you are moving closer and closer to just coin flipping. Jumping from one system to another when it loses money as opposed to lifting the head and seeing what went wrong is just random fitting..
Frosty's auto-trading bot goes live with REAL money
- Thread starter frostengine
- Start date
im curious what % of trades go into positive profit territory during its life span? i always have a gut feeling when thinking about trend following that once profit is obtained, it should be protected somehow.
walter,
Good question i will need to run a test at some point and figure that part out, could be useful
Good question i will need to run a test at some point and figure that part out, could be useful
Here is a longer PL chart than what I showed before.. this is for my origional method #1..... This is slightly more than 1.5 years worth of trades....
Looking at this chart makes me wonder why I am getting so bent out of shape.. looking at the longer term I should be solid.. perhaps I should just stick to method #1 and stop sweating the little stuff....
Looking at this chart makes me wonder why I am getting so bent out of shape.. looking at the longer term I should be solid.. perhaps I should just stick to method #1 and stop sweating the little stuff....
On what time frame? Where are you getting all this data?
I was running my trading system on 1m charts and the best I can do is go back to September for data. I started having to archive 1m data and store the gigs and gigs of data myself. I lost a few hard disks running programs that were doing nothing but non-stop abuse of the disks while I was testing some of my ideas. I'm sure it was just a coincidence, but data sources are a huge problem for me right now.
fatrat,
Several months ago I purchased approx 5 years of data for various e-mini's.. some I got more data some I got less.. This data I have is all 1 minute data.. which sucks because I have some strategies where I need tick by tick data to test.... but can't due to this data limitation I have...
I have however been storing tick by tick data that I capture real time since august 3rd...... So one day I'll have plenty of tick by tick data to test with but for now I have to settle for this crappy 1 minute data.. which I mostly use just to see how the strategy looks going back to then.. the results are fairly unreliable I am sure compared to the results I get from my captured data.. but still serves as a general barometer for testing over a longer period....
Several months ago I purchased approx 5 years of data for various e-mini's.. some I got more data some I got less.. This data I have is all 1 minute data.. which sucks because I have some strategies where I need tick by tick data to test.... but can't due to this data limitation I have...
I have however been storing tick by tick data that I capture real time since august 3rd...... So one day I'll have plenty of tick by tick data to test with but for now I have to settle for this crappy 1 minute data.. which I mostly use just to see how the strategy looks going back to then.. the results are fairly unreliable I am sure compared to the results I get from my captured data.. but still serves as a general barometer for testing over a longer period....
You can't compare the 2-week trading with the 1.5-year testing. If you have to, zoom in the 1.5-year into 2-week period, in that you may find situations that are worse than what you currently experiencing.
Long term PL looks good only because it is LONG term, the details could scare you and turn you off.
IMHO.
Looking at your P & L chart, I only wish to point out 2 things.
Before I do that, let's assume you have a constant number of trades/day. You have approx. 1600 bars/trades over 18 months. , which is around 89 per month or approx 4.25 per 21 day month.
Point 1 - Between Bar 470 and 600 you lost over $2000. Say that's 2k over 6 weeks.
Point 2 - Between Bar 1050 and 1090ish, you lost close to $4000. So $4k over 2 weeks.
The chart shows that you are in a drawdown at the moment (near bar 1600).
Rather than playing with Systems you need to be playing with your backtest and write the following items on a piece of paper and stick them over the mouse on your computer.
How long was the longest drawdown?
How large was the largest drawdown?
ONLY when you have surpassed one or both of those points should you take the piece of paper off the mouse and even consider tinkering with the system.
Seriously, you need to get out of this constant tinkering. Read a trading psychology book like 'In the Zone' by Mark Douglas. To me it looks like system 1 is peforming WITHIN expectations.
Why did you pick
System 1 in the first place? Has the reason why changed? I recall you saying that a backtest across your real trades produced the same results. If so, that's a great validation of your original logic.
Before I do that, let's assume you have a constant number of trades/day. You have approx. 1600 bars/trades over 18 months. , which is around 89 per month or approx 4.25 per 21 day month.
Point 1 - Between Bar 470 and 600 you lost over $2000. Say that's 2k over 6 weeks.
Point 2 - Between Bar 1050 and 1090ish, you lost close to $4000. So $4k over 2 weeks.
The chart shows that you are in a drawdown at the moment (near bar 1600).
Rather than playing with Systems you need to be playing with your backtest and write the following items on a piece of paper and stick them over the mouse on your computer.
How long was the longest drawdown?
How large was the largest drawdown?
ONLY when you have surpassed one or both of those points should you take the piece of paper off the mouse and even consider tinkering with the system.
Seriously, you need to get out of this constant tinkering. Read a trading psychology book like 'In the Zone' by Mark Douglas. To me it looks like system 1 is peforming WITHIN expectations.
Why did you pick
System 1 in the first place? Has the reason why changed? I recall you saying that a backtest across your real trades produced the same results. If so, that's a great validation of your original logic.
Hey Frosty,
Your journal is the most enteraining on ET at the moment. In fact it reads like a day-time soap opera. We've got a likable hero with a worthy objective who is in grave peril of blowing up (in this case his account). As the drama unfolds our hero is victim to his own failings and makes one fundamental mistake after another. Then we've got the goofy friend cheerleading our hero whose advice is invariably stupid and theratens to imperil our hero even further. We also have a series of sage would-be protectors of our hero who's noble advice is ignored or dismissed out of hand. The rest of the action is padded out by a cast of wannabes in a series on mindless vignettes which adds no discernable value to the plot whatsoever.
Don't get me wrong Frotsy, i'm rooting for you. I hope you make it. By the same token I can't wait for the next episode.
hang in there,
bolter
Your journal is the most enteraining on ET at the moment. In fact it reads like a day-time soap opera. We've got a likable hero with a worthy objective who is in grave peril of blowing up (in this case his account). As the drama unfolds our hero is victim to his own failings and makes one fundamental mistake after another. Then we've got the goofy friend cheerleading our hero whose advice is invariably stupid and theratens to imperil our hero even further. We also have a series of sage would-be protectors of our hero who's noble advice is ignored or dismissed out of hand. The rest of the action is padded out by a cast of wannabes in a series on mindless vignettes which adds no discernable value to the plot whatsoever.
Don't get me wrong Frotsy, i'm rooting for you. I hope you make it. By the same token I can't wait for the next episode.
hang in there,
bolter
oh oh. which character am i? i hope i'm one of the mindless wannabees.
