Magix, Martingales, and Market Myths

I've coded my EA, that has a successful strategy, to take martingale type trades, lets see how long it lasts before it blows up.

http://www.myfxbook.com/members/thes...-mgish/1007561

Asw

Hi!

Recently I saw a reference of the alarming amount of busted demo accounts I have on my MyFXBook (see post here: http://www.forexfactory.com/showthre...24#post7702224) ... and I really enjoy having that many busted ones!

Why? Let me put an example ... How do car builders know the capacity of a car to withstand an impact? They make their calculations (backtesting) and then the actually crash them (forward testing on demo) ...

The kind of grids I trade are a mix between a basket trade and a pseudo-martingale trade (grid trading). So many of the problems and risks the account will be facing can NOT be backtested due to:

1. Real spread from the broker at the time the simulation is testing is not known
2. Almost all my entry exit signals trade below the 1 hour timeframe ... so they became broker feeder dependent
3. I trade a multi-symbol strategy and crappy MT4 backtester cannot simulate them
4. And many times it is a multi-strategy, multi-symbol strategy ... so even more complicated to simulate

So backtesting is only useful to check if the triggers are performing correctly and to have a general idea if the strategy could be profitable.


So I need to use many forward testing account to measure the performance and extremes of the strategy to know where put the limits on the real ones. So a demo busted account is a forewarned problem on a real one ....

When trading multiple symbols you have the advantage than DD tends to equilibrate but if symbols are too correlated they just give you a different effects ... so forward testing is the only way to go.

How many demo accounts I can be running at the same time? At this moment, between 20 to 40, plus the ones fellow trades also kindly share with me ...

How many real accounts I trading on real now? At this moment, none. Yup! None! I'm still on my summer break from trading.

How do I trade a real grid account? Let me check ...

1. Predefined account, EA, position, grid and order risks depending of what the forward testings of the strategy gave me ...
2. Manual intervention is something smells awful ...
3. Predefined target profit of the whole account where everything is closed down in order to rinse and repeat ...
4. No compounding at order level ... compounding is calculated at account level every time after the account is reset and taking into account my full trading equity (not the amount used to fund the account)

So reading this you can say as a nobby ... "Hey! That seems a lot of work to run a marty ..." ...


EA grid trading requires a lot of efforts and work ... if you want to be profitable.

Cheers,
Sergien

EA grid trading requires a lot of efforts and work ... if you want to be profitable. <== I think any form of trading is very difficult and I think people are trying to trade from the bullshit they have either read or trying to trade techniques that the so called gurus are teaching people, apparently 95% of people lose, trading forex, I would say the number is higher.

Biggest mistake at potential trader can do is trade low timeframes and trade against the trend, low timeframes are just noise.

Asw

What do you mean by that, maybe I've missed something.
Are either of these applicable:

NOISE
3. Physics A disturbance, especially a random and persistent disturbance, that obscures or reduces the clarity of asignal.
4. Computer Science Irrelevant or meaningless data.

I don't get your logic, can you tell me on what timeframe you think the 'noise' stops?

Well I have been unable to get any constructive exchange of ideas in the Martingaling - Looks so good until the inevitable thread. Just the usual individuals repeating the same "wisdom" of the death trade. So maybe some discussing and exchanging of ideas on here.

A trend for me is a confluence of 3 time frames agreeing, H4, D1 and W1, Dow theory.

I do trade into the D1 or W1 timeframes, lower timeframes range before heading into D1 and W1.

Asw

ok...

Dead thread that needs more views. Martingale is frowned upon for the most stupid reasons. I've tried lot of variations and finally have a style the works well for me.

I start with $5000 and double my lot size every time I win and half it when I lose to make martingale a little less risky. I've heard that this is known as reverse martingale but I might be wrong.

Martingale needs more discussion. There are so many variations for different trading philosophies that anyone who attacks it comes does come across as naive.

For the risk averse crowd you can micro lots and what not.


I have traded a total of sixteen $5000 deposits and never once blew my account. Three times it took me more than 20 trades to reach my target i.e. martingale 5 times assuming that every trade will be successful. These three times were my only failures and so my failure isn't loss but time invested.


I can't trade any other way and every time I see someone on a thread bashing someone who claims he can go from rags to riches pisses me off to no end.


Step 1. Create a successful system with more than 70% winning ratio in the past(past does not equal present but this is you best bet)

Step 2. Find a way to make stop loss small and take profit large to achieve a risk to reward of more than 1:3.
My current risk to reward is 1:4 for one lot and an exit strategy for other. 3 out of 5 days in a week I am able to achieve that ratio on that lot and one out of 5 times a week I am able to achieve a 1:10 on the exit strategy type lot.

Step 3. Apply martingale sensibly

Step 4. Reach your target but never trade beyond that. You could try and keep doubling your money but I personally prefer to start again to reduce my risk further.

The goals is to earn atleast $30,000 a month but the reality is that I end up with 18-22. I haven't been able to break the 30,000 without breaking the 5 trades maximum rule.

The first trade is extremely important. If you lose in that trade you need to half your investment amount next time. If you lose that too then again you half and so on. Never have I wiped out an account after fully fine tuning my strategy. Before I use to not take it seriously and ended up losing shit loads but after taking it professionally I've have turned around my trading experience and come to the conclusion that the only way to actually get rich on forex is aggressive compounding.

Many thanks for sharing your strategy. I like your approach. Am I right to assume, that you take your trades independently from each other? So you don't add to a losing trade and trade instead on different pairs whereever a signal is given by your strategy? What is your starting lot size for a 5000 dollar account?

Independent.

1% of 5000 because we cannot predict if we will win the first trade. After 1:2 both lots are set to breakeven.

That means I am willing to lose $50 on my first trade.

Average stoploss for me is 15pips so $150 are invested. My most recent $5000 had one new rule. No 2 lots. I didn't like watching the market for too long. Now it's just risk to reward and I like it this way. It's much safer than before. So now it's one lot which is set to breakeven(including spread coz I'm a miser).

Also I've increased my TP to 5 times the loss including spread(average of last 6 months for the traded currency pair). This has made my martingale a bit slow and so only 5 total trades are taken and then money is withdrawn.

Hypothetical if all trades are profit trades(obviously this does not happen but it's important for me to set a target and keep trading martingale until it's achieved.)

1. $3 per pip

Target at 75 pips

75*3


2. $6 per pip

and so on.................


I use different brokers every time I start doing this because I know in true ECN this is not possible so I just want them to think I was lucky.