolutely in for dynamic and adaptive trade management, as you stated earlier.
But did I get you right: you wouldn’t suggest to synchronize entry strategy and Perfect Trader?
A good trader is a realist who wants to grab a chunk from the body of a trend, leaving top- and bottom-fishing to people on an ego trip. (Dr. Alexander Elder)
Entry and my idea of trade management are two separate issues in my trading. Entry is decided based upon current known market data. Now you decide to enter. As time goes by, market is developing and changing – and at this point the “perfect trade manager” should adapt to changes – even if the trade is going “your way” or against it. Leaving it to entry fixed setting at initial entry time, that was some time ago, is wrong IMHO, and not the way I manged “on going” open trades.
Yes, I think I got that. I’m absolutely in for dynamic and adaptive trade management, as you stated earlier. But did I get you right: you wouldn’t suggest to synchronize entry strategy and Perfect Trader?
Entry point decision was based on known data at that time. Why should I “synchronize” my trading on history instead of what is going on right now?
Doing so is like “let’s synchronize our watches to the time 4 hours ago”…
G.
Doing so is like “let’s synchronize our watches to the time 4 hours ago”
That’s not what I meant. What about “Let’s synchronize our strategies to current market conditions.” ?
Consider following example when entry strategy and Perfect Trader are not synchronized:
10:00 – Entry strategy says: Market is trending upwards, though momentum is down, pullback entry condition satisfied, no long trade opened. Enter long.
10:01 – Perfect Trader says: Market is ranging, current swing is down.
10:04 – After 3 minutes of trying to make the best out of the trade, PT decides to exit this long trade.
10:05 – Entry strategy says: Market is trending upwards, pullback entry condition satisfied, no long trade opened. Enter long.
10:06 – Perfect Trader says: Market is ranging, current swing is down.
10:10 – After 4 minutes of trying to make the best out of the trade, PT decides to exit this long trade.
… and so on, until at least one of the conditions for entry or trade management will change.
Personally, I would try to avoid such looping between any two strategies. And trying to avoid that is what I mean by synchronization between the two EAs involved. This is what I believe is the minimum synchronization needed between them.
Depending on the particular implementation of both EAs, I can easily imagine scenarios where the above loops don’t occur once every 3 to 5 minutes, but several times per minute instead. I wouldn’t like that on my account.
A good trader is a realist who wants to grab a chunk from the body of a trend, leaving top- and bottom-fishing to people on an ego trip. (Dr. Alexander Elder)
Doing so is like “let’s synchronize our watches to the time 4 hours ago”
That’s not what I meant. What about “Let’s synchronize our strategies to current market conditions.” ? Consider following example when entry strategy and Perfect Trader are not synchronized: 10:00 – Entry strategy says: Market is trending upwards, though momentum is down, pullback entry condition satisfied, no long trade opened. Enter long. 10:01 – Perfect Trader says: Market is ranging, current swing is down. 10:04 – After 3 minutes of trying to make the best out of the trade, PT decides to exit this long trade. 10:05 – Entry strategy says: Market is trending upwards, pullback entry condition satisfied, no long trade opened. Enter long. 10:06 – Perfect Trader says: Market is ranging, current swing is down. 10:10 – After 4 minutes of trying to make the best out of the trade, PT decides to exit this long trade. … and so on, until at least one of the conditions for entry or trade management will change. Personally, I would try to avoid such looping between any two strategies. And trying to avoid that is what I mean by synchronization between the two EAs involved. This is what I believe is the minimum synchronization needed between them. Depending on the particular implementation of both EAs, I can easily imagine scenarios where the above loops don’t occur once every 3 to 5 minutes, but several times per minute instead. I wouldn’t like that on my account.
I see what you mean.
The “bug” in the logic is that your Entry logic is using different measurements for “Trend” and “Range” than the PT.
My trading style is built around “fixed” way of determinig “Trending” or “Ranging” market condition. It is always the same and depends on the entry TF (check 2 higher TF’s and 1 lower TF).
If you call this “Synch” – yes, they should use same methodology and same “language” to describe and determine general market condition.
G.
If you call this “Synch” – yes, they should use same methodology and same “language” to describe and determine general market condition.
I’m happy that we found some common ground to stand on about what I call ‘synchronization’. Now we can start over to plan the foundation for Perfect Trader.
I can’t go to greater detail right now (leaving the house for some sports), but I will certainly be in for it later this evening or maybe tomorrow morning.
s.
A good trader is a realist who wants to grab a chunk from the body of a trend, leaving top- and bottom-fishing to people on an ego trip. (Dr. Alexander Elder)
After reading all this, I was finally able to write a simple EA that aims to trade only in trending markets. Enters when the trend is up (regardless of the News just broadcast) or down, stays out when the market moves sideways. I have partial success with Strategy tester yet I still cannot 100% false signals. I’m working on it.
There are strange resemblances with what I thought as a newbie trader with what you experienced ones stated:
”
3. Sideways/Ranging module, that do the ‘trades’ according to our sideways trading rules.
4.Uptrend module, that do the ‘trades’ according to specific characteristic of uptrend (less risk)
5.Downtrend module, that do the ‘trades’ according to specific characteristic of downtrend (more risk)
“
Nothing has ever motivated me more than this...
Now I’d like to come back to the preconditions from gg53’s first post and just note what comes to my mind. Let’s consider this a quick brainstorming, not more.
1. It should adapt to current currency-pair and market condition.
CSS or similar slope strength
2. It should obey strict risk-management (SL & Lot size)
x%-rule or similar, ATR-based SL
3. Set initial SL by market and pair – not by “fixed” pips.
ATR-based SL or TZ/RZ as an alternative
4. It should minimize the risk as the trade continues (move SL by market condition).
ATR-based TS
5. It should enforce a Break-Even at the right time.
Based on ATR or probabilities from TZ/RZ
6. It should scale the position while in trend (most important and overlooked feature by most)
dynamic variation of x%-rule from above
7. To trail profit or not to trail (again, according to market conditions and not “fixed” pips)?
ATR again as a classic algorithm, TZ/RZ as an alternative
8. Exit the trade according to pair and market conditions.
ATR again can play a role, but TZ/RZ also. When an extreme spike occurs to our favour, switch to lower TF instantly and narrow TS. When it occurs against our favour, depends on where TS is at the moment.
Have to leave now for several hours, but will check back later.
CU, simplex
A good trader is a realist who wants to grab a chunk from the body of a trend, leaving top- and bottom-fishing to people on an ego trip. (Dr. Alexander Elder)
Now I’d like to come back to the preconditions from gg53’s first post and just note what comes to my mind. Let’s consider this a quick brainstorming, not more.
1. It should adapt to current currency-pair and market condition.
CSS or similar slope strength
2. It should obey strict risk-management (SL & Lot size)
x%-rule or similar, ATR-based SL
3. Set initial SL by market and pair – not by “fixed” pips.
ATR-based SL or TZ/RZ as an alternative
4. It should minimize the risk as the trade continues (move SL by market condition).
ATR-based TS
5. It should enforce a Break-Even at the right time.
Based on ATR or probabilities from TZ/RZ
6. It should scale the position while in trend (most important and overlooked feature by most)
dynamic variation of x%-rule from above
7. To trail profit or not to trail (again, according to market conditions and not “fixed” pips)?
ATR again as a classic algorithm, TZ/RZ as an alternative
8. Exit the trade according to pair and market conditions.
ATR again can play a role, but TZ/RZ also. When an extreme spike occurs to our favour, switch to lower TF instantly and narrow TS. When it occurs against our favour, depends on where TS is at the moment. Have to leave now for several hours, but will check back later. CU, simplex
#1 – Ok.
#2 – % of balance method is debatable… Consider losing 50% of your account, recovery will take ages because of smaller lots.
#3 – Ok.
#4 – Initial Stop by %ATR is Ok, adapting SL should reduce initial risk and not based on ATR. Maybe variation on PSAR or dynamic reduction of ATR factor.
#5 – BE should be based on risk factor (R:R) and TF, not ATR
#6 – Scale (add to position) should be based on current profit vs. added risk and only after BE of initial position. Partial close should also be considered.
#7 – Trailing should be dynamic. Amount of trailing should constantly reduced as we approch target.
#8 – Trailing as per #7 should solve part of this issue. Changing market condition (reversal signs, etc.) should close (or partial close) the position.
G.
My personal “strategy” for exit a trade with “no reason”:
If the trade is not in profit after 4 bars – I exit the trade.
Reason: My entry signal was probably false, Market or conditions might have changed.
Staying for longer than that is landing me in “La-La Land” and guess work, not trading per my entry logic anymore.
My Ego is bruised, but I’m out.
G.
I read this thread with great interest as I’m searching for something along these lines too. About the management of the losing trades it was suggested 3 solutions: 1/ take the loss 2/ average down 3/ hedge. I found another approach: using the fractal dimension of the market. If you dislike your trade on the TF you opened it, perhaps you would like it on a higher TF. It goes against the “never widen your SL” but if you move to a higher TF your target will adjusts to this new environment as well. It’s not about closing at BE! The RR and the probability stay roughly the same because you decide based on the same strategies. The duration of the trade increases (a lot). If none of the higher TF like the trade, close it. Of course this forbids an aggressive MM. You have to be able to absorb a 300 or 500 pip DD. It shall hurt, not kill. The PT becomes a hierarchy of PTs.
I read this thread with great interest as I’m searching for something along these lines too. About the management of the losing trades it was suggested 3 solutions: 1/ take the loss 2/ average down 3/ hedge. I found another approach: using the fractal dimension of the market. If you dislike your trade on the TF you opened it, perhaps you would like it on a higher TF. It goes against the “never widen your SL” but if you move to a higher TF your target will adjusts to this new environment as well. It’s not about closing at BE! The RR and the probability stay roughly the same because you decide based on the same strategies. The duration of the trade increases (a lot). If none of the higher TF like the trade, close it. Of course this forbids an aggressive MM. You have to be able to absorb a 300 or 500 pip DD. It shall hurt, not kill. The PT becomes a hierarchy of PTs.
I’m against this kind of “strategy”.
1. It is against your initial Entry logic, MM, and risk management – introducing “hope”, expectation and guess work.
2. It will keep you in the market for much longer – consuming margin, and exposing you to yet another market changes.
G.
I found another approach: using the fractal dimension of the market.
I have similar ideas in mind, just the other way round: not shifting SL to a higher TF if I ‘dislike’ it, but shift my TP to a higher TF instead after price has moved in my favoured direction to a certain extent. I would consider price coming close to my initial TP a good point to start such an algorithm. I considered implementing this based on stacked TZ/RZ probabilities, yet haven’t figured out any details until now.
Widening my SL under such circumstances appears too risky to me. I would fully support gg53’s arguments in this case.
A good trader is a realist who wants to grab a chunk from the body of a trend, leaving top- and bottom-fishing to people on an ego trip. (Dr. Alexander Elder)
On position sizing:
The main target of money management is on how to MAXIMIZE the geometric return in our account.
Using %Balance or %Equity is too simplistic and reduces your earning potential.
My favorites are the trade-size techniques introduced by the “Kelly Criterion” (either half or full Kelly) or the “Fixed Ratio concepts” by Ryan Jones (Google it and watch the results).
It requires logging of previous trades results and calculating positions based on it – but it worth the extra one-time effort to program it into the “PT” EA.
G.
I’m against this kind of “strategy”
It is not strategy it is tactic.1- It is against your initial Entry logic
Per definition, of course, as it is a recovery trade!1- It is against your… MM
It has to be included into. But the losses are linear in the number of pips. An average down is quadratic and a marty is exponential.1- It is against your…risk management
It *IS* the risk management!1- …introducing “hope”, expectation and guess work
Not at all. Any decision comes from a statistically based decision. The trading decision comes from the higher TF. Also the decision includes the option to decline the adoption of the trade and stopping immediately. The hedging is much more a guess work and hope with no logic. Fighting to reduce the gap between the long and the short in order to only manage BE is in my opinion a waste of energy since the RR drops to 0. The multi-dimensional approach maintains the expectancy of the system.2. It will keep you in the market for much longer
Absolutely correct. The time is super linear in the pips to recover.2. consuming margin
No more than at the open time since the lot size doen’t change.2. and exposing you to yet another market changes.
Isn’t it the very exact reason of the idea? Except if it changes on the higher TF too, in this case it becomes a loser on the higher TF. The next-higher level can now accept or reject the adoption.I found another approach: using the fractal dimension of the market.
I have similar ideas in mind, just the other way round: not shifting SL to a higher TF if I ‘dislike’ it, but shift my TP to a higher TF instead after price has moved in my favoured direction to a certain extent. I would consider price coming close to my initial TP a good point to start such an algorithm. I considered implementing this based on stacked TZ/RZ probabilities, yet haven’t figured out any details until now. Widening my SL under such circumstances appears too risky to me. I would fully support gg53’s arguments in this case.
I do it as well. The problem is that the SL becomes too small for the higher TF volatility. You have to bank some pips before shifting. My way is to add to the winner (perhaps too aggresively…). When it is time to take profits I close all the pyramid but the best one. The high TF may adopt it or not.
I’m against this kind of “strategy” It is not strategy it is tactic. 1- It is against your initial Entry logic Per definition, of course, as it is a recovery trade! 1- It is against your… MM It has to be included into. But the losses are linear in the number of pips. An average down is quadratic and a marty is exponential. 1- It is against your…risk management It *IS* the risk management! 1- …introducing “hope”, expectation and guess work Not at all. Any decision comes from a statistically based decision. The trading decision comes from the higher TF. Also the decision includes the option to decline the adoption of the trade and stopping immediately. The hedging is much more a guess work and hope with no logic. Fighting to reduce the gap between the long and the short in order to only manage BE is in my opinion a waste of energy since the RR drops to 0. The multi-dimensional approach maintains the expectancy of the system. 2. It will keep you in the market for much longer Absolutely correct. The time is super linear in the pips to recover. 2. consuming margin No more than at the open time since the lot size doen’t change. 2. and exposing you to yet another market changes. Isn’t it the very exact reason of the idea? Except if it changes on the higher TF too, in this case it becomes a loser on the higher TF. The next-higher level can now accept or reject the adoption.
I’m confused…
If the entry was based on higher TF – than looking at a higher TF is no change in “tactics”.
Same goes for MM & Risk.
But you suggested, or my understanding was, to CHANGE the initial entry logic to another logic, while in the tade- and my response was based on this “tactic”.
G.
- You must be logged in to reply to this topic.