Estimating Future Risk…

[DPH]

Here, is a scenario to consider. Two money managers have identical MAR ratios. They have the same CAGR and the same maximum drawdown. The investor’s job is to pick the better manager. What tools other than MAR can the investor use?

For many, industry standards like Sharpe Ratio, Sterling Ratio and Sortino Ratio come to mind. Although, it is my opinion that these ratios represent more hindsight than foresight.

Take, for example, the S&P option writing programs of the early to mid 2000’s. One of the best known was ACE Investment Strategies. ACE had performance ratios off the charts (Sharpe, Sterling, Sortino, MAR) with smooth, consistent growth and barely noticeable drawdowns. From just about any measure, investors could not have asked for more. It was the darling of retail brokerage firms with small minimum account size requirements. (See graph below)



But, just like Long Term Capital Management, Ace had a sudden and for most, unexpected implosion! After years of flawless growth, they had close to a 70% drawdown in several months! (see graph below)



In hindsight, ratios such as Sharpe were worthless. They lured many people into the lion’s den!

But, there was one group of risk analysts who did have an excellent handle on the potential for Ace’s implosion. Who was it? The Chicago Mercantile Exchange, that is who.

It was the CME that set the margin requirements, and Ace was trading at high margin to equity ratios of 50% and more. This compares to many CTAs who trade at less than 20% and some less than 10% margin to equity ratios.

It is my opinion that margin usage is a leading indicator of potential risk and drawdowns, but there is one school of thought that thinks using more margin is not necessarily dangerous because it can add to diversification. To some degree I agree with this, but traders quickly reach a point of diminishing returns.

In the following graph, I have plotted average margin usage against average drawdowns for several hundred CTAs.



Traders can clearly see from the data and trend line that using more margin on average leads to higher drawdowns.

Some investors might say this is obvious, more leverage = more risk, but here on this forum I had a debate with someone who referred to himself as a “mathematical geniusâ€

[AFJ Garner]

Well I for one have always enjoyed your contributions to the Forum. Yes, I remember the self styled maths genius. I seem to recall finding him a little rude and self opinionated but no doubt I suffer from the same flaws on occasion.

A most interesting chart. I do not intend to argue the toss one way or the other but my colleague and I have certainly adopted a low margin to equity ratio following the same beliefs as you outline.

Sadly, I do not believe such a policy will always keep us at DD levels we will be happy with but I believe it will certainly help.

In my book on ETFs I outlined a momentum system which re-allocated the portfolio once a month. By way of example such a system had the investor heavily in stocks in the early part of the 2010 run up and suffered a drawdown in May 2010 as the markets turned. Such a system then switched heavily into bonds which proved fortuitous. However, the pendulum could have swung the other way - having suffered a month of losses on stocks, the markets could have reversed just as the system moved to bonds.

Despite a low margin equity ratio therefore such a system could incur compounding losses each period if its monthly asset re-allocation results in the portfolio being churned in and out of the markets at what turn out to be the wrong points.

This may sound a bit garbled and I do not believe that such a system would always "get it wrong". But when you are sitting on a turning point, when you have been using 10% of funds as margin for stock and commodities and suffered a loss and then switch to using 10% of funds to invest in bonds and defensive plays, you are only too aware that it might go wrong and that your losses could be doubled over the ensuing month!

But of course exactly the same goes for any system getting chopped. And the situation would be a whole lot worse using an ME ratio of 20%. So, yes, quite right Dean.

[kianti]

It is my opinion that margin usage is a leading indicator of potential risk and drawdowns, but there is one school of thought that thinks using more margin is not necessarily dangerous because it can add to diversification. To some degree I agree with this, but traders quickly reach a point of diminishing returns.....

Not just your opinion, I would add that so far the clearing houses margining system could be regarded as the best risk tool around. Apart from some COMEX trouble long time ago, I don't know of any clearing house default. When calculating the classical ratios I account also for margin/equity.

Best regards, as ever

[Zoso]

I hope we are not trying to make an argument for a strong relationship between:

- Current Margin to Equity (and)
- Future efficacy of a manager's program

[Wisdom]

Margins on index options went up substantially after the October 1997 market crash. Prior to 1997 margins for index options were a fraction of what they are today.

After that event, the CME realized the theoretical risk and raised margins substantially. Nearly all index option writers are running at 50% m/e these days in order to generate reasonable returns. They don't have much choice.

ACE was running at 50%+ from the start. I'm not really sure I see the relationship between their margin exposure and the drawdown. It seems like its just a matter of time if that's the game you play.

[Chelonia]

I keep daily track of M/E ratio in a seperate file based on broker statements. TB provides this as well but doesn´t seem to work. In TB M/E is always 0.00%

Margin to Equity Ratio 0.00%

I never gave this much attention because real life happens on the statements, but am i missing something. Should this work in TB?

Thanks

[marriot]

Have you set Margins one by one in the futures dictionary?

[Chelonia]

Thanks Marriot

[LeapFrog]

Interesting topic for sure. Regarding Dean's hypothesis above, might this be a case of correlation not implying causation? The public investor does not usually have access to a CTA's actual leverage (Total Assets controlled divided by Total Funds in house, e.g.) and so tend to rely on the margin to equity ratio as a proxy - hence a desire for low numbers. So all else being equal (CAGR, DD, etc.) then why not pick the CTA with a lower M/E?

This makes perfect sense for a CTA running public funds, but why would an independent trader managing their own funds want to be limited by the M/E ratio? In my view this ratio should be maximized. I for one want to get the best ROA I can, within my own risk tolerance parameters. Having a very large portfolio across multiple geographic locations tends to chew up margin requirements, but also tends to produce more desirable equity curves with pretty VAMI's and r-squareds.

I gave up using TB to "control" my margin usage in backtesting since margin requirements are like shadows that keep moving and there is no reliable historical margin databases that I've been able to get my hands on for 150+ worldwide futures. I track it in other ways and manage it daily. Never been close to a margin call, but I have run as high as 80 percent margin to equity intraday and often average around 40 to 50 percent.

[AFJ Garner]

Interesting topic for sure. In my view this ratio should be maximized.

In many ways, for the private trader it's a question of semantics. Do I devote $100,000 to my trading and adopt a very high ME ratio or $1,000,000 and use less leverage. If the former, then as long as I am willing to add capital in a wipe-out (up to the balance of the $1m) it comes to much the same thing.

It is the same for an investor with a well known CTA. Provided the investor has faith in the CTA and his program(s) then the investor can adopt exactly the same attitude using either a managed account and notional funding or a more highly leveraged fund based on the relevant manager’s underlying program.

[Moto moto]

In house office discussion this morning ....
just because you can do a deal that will make you 5%, and you think therefore you should use extra leverage to make multiples of 5%, does not change the risk profile of the original deal...... the base question still remains of what is the risk of the original deal.

I guess this also applies to this thread.

[DPH]

Leapfrog,

In your case, I agree you want to maximize margin usage, which is why given two programs that make the same return (and all other things being equal), you still choose the one with lower margin. Why tie up $30,000 to make a $30,000 return, if you only need to tie up $15,000 to make the same $30,000 return.

Your return on margin is higher in the second case and you can invest the difference elsewhere. For example, you could trade the second program at twice the level and make double the profit with the same investment size.

What we are talking about here is return on margin, verses return on account size.

[Wisdom]

just curious...

[Chelonia]

It´s all good old leverage

[Chelonia]

@Dean Hofman

Curious to know how you can trade your full diversified portfolio at 8-10% margin with your min. acct of 125-200k?

[DPH]

Wisdom, the guys you asked about are Beechdale Capital and Two Sigma Investment. Some stats below…



Feel free to private message me if you’d like a full blown 5 page (or so) report on both of them, just takes me a second to PDF them…

Something else to keep in mind is that the data I used in the chart suffers from survivorship bias. All the managers who “blew upâ€

[DPH]

@Dean Hofman

Curious to know how you can trade your full diversified portfolio at 8-10% margin with your min. acct of 125-200k?

I do it through a sort of dynamic portfolio selection process. In other words, even though I am tracking over 70 markets, my dynamic portfolio may only be permissioning 10 of those markets today. My trading systems will then generate signals in only those 10 markets. I end up with concentrated bets in just a few markets. I find it is a way to stay widely diversified while maintaining low-margin requirements.

[Chris67]

Its interesting to note how neither of these funds DH has mentioned are trend followers - but they are CTA's - Once again paying homage to the fact that CTA Does nOt equal trend follower and therefore you cannot compare their M/E ratio's with somebody like Abraham as its comparing apples to oranges

[LeapFrog]

Leapfrog,

In your case, I agree you want to maximize margin usage, which is why given two programs that make the same return (and all other things being equal), you still choose the one with lower margin. Why tie up $30,000 to make a $30,000 return, if you only need to tie up $15,000 to make the same $30,000 return.

Your return on margin is higher in the second case and you can invest the difference elsewhere. For example, you could trade the second program at twice the level and make double the profit with the same investment size.

What we are talking about here is return on margin, verses return on account size.

Yes, I think we are in agreement here Dean. I see M/E as one of my system "constraints" to be maximized along with CAGR, while keeping DD and Total Risk within my desired risk boundaries. Margin requirements, being exchange imposed, only act as a limitation of my system parameters, but are not themselves essential or integral parts of the system - they are real, but external to system optimization.

For you they are real in that investors measure you by them and the M/E is therefore something you would want to make primary in your sytem design.

In your example above, I agree that if I'd already reached my desired CAGR/DD ratio, Total Risk level and R-squared on two different systems, and after looking at frequency of trading and order type (stops versus MOO or MOC) I would then pick the one with lower margin usage, assuming I could gestimate the likely historical margin usage given current margin rates input into TB.

Good discussion.

[LeviF]

...I have run as high as 80 percent margin to equity intraday and often average around 40 to 50 percent.

What sort of drawdowns are you realizing with this level of heat?