Margin To Equity Ratio - useful or not!???

Discussions about the testing and simulation of mechanical trading systems using historical data and other methods. Trading Blox Customers should post Trading Blox specific questions in the Customer Support forum.
Austrian
Roundtable Knight
Roundtable Knight
Posts: 223
Joined: Sat Oct 25, 2003 12:35 pm
Location: Austria, Vienna

Margin To Equity Ratio - useful or not!???

Post by Austrian » Fri Jan 06, 2006 2:40 am

As far as I see all examples of the performance of trading systems (created with trading blox) presented in this forum are with the global parameter leverage set to 1. As my experience shows that even the total risk profile shows you a tolerable heat, the margin to equity ratio could reach values professional money managers would not dare to market to their customers!??

What margin to equity ratios do you use for trading your own money compared to managing customers money??

Old European
Roundtable Knight
Roundtable Knight
Posts: 144
Joined: Mon Apr 19, 2004 3:04 pm
Location: Old Europe

Post by Old European » Fri Jan 06, 2006 7:56 am

Austrian,

The margin-to-equity ratio I personally feel comfortable with is somewhere between 10% and 20%.

I however have to admit that from a theoretical risk management point of view this ratio doesn't make a lot of sense. I still don't understand why consultants are still attaching so much importance to this figure.

Cheers,

Old European

kianti
Roundtable Knight
Roundtable Knight
Posts: 335
Joined: Fri May 02, 2003 6:10 am
Location: Florence - Italy

Post by kianti » Fri Jan 06, 2006 8:27 am

If you're familiar with the Kelly criterion, you can think margin/equity as the optimal bet size, the article below maybe can help.

Understanding the relationship between risk, reward and
margin/equity ratio in managed futures

http://www.aima.org/uploads/2002%5CJune%5Cwinton.pdf

sluggo
Roundtable Knight
Roundtable Knight
Posts: 2986
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo » Fri Jan 06, 2006 9:08 am

The margin-to-equity ratio people are hoping to calculate in TBB, has two components:
  • Numerator = For all currently open positions, the sum of (Exchange mandated Initial Margin Deposit per contract, times number of contracts in the position)
  • Denominator = Total account equity (Cash + Tbills + Open trade equity)
It's important to note that the Numerator consists of numbers pulled out of the hindquarters of exchange officials. These figures are infrequently updated and have very little relevance to current market conditions. They are "Cover Your Ass" protection for the exchange and for brokers; they are certainly not sensitive indicators of market risk.

Hence we realize that the margin-to-equity ratio is (baloney / total_equity), which means it's baloney. I only use it to assess how likely or unlikely it is to hit a margin call: If max(M/E) is less than 0.5, there's very little chance of a margin call. But if max(M/E) is 0.9. there's a pretty good chance of a margin call.

On the other hand, institutional managed money clients insist upon seeing graphs of M/E ratio vs. time (their daddies said to ask for them), so there are lots of these graphs lying around. Comparing two of these graphs to each other, you can (very) roughly estimate which manager uses more heat. As economists say just before they tell huge whoppers, All other things being equal, if two managers have equal returns, the one with the lower (M/E) ratio is taking less risk.

But the only plausible reason to want an M/E ratio graph from TBB, is "because my client(s) asked to see it".

kianti
Roundtable Knight
Roundtable Knight
Posts: 335
Joined: Fri May 02, 2003 6:10 am
Location: Florence - Italy

Post by kianti » Fri Jan 06, 2006 10:49 am

sluggo wrote: As economists say just before they tell huge whoppers, All other things being equal, if two managers have equal returns, the one with the lower (M/E) ratio is taking less risk.
Also, if a manager trades intra-day only the end-of-the-day margin will be zero; that doesn't mean that the risk is close to zero. It seems to me that margin/equity could only taken as a comparative indicator when the trade generators are similar. Comparing margin/equity ratios of a trend follower with the one of a naked options seller for istance could be very misleading.....I guess.

best regards, as ever

Austrian
Roundtable Knight
Roundtable Knight
Posts: 223
Joined: Sat Oct 25, 2003 12:35 pm
Location: Austria, Vienna

Post by Austrian » Sat Jan 07, 2006 6:41 am

Thank you all for your time and effort!

I attach a graph showing max margin/equity ratio %, avaerage margin/equity ratio and CAGR% of one of my systems as a function of leverage.
Attachments
margin.jpg
margin.jpg (40.98 KiB) Viewed 16208 times

edward kim
Roundtable Knight
Roundtable Knight
Posts: 344
Joined: Sun Apr 20, 2003 2:42 pm
Location: Silicon Valley / San Jose, CA USA
Contact:

Post by edward kim » Sat Jan 07, 2006 4:58 pm

i agree with everyone: margin/equity doesn't help very much.

(core equity minus skid) / total equity

these ratios show what kind of drawdowns you can experience if everything is stopped out now.

LeapFrog
Roundtable Knight
Roundtable Knight
Posts: 695
Joined: Mon May 17, 2004 4:18 pm
Location: Boston, MA

Post by LeapFrog » Wed Apr 16, 2008 9:38 am

Thought I would try running this little chestnut up the flag pole again to see if anyone else is in a saluting mood.

Since about early last Fall, I've seen around a doubling of margins from my brokers - understandable given the volatility, etc.

From the above thread and other comments, I seem to take the measurement of margin to equity more seriously than most. It is one of the key things I focus on in my trading because I believe in a concept Bob Buran called "margin efficiency"...the idea that we should get the most out of our margin, basically. I see no reason to trade only 30% of my equity if I can trade more of it - and not get a margin call. Why have all that money sitting around wasting time? Or put another way, I like to take the highest managed risk I can in order to get the greatest returns possible, as long as I don't get margin calls or go bust.

Anyway, I digress. The problem we/I have is that I know my broker margins today, and maybe recent years to the extent I've made a note of them, but I sure don't have them going back 30 years and TB can only handle one margin/intrument in the dictionary.

So plugging in double margin numbers today versus 6 months ago, makes a huge difference in our/my backtesting results WHEN I USE A MARGIN FILTER (as I do) to NOT add trades that would put me over an acceptable margin/equity percentage (pick you own number - same effect).

Apart from being aware of it (I test both old margin numbers and current ones), what thoughts do others have about this issue?

Thanks.

sluggo
Roundtable Knight
Roundtable Knight
Posts: 2986
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo » Wed Apr 16, 2008 10:36 am

LeapFrog wrote:... plugging in double margin numbers today versus 6 months ago, makes a huge difference in our/my backtesting results WHEN I USE A MARGIN FILTER (as I do) ...
Perhaps it would be useful to substitute "Volatility" for "Margin", based on these assumptions
  • When exchange officials lower and (especially) when they raise margin requirements, they are responding to changes in volatility.
  • Unlike Margin, Volatility can be measured simply from a price series, so you've got it going back 30+ years.
  • For each market or "Market Complex" (the beans/meal/beanoil complex, the crude/heating/gasoline complex, the hogs/corn complex, ...), you can do a data-fitting regression in Excel to see what the historical relationship has been, between Margin Requirements and Volatility. Maybe the volatility measurement that correlates best with Margin, is 10-day-ATR. Maybe it's HH(20) - LL(20), the height of the 20 day Donchian Channel.
It isn't perfect but it IS objective and it'll get you some of what you want, perhaps most of what you want, in the absence of historical data for margin requirements. Once you find the volatility measure that fits best (has the highest goodness-of-fit r2 coefficient), you automatically get a two parameter linear model (predictor) of margin requirement vs volatility. Done. Voila.

LeviF
Roundtable Knight
Roundtable Knight
Posts: 1428
Joined: Mon Dec 22, 2003 12:24 pm
Location: Des Moines, IA
Contact:

Post by LeviF » Wed Apr 16, 2008 12:05 pm

Sluggo, you are a clever one...

LeapFrog
Roundtable Knight
Roundtable Knight
Posts: 695
Joined: Mon May 17, 2004 4:18 pm
Location: Boston, MA

Post by LeapFrog » Wed Apr 16, 2008 1:32 pm

sluggo wrote: Done. Voila.
Yes, I know the exchanges dream up the margin numbers based on their feelings about volatility - so your suggestion makes sense.

It would be tricky to get to the Voila stage, at least for me, but maybe worth a go. I imagine I would need to build it in the Risk Manager and have the margin filter be a "volatility filter" instead. This probably means building arrays and things which are beyond my meager programming skills right now - but I like the way you think.

sluggo
Roundtable Knight
Roundtable Knight
Posts: 2986
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo » Thu Apr 17, 2008 7:58 am

You could do all the calculations off-line (i.e. using software other than Blox) and store your calculated-from-volatility PseudoMargin numbers in files. What Blox's esteemed competitor "M" calls companion files. This way your Blox systems don't have to do all those messy calculations, so your systems will run faster. Especially useful for large Stepped Parameter simulations comprising hundreds or thousands of Tests.

Your Blox code would read the companion files (one per instrument) using the instrument.LoadExternalData() function.

It's exactly the same mechanism you would use if you did have historical data of actual margin requirements for each date of past history. You're just plugging in estimated numbers rather than actual numbers.

Angelo
Roundtable Fellow
Roundtable Fellow
Posts: 90
Joined: Fri Apr 29, 2005 4:31 am
Location: Italy

Post by Angelo » Thu Apr 17, 2008 10:16 am

LeapFrog wrote: Anyway, I digress. The problem we/I have is that I know my broker margins today, and maybe recent years to the extent I've made a note of them, but I sure don't have them going back 30 years and TB can only handle one margin/intrument in the dictionary.

So plugging in double margin numbers today versus 6 months ago, makes a huge difference in our/my backtesting results WHEN I USE A MARGIN FILTER (as I do) to NOT add trades that would put me over an acceptable margin/equity percentage (pick you own number - same effect).

Apart from being aware of it (I test both old margin numbers and current ones), what thoughts do others have about this issue?

Thanks.

Hi LeapFrog,

your remarks are always interesting.

As you are certainly aware, margins don't change just by the variations in volatility (that has - normally - short term spurs on the upside but in the long term seems to be mean-reverting, and the same should be its effect on margin's amounts) but also because changes the total amount of the contract.

For example, I can trade 1 lot of the DAX future for about 7-10% of the notional value, so it makes a huge difference if the DAX index is in the 3000 range (as in 2003) or in the 8000 range (year 2000) or in the 6500 range (as in 2008).

Similarly, I don't trade soybeans but I will be much surprised to know that - supposing constant volatility - one was asked the same margin when price was above 1000 or around 500.

That's the reason why, just two days ago I made this little suggestion in the customer’s area of the forum:

viewtopic.php?t=5215

Maybe others can give their opinion on my proposal.

LeapFrog
Roundtable Knight
Roundtable Knight
Posts: 695
Joined: Mon May 17, 2004 4:18 pm
Location: Boston, MA

Post by LeapFrog » Fri Apr 18, 2008 7:39 am

sluggo wrote:You could do all the calculations off-line (i.e. using software other than Blox) and store your calculated-from-volatility PseudoMargin numbers in files. What Blox's esteemed competitor "M" calls companion files. This way your Blox systems don't have to do all those messy calculations, so your systems will run faster. Especially useful for large Stepped Parameter simulations comprising hundreds or thousands of Tests.

Your Blox code would read the companion files (one per instrument) using the instrument.LoadExternalData() function.

It's exactly the same mechanism you would use if you did have historical data of actual margin requirements for each date of past history. You're just plugging in estimated numbers rather than actual numbers.
I'm not sure either a single percentage or fractional approach nor a simulated number based on price volatility will give me the accuracy I desire.

I think this is the way to go - and thank you Sluggo for pointing me in this direction. With the help of my friendly broker (OEC) I'm building up a database of actual historical margins. For example, the CBOT kindly publish theirs right here:

http://www.cbot.com/cbot/pub/page/0,3181,1859,00.html

Once assembled, I will have a go at coding up a instrument.LoadExternalData function within this Blox:

viewtopic.php?t=3524

I'll report back with the effects if it works out. I expect it should actually improve my historical back testing results because as of now, I use the most current margin rates which are high by historical standards and thus penalize my results by excluding trades which would not otherwise be excluded if lower margins were used.

Dean Hoffman
Roundtable Fellow
Roundtable Fellow
Posts: 87
Joined: Wed Jul 02, 2003 3:04 pm

Post by Dean Hoffman » Tue Jul 08, 2008 5:02 am

Where the margin to equity ratio is particularly useful is when evaluating CTA’s. The reason for this is because most CTA’s don’t report their portfolio heat; however they do report their MTE ratios. I know in my own trading my portfolio heat and margin-to-equity ratio come pretty close to being the same. Any manager using a volatility calculation for risk is going to have high correlation between his portfolio heat and his MTE ratio. As sluggo said “All other things being equal, if two managers have equal returns, the one with the lower (M/E) ratio is taking less riskâ€

Mathemagician
Full Member
Full Member
Posts: 20
Joined: Thu Jun 28, 2007 1:46 pm
Contact:

Post by Mathemagician » Tue Jul 08, 2008 2:46 pm

There are far more direct and effective proxies for risk (even relative risk) and suggesting that we use margin simply because it's there is not a terribly compelling argument, even when so eloquently expressed.

jj

LeapFrog
Roundtable Knight
Roundtable Knight
Posts: 695
Joined: Mon May 17, 2004 4:18 pm
Location: Boston, MA

Post by LeapFrog » Tue Jul 08, 2008 7:54 pm

I thought Dean's remarks made a lot of sense and am glad to see him posting here again.

He is talking about evaluating public funds which disclose very little when it comes to their risk adjusted returns.

For the individual trader, I believe, things are quite different. We can use any number of ways to measure our own risk adjusted returns and the mathematical types have plenty I am sure. Personally, I like to use the greatest amount of margin possible in my trading because I see no point in having a stack of cash goofing off earning treasury rates when I can get much higher returns on it trading - means higher volatility of those returns of course and that is where the fun is. This is just me, but I would rather earn outsize returns while running a high MTE ratio (say 50 to 80 percent) than a 12 - 20 percent return while using only 10 to 20 percent margin to equity.

Funds are satisfying different requirements and in that context Dean's comment makes perfect sense to me.

Dean Hoffman
Roundtable Fellow
Roundtable Fellow
Posts: 87
Joined: Wed Jul 02, 2003 3:04 pm

Post by Dean Hoffman » Wed Jul 09, 2008 7:49 am

Yes LeapFrog, my point is more directed towards evaluating CTA’s. Indeed I have all sorts of additional measuring tools available to me in my own systems trading than I do if I’m selecting a CTA. And by no means did I ever imply that MTE was the ONLY tool one would use. To me it’s just a “quick and dirtyâ€

sluggo
Roundtable Knight
Roundtable Knight
Posts: 2986
Joined: Fri Jun 11, 2004 2:50 pm

Post by sluggo » Wed Jul 09, 2008 9:48 am

Dean, the Sharpe Ratio is usually interpreted as a Reward/Risk measure. Not a measure of risk itself. Good old Sharpe himself, the Nobel prizewinner, has a webpage on his department website at Stanford U, devoted to the Sharpe Ratio. Have a look: http://www.stanford.edu/~wfsharpe/art/sr/sr.htm .

I think a big part of the problem is that the word "risk" is ill-defined or undefined. We recognize that there is a probability distribution of possible outcomes, an infinite number of possible outcomes, and we try to boil this down to a single metric called "risk". Some people take the left tail three sigma point (99.5% of possible outcomes have bigger gains / smaller losses) and call it "the risk". These people aren't stupid, they are fully aware that bigger losses than "the risk" are possible and will eventually occur, but they have decided that defining "the risk" to be three sigma is more useful in practice than saying "the risk is infinite" or "there WILL be another crash, worse than 1929, we just don't know when"

Then when a rare and truly whopping loss occurs, way waaay WAAAY off in the left tail of the distribution, self satisfied schmucks start bleating "Black Swan, Black Swan!" and telling people that nobody understands "risk" but them.

Who is right? What is the "true" value of "risk"? Neither is right because "risk" is a probability distribution of unfavorable outcomes (and risk-in-the future is an unknown distribution), which cannot be captured in a single number.

You'll notice that Prof. Sharpe tries to confine himself to talking about "variability" rather than "risk". But he fails (!), as you can see for yourself by doing a text search for "risk" on his Sharpe Ratio page.

blinkybill
Roundtable Knight
Roundtable Knight
Posts: 188
Joined: Thu Mar 27, 2008 6:44 pm

Post by blinkybill » Thu Jul 10, 2008 2:14 am

This is an area that has been occupying my mind so I thought I would throw in my 2 cents. Most of the goodness measures proposed on this forum seem to define risk in terms of drawdown. I am a little unsure about this for precisely the reasons put forward by Dean (LTCM et al). Shouldnt we be looking for a measure which calculates the ROI of a strategy, which is how a trading proposition might be looked at in the real world. e.g I committ x dollars to the strategy to receive y dollars in return. In this case CAGR in equity needs to be seen in the light of how many dollars were committed to achieve the return. One could say drawdown is the best measure of that but perhaps not. Perhaps it would be better to use the portfolio heat measure or open risk to equity instead. I would like to know if anyone else has played around with these types of measures as an alternative? I know a little about fund management and the measure we always focused on was the annual returns gained and how much leveraged was used to achieve them? i.e the leverage could be 5-10x or more.......on gross positions

Post Reply