- #143
- Sep 20, 2015 12:20am Sep 20, 2015 12:20am
Hi dude!
Great to see that someone is working with martingale strategy. It's like the taboo of the trading world. People are so scared and will start lecturing you on how it will not work bla bla bla.
Any who I will just try to point out how I use it myself.
Firstly I use a very conservative mm for this strategy. The goal is to make money over the long term and not to be rich overnight.
I choose the direction using fundamentals. After I have selected the direction I will enter my first position. I usually re-average after 300-700 pips depending on the pair and its current range in the market. For some reason if the fundamentals change I take the hit immediately. Let me give a reason why I do that. Say u entered short when USD/JPY was under 85. If you tried to use martingale to recover that trade you would have blown ur account several times lol. So its good to have an uncle point.
So basically that is it. I tried demoing this strategy for couple of months. Results as follows:
https://www.myfxbook.com/members/imt...profit/1172157
No forward testing it with a small live account. You can check the result on my profile. Its the second system called solidfx. The results are not identical as I am also trying to do some short scalp plays and trying to improve it a bit. After another 6 month will increase my bet sizes.
Hope this helps. Will try to pop by here from time to give and give my feedback if u find it useful or else will just shut up.
Great to see that someone is working with martingale strategy. It's like the taboo of the trading world. People are so scared and will start lecturing you on how it will not work bla bla bla.
Any who I will just try to point out how I use it myself.
Firstly I use a very conservative mm for this strategy. The goal is to make money over the long term and not to be rich overnight.
I choose the direction using fundamentals. After I have selected the direction I will enter my first position. I usually re-average after 300-700 pips depending on the pair and its current range in the market. For some reason if the fundamentals change I take the hit immediately. Let me give a reason why I do that. Say u entered short when USD/JPY was under 85. If you tried to use martingale to recover that trade you would have blown ur account several times lol. So its good to have an uncle point.
So basically that is it. I tried demoing this strategy for couple of months. Results as follows:
https://www.myfxbook.com/members/imt...profit/1172157
No forward testing it with a small live account. You can check the result on my profile. Its the second system called solidfx. The results are not identical as I am also trying to do some short scalp plays and trying to improve it a bit. After another 6 month will increase my bet sizes.
Hope this helps. Will try to pop by here from time to give and give my feedback if u find it useful or else will just shut up.
- #146
- Edited 2:20pm Sep 20, 2015 9:10am | Edited 2:20pm
- Joined Sep 2006 | Status: ... | 8,112 Posts
It appears that you are a strict adherent to EMH
Where I differ is the notion that MM cannot be used to enhance performance having achieved positive expectancy
The attached XLS was used to generate the following screenshots (changes from screenshot to screenshot in green highlighter):
Scenario #1: No edge from entries/exits (win probability=50%). Flat betting. Result: Expectancy = 0 (red arrow)
Scenario #2: No edge from entries/exits (win probability=50%). Martingale betting. Result: Expectancy still = 0 (hopefully no explanation required)
Now let's see what happens when we have an edge (but trials remain independent):
Scenario #3: Positive expectancy from entries/exits (win probability=60%). Flat betting. Result: Expectancy = 38.4
Scenario #4: Positive expectancy from entries/exits (win probability=60%). Martingale betting. Result: Expectancy = 86.4
On face value it appears that using martingale has improved expectancy.
However, in reality all martingale has done is is increase the average bet size, and we can achieve the same expectancy using flat betting, thus:
Scenario #5: Positive expectancy from entries/exits (win probability=60%). Flat betting, bet size increased to 2.25. Result: Expectancy = 86.4 (same as in scenario #4)
Martingale is merely used as an example here; the same principle applies with any progressive betting system, i.e. the same expectancy can be achieved with flat betting, by merely adjusting the bet size.
All of the spreadsheets that I posted assume that trials (trades) are independent. I understand that, to whatever extent correlation/dependence exists, it will dilute the rationale in the above scenarios, and make the results inaccurate. I take a default view of independence simply because I know that it's the most 'extreme' scenario, and because market interrelationships are so complex that it's too difficult for me (given that statistics is not my specialist area) to calculate dependencies.
- #147
- Sep 20, 2015 9:53am Sep 20, 2015 9:53am
- Joined Oct 2009 | Status: . | 4,038 Posts
Congratulations ScalaFX for coming out of the closet!
- #154
- Edited 8:40pm Sep 20, 2015 6:11pm | Edited 8:40pm
- | Commercial User | Joined Apr 2013 | 4,366 Posts
{quote} No way! I believe that it's possible to find inefficiencies, and attain positive expectancy, using entries and/or exits. If I believed that markets were perfectly efficient at all times, I'd quit trading and play poker instead.
After reading your comprehensive response I am onboard with your viewpoint entirely. Thanks for the detail as It really helps to isolate the critical elements of trading that actually do contribute to a positive edge in the market which are so hard to feret out. Your post eloquently describes what doesn't contribute to an edge which is very useful in itself.
PS. Your post describes how all things being equal, you can match a Martingale progression using flat betting just by increasing position size, but I think the kicker comes in when you compare maximum drawdowns of each scenario. I don't have a spreadsheet at hand to display this but from my understanding of the risk profiles of both forms of betting, the equity profile of the flat betting scenario is far less volatile despite being able to match expectancy performance through position sizing.
Martingales serve to increase risk volatility and execaerbate drawdowns. In this way, I refer to Martingale progressions as being far less risk efficient than alternative MM approaches under positive expectancy. What this achieves when applying this MM to a portfolio is to significantly reduce your capability of diversifying due to risk exposure constraints applied to a finite capital allocation.....hence why fund managers steer away Martingales, averaging down and also 'pyramiding' as a form of MM and look for alternative more efficient less capital intensive ways to flatten the equity curve through diversification.
I was fortunate enough to approach trading with two inherent biases in mind. Biases can be bad, but these two have saved me from being too enamored by money management styles, too soon:
1. I learned early on, that past behavior does not predict future behavior (and I'm talking, getting this mentally hardwired when I was about 10)
If you flip a fair coin, and get 10 heads in a row, what is the next result going to be - heads or tails?
2. Another early-age lesson: "probability" means nothing, when the event actually happens.
The chances of getting struck by lightening, or winning the lottery are very very low. And yet, both those things do happen.
So if there's a chance that you might have 100 losers in a row, that might just happen.
Or you could phrase it in the way Murphy's law is typically stated: "Anything that can go wrong, will".
[I am such an optimistic that I prefer to say, "Anything that can go wrong, might".]
Regarding Bias 1: yes, there might be strategies, where the result of Trade n is somewhat dependent on the result of Trade (n-1); however, this is very hard to prove, and I haven't come across such strategies myself.
Regarding 2: optimistic thinking might say otherwise, but most strategies involve some form of "what to do when things go differently from what you expected".
Again, I'm not trying to be "negative" or "discourage" people from exploring Martingale. But if graphs and excel worksheets don't help, then you might try to explain logically how a martingale can overcome Murphy's law.
1. I learned early on, that past behavior does not predict future behavior (and I'm talking, getting this mentally hardwired when I was about 10)
If you flip a fair coin, and get 10 heads in a row, what is the next result going to be - heads or tails?
2. Another early-age lesson: "probability" means nothing, when the event actually happens.
The chances of getting struck by lightening, or winning the lottery are very very low. And yet, both those things do happen.
So if there's a chance that you might have 100 losers in a row, that might just happen.
Or you could phrase it in the way Murphy's law is typically stated: "Anything that can go wrong, will".
[I am such an optimistic that I prefer to say, "Anything that can go wrong, might".
]Regarding Bias 1: yes, there might be strategies, where the result of Trade n is somewhat dependent on the result of Trade (n-1); however, this is very hard to prove, and I haven't come across such strategies myself.
Regarding 2: optimistic thinking might say otherwise, but most strategies involve some form of "what to do when things go differently from what you expected".
Again, I'm not trying to be "negative" or "discourage" people from exploring Martingale. But if graphs and excel worksheets don't help, then you might try to explain logically how a martingale can overcome Murphy's law.
