0_0
When news hedging goes bad; how the last few hours of January sorta sucked.
http://i.imgur.com/r44merV.png
To be upfront: there are no excuses, the technical reasons and events surrounding the equity hit were entirely avoidable if I had taken better precautions or made a more appropriate choice given the risk of 3 high impact news events hitting the target pair within a 6 hour span.
No matter what the circumstances may be that led up to realized loss. I will still take complete ownership of the result.
I'm sour about it, since this morning I thought the monthly update was going to be awesome with a 23% gain to report. But it looks like the monthly update will have to wait til this weekend when I'll have the time to properly review what happened (blow by blow) so I can properly break down what's going to be done to avoid this in the future.
In the mean time, the basic run down was this: The algo thought it was in a heavy draw down thanks to a market shock post news release and started cutting losses accordingly. However, since a hedge position was on to flatten out risk during a news release, no actual draw down was taking place.
By the algo removing its positions, the account gained unintended directional risk at the worst possible time, since a 2nd news release was to take place shortly after. The algo then started trading again and reversed its previous position, which added to the directional risk on the account. Then the 2nd news release hit and didn't favour the positions at all..
The combined drawdown triggered the algo to, again, start cutting losses. And finally the hedge script (more of a conditional order based on time and price) figured out it was time to flatten itself too.
I didn't notice most of this even happening since I was at a night class (I'm taking a beginners course in German a few nights a week now, it's been fun) and by the time I could step in to manage things by hand there was a lot of damage done.
Then in trying to manage the exit positions left over from the EA by hand, I botched up a few executions since I was attempting to do so from my phone, in the middle of class, through a VPN app to my VPS server (because I had to manage the EA itself, not just the orders, else the EA would just reopen trades after I freed up exposure.) So frustrating to say the least. (My impact on the drawdown was very minimal, but it still didn't help things.)
What I should have done was recognized that my script to hedge on news wasn't ready to take on multiple events that overlapped each other (3 major ones that affected fundamentals within a 6 hour span,) and I should have just turned the EA off and flattened out hours before the first news release. not only would that have been more simple, but it was the better choice to make as a risk manager of my capital. That's the reality of it, I knew I was going to be preoccupied with a class, and have limited accessibility to the algo if something went wrong, but I felt overly comfortable with my EA, scripts, and setup (having just spent the last week testing them on demos) and figured trial by fire was a good idea.
All that being stated, the next step is to press pause, collect as much data on the event as possible, and start learning from what happened.
In the grand scheme of things, the drawdown wasn't major. My initial backtesting of the strategy tells me that 30-40+% drawdowns could happen easily if news or major market events weren't managed properly, so this isn't on the high side of that possibly either. And, the account is still in profit overall (even with a 1.11 PF after this mess,) the damage just killed January's profit, which is a setback, but not a death sentence.
There's more to this, not just 'should haves' and remarks from Cpt Hindsight, but the whole point is that there's no excuse: Today I failed to mange risk properly. The account should never be in a position like this without me being ready to pull the plug earlier on at the first sign of issues.
I have more details to review, which I'll be journaling here, but the plans I had to add more to the account given the good performance lately have been shelved for now.
*grumble*
When news hedging goes bad; how the last few hours of January sorta sucked.
http://i.imgur.com/r44merV.png
To be upfront: there are no excuses, the technical reasons and events surrounding the equity hit were entirely avoidable if I had taken better precautions or made a more appropriate choice given the risk of 3 high impact news events hitting the target pair within a 6 hour span.
No matter what the circumstances may be that led up to realized loss. I will still take complete ownership of the result.
I'm sour about it, since this morning I thought the monthly update was going to be awesome with a 23% gain to report. But it looks like the monthly update will have to wait til this weekend when I'll have the time to properly review what happened (blow by blow) so I can properly break down what's going to be done to avoid this in the future.
In the mean time, the basic run down was this: The algo thought it was in a heavy draw down thanks to a market shock post news release and started cutting losses accordingly. However, since a hedge position was on to flatten out risk during a news release, no actual draw down was taking place.
By the algo removing its positions, the account gained unintended directional risk at the worst possible time, since a 2nd news release was to take place shortly after. The algo then started trading again and reversed its previous position, which added to the directional risk on the account. Then the 2nd news release hit and didn't favour the positions at all..
The combined drawdown triggered the algo to, again, start cutting losses. And finally the hedge script (more of a conditional order based on time and price) figured out it was time to flatten itself too.
I didn't notice most of this even happening since I was at a night class (I'm taking a beginners course in German a few nights a week now, it's been fun) and by the time I could step in to manage things by hand there was a lot of damage done.
Then in trying to manage the exit positions left over from the EA by hand, I botched up a few executions since I was attempting to do so from my phone, in the middle of class, through a VPN app to my VPS server (because I had to manage the EA itself, not just the orders, else the EA would just reopen trades after I freed up exposure.) So frustrating to say the least. (My impact on the drawdown was very minimal, but it still didn't help things.)
What I should have done was recognized that my script to hedge on news wasn't ready to take on multiple events that overlapped each other (3 major ones that affected fundamentals within a 6 hour span,) and I should have just turned the EA off and flattened out hours before the first news release. not only would that have been more simple, but it was the better choice to make as a risk manager of my capital. That's the reality of it, I knew I was going to be preoccupied with a class, and have limited accessibility to the algo if something went wrong, but I felt overly comfortable with my EA, scripts, and setup (having just spent the last week testing them on demos) and figured trial by fire was a good idea.
All that being stated, the next step is to press pause, collect as much data on the event as possible, and start learning from what happened.
In the grand scheme of things, the drawdown wasn't major. My initial backtesting of the strategy tells me that 30-40+% drawdowns could happen easily if news or major market events weren't managed properly, so this isn't on the high side of that possibly either. And, the account is still in profit overall (even with a 1.11 PF after this mess,) the damage just killed January's profit, which is a setback, but not a death sentence.
There's more to this, not just 'should haves' and remarks from Cpt Hindsight, but the whole point is that there's no excuse: Today I failed to mange risk properly. The account should never be in a position like this without me being ready to pull the plug earlier on at the first sign of issues.
I have more details to review, which I'll be journaling here, but the plans I had to add more to the account given the good performance lately have been shelved for now.
*grumble*
FXGears.com