- This topic has 29 replies, 3 voices, and was last updated 9 years, 8 months ago by
.
- Posts
Recently I stumbled upon a thread over at FF with the very same title. The leading post in that thread contains a pdf, which IMO is one of the rare pearls you may find surfing around in FX forums. Highly recommended reading!
The author (PipMeUp) examined well-known general trading techniques like trailing stops, moving SL to BE, partial closure of trades, and so on from a purely probabilistic perspective. He aims to separate techniques which add a positive gain to trading results from the rest.
The paper contains a considerable dose of math, but you can also take note of the conclusion of each paragraph while skipping the formula stuff.
Also particularly interesting is the 2nd part of the paper dealing with multi-position trades (grids).
I will not repost the paper here without the permission of the original author, so please get it from FF:
http://www.forexfactory.com/showthread.php?t=411923
However, I’m interested to discuss its content here.
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)
I will have a read over this as I am trying to put together my first EA and MM/RM/EM is as I am sure all agree rather important so there may some relavant material in there, thank you Simplex
skype : pipatronic
Side note, sorry if it is slightly of course but hedging for me is the answer to SL, material for another video
skype : pipatronic
A very interesting paper and I think that PipMeUp made some good conclusions. I think he’s right with all scenarios.
Nevertheless, I believe that the risk/reward calculation is only half the truth. IMHO the risk-reward ratio is not fixed as it is usually discussed as this would imply same probabilities for moving up and down at any point in time. I think we all would agree that there are (micro-)trends and thus times where one market direction is biased (increased up-probability, less down-probability and vice versa). Thus, the missing part of PipMeUps calculations is directly related to ones trading decision (method). If you always guess the wrong drection of the market you’ll never win, even if your risk-reward ratio is 1:1,000.
Your position handling and trade management will only influence your results if you overcome the 0.5 threshold (you make more than wins than losses).
Side note, sorry if it is slightly of course but hedging for me is the answer to SL, material for another video
Good point! The old discussion: why hedging if you could also exit your original trade and re-enter it later, when you would close your hedge?
Talking about EAs: when looking at hedged trades from a system developer’s perspective, which technique is easier to set up? Managing a main trade and its hedge or just exit and re-enter?
I do not consider this off topic at all!
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)
Your position handling and trade management will only influence your results if you overcome the 0.5 threshold (you make more than wins than losses).
Absolutely correct! Trade management and money management will not turn a losing strategy in a winning one, on the long run. If the expectancy of your pure entry and exit strategy is below 0.5, you got to modify the strategy itself.
Trying to ‘optimize’ MM and TM details like partial closure, trailing stops, etc. will not do any good in that case. Knowing this, time consuming irrelevant loops during the process of system development can be avoided.
Of course R:R will vary over time for particular trades, but I don’t think that this insight will challenge PipMeUp’s general conclusions.
I’ll come back to R:R later, referring to another link.
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)
Good points Simplex, video required a comparison – hedging vs SL – very interesting area that
Hedge strategy – reversal of entry method ?? though there are lots of variables there and it has to be carried out in the correct place or you could end up with a loss (video required)
Birth of an interesting thread ??????
skype : pipatronic
Once I’ve thought that hedging would be the solution for everything. My strategy was simple:
- Open two positions at the same time, one long and one short (regardless any strategy).
- Wait until one position reached your profit traget (e.g. 50 pips).
- Close the trade that reached the target.
- Open a new trade in the same direction of the trade which won and leave the opposite trade running.
- Re-invest your money whenever you earned one by starting with #1 again.
That would work fine for some time. But in the long run the market will reach an extreme state (very low prize or very high prize). And this is the crucial point of the strategy. If you reach a very rar state, it may take a lot of time until you can close your position. Maybe we could handle the loss during those rare states and compensate it with the wins we made.
Before addressing R:R, like announced in my earlier post, let me come back to expectancy with another link. Hanover @ FF put together some conclusions on the ‘area of tension’ between expectancy and possible attempts to increase it by MM / TM methods here:
http://www.forexfactory.com/showthread.php?p=2240127#post2240127
The rest of that thread is not so relevant, IMO. What I’m referring to is Hanover’s post and the links he’s referring to in particular.
Amongst the links Hanover is referring to these got my special interest:
http://www.forexfactory.com/showthread.php?p=2093339#post2093339
http://www.forexfactory.com/showthread.php?p=2093895#post2093895
Let me quote his conclusion from the first one partially:
The idea that MM can provide an edge is one of the biggest trading myths.
I don’t care if the best selling authors disagree: they are ALL WRONG.
I challenge anybody to post a legitimate mathematical proof that proves otherwise.So, as long as the overall expectancy of your very basic strategy is not sufficient, you just don’t have to worry about MM / TM. Refine your strategy!
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)
To hedge or not to hedge?
- Pro hedging: http://www.forexfactory.com/showthread.php?p=6544091#post6544091
- Contra hedging: http://www.forexfactory.com/showthread.php?p=8868124#post8868124
My personal conclusion: from a pure system developer’s perspective, I would always follow the contra arguments. When it comes to highly leveraged trades in practice, hedging can make sense just to remain in business (see the ‘pro’ link).
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)
Once I’ve thought that hedging would be the solution for everything. … That would work fine for some time. But in the long run the market will reach an extreme state …
That’s the point: relying on hedged grids consisting either of limit or stop orders, is a pseudo-strategy heading straight for the annihilation of the account. If your entry system has no edge in guessing the price moves ahead correctly more than 50 % of the time (loosely speaking), hedged grids won’t rescue you. And if your system has an edge, you don’t need hedged grids for your entries.
I found posts and articles claiming that hedged grids (stop or limit orders!) were a market-neutral ‘strategy’, and therefore a key to guaranteed success. This is dangerous nonsense, and I can only assume that the authors of those articles were somehow affiliated with market makers aiming to drain retail traders’ accounts.
A hedged grid made of limit orders can only be profitable as long as the market stays in a well-defined range and prices will return to the center reliably, and a hedged grid made of stop orders will only work profitably if the market is clearly trending and price will not return to the opening price for a certain period of time. This is definitely not what I would call a ‘market-neutral strategy’.
But I think that grids indeed are interesting to improve your trading, yet I’m talking about non-hedged grids here. I’m still in an early stage of development here, but running backtests with directed grids of limit orders seem to improve a strategy slightly by smoothing short term drawdown caused by volatility when placing single trades. See last page of the pdf I referred to in the 1st post for a conceptual background.
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)
“A hedged grid made of limit orders can only be profitable as long as the market stays in a well-defined range …”
The question is, can’t we take advantage from it? For instance, we could define that range and will only use a hedging strategy whenever prize is inside that range?
“… and prices will return to the center reliably, and a hedged grid made of stop orders will only work profitably if the market is clearly trending and price will not return to the opening price for a certain period of time. This is definitely not what I would call a ‘market-neutral strategy’.”
Here I’m not so sure. I think one can also make money if market is range-bound/sideways/consolidating. But your statement seems to be true whenever we use hedging in common sense (open another position with a bigger size whenever pize goes to opposite direction).
Now think about using two equal sized opposite positions oppened at the same time …
I’m not sure if this could happen, but I think if at all this is the better way of applying hedges. The question is, can’t we take advantage from it?
Sure we can, as long as we are correct about predecting that the market will stay in that range while our trade is open. My point is, that this is not ‘market-neutral’, like claimed by some authors. If price leaves the range and starts to trend, we loose.
Now think about using two equal sized opposite positions oppened at the same time …
I did. For my conclusions, see posts above. If your strategy does not have an edge predicting market state and / or direction, it’s doomed to fail. The grid won’t change that, it will only accelarate or slow down draining the account, depending on the prevalent market conditions.
I never before thought about combining a hedged grid with a martingale technique. I would say that’s pouring in gasoline in the fire burning your account … Inevitably the day will come when your remaining margin is not sufficient for your next hedge, right?
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)
I go out for a couple of hours and come back to find that I suddenly have lots of homework and catching up to do
skype : pipatronic
Okay see attached on hedging, I have done this in the past and managed to break even/profit from it, very clear from a drawing but quite different in real time so you need to have your wits about you and be sure of what you are doing and trust the system/method you trade by.
skype : pipatronic
Just to add if it does not work out hedge again, might be an idea if I can do a screen shot video of a hedge in real time as they are never as clean as in my drawing.
skype : pipatronic
Okay see attached on hedging …
Good point! Now let’s put on our accountant’s glasses and change our action on your scenario slightly, without changing the scenario itself:
- Replace ‘open short hedge’ by ‘close trade’
- Replace ‘short closed’ by ‘re-enter trade’
Taking the nit-picking accountant’s view: what would be the difference between your proposal and mine?
There would be absolutely no difference, correct? For the counter trend trade to be net profitable, it is irrelevant whether you execute it by placing a hedge or by exiting and re-entering the original trade. For both variants, timing is equally essential.
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)
Yes I see, I have just been reading over Hanovers notes on exactly this, I wonder if it is a physiological thing (for me)
Good post
skype : pipatronic
As announced earlier, I would like to come back to the topic Reward to Risk Ratio, RR.
To kick off a discussion, let me start with two links to external sources this time, which at least at a first glance seem contradictory in their respective conclusions.
First Hanover @ FF again, see here, and one quote from that post:
RR (in the conventional sense) is effectively irrelevant. … However, expectancy is the product of win size and win rate, and the latter is determined by how frequently our assumption of a reversal is correct. That relates to the efficacy of our analysis and forecasting. It’s the extent to which the forecasting goes beyond the inverse balance of RR and win rate that provides the edge.
Second, the AuTraSy blog, see here:
Trend followers position themselves to profit from and capture the “fat tails” exhibited in market returns distribution. In a fat-tail distribution (Power law, Levy or Mandelbrotian distributions), extreme occurrences occur with a probability greater than normal.
…
The basics of trend following is to ride the trend until the end (when it bends) and to protect yourself on the downside by cutting your losses. This ensures that the location of your trades in the returns distribution will:– Never venture on the left fat-tail (i.e. no extreme negative return)
– Not be bounded on the right-hand side of the distribution (i.e. allow for extreme positive returns)So on one hand we have the probabilistic point of view, stating that the expectancy of a system contains two factors, win rate and RR, which are created antagonistically to each other, i.e. if one of them increases, the other one will decrease, if the quality of our price predictions remains constant.
On the other hand, we have a statement about non-normal distributions of price moves in the FX market, especially Lévy-Distributions, which tend to overweigh extreme movements (both win and loss) as compared to a normal distribution. Here the conclusion is, that a strategy to cut losses, the ‘fat tail’ on the left side, very quickly and give profits the chance to ride that ‘fat tail’ on the right hand side of the distribution effectively. This means aiming for rare trades with high RR would be the most promising strategy, while approving that the majority of our trades will end up with minimal profits or losses in the sharp-faced center of the distribution. This is mostly because most entry strategies will not be able to predict how persistent a particular movement will be in a mostly chaotic market. Those ‘center-trades’, should effectively even each other out. Our strategy has to be precise enough to provide for this.
Are both statements really contradictory, or do they present two different faces of the very same token?
Maybe I missed something here.
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)
I wonder if it is a physiological thing (for me)
Hmm, do you mean ‘psychological’ ?
‘Physiological’ might mean that your trades are causing you a stomachache, or worse. I don’t hope so!
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)
At Binary Options Edge our friend @MTH2014 wrote:
using high frequency grid entries is more safer compare to use martingale concept
Would be intereting to calculate if this is true. As this is not a question of money management but of good entries (binary options usually have a pre-determined ending time), I’m not sure how to calculate it … Any ideas, pros or cons?
I do not know in detail how @MTH2014 sets up his grids. But in general this fits perfectly the arguments PipMeUp @ FF lays out in the last chapter of his paper, as stated above. It’s some kind of diversification of your entries trying to even out potential drawdowns caused by short term volatility.
When entering a trade, you do not know in advance how much drawdown that particular trade will face before moving in the desired direction (assuming that the signal we were following was correct). A non-hedged gridded entry will smoothen the potential drawdown caused by volatility and increase profit expectancy to a higher rate than potential loss, if every trade in the grid gets the same SL level.
Additionally we might follow the alternative approach PipMeUp described:
If the probability of a little pullback after the entry is small, that is we expect the price to go directly to the target, we can open two positions at the entry price, the second one with a smaller SL, say half the SL. The risk is increased by 50% while the maximal reward is doubled. R is multiplied by (2 / 1.5), or 33% increase, in the good case. In the bad case R is reduced only of 0.5. This is typical of a “pin bar with the trend” entry.
Some experiments with this alternative approach are on my schedule for the next days to come.
Addendum: I just won’t discuss suicidal tactics like a Martingale approach.
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)
I’m not sure if this could happen, but I think if at all this is the better way of applying hedges. 
- You must be logged in to reply to this topic.