Forex Factory (https://www.forexfactory.com/forum.php)
-   Trading Systems (https://www.forexfactory.com/forumdisplay.php?f=71)
-   -   Capital Weightlifting Championship (https://www.forexfactory.com/showthread.php?t=557430)

Matts Micro Sep 15, 2015 6:35am | Post# 41

Currently short on UJ and close to trade TP. It will be the 3rd net trade of the series and if it completes I will be able to top up my Principle acc. again. Currently at tough support but a nice comment from Kuroda sped things along a bit .
TP. That was the 3rd net trade so transferred winnings minus deposit back to Principal account. I am still on my previous Level but one more series of 3 net trades will get me to the next.

FXEZ Sep 15, 2015 3:21pm | Post# 42

I think I have good news for you!! I did a monte carlo simulation, using 10,000 iterations, repeated the test 3 times. Here are the results Mean Variance Lower CL Upper CL Valid iterations Time taken 4.668 30878 -420 280 10000 23 sec (Test 1) Mean Variance Lower CL Upper CL Valid iterations Time taken 2.544 31697 -430 270 10000 23 sec (Test 2) Mean Variance Lower CL Upper CL Valid iterations Time taken 6.275 29416 -420 270 10000 23 sec (Test 3) I used 0.67 losses 0.33 wins (dice method, no knowledge edge). As you can see you still have a positive...
I can't confirm your results. After running lots of tests and many various iterations of the logic I can't see that there is any advantage. It's likely that you made a mistake in your code that created overly optimistic results. I made a few of these and after fixing, I get no consistent advantage. I get some positive, some negative with lesser/greater extent depending on which specific rule set I'm using.

The other possibility is that I don't understand the rules well enough to implement them in code.

GoldTheHun Sep 15, 2015 3:45pm | Post# 43

{quote} I can't confirm your results. After running lots of tests and many various iterations of the logic I can't see that there is any advantage. It's likely that you made a mistake in your code that created overly optimistic results. I made a few of these and after fixing, I get no consistent advantage. I get some positive, some negative with lesser/greater extent depending on which specific rule set I'm using. The other possibility is that I don't understand the rules well enough to implement them in code.
There is always a possibility that I made an error somewhere..

The rules are, as I understand:

TP= 2, SL=1.
Start trading with a certain risk, say $10
win = double risk
loss = go back 1 step on risk until initial risk, than if you keep on losing the initial risk stays the same.
every time you win, counter = +1
every time you lose, counter = -1 (the counter never goes below 0)
if the counter = 3, bank all money and start from scratch

Thats it.
I checked my sheet and the rules are correct, no error there. You have to make profitable trades %33, losing trades %67, using binomial random number distribution.

PipMeUp Sep 15, 2015 4:42pm | Post# 44

1 Attachment(s)
If I understand the rules correctly that's just the good ole reverse martingale limited to the 3rd level.
If the system has no edge it will lose because it will randomly hit the bankruptcy level at some point (RW).
If the system does have an edge 1% MM will always beat it.
In the excel I offer an edge to the trader of 40% winrate with RR=2 (expectency of 0.2). Not realistic but clearly shows the point.
revMarty.xlsx

GoldTheHun Sep 15, 2015 5:31pm | Post# 45

If I understand the rules correctly that's just the good ole reverse martingale limited to the 3rd level. If the system has no edge it will lose because it will randomly hit the bankruptcy level at some point (RW). If the system does have an edge 1% MM will always beat it. In the excel I offer an edge to the trader of 40% winrate with RR=2 (expectency of 0.2). Not realistic but clearly shows the point. {file}

Well I apologize everybody. The simulation PipMeUp made is the correct one. I messed up one "IF" command in rules. Sorry for giving a false hope... Without the edge the mm becomes a straight loser...
Thank you Pip for posting this ..

FXEZ Sep 15, 2015 11:31pm | Post# 46

If I understand the rules correctly that's just the good ole reverse martingale limited to the 3rd level. If the system has no edge it will lose because it will randomly hit the bankruptcy level at some point (RW). If the system does have an edge 1% MM will always beat it. In the excel I offer an edge to the trader of 40% winrate with RR=2 (expectency of 0.2). Not realistic but clearly shows the point. {file}
Thanks for posting this. If this is just a reverse martingale then I guess my martingale radar may be broken. But it appears that the transferring back and forth between subaccounts is either critical or just sleight of hand?

Could the OP please comment on whether PipMeUp's spreadsheet faithfully reproduces the betting progression or to what extent it departs from it?

By the way, this exercise sort of reminded me of Shannon's Daemon, also Volatility Harvesting or more recently Cover's Universal Portfolio (post #3 on NuclearPhynance thread). But you need something that doubles/halves with 50% probability for Shannon's Daemon to work, perhaps a unicorn. In other words, a 2:1 reward:risk strategy with a 50% win rate. That alone is a winner so you don't really need the fancy MM? Is the proof of this already in PipMeUp's Excel sheet?

Matts Micro Sep 16, 2015 1:05am | Post# 47

If I understand the rules correctly that's just the good ole reverse martingale limited to the 3rd level. If the system has no edge it will lose because it will randomly hit the bankruptcy level at some point (RW). If the system does have an edge 1% MM will always beat it. In the excel I offer an edge to the trader of 40% winrate with RR=2 (expectency of 0.2). Not realistic but clearly shows the point. {file}


Firstly, thankyou guys for this fantastic work.

I am just nutting over the formula now. I am unfamiliar with the RAND function but I think I can see how it's working.

From the outset I can see something isn't right however. Every time the +3 net trades are made (or 8 in column B I think in this example) then the process begins again and there is at least a 50% gain (that is with my system but in formula it is an 80% gain) in capital.

I will have a look to see if I can adjust myself.

Thanks again guys I appreciate it.

Matt.

Matts Micro Sep 16, 2015 6:48am | Post# 48

1 Attachment(s)
I have finished my adjustments to the test. It now works as per strategy. I have to apologise for my lengthy formula as Ima self taught noob on spreadsheets and the only way I know

My strategy is volatile but highly successful using this test, which was to be expected (large lot trading is rarely not volatile).

What to look for is the "risk" column. It is aiming for +3 winning trades or a -1 losing trade from the starting position. It will deposit when +3 trades or withdraw when -1. It withdraws 5% of the Level column and deposits 80% which is the correct risk system.

I tried my theory that I haven't really discussed much of withdrawing just 50% but it didn't work well, which is food for thought for me.

Thankyou again PipMeUp and others for helping with this.

Cheers
Matt.

Ps. If anyone spots any mistake pls let me know so I can have a look at it.

Pss just changed a bit of code because of a mistake. It doesn't seem to change results by a huge amount (perhaps a little more volatile maybe).
Capital Weightlifting Equity.xlsx

Matts Micro Sep 16, 2015 7:15am | Post# 49

{quote} Thanks for posting this. If this is just a reverse martingale then I guess my martingale radar may be broken. But it appears that the transferring back and forth between subaccounts is either critical or just sleight of hand? Could the OP please comment on whether PipMeUp's spreadsheet faithfully reproduces the betting progression or to what extent it departs from it? By the way, this exercise sort of reminded me of Shannon's Daemon, also Volatility...
Hi FXEZ, thanks for links. I don't agree with martingale systems (at all) and this isn't martingale. I am only using 5% risk of the account (I use this higher risk but it can be used for any % of the account).

It is the activity in the trading account (that is separate from your main capital) that makes the difference. Basically I am using 5% of my capital just as I would in one of 50/100% trading strategies. I try to accomplish this 3 times. If I succeed 3 times I withdraw the profits and that 5% has now become 80% of my overall capital and start again with 5%. If I win one trade and then lose the next I am still left with the original 5% risk. I do this until I either hit +3 trades or lose that 5% and deposit another 5%. So one 5% deposit can often harvest around 5-20 up and down trades until I hit either the bottom or top.

Matts Micro Sep 16, 2015 7:39am | Post# 50

On top of the simulator I also have to mention (again) that I don't believe that my winning streaks and losing streaks are entirely random. I have noticed that when the market is moving with confidence I can easily put together 3+ trades but in uncertainty, a time I try to avoid (like the next couple of days) it is difficult to put together more than a couple of winning trades. Next week however after FOMC people will be more confident to position themselves... depending on the outcome of course. So my system is built to accommodate that wave like of winning and losing. When losing only minimal risk is involved but when winning maximum profits are achieved.

Matts Micro Sep 16, 2015 8:42am | Post# 51

1 Attachment(s)
I noticed a slight error in the code. It wasn't much and doesn't noticeably change the results.

I will replace the original too.
Capital Weightlifting Equity.xlsx

FXEZ Sep 16, 2015 10:01am | Post# 52

{quote} Hi FXEZ, thanks for links. I don't agree with martingale systems (at all) and this isn't martingale. I am only using 5% risk of the account (I use this higher risk but it can be used for any % of the account). It is the activity in the trading account (that is separate from your main capital) that makes the difference. Basically I am using 5% of my capital just as I would in one of 50/100% trading strategies. I try to accomplish this 3 times. If I succeed 3 times I withdraw the profits and that 5% has now become 80% of my overall capital...
Thanks for updating the spreadsheet and your posts. I will follow up after I can get a clear look at the changes and have some time to study them.

I should have said reverse Martingale, or adding into winners to be more precise.

Matts Micro Sep 16, 2015 11:08am | Post# 53

{quote} Thanks for updating the spreadsheet and your posts. I will follow up after I can get a clear look at the changes and have some time to study them. I should have said reverse Martingale, or adding into winners to be more precise.
Yeah I missed the reverse part. I had to look it up but by looks of it, it may be similar.

On the spread sheet if you wanted to change the risk from 5% (is at high) it can be changed on the P/L column in the formula. 1% Risk = 16% reward. 1% (or what ever risk your using) is always kept in the account as a buffer zone for margin so the deposit of +1% from the Capital (Principal) account will = 2%. 3x 2:1 trades (from 2% deposit) will be 4,8,16%.

set on 1% it almost always beats the 1% static account and quite often a great deal more with obviously a cleaner equity line.

Like I said in the opening, I'm sure it's not a new strategy and even a check with google seems to bring up similar systems but I just think it suits my style of trading with multiple wins followed by a couple of losses.

Cheers,
Matt

FXEZ Sep 16, 2015 11:31am | Post# 54

1 Attachment(s)
{quote} Yeah I missed the reverse part. I had to look it up but by looks of it, it may be similar. On the spread sheet if you wanted to change the risk from 5% (is at high) it can be changed on the P/L column in the formula. 1% Risk = 16% reward. 1% (or what ever risk your using) is always kept in the account as a buffer zone for margin so the deposit of +1% from the Capital (Principal) account will = 2%. 3x 2:1 trades (from 2% deposit) will be 4,8,16%. set on 1% it almost always beats the 1% static account and quite often a great deal more with...
Hi Matt, yes I've been playing around with various values. I think the pnl column in your sheet may overstate profit due to not deducting all the amounts for losses. I just edited / simplified that column's formula and re-titled "Principle [sic] Account" as "Bank" to more accurately reflect its role. So now after each trade, profits/losses are swept to the Bank account, and because trade size is based on Level, it should still be correct.

So to summarize, I think the concept can be simplified it to just use the Bank (equity) instead of the idea of principal / trading accounts and the complexity of transferring back and forth. Since we compute the Level from Bank, and the risk from past wins/losses, we can compute trade size and P&L properly. So I edited the pnl column to keep track of all +/- wins/losses and each time it is accumulated back to the bank, so the equity curve shows the true account value.

By pressing F9 repeatedly in Excel the sheet will recalculate and it's possible to see how much variation there is in different equity curves, effectively producing a different Monte-Carlo run each time.
Capital Weightlifting Bank.xlsx

Good Lookin Sep 16, 2015 11:46am | Post# 55

The excel that PipMeUp has posted is indeed a reverse martingale system and is different from what Matts Micro has posted. The main difference is that when there is a loss, reverse martingale will have you start back at the original risk while Matts Micro will have you back at your last risk.

Back when Rags2Riches was 1st active with his thread, i tested this exact same money management. When compared to a win 2 lose 1 50% win rate compounding system i found that this system did perform slightly better. But of course, the trade off is larger draw downs and swings into appreciation. Of course you must also take into account that your trading system is largely the cause of success in this money management system. If you have more streaks of losses or wins then this system will outperform regular 2 win 1 loss compounding in a good or a bad market.

Matts Micro Sep 16, 2015 12:07pm | Post# 56

{quote} Hi Matt, yes I've been playing around with various values. I think the pnl column in your sheet may overstate profit due to not deducting all the amounts for losses. I just edited / simplified that column's formula and re-titled "Principle [sic] Account" as "Bank" to more accurately reflect its role. So now after each trade, profits/losses are swept to the Bank account, and because trade size is based on Level, it should still be correct. So to summarize, I think the concept can be simplified it to just use the Bank (equity) instead of the...
Hi FXEX,
I really appreciate the time your putting toward this. I think you may of misread my sheet (it's a mess I know heh). I am going to try to add another column with the trading account balance to show you what I mean (a bit slow with these thing

I represented a 3rd net win (3 wins in total) with a zero (0). I used this so I could formulate it. I think your sheet references that 0 as a loss and hence it is only able to produce a two winning streak. As I said I will try another column to show you what I mean.

Cheers
Matt.

Edit: Lol I think I was about there and my spread sheet crashed... still have partial formula....

24hrs Sep 16, 2015 12:36pm | Post# 57

My question is ..

IS IT POSSIBLE TO USE THIS STRATEGY IN BINARY OPTIONS TRADING? PLS EXPLAIN IF SO.

FXEZ Sep 16, 2015 1:23pm | Post# 58

{quote} Hi FXEX, I really appreciate the time your putting toward this. I think you may of misread my sheet (it's a mess I know heh). I am going to try to add another column with the trading account balance to show you what I mean (a bit slow with these thing I represented a 3rd net win (3 wins in total) with a zero (0). I used this so I could formulate it. I think your sheet references that 0 as a loss and hence it is only able to produce a two winning streak. As I said I will try another column to show you what I mean. Cheers Matt. Edit: Lol...
Matt, I was reading when risk goes 1, 2, 4 as the 3 sequence, then the zero is the reset. If risk goes 1,2,4,8 isn't that 4 levels instead of 3?

sinqua Sep 16, 2015 1:25pm | Post# 59

My question is .. IS IT POSSIBLE TO USE THIS STRATEGY IN BINARY OPTIONS TRADING? PLS EXPLAIN IF SO.
Err, I think you have jumped the gun, this is not a trading strategy, it is just one part, the Money Management part, if you do not have sound price analysis you will still end up -ive.

But yes you can use it for whatever type of trading you desire.

Matts Micro Sep 16, 2015 1:39pm | Post# 60

{quote} Matt, I was reading when risk goes 1, 2, 4 as the 3 sequence, then the zero is the reset. If risk goes 1,2,4,8 isn't that 4 levels instead of 3?


you're right, something is very wrong. Will have to get back to you on this..


© Forex Factory