New EA - "Predator" - page 3
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
Oidunno - very eloquently written...some great ideas and thoughts.
Janus - your feedback is prized.
It's fun to watch and participate in the evolution and endless discussions of this 10Point3 based EA and it's hybrids. The results of this vigilante tradebot can be both amazing and terrifying. I got involved with this thing first because I was surpised to see 'ol standard MACD at the core of the trade decisions and I knew that part could be improved. Of course Terminator opened things up by offering multiple trade logic mechanisms which Predator is a continuation of. I've been studying adapative moving average based indicators and thought ripping out the old MACD would be a good place to start with these Jurik driven indicators. That helps address one of the three major sore spots with 10Point3. At least now, we have a better chance to be on the right side of the trade.
That leaves the other two areas which are (1) Risk-Reward Management and (2) Flexible Order Processing options. Maji and I have been working on a few EA's offline and he's got a great Order Processing foundation I'd like to steal (with his approval). The Risk piece will be a total overhaul because that backdoor exists in 10Point3 to wipe you out on news fast-bars or trend runs without retracement (GBPUSD)...the average trader here probably has $5K - $10K to play with at best and there's just no room for runaway drawdowns and margin calls not to mention huge over-leveraging. One of my favorite strategies is the idea of dumping 1/2 of the lot at breakeven and letting the other half run for gravy...or dynamic stoploss/takeprofit sizing on incremental orders based on how the trade is running. Heck, when it's all said and done, all traces of 10Point3 will be gone. This is something I've dedicated my time to and it will happen, especially if I can keep getting great feedback from guys like you. I'm keeping notes on every idea tossed out here.
Sl
Hi guys, I have a novice question....I see that the SL is default 0 therefore is the stop then handled by the martingale or is there a SL built into the code vis a vis an exit strategy? Just wanted to know in order to know if I need to do some tests by changing the default 0. Also mm is default 0, should that be changed to 1 ?? Thanks for any feedback.
Martingale Strategy
Here are my 2 thoughts on that:
1.The Terminator setup is mainly based on the positive swap trade idea.
This means to trade the EA only in a single direction on each currency pair.
The investment assumption is simple:
Over the longer term, all interest paying pairs will swing back, allowing the position exit.While you wait you get paid the swap each day.
(Good example from today .......look at NZD/JPY at 4:55 EST)
Folks simply do not want to "pay" swap for their shorts.
Additional setup i use.
I use a medium term swing trading model to simply see when it is time to "turn on" the EA.Rather than just taking static position with possible average ins, i let the EA trade and seek multiple trades until my final profit target is reached(can take several weeks)
Suggestion: Use different TP,SP,OTP and pipsteps for each pair.If you trade cable short, you must trade maddog long as well.MHO only.
On the other hand, a trader must know before, how large the maximum drawdown can be, if the EA takes 10 trades, plus from the historical data, determine the maximum swings.This leads to a caculation for each pair you wanna trade and gives you a rough equity figure.
Fancy numbers like 250000:1 for EUR/USD are just "guesstimates" and do not bear any calculation in it a real trader would use.
Too much underleverage is as bad as the opposite.(IRR figure gets out of whack)
Inplementing this into an EA will get fairly complicated.One solution might be to pick the "right" correlated pairs in the first place.Look at the Terminator demo pairs and draw your risk and "hedge" analysis.
Predator and Juriks indicators might acomplish two things:
1. more trades in both directions
2.Getting into position quick and at the right time
This should provide extra risk management and allow for faster and higher quality trades.
IMHO the swap setup replaces simply the "unknown" trend component in the first place.With an indicator like Juriks MACD/JRSX this might(and this is a to be tested "Might") solve this problem.
Lastly the trade management and account equity component needs to be integrated into each and every Martingale setup.Note that all martingale setups are almost always in a permanent drawdown, but hopefully equity rises each trade.Further a trader wants some "wrong" trades, because that increases overall performance.Look at the Terminator Demo statements.Whenever there were bigger trades, the bigger was the monthly result.Martingale requires nerves of steel.
If you have a full position or close to, the position size on new trades has to be adjusted down by the EA.
This is fairly easy to integrate(see Phoenix 5 version)
Janus
Hi guys, I have a novice question....I see that the SL is default 0 therefore is the stop then handled by the martingale or is there a SL built into the code vis a vis an exit strategy? Just wanted to know in order to know if I need to do some tests by changing the default 0. Thanks for any feedback.
That initial stop default is a throwback to the original 10Point3. If you don't put something in there, you won't have a stoploss! The dynamic stoploss feature added back on 09/19/06 by tururo in the 10Point3 thread is in there...these stops kick in and shut the whole cycle down when maxtrades hits the final loss point.
One of my favorite strategies is the idea of dumping 1/2 of the lot at breakeven and letting the other half run for gravy...or dynamic stoploss/takeprofit sizing on incremental orders based on how the trade is running.
1/2 lot at breakeven for the whole trade always sounds appealing, but that means we need a 2:1 TP:SL for that half. After we TP half of the trade, the other half needs to be converted to a TS at that point. Without per tick data, TS profits are going to be very exaggerated. Maybe the new Build 200 will have enough improvement that we don't see craziness.
I detailed a few ideas in a thread ( https://www.mql5.com/en/forum/175481 ) and have worked out most of the code.
If TP:SL balance of at least 2:1 is not used, then "Winners don't loose" might be a better strategy since it performs loss prevention.
Another variation of "breakeven on half and let the rest ride" is to cash out some of the trades with TS, and convert the others to a set SL with a higher TP. The SL value would guarantee some profit, the TS would guarantee more. Now both sets of trades will profit. The TS trades might cash out at a quick drop, and be out of the trade for the longer haul. The TP is kept trying to weather out the short counter trend storms.
A few more strategies are in the thread, along with some ideas on how to manage groups of trades without setting a GroupID.
1/2 lot at breakeven for the whole trade always sounds appealing, but that means we need a 2:1 TP:SL for that half. After we TP half of the trade, the other half needs to be converted to a TS at that point. Without per tick data, TS profits are going to be very exaggerated. Maybe the new Build 200 will have enough improvement that we don't see craziness.
I detailed a few ideas in a thread ( https://www.mql5.com/en/forum/175481 ) and have worked out most of the code.
If TP:SL balance of at least 2:1 is not used, then "Winners don't loose" might be a better strategy since it performs loss prevention.
Another variation of "breakeven on half and let the rest ride" is to cash out some of the trades with TS, and convert the others to a set SL with a higher TP. The SL value would guarantee some profit, the TS would guarantee more. Now both sets of trades will profit. The TS trades might cash out at a quick drop, and be out of the trade for the longer haul. The TP is kept trying to weather out the short counter trend storms.
A few more strategies are in the thread, along with some ideas on how to manage groups of trades without setting a GroupID.Thanks, daraknor!
Getting it right
Janus,
Something is misleading to me. Indicators are not used to determine when EA trades but the code in EA alone determines this. So what matters what indi one attaches to the chart? I've got to be missing something here because you and Bluto make it sound like certain indi's are creating/capturing certain trades. Please set me straight.
For now till I'm clear on this; I've left inputs at default except for:
T/P =30
TrailingStop=0
Max Trades= 7
Pips(step) =20
SecureProfit=22
OTP=3 (GbpJpy),(GbpChf),(EurJpy)
I've left the 4 indi's in the 'input' at default.
Respectfully, Kevin
Janus, Something is misleading to me. Indicators are not used to determine when EA trades but the code in EA alone determines this. So what matters what indi one attaches to the chart? I've got to be missing something here because you and Bluto make it sound like certain indi's are creating/capturing certain trades. Please set me straight.
The indicators are used to find a good place to make the initial trade.. they're supposed to put you in the market on the right side of the trend. If they're right, you simply take your profit when the market moves in the direction of your trade.
The EA code adds trades only in the case (fairly often) that the indicator was wrong, and the market moves against your initial trade. Instead of closing at a loss, it sets up a series of trades against the trend and looks for an opportunity to take a profit when it sees a sufficient pullback from the trend. The significance of the indicators, then, is to help you trade on the right side of the market. The EA trading code helps you make money when the indicators are wrong.
I hope that helps...
Indicators
Kjenkins,
i will try to go a bit further into detail on the indicators:
Terminator gives 5 options to choose from, based on which indicator signal trade entry is determined.
Developer Tom uses i-trend(default=5) and others(mr.trader) use MACD.
Additionally almost all results show, that trading Terminator only in 1 direction(long OR short, but not both) on selected currency pairs ONLY in positive swap direction gives the safety net to trade it.
Bluto has added the Jurik MACD to Terminator, which is much more responsive and precise than the standard MACD. This has so far increased trade results by a few digits on performance
Terminator live account(i-Trend) up 1.9% vs Terminator Demo with Jurik MACD 2.16%. since last sunday.
Predator is designed with additional , custom added fine tuning by Bluto, to allow Predator not only to trade safe in BOTH direction, but also overcome certain "wrong" direction decisions or "missed" trades on the Terminator setup.
In theory and so far also in Demo trading, Predator is up 5% on equal levarage since Sunday night as i write.
A case can be made so far, that trade entry quality and frequency is far superior on Predator with Jurik Turbo JRSX.
For further info of "what is Jurik" please visit Marc Juriks webpage
http://www.jurikres.com/
to understand the "no lag" edge, Jurik indicators have.
Hope that clears up some questions.
Janus
I really like the idea of a positive swap trading EA. If the EA is only able to break even, you still walk away with a nice profit compared to other passive investments. In researching this, the price volatility is too high in relation to the swap payments to be significant. For example, the range on USDJPY is about 300 pips for the past 30 days. At 1 lot, that is $2500 or so. The swap payments for one of the most lucrative pairs varies drastically broker to broker. Some brokers pay $12, $10, or as low as $2 per day on swap. The volatility over time is insane vs the swap payments of about 1/8th the amount over the same period. Even with a broker that has tight swap spreads, there is a large issue with hedging, etc.
I don't think that this is necessarily an Achilles heel, I just think it is only a partial solution. If the trades were made and the drawdown from our position was less, that wouldn't be so bad. If the EA is overall profitable, maybe we can ride it out by continuing to trade. It might take nerves of steel, but it is a possibility.
I also looked into the 90 day currency correlations on BUY swap paying and SELL swap paying currencies. They are all inversely proportional. Which means if you do any trade where all of the currencies pay you, ALL of the positions move against you. Now THAT is an achilles heel. To reiterate, BUY swap paying positions have a negative currency correlation to SELL swap paying positions. Sometimes the correlation is low, but it is always in that direction.
If a buy position is against you on a large swap paying currency, you may want to take a position that charges you swap, but a very small amount. I recommend looking at the Oanda interest rate calculator to see the actual differences in interest rates. That said, sometimes brokers charge you interest both ways
I have been collecting swap data from dozens of brokers, they vary so so much.