Recovery system with a different twist

User avatar
SteveHopwood
Owner
Posts: 9754
Joined: Tue Nov 15, 2011 8:43 am
Location: Misterton - an insignificant village in England. Very pleasant to live in.

Re: Recovery system with a different twist

Post by SteveHopwood »

gaheitman wrote:The algorithm you describe is almost exactly how it is implemented in ChangeTheColor. And it is written in such a way that it could be changed to support other compounding systems.

George
Have we got the source for ChangeTheColor? Did I miss it whilst speed-reading this thread?

:D
Read the effing manual, ok?

Afterprime is the official SHF broker. Read about them at https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?p=175790#p175790.

I still suffer from OCCD. Good thing, really.

Anyone here feeling generous? My paypal account is always in the market for a tiny donation. [email protected] is the account.

To see The Weekly Roundup of stuff you guys might have missed Click here

My special thanks to Thomas (tomele) for all the incredible work he does here.
User avatar
gaheitman
Trader
Posts: 655
Joined: Tue Nov 15, 2011 10:55 pm
Location: Richmond, VA, US

Re: Recovery system with a different twist

Post by gaheitman »

SteveHopwood wrote:
gaheitman wrote:The algorithm you describe is almost exactly how it is implemented in ChangeTheColor. And it is written in such a way that it could be changed to support other compounding systems.

George
Have we got the source for ChangeTheColor? Did I miss it whilst speed-reading this thread?

:D
Yep, it's there. Look for CalculateLotSizeShake(). There is a lot to read there, Mike's been busy!

George
User avatar
SteveHopwood
Owner
Posts: 9754
Joined: Tue Nov 15, 2011 8:43 am
Location: Misterton - an insignificant village in England. Very pleasant to live in.

Re: Recovery system with a different twist

Post by SteveHopwood »

Doh. :oops:

:D
Read the effing manual, ok?

Afterprime is the official SHF broker. Read about them at https://www.stevehopwoodforex.com/phpBB3/viewtopic.php?p=175790#p175790.

I still suffer from OCCD. Good thing, really.

Anyone here feeling generous? My paypal account is always in the market for a tiny donation. [email protected] is the account.

To see The Weekly Roundup of stuff you guys might have missed Click here

My special thanks to Thomas (tomele) for all the incredible work he does here.
esquire
Posts: 4
Joined: Sat Dec 10, 2011 2:29 pm

Re: Recovery system with a different twist

Post by esquire »

Did some simulation on this topic using R code. Like all these systems, there is a point where the current loss is so big you have to stop. Now let's take the following parameters:

- Chance of winning is 0.49 (i'm using a 1:1 ratio here, so this would be like random trading with spread 2 and sl=tp=100)
- Maxloss is 100 base units. Asumming most traders are trading at a 1% risk, this would be a margin call. (Technically the margin call would lie around 95, but whatever :p)
- Simulate a 1000 trades (or 'cycles' as they are called here)
- Simulate that a 1000 times.

Now, the expected return with this system is -359 base units per 1000 trades, the expected return for normal trading (equal lots) is -20 per 1000 trades.

Now let's turn the odds in our way: we win 60% (meaning average win * percentage win + average loss * percentage loss = 0.6). Using the system would be huge: the expected return will be 1511,809 on a 1000 trades, while normal trading would only deliver 199,37 base units.

(For the clever fellows here: 1511 in a 1000 cycles is possible because sometimes the lotsize will be for example 5 and the currenct count of the cycle only -1, so if you win you win 4 lot's instead of 1)

So give me some realistic parameters for your systems and i'll do the simulation. I need:
- average win * percentage win + average loss * percentage loss
- maximale allowable base units to lose
- estimated trades per timeperiod (e.g. 5 trades a week), so i can give time restricted estimations
User avatar
gaheitman
Trader
Posts: 655
Joined: Tue Nov 15, 2011 10:55 pm
Location: Richmond, VA, US

Re: Recovery system with a different twist

Post by gaheitman »

esquire wrote:(For the clever fellows here: 1511 in a 1000 cycles is possible because sometimes the lotsize will be for example 5 and the currenct count of the cycle only -1, so if you win you win 4 lot's instead of 1)
Actually, the rules have us only trading enough to win one base unit, so in your example you would only trade 2. Not saying what you coded isn't better...

George
esquire
Posts: 4
Joined: Sat Dec 10, 2011 2:29 pm

Re: Recovery system with a different twist

Post by esquire »

gaheitman wrote:
esquire wrote:(For the clever fellows here: 1511 in a 1000 cycles is possible because sometimes the lotsize will be for example 5 and the currenct count of the cycle only -1, so if you win you win 4 lot's instead of 1)
Actually, the rules have us only trading enough to win one base unit, so in your example you would only trade 2. Not saying what you coded isn't better...

George
Yes, just read that. I'll adjust the code tonight and add a ruin probability.
User avatar
Iamshakey
Trader
Posts: 55
Joined: Wed Nov 16, 2011 7:29 am
Location: Midwest USA

Re: Recovery system with a different twist

Post by Iamshakey »

SteveHopwood wrote:I am astonished by the creativity unleashed at this site. Everywhere I look, someone is developing something fantastic. Had I thought this would happen, I would have pissed of TIT a long time ago. :lol:

Shake, this one sent me scurrying for a bath-lol. :lol:

What we need is a module that can slot into the code shell, then be retro-fitted into existing ea's where appropriate, hived out into a couple of buy/sell scripts for the manual traders, then exported to a simple monitoring ea so the manuals can keep track of the position.

Here is my understanding of all this
  • - each trading cycle aims to make 1 unit of profit.
    - a unit is defined as the take profit multiplied by the base lot size.
    - if we achieve that profit from the first trade, then it is smug, self-satisfied smirk time
    - if not we take the next trade at base lot size
    - when a trade wins, take the next trade at the winner lot size + base lot
    - keep a tally of the entire 'basket' and when the upl reaches 1 unit of profit, we can start the cycle again
Forget the technicals of the coding guys - I can see to that. Here is the logic I arrived at in the bath:
  • 1) the EA regards all trades as part of a basket of trades, whether open or closed.
    2) when it sends the first trade in the cycle, the ea needs to leave behind some sort of indication that this is the first. Override the user's trade comment choice and place a '1' in the comment field.
    3) then, whenever the ea spots a trading opportunity, it cycles back through the trades in the History tab until it either reaches a trade with '1' in the comment field. As it goes, it adds up the combined upl of order profit, swap and commission.
    4) if the upl >= I unit of profit, or there are no relevant trades in the History tab, the bot can yell, "Party time" and return to no 2)
    5) if the upl < 1 unit of profit, it can examine the most recent trade in the History tab and:
    • - use base lot size if the trade was a loser.
      - use order lot size + base lot if it was a winner.
Sound ok?

:D
Sorry all you guys, for taking so long to reply. I've been busy, but not THAT busy. I started the thread THUS I should be responsive to the posts on it. This is REALLY SIMPLE, but DIFFICULT to explain. Anyone having difficulty getting my meaning, please post your questions, or PM me. I will be infinitely patient, as most of you have been with me.

STEVE:
I THINK you've got it, but I am not sure if you got what happens after this bit . . .


"- when a trade wins, take the next trade at the winner lot size + base lot
I WOULD ADD
- when a trade loses, return to base lot
continue to trade at base lot until have another win
-AFTER THE 2ND WIN take the next trade lost size X3
-AFTER THE 3RD WIN take the next trade lot sizeX4
- keep a tally of the entire 'basket' and when the upl reaches 1 unit of profit, we can start the cycle again"when the next loss occurs

so the progression looks like this.... we'll just throw some stuff out there regardless of lot size and account balance. Forget the math, and we won't worry about break even. . . This way you can see the progression of trade sizes . . .

PROGRESSION;
WIN======> NEXT LOT 1
WIN======> NEXT LOT 1
WIN======> NEXT LOT 1
LOSS=====> NEXT LOT 1 (THIS STARTS THE CYCLE)
LOSS=====> NEXT LOT 1
WIN======> NEXT LOT 2
WIN======> NEXT LOT 3
LOSS=======>NEXT LOT 1
LOSS=======>NEXT LOT 1
WIN=======>NEXT LOT 4
WIN========>NEXT LOT 5
LOSS=======>NEXT LOT 1
LOSS=======>NEXT LOT 1
WIN========>NEXT LOT 6

and so on. . . NOW, THIS IS IMPORTANT . . . WHEN THE GOAL IS ALMOST REACHED, ONLY TRADE BIG ENOUGH TO REACH THE GOAL. LETS SAY YOU ARE ON THE LOT SIZE X5, AND YOU ONLY NEED LOT SIZE X3 TO REACH GOAL, TRADE ONLY THE LOT SIZE X3. THIS WAY, SAVE A LITTLE OF THE RECOVERY WITHOUT TUMBLING ALL THE WAY BACK DOWN THE HILL. IT IS RARE FOR THIS TO GO GREATER THAN LOT SIZE X3. notice, unlike martingale, you can take successive losses without too much damage to your account.

This strategy is pretty much bullet proof with a decent trade plan. I have used this for a year and a half with my manual trading. If you veiw some of my previous posts, you will also see that I keep track of the pairs I trade, so that I can be running the cycle on more than on pair at a time. I try to focus on making each pair an eventual winner. If this is done, the account balance will take care of itself. The profits can come slowly, but they are nicely consistent.

Thanks Steve, for taking an interest in this. I really hope you and others can find a use for it in some of your EA's. I've gotten a lot from many of you guys, and I hope to give back at lease a little.

Shake

Edit, in re-reading, and further digesting especially the second paragraph about how the coding would basically work, I seems correct to me :lol:
Last edited by Iamshakey on Sun Jan 29, 2012 9:30 pm, edited 1 time in total.
It's Better to Burn Out, than it is to Rust
User avatar
Iamshakey
Trader
Posts: 55
Joined: Wed Nov 16, 2011 7:29 am
Location: Midwest USA

Re: Recovery system with a different twist

Post by Iamshakey »

Oh yeah,

I am sure you have probably thought of this . . . profits and losses can be pretty variable with trading, due to slippage, swaps etc, I think at times you would need this to do the "logical" thing. It would be a shame to come within a few cents of the goal. Perhaps something could be written in that would end the current cycle if balance reached goal + or - "X"

Also might consider a "to hell with it" routine of some sort. This way, if the recovery has gone several cycles deep, and finally fights it's way back CLOSE to BREAKEVEN, the EA says to hell with it, and ends the current cycle with a small loss. This way, accepting the small loss can reset and perhaps the next cycle will move on to gain you some profits.


Just a couple of thoughts

Shake
Last edited by Iamshakey on Sun Jan 29, 2012 9:34 pm, edited 2 times in total.
It's Better to Burn Out, than it is to Rust
User avatar
Iamshakey
Trader
Posts: 55
Joined: Wed Nov 16, 2011 7:29 am
Location: Midwest USA

Re: Recovery system with a different twist

Post by Iamshakey »

Khalaad wrote:In my early trading years I was eager to do everything. For a while in 1986 (I think) I subscribed to Phyllis Kahn’s Gann Angles. After I had been a subscriber for a few months I had an occasion to talk to Phyllis over the telephone. The conversation was quite enjoyable until Phyllis asked: “But have you made any money?”

Of course, I hadn’t. Up until then I would simply read the newsletter every month, but hadn’t bothered to act a single advice.

Nowadays, whenever I come across discussions like this thread, I always remember Phyllis Kahn’s words.

Khalid
Khalid, great point.
Ordinarily I agree that with most recovery gimmicks like this, the center just doesn't hold in the long run. Sooner or later, the ONE THING that will destroy a system WILL HAPPEN, at least if left to it's own devices.

However, I have used this recovery system for about a year and a half, and I promise you, it has always worked. It hasn't made me big money, but it has made me small steady money. HOWEVER, and I have maintained this:
1. You still must have a good trade system.
2. You still must trade lot sizes that don't badly damage your account while taking successive losses.
3. Reasonable risk reward is always a good thing.

You are obviously a pretty seasoned trader, and I am sure I can speak for others when I say, we value your input, keep giving us your thoughts. :idea:

Shake
It's Better to Burn Out, than it is to Rust
esquire
Posts: 4
Joined: Sat Dec 10, 2011 2:29 pm

Re: Recovery system with a different twist

Post by esquire »

I did some more calculations and thought experiments, and the conclusion about all lot size ideas is pretty simple:

It works as long as your system is any good

If your system makes profit, and you add a recovery system like this, you will make more money (assumed that you won't trade extremly big lot sizes). It's simply because you trade bigger lot sizes, multiplied with an average win = more money.

It's rather rediculous to put efford into developing things like this. This is assuming you should look at trades as a bundle but you shouldn't (well maybe for psychology but not for the math side). All trades are indepentent and that's the way they should be threated. You should only increase the lot size if you reduce the risk or - for whatever reason - believe there is a bigger probability of winning.
Post Reply

Return to “Trading Psychology & Money Management”