100% Win Math Grid Ea | Page 2 | Forex Factory

Options

Go to
Permalinks
Print Thread

Similar Threads

Grid - New indicator that replaces chart grid 60 replies

Grid (not grid system) or horizontal lines indicator 6 replies

100% win grid EA 45 replies

Grid, Burst Grid, Grid Risk and Grid Profit 1 reply

Double Time Grid and Pip Grid 0 replies

Attachments: 100% Win Math Grid Ea
Exit Attachments

100% Win Math Grid Ea

  • #21
  • Sep 27, 2008 5:50am Sep 27, 2008 5:50am
Quoting FOREXflash
There you go......

external variables:

extern bool CONTINUE=true;\\use this if you want to stop trading(false) after TP are hit
extern bool UseEntryTime=false; \\ set this to true
extern int EntryTime=0; \\ hour you want trades to be executed
Incredible fast answer!!!! congrats and thanks again.
  • #22
  • Sep 27, 2008 6:35am Sep 27, 2008 6:35am
thats weird ive got the ea grid indicator on my mt4 platform by the way thanks for sharing it but for some reason i cant attach it to a chart it just wont respond when i double click on it the normal way i would add any indicator can someone help me out ive used these charts for years
  • #23
  • Sep 27, 2008 6:48am Sep 27, 2008 6:48am
This is not indicator, it is a expert advisor, you must put it in experts folder,
then restart mt4, and then go to experts and double click on mGRID EA.mq4!
forexflash
  • #24
  • Sep 27, 2008 6:54am Sep 27, 2008 6:54am
good man yourself
  • #25
  • Sep 27, 2008 8:25am Sep 27, 2008 8:25am
Explanation of external inputs:

INCREMENT=35; \\ pips betwen each level
LOTS=0.1; \\ size of starting lots
LEVELS=3; \\ number of levels below and above the price
MAX_LOTS=99; \\ if opened trades exceds 99lots all trades will be closed
MAGIC=1803; \\ mag. number
CONTINUE=true; \\ will ea continue to trade after closing of first set
MONEY_MANAGEMENT=false; \\ money managment
RISK_RATIO=2; \\ if used with true MM, risk ratio works..it can be from 1-10
UseEntryTime=false; \\ do you want to trade at specific time of day
EntryTime=0; \\ on what time do you want trades to be executed (0-24)
forexflash
  • #26
  • Sep 27, 2008 10:49am Sep 27, 2008 10:49am
What is the stop for this method, if any? Also, you recommend eurusd because you said it is volatile, but there are other pairs even more volatile, such as gbpusd - so why not trade them too with this EA?
  • #27
  • Sep 27, 2008 10:54am Sep 27, 2008 10:54am
eagle4x is back
Hey man whats up?
You can use this on any pair you want, just need to be a pair with big range.
Every trade has a TP and SL!
forexflash
  • #28
  • Sep 27, 2008 11:19am Sep 27, 2008 11:19am
Not much, just looking at different strategies. I don't see a SL amount mentioned in this thread, and I didn't see it in the EA. What is the SL amount?

Quoting FOREXflash
eagle4x is back
Hey man whats up?
You can use this on any pair you want, just need to be a pair with big range.
Every trade has a TP and SL!
  • #29
  • Sep 27, 2008 11:23am Sep 27, 2008 11:23am
Just execute ea in tester on visual mode and all will be clearer to you!
Forget about changing the SL and TP, becouse ea will not do what it has to do,
and that is taking profit everytime.
forexflash
  • #30
  • Sep 27, 2008 12:13pm Sep 27, 2008 12:13pm
Quoting eagle4x
Not much, just looking at different strategies. I don't see a SL amount mentioned in this thread, and I didn't see it in the EA. What is the SL amount?
Eagle, please see my diagram in post #8.
-- The TP for the sell orders, and SL for the buy orders, are at the same point: 140 pips (i.e. 4 x the increment) below the price at the time the original buy/sell orders were created.
-- The TP for the buy orders, and SL for the sell orders, are at the same point: 140 pips (i.e. 4 x the increment) above the price at the time the original buy/sell orders were created.

Like ForexFlash says, the TP and SL is kind of irrelevant:
-- When price reaches the upper TP/SL point, there will always be more buy orders open than sell orders, hence profit is guaranteed.
-- When price reaches the lower TP/SL point, there will always be more sell orders open than buy orders, hence profit is guaranteed.

As long as price reaches one of these levels before the account runs out of available margin, each set of trades is guaranteed to win.

David
  • #31
  • Sep 27, 2008 12:36pm Sep 27, 2008 12:36pm
Quoting FOREXflash
Hanover thanks for investing your time in this!
You got the method right.........
FF, I've reached the conclusion that this EA is basically a cleverly disguised Martingale. The Martingale will always win as long as the trader has infinite capital in his trading account; likewise with this method.

Here's the 'problem'. There's no guarantee that price won't whipsaw indefinitely before it reaches either TP/SL level, creating vast numbers of orders, and eventually exceeding the available margin. This is analagous to the Martingale's 'death trade'. No matter how you try to vary the parameters to reduce this risk, the risk will never be zero, which is exactly the same as with the Martingale.

Of course you can reduce the initial position sizes - e.g. find a broker who allows microlots - but then you're reducing your returns by a factor of 10 also. As with all MM systems, I'm afraid there's no way of decreasing risk without decreasing return, and vice versa. They are inherently proportional to one another.

If forex was this easy, we'd all be millionaires.

Ingenious approach, though.

David
  • #32
  • Sep 27, 2008 12:48pm Sep 27, 2008 12:48pm
Thanks
forexflash
  • #33
  • Sep 27, 2008 2:52pm Sep 27, 2008 2:52pm
Hi there

Thanks to FF for the EA, been having some fun with it over the weekend. And to Hanover for your insights into the potential pitfalls.

It would seem that there are a number of ways to mitigate the dangers outlined by Hanover. A progressive narrowing of the take profit positions as available margin is reduced could be one way, or a progressive reduction of the number of levels. Or a combo of the 2.

Having not tested it forward i can't say for sure, however I imagine that some of the positions could be manually closed thus freeing up trading equity.

It will be interesting to see what solutions evolve as this one gets kicked around

Thanks again for sharing.

James
  • #34
  • Sep 27, 2008 3:49pm Sep 27, 2008 3:49pm
Quoting hanover
Eagle, please see my diagram in post #8.
-- The TP for the sell orders, and SL for the buy orders, are at the same point: 140 pips (i.e. 4 x the increment) below the price at the time the original buy/sell orders were created.
-- The TP for the buy orders, and SL for the sell orders, are at the same point: 140 pips (i.e. 4 x the increment) above the price at the time the original buy/sell orders were created.

Like ForexFlash says, the TP and SL is kind of irrelevant:
-- When price reaches the upper TP/SL point, there will always be more buy orders open than sell orders, hence profit is guaranteed.
-- When price reaches the lower TP/SL point, there will always be more sell orders open than buy orders, hence profit is guaranteed.

As long as price reaches one of these levels before the account runs out of available margin, each set of trades is guaranteed to win.

David
Do you continue to double the lots at the previous levels in the event whereby price almost got to a TP level but turn around only to hit the opposite TP level?
  • #35
  • Sep 27, 2008 5:57pm Sep 27, 2008 5:57pm
Quoting aelimian
Do you continue to double the lots at the previous levels in the event whereby price almost got to a TP level but turn around only to hit the opposite TP level?
Aelimian, assuming the starting position size is 1 unit, then the maximum size on any individual position is only ever 3 units. But every time a new buystop order gets triggered, additonal sellstop orders (at the original levels) get added; and vice versa. So if price continues to zigzag between the TP/SL levels without actually reaching them, the number of orders that could open is potentially infinite (althoughb I note that there is a MAX_LOTS parameter). I'm currently working my way through examples to try to discover exactly how many orders get open, and when, and what effect that this has on available margin.

David
  • #36
  • Edited Sep 28, 2008 10:28am Sep 27, 2008 7:41pm | Edited Sep 28, 2008 10:28am
Must admit that my first thought was to decrease the INCREMENT value, on the basis that price is more likely to reach a smaller SL/TP more quickly than a larger one. However, this means that the intervening levels (where additional buystop/sellstop orders are placed) are going to get triggered more frequently also.

So the key is to somehow reduce risk by trading when the probability of long, clean moves is greater than 'normal'. (Having said that, I don't believe that it's possible to eliminate risk altogether). Here are some possibilities for your consideration:

-- when prices are moving at greater than normal volatility
-- when the higher timeframe charts are trending strongly (like we had in August)
-- prior to major news announcements
-- when price is apparently breaking out of significant congestion, or through significant S/R

Of course I realize that these ideas are not necessarily easy to code.

IMHO, one point this EA method has in its favor is that it averages up instead of down into positions, and (If I may generalize) I believe that forex trends well enough to favor this approach.

There's an interesting thread here, where a lot of thought has been given to reducing risk in a Martingale style MM scenario. In post #232, Nanningbob outlines some of his ideas. The way I see it, while runaway trends present the greatest danger to Azam's EA, it's prolonged sideways movement that's the greatest threat here. So we could perhaps invert some of the concepts in post #232.

Also in this vein, I wonder if it's possible to reduce risk by somehow 'hedging' opposing (i.e. 'averaging up' versus 'averaging down') Martingale methods against each other. Anybody got any thoughts?

David
  • #37
  • Edited 8:38pm Sep 27, 2008 8:00pm | Edited 8:38pm
Quoting hanover
Here's the 'problem'. There's no guarantee that price won't whipsaw indefinitely before it reaches either TP/SL level, creating vast numbers of orders, and eventually exceeding the available margin. This is analogous to the Martingale's 'death trade'. No matter how you try to vary the parameters to reduce this risk, the risk will never be zero, which is exactly the same as with the Martingale.
An important point, David.
If you don't search the tester log you will not know how often this is happening. This EA will not work near as good as many tester graphs show. Be sure to look in your tester log file for 'no enough money' messages. In a real account being this close to the margin limit would be a problem. In tester it just continues. I've been running backtests on this EA that produced a good earnings graph showing a good gain. But the log is full of trades that could not be opened. The MAX_Lots is certainly not a cumulative lots setting.

Here is an example of what to look for. Just do a search for the word money in the log file.
Inserted Code
19:44:45 2006.01.18 17:42  Tester: not enough money for sell 0.60 EURUSD at 1.2085 sl: 1.2167 tp: 1.2067 [2006.01.18 17:42]
19:44:45 2006.01.18 17:42  Tester: PrevBalance: 34962.56, PrevPL: -4656.00, PrevEquity 30306.56, PrevMargin: 29787.92, NewMargin: 30452, FreeMargin: -145.12
19:44:45 2006.01.18 17:42  Tester: pending order is deleted [no enough money]
EDIT:
I want to thank FOREXflash for posting this EA. Even if I have doubts about it I enjoy testing and attempting to modify EAs that are posted on the forum. Every weekend I try another couple. I always hope every EA will become profitable after the correct modifications. This post is meant as a warning not a statement that the EA is unprofitable.

I also appreciate Hanover's posts. I wish I had his writing skills to explain ideas.
  • #38
  • Sep 28, 2008 1:11am Sep 28, 2008 1:11am
Guys thank for your time and ideas!

Nvwine I must say that I dont agree with you.
I tested this live for almost 2 months on 2 accounts using
different lots and sizes of acc.........and both of them survived.
One account was 500 usd using 0.01 lots!!!
The reality is that this method proved to be beter on forward tests.
Backtest isnt so good becouse of the tester interpolation......candles
tend to whipsaw more......
It is relatively easy to see is the tester ok.....just look at the graph, after every drawdown
there must be a oposite up wich is above drawdown.......if some of trades isnt executed, the graph
will show that.......if any of the trade isnt executed becouse of small margin, you will end up in loss.

Hanover i really like your idea of inverted "hedging" ea....
but i dont know is that plausable.....
I allready tried that, I run the ea on 2 reverse
correlated pairs, but I didnt tested enough to know will it work.

Respect
forexflash
  • #39
  • Sep 28, 2008 2:15am Sep 28, 2008 2:15am
I ran Strategy Tester on USDCHF using an initial equity of $10k and with LOTS=1 (which means that position sizes are 1, 2 or 3 lots). I searched the text in the log file (....\tester\logs\20080928.log) for "no enough money" and couldn't find it, so I assume (well, I hope!) that the available margin was never exceeded. As can be seen in the attachment, the result is a about a 600% gain in 3 months. I don't know how Strategy Tester calculates its maximal drawdown, but, by taking values from peak to trough in the curve it's nothing like the $17,033 stated in the report. The loss at the end of the curve is due to the automatic closure of the incomplete positions when the data runs out.

Given that we're not compounding position sizes, the risk of an 'insufficient margin' scenario gets progressively less, as the account balance grows. Hence, provided that we have some luck to begin with, the balance will eventually reach a point where the risk becomes very small. But of course this also means that our returns will be linear rather than exponential.

We could perhaps compromise by using a more gradual type of compounding. Alternatively, we could backtest across a huge sample to establish a worst case scenario, and then calculate whether the position size needed to avert this would, given our starting equity, provide a satisfactory level of income. Of course we must realize that no amount of back-testing is exhaustive, and that a 'death trade' type situation, however improbable, is always possible.

'Prudent' trading is about employing a positive expectancy method of entries and exits. Unless we are using some kind of TA/FA that provides an edge sufficient to overcome costs, trading will always be a negative expectancy exercise, and no amount of MM (e.g. Martingale) can avert this. With Martingale we are only ever one trade away from ruin. To make outsize returns with a near perfect win rate, that is the risk that we must be willing to take.

I assume that, the higher the leverage offered by the broker, the greater the probability of survival, for a given position size. Somebody please correct me if I'm wrong.

David
Attached File(s)
File Type: zip StrategyTester GridEA.zip   30 KB | 5,508 downloads
  • #40
  • Edited 3:24am Sep 28, 2008 2:26am | Edited 3:24am
Quoting hanover
Here are some possibilities for your consideration:

-- when prices are moving at greater than normal volatility
-- when the higher timeframe charts are trending strongly (like we had in August)
-- prior to major news announcements
-- when price is apparently breaking out of significant congestion, or through significant S/R
...
Also in this vein, I wonder if it's possible to reduce risk by somehow 'hedging' opposing (i.e. 'averaging up' versus 'averaging down') Martingale methods against each other. Anybody got any thoughts?

David
Hi David,
I would concur. I see the following ways to mitigate risks:

1. Do not automatically re-enter setups after closing all orders. 2 obvious options
1a. Time filtering, do not enter in the end of the day, until before starting London session
1b. Do not re-enter on decreasing volatility. Can be measured by comparing with previous entry point's StdDev on H1 or D1. Plus take into account the reserved margin and the lifetime of the recent setup.

2. As you already mentioned, on prolonged and consuming margin fills change the strategy to decrease opposite opened orders instead of opening new (2a). That essentially would mean taking emergency losses to stabilize the situation. Another alternative to that would be closing all on such critical points (2b).


Here is the result from MetaQuotes history center. Tested on EURUSD M1 from 2007.01.01 to 2008.09.26 with all default parameters. 21 months test took about an hour. In this first EA version I got more than 100 positions opened at a time, irrespective to MAX_LOTS setting. Even with more than doubled equity we got margin call. Although the idea is quite interesting Martingale implementation, it wouldn't work on its own. There is definitely some room for making the idea more resilient.

Below is the last setup that caused the failure on 2008.04.14. As we see the amount of open positions caused margin overflow on that specific day. Search for Stop Out in attached tester log file.
Attached Image(s) (click to enlarge)
Click to Enlarge Name: EURUSD M1 2007.01.01-2008.09.26.png Size: 9 KB
Click to Enlarge Name: EURUSD 2008.04.14 - MarginOverflow2.png Size: 20 KB
Attached Image
Attached File(s)
File Type: zip EURUSD.M1.2007.01.01-2008.09.26.zip   112 KB | 3,774 downloads
File Type: zip 20080928_log.zip   47 KB | 2,282 downloads
0 traders viewing now