Hold and Hope changed to APD Ratio


Moreover, Matthew has bigger fish to fry. Check out this C2 system:



Cumu $ ($11,032)

after typical commission ($20,993)

Avg Win $3,820

Avg Loss $6,412

W:L Ratio 1.0:1

P/L per unit $155.67

after typical commission $153.43

Avg Trade Length 14.0 days

Compound Annual % -5.5% over 584 days

Sharpe Ratio -0.045

Max Drawdown 85.16% (20050627 to 20050720)

Risk of 20% account loss 58.4%

Risk of 50% account loss 27.4%

Risk of 100% account loss 0.0%



Please note it has a net loss: Cumu $ ($11,032)

Yet: P/L per unit $155.67

P/L per unit should be (using industry standards) a $30 loss.



Also:



Max Drawdown 85.16% (20050627 to 20050720)

Risk of 20% account loss 58.4%

Risk of 50% account loss 27.4%

Risk of 100% account loss 0.0%



Please note the DD occurred right when the system came

on line. IOW, there was an 85% account loss, yet there is

only a "27% risk of a 50% loss". The system is currently

down 80% off its equity peak. How is there not a 100%

risk of a 50% loss? There was a 50%+ loss and the account is

down 80% (based on its peak) right now!

Well, I don’t want to be seen as beating a horse that’s been dead for sometime, but my point about bringing up the “average” system ADP is…isn’t that the point at which a judgment should be based? There are enough systems on this site, that it seems to me that simply < or > the average would be where a subscriber would begin to make a judgment of concern or confidence.



As both a vendor and a subscriber, I would find the average ADP to be both valuable and interesting.



The P/L per unit is because of the way the calculation is done.



Say you have a system with the following two trades:



Trade 1: 2 contracts for a loss of -120.

Trade 2: 1 contract for a profit of 100.

Total profit: -20



The P/L per unit is calculated as follows:



Trade 1: P/L per Unit = -120/2 = -60.

Trade 2: P/L per Unit = 100/1 = 100.



P/L per unit = (-60 + 100)/2 = 20.



Technically a value of 20 is correct when you look at it from the viewpoint of someone who is trading fixed contracts per signal and ignore the position size from the vendor. If someone was trading just a fixed number of contracts for each signal, the P/L per unit would be 20 and the system actually would have been profitable.



With regard to the Risk of Account loss of x%, I think this is calculated based on Monte Carlo simulations. For the example above, 58.4% of Monte Carlo simulations resulted in a 20% drawdown, 27.4% in a 50% drawdown, etc…



Regards

- Fanus

> P/L per unit should be (using industry standards) a $30 loss.



I assume that you divided the net profit by the total number of contracts? Then I agree, but I remember a thread where I was trying to communicate that C2 doesn’t compute it that way, and that you denied that.

Fanus,

As Dustin pointed out a while ago, even if you trade at fixed size the formula isn’t correct when the trades are multi-legged.

We’ve been through all this. I understand. I just find

it curious that vendors rain down because ADP is

not an industry standard, yet these computations,

which are clearly VERY gentle regarding vendor risk,

and very non-standard, and not really good depictions of

reality, don’t get a peep. Indeed, they are defended.



As an example, why use theoretical risk (monte carlo) when

you have the actual risk in the same table!


We've been through all this. I understand.

Why bring it up again then?


As an example, why use theoretical risk (monte carlo) when
you have the actual risk in the same table!


Because Monte Carlo makes a projection of possible outcomes and risk based on past results and is not based on just one result. Assuming the system's characteristics don't change, the past results of is just one possible outcome of many and Monte Carlo simulations projects many different ones and show what other outcomes were also possible. Just as a low drawdown based on past trades can mean a system got lucky to obtain the perfect combination of trades which resulted in a the low drawdown, a high drawdown also can means that the system got unlucky to run into a bad combination of trades resulting in a high drawdown.

I don't really understand what you have against it. The actual result is there for everyone to see and Monte Carlo shows what else was possible.

You can also turn your argument around and look at a system with 10% max drawdown in the past, but Monte Carlo simulation show there is a 30% chance of a 50% drawdown based on the characteristics of the system. Would you rather that C2 then show there is a 0% chance of a 50% drawdown, because it didn't happen yet?

- Fanus

Yes, I remember now. Thank you for reminding me. At least I got the part right of explaining why total profit can be negative, but P/L per Unit is positive? Do I get bonus points for effort? :slight_smile:



- Fanus

> Why bring it up again then?



Because they are still broken. And Pal is asking for other fixes

that go against the way 99.9% of the investment world does

business.



> I don’t really understand what you have against it. The actual result is there for everyone to see and Monte Carlo shows what else was possible.



It doesn’t say it’s the Monte Carlo, it says:



" Risk of 50% account loss 27.4%"



When in fact there have been FIVE actual 50%+ losses in less

than two years.



> You can also turn your argument around and look at a system with 10% max drawdown in the past, but Monte Carlo simulation show there is a 30% chance of a 50% drawdown based on the characteristics of the system.



If the reverse were ever true in C2 you could. I have yet to see where

the “Risk of 50% account loss” was over stated. Show me some. You usually see systems with 5-10% DD’s have a 0% “Risk of 50% account loss”.

>You didn’t like the ADP because it wasn’t SOP



It would not be a SOP if it is based on closed trades and the open unrealized drawdowns are not full of errors which necessitates us to to base drawdowns and hence ADP on realized losses, i.e., drawdowns based on objective evidence which we can all trust…

> unrealized drawdowns



Using unrealized DD’s is SOP (standard operating procedure).

Every CTA and hedge fund and broker in the world looks

at open equity when judging performance. You should to.

Sorry, I thought SOP is Shi* Anyway, I will also look at it, if and when there is conclusive evidence for its existence which we wont have unless the position is exited. Until then…As EU says, the highs and lows of the day for stocks do not even officially exist…

As EU says, the highs and lows of the day for stocks do not even officially exist…

May I ask where did I say it?

Eu

> Anyway, I will also look at it, if and when there is conclusive evidence for its existence …



Open a brokerage account. Trade it the way you do your systems.

Your broker will give you conclusive evidence in the form of margin calls,

forced liquidation, and account closure if you trade as you do: doesn’t

matter if the position is open or closed. Just trade real money. You will see

PDQ.

sorry, not evidence enough…

Maybe, maybe not… but I do remember in the chat you mentioned something similar…

ps: not "conlusive" evidence, unless the position is exited…

> sorry, not evidence enough…



How would you know? Clearly you have never traded real

money in the real world.

Maybe, maybe not… but I do remember in the chat you mentioned something similar…

In the case I would appreciate if you won’t use my name for your fantasies.

Thank you in advance,

Eu

>In the case I would appreciate if you won’t use my name for your fantasies.



fantasy or not, you did mention it. Now you say that is a fantasy, so I am sure now that I can’t trust your word. I’m sorry that I ever engaged in a converstation with you. In future, remind me not to do so. Thanks in advance…