On some C2 trading systems, I noticed that a few trades were marked as "Extreme" on the Risk column (over 50% drawdown in some cases) but, oddly enough, the C2 stats report their current "Max peak-to-valley drawdown" as 30% of less, even though the trades responsible for these huge heart-stopping drawdowns were closed more than 3 months ago!
How could it be?
It is possible to imagine a scenario where a single trade has a 90% drawdown from peak to valley (example: buy 1 share of IBM at 10, see it rise to 100, then decrease to 10) while the entire system itself on an aggregate level does not have such an extreme peak-to-valley drawdown. One is the single-trade adverse excursion, the other is the measurement of the drawdown of total system equity.
Thanks for your prompt reply Matthew.
So using your example, if I buy IMB at 10 and the stock immediately goes to 100 the same day and then comes back to 10 and I close the trade, potential subscribers will see a huge 90% drawdown on my stats even though the system did NOT lose any real money at all ??
And by the way, the maximum negative adverse excursion in this example is zero, not 90 dollars (the stock did not move against my position at all, and this from the start).
Something does not seem right in your example.
Can you please further explain and expand on it.
Simply put: Trade drawdowns (i.e. peak to valley max excursion on a single open trade value) are not the same thing as overall system equity drawdowns.
So in your example, if a system starts brand new with a capital of $100 and on the first day buys a stock at $10 which goes up to $100 but ends the stock at $10 what would be the Trade Drawdown and what would be the System Drawdown?
These guys are right. In this example, the system draw down is $0, whether or not C2 agrees. If you don’t make it $0 you’re doing it wrong. Only the trade has a draw down.
The answer to your question is: $90
It’s a fallacy to say: "I bought a stock at $10, so if it winds up at $10, then I have lost no money."
If you buy a stock at $10, and it goes up to $100, and then goes back to $10… guess what? – you have lost $90 from some point in time.
The single trade drawdowns would be better labeled as maximum adverse excursions, which is what they really are.
Are you saying than that in this case both the Single Trade Drawdown and the System Drawdown are $90 and if not why should the same logic not apply to the system as a whole?
I can understand that the percentage drawdowns will be different if the system has already accumulated some gains.
Yes but there is a huge difference between $90 open profit drawdown and $90 drawdown in capital… Something c2 does not capture.
No one in their right mind would enter a trade midway and thus to say at some point in time you have lost $90 is totally incorrect, other than you have lost $90 of open profit which is not a drawdown because you have not eroded your starting capital.
Ex1. Trader1 enters long at 10 and at exits 50 with a profit of 40 and drawdown nil.
Ex2. Trader2 enters long at 10 but does not exit at 50, however price goes to 25 before rising again and exits at 75.
c2 will reward trader1 with 0 drawdown.
However, c2 will penalise trader2 with open profit drawdown of $25 even though starting capital was never eroded and trader2 exited with greater profit.
I know there are fancy names and calculations for all types of different drawdown but "Real" life drawdown is any erosion of starting capital including open trades and but not trades which never erode the starting capital.
If an account has $100 and a trade goes up from $100 to $200 then back to $150,the system makes 50% profit, but C2 will rate it as $50 drawdown.C2 score will give this kind of trade a negative.That was what happened when (I am surprised to find) my account kept going up and my C2 score kept going down.
I just find my drawdown data increased from 11% to 13.5% overnight.This is not correct.
The overnight futures price from yesterday night to today morning so far didn’t have so much drop.
Your drawdown increased between March 25 and March 27 because the markets opened on Sunday night and your system equity when measured on Sunday night went down.
Karl writes:
"Are you saying than that in this case both the Single Trade Drawdown and the System Drawdown are $90 and if not why should the same logic not apply to the system as a whole? “
I am not saying this.
I am saying one measurement describes a single trade a time.
The other measurement describes the overall system equity.
These are different. To demonstrate how they are different, imagine this scenario:
You open two trades. One long gold. One short silver. Your Gold trade goes from 100 to 150 and back to 100. At the same exact instant, your silver trade goes from 100 to 150 to 100.
Your total system equity will be a flat line. Your gold trade will show a 50% drawdown.
Now, to all you who say, “Matthew, it’s not really a 50% drawdown, because I was playing with ‘house money.’ I had already earned $50 in profit on that trade, and even though I lost $50, it shouldn’t be counted because it was profit I had earned” - here is my response.
It may be a useful mental tool to tell yourself you are playing with house money, but you still lost $50. Is this drawdown “better” than a $50 drawdown that starts the moment you open the trade? In the sense of aggregate system performance, of course it is. (Think about it visually: a system equity chart showing an immediate $50 dollar decrease looks ‘worse’ in some aesthetic sense than a chart that shows a $50 gain followed by a $50 loss.”)
But from an accounting, reporting, and capital efficiency perspective, it’s still a loss of $50. To take an unlikely example: if you ran a fund that contained only that single position, and it represented your entire capital, and the trade reached its apex on Jan 31, and then the trade was closed on Feb 28, well, then you would be required by law to report a “loss” in February of 50%. No one will say, "Don’t report it, because you were playing with house money."
Now, look, it’s fine if you want to disregard trade by trade drawdown stats. I personally think they are extremely useful and they tend to identify systems that let big drawdowns run. If you don’t think they’re helpful, then ignore them. But there is no denying facts: a decrease of $50 dollars in your account is an inefficient use of capital. You had the $50. You lost the $50. If you had closed the position at its apex, you would have kept the $50.
A system which can predict the future and close each trade at its apex will use capital more efficiently than one that does not. (Of course, no such system exists; I say this only for the sake of example.)
Matthew
You said
"But from an accounting, reporting, and capital efficiency perspective, it’s still a loss of $50. To take an unlikely example: if you ran a fund that contained only that single position, and it represented your entire capital, and the trade reached its apex on Jan 31, and then the trade was closed on Feb 28, well, then you would be required by law to report a “loss” in February of 50%. No one will say, "Don’t report it, because you were playing with house money."
If on Jan31 you started with $50 and Feb28 ended at $50, is not that a gain/loss of zero percent? If you claim a loss of $50 then all the power to you because it’s a useful tool to offset other taxable income.
Sorry. I should have been more exact. What I should have said was that the monthly returns between Jan 31 and Feb 28 (i.e. what would be reported to fund investors as "February Results") would be minus 50%.
Obviously the aggregate system performance from fund inception to Feb 28 would be net zero.
Matthew, you wrote : "Now, look, it’s fine if you want to disregard trade by trade drawdown stats. I personally think they are extremely useful and they tend to identify systems that let big drawdowns run. If you don’t think they’re helpful, then ignore them."
That’s precisely the problem Matthew, potential subscribers will NOT simply ignore these (imaginary) drawdowns! And both the vendor AND Collective2 will lose (potential) money in the process.
Let me explain.
First here is the real and only definition of MAE (Maximum Adverse Excursion), according to Traderpedia for example :
"The MAE is a statistic provided as a result of backtesting a mechanical trading system. It represents the largest loss suffered by a single trade while it is open. For example, a trade may close with a loss of -2, but while it was open if it was at one point down a maximum -5, then that would be the MAE, provided of course no other trade in the backest exceeded this loss."
Anyone googling for “Maximum Adverse Excursion” (MAE) will find the same exact definition. In fact an entire book has been devoted to that fascinating subject alone (Maximum Adverse Excursion: Analyzing Price Fluctuations for Trading Management).
That said, let’s look at your example again. Stock A goes from 10 to 100 and then all the way down to 11. Collective2 will list the trade as “Extreme” in the RISK column and report an (imaginary) 89% drawdown. Sorry but this trade just made a 10% profit with ZERO drawdown, not 89%, period. From the start the trader did not lose a penny, simple as that.
Now here is my point Matthew : the 89% drawdown you are referring to in this example is the maximum PEAK to Valley drawdown and will be reported ANYWAY in the C2 Stats, under the heading “Max peak-to-valley drawdown (historical)”.
When calculating the maximum drawdown (for the RISK column) after the position is closed your C2 software should only be concerned with this question and this question only : "How much INITIAL capital did the system lose while the trade was still open?"
Now here is another example if we follow the same logic you use for calculating the RISK. Stock A goes from $10 to 100 to 1 and the trade closes at $10. Now how would C2 staff react if I go to this forum and say “hey look, it’s amazing, my system just went from 1 to 10, the capital has increased tenfold, wow, is my system great or what!” Surely you would think I urgently need a shrink.
Now here is the only way the maximum drawdown (RISK) column should be calculated. We will assume the system made only 1 trade so far.
Stock A goes from 10 to 100 and trade is closed at 10. The Maximum drawdown should read zero in the RISK column (NOT 90%), while the maximum peak to valley drawdown should be set at 90%.
Stock A goes from 10 to 9 and trade is closed at 100. Maximum drawdown is 10% in the RISK column and the maximum peak to valley drawdown is also 10%
Matthew, I strongly urge you to change the way your software compute the maximum drawdown in the RISK column, otherwise just one or two (imaginary) “EXTREME” in the risk colum and a lot of potential subscribers will simply discard the system! And both the vendor and C2 will lose BIG money in the process. The vendor’s system will (incorrectly and unfairly) be seen as way too “risky” and the vendor will lose a lot potential subscribers. And because he is not making money from his subscribers C2 is also not making any money, simple as that.
In fact because of this huge problem C2 is probably losing (potential) money as I write this.
I absolutely agree:
RISK for a trade = difference between entry and the worst price during that trade.
I’m relatively new to C2, so please take what i’m telling you as constructive criticism as coming from a new prospective customer of C2.
I’ve been looking around C2 for a few days, looking to pick a system.
I particularly look at the Trade Entries and Exits and the trade drawdown (risk).
I disregarded MANY systems on the basis that the RISK column in the trade-by-trade data, contains the amount the market moved against you from ENTRY. To me that’s the RISK !
My RISK would be if i don’t have enough capital to cover that intra-trade drawdown. That’s what my common sense would tell me. And that’s the important number - it tells me how much money i would’ve had to have in my account to achieve THAT trade result.
Now. i’m finding out that it’s NOT SO on C2…
I’m not happy, but I consider myself LUCKY that i found this thread now and only wasted a few days of my time on C2 so far…
Now i have to go back and re-assess…
And…i have no idea how to determine the trade drawdown (the risk) on a per trade basis… Is that even shown anywhere? Is that what “worst price” is ?
(sorry if it’s a noob question, it’s just not clear enough).
I come from Tradestation where everything is very clear in the performance reports of a system.
Also I wish C2 would show the risk (as i defined it above) on ALL trades, not just some… some show “N/A”.
How can it be “not available”? After reading this now, maybe “N/A” means “not aplicable” . What does that mean? Market never moved against the trade at all… not even by 1 tick?
I’m confused by “N/A”. In cases where RISK column shows “N/A” the mouse over, shows nothing… so…i’m still missing the all-important intra-trade drawdown number. Where is that number? That’s really what people want to see.
Hi All,
Can’t help but agree with Matthew on this one as a FT trader myself.
Max Adv Excursion is nice in theory however the duration on the drawdown period can have a substantial impact on your capacity to enter any other trade setups during that period.
Rookie traders routinely make the mistake of considering profits as “house money” however there are few more destructive attitudes in trading than this mentality - it essentially changes you from trader to gambler.
Trading is an income generating activity with a lot of time allocated over long periods to achieve profitable returns of equity and a reliable longterm revenue stream. Time has a value regardless of which theoretical concepts you may espouse arguing this fact. You can never get your time back and it has an inherent value without question . Some may have a lesser value associated with their time however there is still always a value.
For this reason, how can a trader (defined as a person allocating their time to analyze the markets and devise methods to take advantage of supply/demand inefficiencies in price action in return of profiting from said inefficiencies) ever be playing with house money? The concept of house money is that of a gambler in a casino and has nothing to do with the markets.
If you have made money over and above your initial capital, this is INCOME and is your return for the time spent analyzing the markets or writing the code for your automated trading robot etc.
It is always therefore YOUR money that you are trading and this cannot be questioned as should you pull the plug on your trading account at any point in time the money you will receive is what is sitting in your account under the equity column.
People trading with demo accounts acknowledging theories on the concept of drawdown that are willing to completely discount the value of the time invested to make returns in trading are not valid representations of real traders risking real money of their own.
Drawdown is not just the missing equity in your account since opening a trade that goes against you but the lost opportunity cost of not having that same capital available to allocate into other profitable trade setups.
This C2 policy without doubt makes life more difficult for the system providers however as a real trader and system provider it allows me to discern quickly and effectively those who are using good money management in conjunction with profitable methods. It also quickly exposes those who are gamblers and people playing with demo accounts.
Anyone willing to ignore the level of drawdown between opening and closing a trade is either in need of further education or a system provider whose poor risk management has finally exposed the inherent dangers in their trading method.
Real system providers do not want clients/subscribers who think this kind of attitude to drawdown is ok because they will blow up their accounts swiftly and not be long term clients.
C2 is not without it’s quirks and I am the first to acknowledge this however Matthew sets himself apart from the masses by remaining vigilant on the level of drawdown recorded and hence protecting the best interests of his subscribers and best providers.
Regards
David.[LINKSYSTEM_54871675]
Hello,
You wrote : "I disregarded MANY systems on the basis that the RISK column in the trade-by-trade data, contains the amount the market moved against you from ENTRY. To me that’s the RISK ! "
Just what I thought, potential subscribers see a couple of (imaginary) EXTREME-ly “Risk-y” trades and they immediately discard the system, not knowing that the “huge” 60% drawdown they see is not necessarily a drawdown from the starting capital but a “drawdown” from the high of the equity curve while the trade was still open, hurting both the vendor and Collective2.
Personally I have never seen any trader, website, book, author, software or trading expert reporting “drawdown” that way.
