The Max Intraday Drawdown is incorrect

Questions about MultiCharts and user contributed studies.
ppan
Posts: 106
Joined: 31 Jan 2007
Has thanked: 2 times

The Max Intraday Drawdown is incorrect

Postby ppan » 20 Aug 2008

The Max Intraday Drawdown in Optimization Report of Portfolio Backtester is incorrect.

khalaad
Posts: 323
Joined: 07 Jan 2007
Location: Lahore, Pakistan
Has thanked: 64 times
Been thanked: 57 times

Postby khalaad » 21 Aug 2008

ppan,

Thank you for raising this matter. I am afraid you are unlikely to get remedy!

On August 6 -- http://forum.tssupport.com/viewtopic.php?t=5420 -- I wrote:
in my estimation does not calculate the drawdowns correctly, which is probably why it does not provide the Sharpe;
To this (and a few other complaints) Marina replied;
Could you elaborate on the above please?
I did not bother to elaborate the self evident.

Khalid

User avatar
Marina Pashkova
Posts: 2758
Joined: 27 Jul 2007

Postby Marina Pashkova » 22 Aug 2008

Khalid,

If you re-read my answer you will see that I wrote:
Sharpe ratio is calculated, however it is not displayed properly (as well as some other ratios). This will be fixed shortly.
As for the drawdown calculation we need more information. This is why I asked you to elaborate. You can contact us via LiveChat to point out the wrong values.We can only help you if you provide us with the necessary information. The figures that you cited are not enough to make any comments.

Regards.

khalaad
Posts: 323
Joined: 07 Jan 2007
Location: Lahore, Pakistan
Has thanked: 64 times
Been thanked: 57 times

Postby khalaad » 22 Aug 2008

I have trashed my MultiCharts today so I shall address only the two pertinent matters and not bother TS Support about what else I think is wrong with Optimization Report:

To Marina's argument "SHARPE ratio is calculated, however it is not displayed properly ...", I ask: is a Rolex which does not display time any good?

I think MultiCharts carries one days intraday drawdown into the next day, and continues to do so. I fail to understand how else could MAXIMUM INTRADAY DRAWDOWN can be 3 or more times than the maximum closing drawdown. If I had MultiCharts, I would have shown an example here. Once a market is closed for the day the intraday drawndown should cease to matter, what should matter is the closing drawdown.

Khalid

khalaad
Posts: 323
Joined: 07 Jan 2007
Location: Lahore, Pakistan
Has thanked: 64 times
Been thanked: 57 times

Postby khalaad » 22 Aug 2008

Marina,
As for the drawdown calculation we need more information. This is why I asked you to elaborate. You can contact us via LiveChat to point out the wrong values.We can only help you if you provide us with the necessary information. The figures that you cited are not enough to make any comments.
I understand Andrew has already accepted on another forum the drawdown problem exists.

Does your reply not show MultiCharts IS NOT BOTHERED ABOUT, AND UNWILLING TO, correct this thing?

For your sake I sincerely hope you folks realise it is the TS Support users who are your real paymasters.

Khalid

khalaad
Posts: 323
Joined: 07 Jan 2007
Location: Lahore, Pakistan
Has thanked: 64 times
Been thanked: 57 times

Postby khalaad » 22 Aug 2008

ppan,
I am afraid you are unlikely to get remedy!
Instead of remedy, you have got prosecutorial arguments!!!

Khalid

Tresor
Posts: 1104
Joined: 29 Mar 2008
Has thanked: 12 times
Been thanked: 53 times

Postby Tresor » 23 Aug 2008

Marina,

Just optimize any strategy you want and apply to this strategy a fixed amount of stop loss of e.g. 100 dollars. To make it short, the intraday drawdown should not exceed $100. When you make a drawdown of $100 you shoud be automatically stopped out.

However when you receive the optimization report, you will see that the max intraday drawdown of this strategy can be as high as $ 500 / $1,000 / $10,000 whatever, even if only 1 trade was made during one day.

Regards

khalaad
Posts: 323
Joined: 07 Jan 2007
Location: Lahore, Pakistan
Has thanked: 64 times
Been thanked: 57 times

Postby khalaad » 23 Aug 2008

Tresor,

TS Support knows EXACTLY what is WRONG with MultiCharts. If they were such nincompoops they would not have come this far; they smarter than you and I and have answers to questions we haven't even thought of.

You have read the Elitetrader thread: Andrew accepted problem existed only when he was left with no other option.

All the claims about:
true strategy back-testing;
state-of-the-art integrated auto-trading;

here

http://www.tssupport.com/multicharts/

are untrue until TS Support delivers at some future unspecified far-off date.

Khalid

User avatar
Marina Pashkova
Posts: 2758
Joined: 27 Jul 2007

Re: The Max Intraday Drawdown is incorrect

Postby Marina Pashkova » 25 Aug 2008

The Max Intraday Drawdown in Optimization Report of Portfolio Backtester is incorrect.
ppan,

The above information is not enough. We need further details to understand what problem you are talking about.

To begin with we'll need

1. The description of the problem
2. Screenshots illustrating it

After that we might need further information such as the exported report, the strategy and the data that you were using.

For example, we have just finished working on this problem (maybe this is what you're talking about):

The value in the Optimization Report used to be extremely high as compared to the Strategy Performance Report (see the attached screenshot).

Once again, until you have provided us with at least the description of the problem and a screenshot I cannot make any comments at all.
Attachments
MaxIDDrawDown.JPG
(236.18 KiB) Downloaded 1319 times

User avatar
Marina Pashkova
Posts: 2758
Joined: 27 Jul 2007

Postby Marina Pashkova » 25 Aug 2008

I have trashed my MultiCharts today so I shall address only the two pertinent matters and not bother TS Support about what else I think is wrong with Optimization Report:

To Marina's argument "SHARPE ratio is calculated, however it is not displayed properly ...", I ask: is a Rolex which does not display time any good?
I already said that while Sharpe and some other ratios are internally calculated correctly (which was double-checked by our engineers) there is an error related to the display. Your Rolex argument is, to put it mildly, not quite relevant in this case.
I think MultiCharts carries one days intraday drawdown into the next day, and continues to do so. I fail to understand how else could MAXIMUM INTRADAY DRAWDOWN can be 3 or more times than the maximum closing drawdown. If I had MultiCharts, I would have shown an example here. Once a market is closed for the day the intraday drawndown should cease to matter, what should matter is the closing drawdown.
Khalid
Maybe the definition of the maximum intraday drawdown will help you understand how it's possible:

---------------------------------------

Displays the greatest loss drawdown, from the previous highest equity run-up, bar to bar looking across all trades, during the specified period. If a new bar equity run-up high occurs, the low equity value is reset to 0 so that the next maximum drawdown can be calculated from that point.<br>You can roughly see this value on a detailed equity curve graph by looking from the highest peaks to the lowest peaks moving forward.

----------------------------------------

Fianally, when I asked you to elaborate on the issue, it was not just to write something: every problem should be clearly described with all the relevant information cited. For example, when you are talking about errors or wrong calculations in the strategy performance report we at least need screenshots with the wrong values. In most cases we'd also need exported reports, a strategy etc. Our engineers are solving very specific problems and need very specific details. A problem reported without any specific information is meaningless. If you want to have something attended to quickly, contact us via LiveChat just as our customers usually do when they think they have a serious problem: they show us what's wrong and have it fixed right then and right there. If it can't be fixed right then and right there, the problem is added to our bugs-to-be-fixed list.

If you want us to help you just give us the information that we ask for. Those aren't idle requests: if I ask for information, it means that our engineers need it. If I say that a certain statement is not enough to identify the problem, it is because our programmers are saying so. We can't help you unless you co-operate.

User avatar
Marina Pashkova
Posts: 2758
Joined: 27 Jul 2007

Postby Marina Pashkova » 25 Aug 2008

Marina,

Just optimize any strategy you want and apply to this strategy a fixed amount of stop loss of e.g. 100 dollars. To make it short, the intraday drawdown should not exceed $100. When you make a drawdown of $100 you shoud be automatically stopped out.
Please see the definition of the max intraday drawdown below, and you'll see how it's possible:

-----------------------------------

Displays the greatest loss drawdown, from the previous highest equity run-up, bar to bar looking across all trades, during the specified period. If a new bar equity run-up high occurs, the low equity value is reset to 0 so that the next maximum drawdown can be calculated from that point.<br>You can roughly see this value on a detailed equity curve graph by looking from the highest peaks to the lowest peaks moving forward.

-------------------------------------

If you have a stop loss of $100, it simply exits your trade with that loss. But if you had several losing trades, drawdown will certainly be more than that.

However when you receive the optimization report, you will see that the max intraday drawdown of this strategy can be as high as $ 500 / $1,000 / $10,000 whatever, even if only 1 trade was made during one day.
That's exactly how it's supposed to work.

User avatar
Marina Pashkova
Posts: 2758
Joined: 27 Jul 2007

Postby Marina Pashkova » 25 Aug 2008

Tresor,

TS Support knows EXACTLY what is WRONG with MultiCharts. If they were such nincompoops they would not have come this far; they smarter than you and I and have answers to questions we haven't even thought of.

You have read the Elitetrader thread: Andrew accepted problem existed only when he was left with no other option.

All the claims about:
true strategy back-testing;
state-of-the-art integrated auto-trading;

here

http://www.tssupport.com/multicharts/

are untrue until TS Support delivers at some future unspecified far-off date.

Khalid
Khalid,

MultiCharts is comprised of millions of lines of code. It allows so many scenarios of usage that we cannot test them all to find out what might go wrong in a particular situation.

If you have encountered a problem, just report it providing all the relevant details of how it can be reproduced. Otherwise, we simply won't be able to help you.

khalaad
Posts: 323
Joined: 07 Jan 2007
Location: Lahore, Pakistan
Has thanked: 64 times
Been thanked: 57 times

Postby khalaad » 25 Aug 2008

Marina,
I already said that while Sharpe and some other ratios are internally calculated correctly (which was double-checked by our engineers) there is an error related to the display. Your Rolex argument is, to put it mildly, not quite relevant in this case.
VERY TRUE. MultiCharts is no way near being a Rolex. In 2006 and in 2007 I thought MultiCharts will one day be a Rolex. In 2008.....

AND I do not understand why do you keep on insisting Sharpe is calculated correctly? If a user is not able to see and know the Sharpe value, what good is the internal correct calculation??

Whenever I wrote about drawdown I wrote "I think" or phrases with similar meaning. All I know there is/was a drawdown problem backtesting -- probably, looking at you screenshot, resolved.

I know MultiCharts is work in progress and is likely to remain so for the forseeable future. In my opinion this is least of the problem.

My problems with MultiCharts are more fundamental and described elsewhere in the forum.

Khalid

User avatar
Marina Pashkova
Posts: 2758
Joined: 27 Jul 2007

Postby Marina Pashkova » 25 Aug 2008

Fanus and Khalid,

We are open for criticism and suggestions, however your statements are highly subjective and have nothing to do with objectivity. From now on we are not going to comment them.

If you think our platform is not good enough, stop using it but do not accuse us of lying on our website.

fs

Postby fs » 25 Aug 2008

Fanus and Khalid,

We are open for criticism and suggestions, however your statements are highly subjective and have nothing to do with objectivity. From now on we are not going to comment them.

If you think our platform is not good enough, stop using it but do not accuse us of lying on our website.
I have not even posted in this thread and I have NEVER accused you of lying on your website, even though this was pointed out numerous times that the features listed are just paper features and not real working features.

Your backtesting and optimization is not accurate. You confirmed so yourself. The same goes for portfilio backtesting. Also for autotrading. The numerous posts here shows that. I can post the e-mails sent to you over the last few months not being resolved as of yet as prove. There is nothing subjective about this as you claimed.

You don't have to respond. There are numerous other forums where I can post my "subjective" comments and potential customers can "comment" themselves by deciding not to purchase your beta product marketed as an official release.

Just one thing to remember... Ignoring issues has never and will never work to make them go away.

User avatar
Marina Pashkova
Posts: 2758
Joined: 27 Jul 2007

Postby Marina Pashkova » 25 Aug 2008

Sorry Fanus,

It was meant to be addressed to Khalid only.

Regarding the earlier issue with drawdown:

It took us a while to understand where the problem was but once we did it was addressed. Do you have anything else to report?

Tresor
Posts: 1104
Joined: 29 Mar 2008
Has thanked: 12 times
Been thanked: 53 times

Postby Tresor » 25 Aug 2008

Hello Marina, Khalid and Fs,

The discussion is becoming more and more fierce. Let's have a rest from this and let's give the issue a fresh and unemotional start :)

Marina has just started the thread on the bugs in the Portfolio Backtester.

Fs, please copy your e-mails here in public: http://forum.tssupport.com/viewtopic.php?t=5493

Thanks

khalaad
Posts: 323
Joined: 07 Jan 2007
Location: Lahore, Pakistan
Has thanked: 64 times
Been thanked: 57 times

Postby khalaad » 25 Aug 2008

Tresor,

You are right.

After Marina posted OPTIMIZATION AND BACKTESTING BUGS at 4:57 GMT, I should have had the good sense to stop and see if progress is made.

I did not, for which I apologise to everybody. Anyway the posts have now been removed.

Marina,

I have been scathing in my comments; had I wanted to say you are lying I would have said so.

In my opinion MultiCharts has been careless; I am upset because carelessness is one of the main causes of early death.

Khalid

User avatar
Marina Pashkova
Posts: 2758
Joined: 27 Jul 2007

Postby Marina Pashkova » 04 Sep 2008

Hi everybody,

Drawdown calculations have been fixed. The fix will be availble in 4.0 beta 2 about to be released.

khalaad
Posts: 323
Joined: 07 Jan 2007
Location: Lahore, Pakistan
Has thanked: 64 times
Been thanked: 57 times

Postby khalaad » 18 Sep 2008

Drawdown calculations have been fixed.
True.

The Version 4.0 Beta 2 (Build 1573) correctly calculates all drawdowns.

This is very good news.

Anything which helps optimum money management must be applauded.


Khalid

User avatar
Marina Pashkova
Posts: 2758
Joined: 27 Jul 2007

Postby Marina Pashkova » 25 Sep 2008

Hi guys,

Let us know if you notice any other problems with the report.

Regards.


Return to “MultiCharts”