Max DD Is A Little Misleading

To show a drawdown (DD) against an open profitable trade/s is misleading. I understand that the DD is shown due to the fact that the current profit is below its prior max profit reached, but I still don’t fully agree that it should be shown as a DD. To give viewers a true picture of system equity performance, the equity curve should display two lines on the equity graph: 1. “open equity” (as current) and 2. “closed equity” using different colours and width to distinguish. The open equity curve will show both true max drawdown and max profit reached when referenced against the closed equity , thus the overall system efficiency. I believe the true Max DD should be calculated as closed.equity-open.equity as % (if open<closed). Critisism not intended and probably difficult to implement technically.

"To show a drawdown (DD) against an open profitable trade/s is misleading"

Why? That part of a traed is called maximum adverse excursion, and is well recognized by many in the industry. The only problem is, that the max DD here is not always accurate…

Fair comment. Could you please clarify your comment “max DD here is not always accurate”. Do you mean max DD inaccuracy due to ‘data errors’ and/or ‘calculation method’ ? At least my post was not a total waste of time…To be honest I did not know such a term existed…“maximum adverse excursion”. I learnt something new today on c2.

there were some posts in the past, where vendors/traders said that the max DD was different than what it actually was. But I think most people would find that over time, it averages out. The APD indicator here sums up all the recorded max DDs and all the recorded net profits for a system, and tells you how much punishment a system inflicts on a subscriber on average, en route to profits.

the opposite of MAE is Maximum Favorable Excursion. That is the highest potential profit level attained during the life of that trade.

Both values give some illumination on what might be decent stops or profit targets.

For example, if 80% of the trades have a max DD of less than $300 (per unit), some people draw a line there ($300 per unit) for future stops.

I’m not against your suggestion to show a closed equity curve too, because it will show how much capital is un-used, if I understand your suggestion.

I am against your suggestion to compute the max DD differently. I think that is not industry standard. If you think it is, please provide references. Besides, I don’t see why it would be more “true” than the current formula.

Perhaps I should have worded the post differently. I think that the the max DD will give different results depending on the data series selected.

That is, max DD based on open equity will be different to the max DD based on closed equity. I can not say which data series is industry standard, I assume both may be equally valid. Which raises the question… if both are valid but differ?

I am suggesting displaying both closed and open equity curves gives onlookers more information to evaluate system performance. This is just my opinion.

Take a look at my method description. It has some info on drawdowns at the bottom…

What exactly do you mean by closed equity curve? Do you mean the amount of cash? Or do you mean the cash + the value that the open positions had at the time they were opened? Or something else?

> most people would find that over time, it averages out.

I doubt it. Remember the Prime Time $400K DD posts? That

should have been a $200K DD. C2 assumed the position was

open all day even though it closed out intraday. Takes a lot

to average out $200K errors.

I want to see open and intraday DD’s, but I don’t want to see

$200K errors.

Most of the big ones will wind up getting reported to Matthew for correction in the permanent record. I had one or two of those, and MK fixed them…

System drawdown is measured from portfolio equity. Trade dradown is for single trade and measures change of value of single trade. So trade drawdown can be higher than system drawdown if trade is NOT entered with 100% of equity.

Also Trade drawdown is not the same as MAE because MAE is measured from ENTRY to TROUGH value of trade, while max. trade drawdown is measured from PEAK to TROUGH value of trade (is higher than MAE)

When using MAE and MFE to set stops and profit targets, if you only

use histograms of their distributions you’ll sometimes get very

surprising results when you re-run the backtest with those stops and

profit targets. This is because the histograms don’t show you the

closed profits/losses of the trades associated with the MAE and MFE

values. So sometimes it can be useful to make simple scatter plots of

the trades, with profit/loss on the Y axis and MAE or MFE on the X

axis to see this relationship. You could then break this down by all,

winning, losing, long and short trades. And if you do, you’ll see

there are fundamental differences in relationships between P/L and MAE

for winning and losing trades, for example.

MAE/MFE relate to SINGLE trade only. Drawdowns relate to portfolio

equity, also Max open drawdown is peak-through distance while MAE is only entry->through distance (smaller) while Max closed drawdown is the realized loss (if any).

The main application of MAE/MFE is properly setting max. loss/profit


Think of MAE as the most a trade has gone against you during its

life and MFE as the most a trade has gone in your favor. MAE is almost

like maxdd except it’s the lowest point on the equity line of a

single trade as opposed to the max “peak to subsequent trough

difference” as with maxdd. MFE is simply the highest point on a trade’s equity line. The graphs show the distributions of MAE and MFE for the

trades in a backtest and can give you an idea of what to expect in the

future, assuming past results are indicative. For example, you might

look at the MAE histogram and see that almost all trades went 5% or

less against you at their worst point. Then, if a large number of

your system’s trades have larger MAEs in real time, it should

raise a red flag. Keep in mind that the vertical bars in the histogram

are actually bins into which the MAE observations fall. So the first

bin would contain observations from -5% to 0, the second bin from

-10% to -5.00001% (or whatever), etc. The same principle applies to the

MFE histogram. You can also use this info to help figure out where

to put stops and profit targets or to compare and contrast systems to see which best suit your temperment. Or again, as a benchmark based

on past performance.

Closed being amount of cash/balance where as open being cash+open positions. Its just personal preference to see both values displayed for informational purposes only.

ps: The MFE graphic is excellent at evaluating the relative

efficiency of individual trades. The sign of a good system is

the clustering of small losing trades with low runup and a

clear line of winning trades in a 45-degree angle. The closer

the winning trades follow the 45-degree line, the more

efficient the trades are — the trade’s unrealized profit or

runup and the realized profit are relatively close. It is this type

of trading characteristic that works best with the MFE risk

management strategy.

To apply the MFE analysis, a trader must go through three

stages. The first stage uses the maximum favorable excursion

graphic to evaluate the characteristics of the system and see

if there are any exploitable tendencies. This stage can help

eliminate systems that trade inefficiently and offer little

potential for exploiting the relationship between runup and

realized profit. The second stage focuses on finding an

appropriate MFE support level at which to add to positions.

Finally, the third stage involves testing specific values to

determine the MFE level that works best with the trading


In all, maximum favorable excursion is a powerful strategy

that can benefit both discretionary and mechanical traders.

If closed equity is the amount of cash, then opening a new position will cause a drawdown in the closed equity. Why is it important to know the maximum drawdown of this type?

Indeed you are right. The one in question has been corrected. Very nice:

BTO 190 @ESH7 E-MINI S&P 500 1444.26

2/26/07 12:36 STC 190 1421.62 2/27/07 14:14 ($246,585)

n/a ($215,124)

As I see it, closed equity should remain static and only updated when an open position is closed, hence I dont understand how opening a new position will cause closed equity to change. I dont claim to be an expert on the ins and outs of DD MAE etc…for me DD somehow suggests loss of capital, but if a trade is closed at a profit but not at its maximum profit reached during its life, c2 lists this as a DD even though my cash/balance has increased. It now seems likely that my interpretation of DD is incorrect?

I think the drawdown calculated in the system performance statistics is correct. However this is a drawdown after position is closed. Another thing is the drawdown while position is held (those who experienced it in real trading will know what I mean :))

It would be interesting to see this kind of “Max. Intraday Drawdown” in the systems’ statistics.

>for me DD somehow suggests loss of capital, but if a trade is closed at a profit but not at its maximum profit reached during its life, c2 lists this as a DD even though my cash/balance has increased.

Yes, that is one of the drawbacks of the present combined equity curve making it misleading, unfair and which makes no sense. This very same point was made by Chris Morse some time back (can’t seem to find that post)…

"As I see it, closed equity should remain static and only updated when an open position is closed, hence I dont understand how opening a new position will cause closed equity to change."

Ah, but then, if I understand you correctly, your answer to my previous question should have been that you define closed equity as cash + the value of open positions at the time that they were opened. Just to be sure that we’re talking about the same: Suppose the account has $250K cash at day x and it opens one position that looses $50K in value at day x + 1 but is not closed, looses another $50K at day x + 2 but is not closed, looses another $50K at day x + 3 but is not closed, looses another $50K at day x + 4 but is not closed, and then relative to this gains $175K at day x + 5 and is closed. So the open equity curve would subsequently display

$250K, $200K, $150K, $100K, $50K, $225.

Am I correct that according to you the closed equity curve would be:

$250K, $250K, $250K, $250K, $250K, $225K?

If that is what you mean, then I don’t see the advantage of it. It looks to me as just a loss of information compared to the open equity curve. For one, it would be impossible to distinguish the system on basis of the closed equity curve from a system that holds $250K in cash all the time, and then looses $25K on the last day without going to $50K first.

If this is your idea of a drawdown, then why bother about drawdowns at all? Why is is closed drawdown more important than an open drawdown? For all you know the system may have $50K profit the next day, so would that not deem the temporary $25K closed loss unimportant as well?

In my view drawdowns are a means to indicate the risk of a system. Even if the trade is not closed, a draw down in the open trade indicates that a risk is taken. If that is not important to you, then it suffices to consider only the current account value of the system, and then neither the open nor the closed draw downs are important. If the path through which the current account value is attained is important to you, however, then I think there is no reason to distinguish between open drawdowns and closed drawdowns.

To elaborate further on this, suppose a second vendor has exactly the same system, except that he briefly closes the trade on day x + 4 and immediately after that opens the same positions again. Then his system would show a $200K closed drawdown while the first system has only a $50 closed drawdown, even though they have essentially the same system.


for me DD somehow suggests loss of capital, but if a trade is closed at a profit but not at its maximum profit reached during its life,

I really like an idea of closed equity curve. Unfortunately, even if I hire a person to explain my broker the idea in poetic/philosophical English, the bastard(broker) will liquidate my positions by open equity curve. Period.

If you’ll find a broker who’ll accept closed equity curve, pls, let me know. I’ll be in :wink:

c2 lists this as a DD even though my cash/balance has increased.

C2 lists DD per position(on track record) and per portfolio(global stats). What’s wrong with that?