My Automated Trading Strategy (3Years+ of Historical Data) Please Critique!
- Thread starter julian0625
- Start date
Julian, if you want some good info on automated trading, check out the Geronimo system over at evilspeculator.com
Heck, if the guy takes a liking to you, he might even take you under his wing. Give it a try.
Heck, if the guy takes a liking to you, he might even take you under his wing. Give it a try.
Quote from short&naked:
Julian, if you want some good info on automated trading, check out the Geronimo system over at ************
Heck, if the guy takes a liking to you, he might even take you under his wing. Give it a try.
hmmm, I suppose it was your intention that, it is never too early to let a young kid experience his first scam, so that he learns some life lessons sooner rather than later?
One being, those that can't do, teach. And there is a whole lotta teaching on that website....
Two being, no free lunch. Geronimo costs $100 a month, must be the goodness of this guru's heart that makes him want to sell something so valuable at only $100.
The last time I checked, people usually sell things when they think the value of what they are selling is less than the price they are selling for, or else they would just keep it to themselves.
short&naked, not trying to pick a fight with ya, you may very well be subscriber and doing very well for yourself, but come on, the system they are selling is not impressive to begin with and they neglected commission and slippage.
If he was to be taken under his wing, it would be for "Guru-ing". We can have the youngest snake oil peddler the internet has seen.
The other issue which I don't believe has been mentioned is the quality of data that Julian is using.
I've not used NT, but I had eSignal for a few months before I realised that their FX tick data is completely unusable. The number of rogue prints, even in something like EUR/USD, was unbelievable.. eSignal would regularly show a price 2-3 pips outside of the market bid/ask.
As such, any mean reversion strategy shows great performance (because it will be trading advantageously at non-existent prices) whilst a breakout system will show poor performance (for the converse reason).
Irrespective of the merits or otherwise of the strategy, if the data is of poor quality, the backtest results could be meaningless.
I've not used NT, but I had eSignal for a few months before I realised that their FX tick data is completely unusable. The number of rogue prints, even in something like EUR/USD, was unbelievable.. eSignal would regularly show a price 2-3 pips outside of the market bid/ask.
As such, any mean reversion strategy shows great performance (because it will be trading advantageously at non-existent prices) whilst a breakout system will show poor performance (for the converse reason).
Irrespective of the merits or otherwise of the strategy, if the data is of poor quality, the backtest results could be meaningless.
Quote from DGunz:
hmmm, I suppose it was your intention that, it is never too early to let a young kid experience his first scam, so that he learns some life lessons sooner rather than later?
One being, those that can't do, teach. And there is a whole lotta teaching on that website....
Two being, no free lunch. Geronimo costs $100 a month, must be the goodness of this guru's heart that makes him want to sell something so valuable at only $100.
The last time I checked, people usually sell things when they think the value of what they are selling is less than the price they are selling for, or else they would just keep it to themselves.
short&naked, not trying to pick a fight with ya, you may very well be subscriber and doing very well for yourself, but come on, the system they are selling is not impressive to begin with and they neglected commission and slippage.
If he was to be taken under his wing, it would be for "Guru-ing". We can have the youngest snake oil peddler the internet has seen.
LOL haha trust me, Ive gone through MANY scams which I regret. I can see through a scam easily now.
Although it is not a scam, Online Trading Academy would have to be at the top for me money wise. $15k (multiple courses) for something I mostly likely could have learnt myself online or by tinkering with platforms.
ALTHOUGH OTA is for some people who are slow learners, and need that instructor to help them on the way. As I am younger, I tend to absorb info quite quickly so $15k went down the drain for my parents... although everything happens for a reason...
So its good for some, waste of time and money for others.
I came into the class with some experience, so thats why it was "worthless" to me.
I do however recommend their basic course for someone JUST getting into the markets.
Among the scams I encountered were basically all the promotion guys on youtube saying MAKE 2 POINTS EASY DAILY.
Then you look at their course and you see your risking 10points to make those 2 points...
GOod statement. I do IT consulting and sometimes IT training onthe side, and most of the time it is a total waste to train people - all the stuff you go through with them is in the manuals. But SOME (most?) people just NEED instructions and are relly bad at extracting this themselves. I personally will nver pay for a formal training course in IT (not saying consulting and some people's high end stuff is really interesting, but it is not one of those certification courses) but I have seen people failing to move to another programming langauge without help.
So, your statement about some people needing instructors is perfectly right, imho.
UPDATE!
After a few months of painstaking trial and error, I have managed to get my average trade from $10 to $150!!
It trades the oil market, and does hold trades overnight....however using 3 years of intraday data, it has managed to make $160 000 INCLUDING COMMISIONS & SLIPPAGE!
I makes on average 1 trade per day, all trading one contract.
Out of the 34 months I have backtested, only 9 months were losers, with an average loss of $3000/m. The average gain for months that make money is about $8000/m.
It uses a protective stop loss of $1500 per trade, so in an unlucky situation of say 5 in a row, that would amount to a $7500 loss. This would then make me comfortable trading this strategy only with an account size of $30 000 to $50 000.
The largest drawdown in the 3 year period that the strategy incurred was $18000 from peak to trough.
We will see 
ALSO from just going back in my data, almost every signal I get goes against me by at least 10-20ticks. So instead of entering at market, which is obviously the safest, I could put in a limit order 1-3 ticks below/above the trigger (depending on long/short) and make another bit of money on top of that. Over time that 1 tick can help.
After a few months of painstaking trial and error, I have managed to get my average trade from $10 to $150!!
It trades the oil market, and does hold trades overnight....however using 3 years of intraday data, it has managed to make $160 000 INCLUDING COMMISIONS & SLIPPAGE!
I makes on average 1 trade per day, all trading one contract.
Out of the 34 months I have backtested, only 9 months were losers, with an average loss of $3000/m. The average gain for months that make money is about $8000/m.
It uses a protective stop loss of $1500 per trade, so in an unlucky situation of say 5 in a row, that would amount to a $7500 loss. This would then make me comfortable trading this strategy only with an account size of $30 000 to $50 000.
The largest drawdown in the 3 year period that the strategy incurred was $18000 from peak to trough.
We will see ALSO from just going back in my data, almost every signal I get goes against me by at least 10-20ticks. So instead of entering at market, which is obviously the safest, I could put in a limit order 1-3 ticks below/above the trigger (depending on long/short) and make another bit of money on top of that. Over time that 1 tick can help.
Quote from julian0625:
UPDATE!
After a few months of painstaking trial and error, I have managed to get my average trade from $10 to $150!!
It trades the oil market, and does hold trades overnight....however using 3 years of intraday data, it has managed to make $160 000 INCLUDING COMMISIONS & SLIPPAGE!
I makes on average 1 trade per day, all trading one contract.
Out of the 34 months I have backtested, only 9 months were losers, with an average loss of $3000/m. The average gain for months that make money is about $8000/m.
It uses a protective stop loss of $1500 per trade, so in an unlucky situation of say 5 in a row, that would amount to a $7500 loss. This would then make me comfortable trading this strategy only with an account size of $30 000 to $50 000.
The largest drawdown in the 3 year period that the strategy incurred was $18000 from peak to trough.
ALSO from just going back in my data, almost every signal I get goes against me by at least 10-20ticks. So instead of entering at market, which is obviously the safest, I could put in a limit order 1-3 ticks below/above the trigger (depending on long/short) and make another bit of money on top of that. Over time that 1 tick can help.
Sounds interesting!
Some questions:
-Have you forwarded tested this or is the results only from backtesting?
-How many trades a day do you get?
-Did you take into account rollover periods in your backtesting? What about spreads between bids and asks, if any?
-average drawdown of 3k and 18k max drawdown sequential? Sounds like the "average" is hiding some important information, what was your max drawdown?
Just a side note, 18k of drawdown is huge. It's one thing to look back at a backtest and see 18k of drawdown and then more winnings, but if you are in the middle of that 18k drawdown and you continue to see pain, it's hard not to think that the system is broken. You'll soon find that it's not finding a winning system that's the hard part, it's sticking with a system through an extended drawdown.
For example, let's say you take 5 losses in a row. You think "There's no way there could be a 6th loss, I've never seen 6 in my backtesting", so you double or triple down, and then you hit another loss. Things like that are what mess you up psychologically.
Anyway good luck, sounds pretty profitable if this actually works and wasn't accidentally over-optimized for your test data.