Musings on (worthless) Past Performance

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.
Chris67
Roundtable Knight
Roundtable Knight
Posts: 1046
Joined: Tue Dec 16, 2003 2:12 pm
Location: London

Post by Chris67 » Tue Jan 31, 2012 7:09 am

rajivm

I dont understand your point - show me a strategy with a long track record of making better returns
I dont mean some little fancy short term strategy that has done well for 2 years as we all know where they end up ?
Agreed the Rentechs of this World may have found the holy grail - they may have not - but if you look back over 5 years the CAGR's of many TF's hasnt been too bad in comparison ? Some have been excellent and some have been dissapointing I agree - but on the whole it depends on if you think things have fundamentally changed ?
C

BuyHigh SellLow
Roundtable Fellow
Roundtable Fellow
Posts: 50
Joined: Wed Apr 27, 2011 12:46 pm
Location: U.S.

Post by BuyHigh SellLow » Tue Jan 31, 2012 9:54 am

Chris67 wrote:rajivm

I dont understand your point - show me a strategy with a long track record of making better returns
I dont mean some little fancy short term strategy that has done well for 2 years as we all know where they end up ?
Agreed the Rentechs of this World may have found the holy grail - they may have not - but if you look back over 5 years the CAGR's of many TF's hasnt been too bad in comparison ? Some have been excellent and some have been dissapointing I agree - but on the whole it depends on if you think things have fundamentally changed ?
C
I think there are plenty of strategies that have weaker historical performance (but still positive) than TF.

But, these strategies have little-to-no correlation with TF.

Ok, here's the weakest of them:

Maintain some long S&P500 exposure. The expectancy is positive, AND...there's a weird form of beta forgiveness out there that most investors have. If the stock market is tanking, they don't mind if their account is slightly tanking.

The other 4 futures strategies don't enjoy this same investor love. So, they have to be a lot better.

drm7
Roundtable Fellow
Roundtable Fellow
Posts: 96
Joined: Sun Apr 20, 2003 9:02 pm
Location: Richmond, VA

Post by drm7 » Tue Jan 31, 2012 11:16 am

Here's my counterpoint to the OP, although it is by no means an exhaustive analysis. The attached spreadsheet shows a portolio that I created using the IASG database - called "old school TF." These are CTAs that most people would have recognized as top performers in 2001. (I don't have performance records prior to that date, but they were all considered "brand name" CTAs - most of them former turtles.

If you had invested a roughly equal allocation into each of these managers in late 2001, you would have had a pretty good experience - 13% 10-year CAGR and a 0.75 sharpe.

Now, this doesn't directly address the OP's original premise that high sharpe ratios dont' persist - I actually agree with him. However, this example shows that you can get "good" performance from a group of "good" managers, if you spread your bets.

Unfortunately, this is beyond the capabilities of most non-institutional investors, given the minimum investment sizes required. I think that it shows how well trend-following has held up, though.

(Note that there is another tab on the spreadsheet (All-stars), which shows a more cherry-picked group of managers with even better performance.)
Attachments
Trendfollowing Performance 2001 to 2011.xls
(19 KiB) Downloaded 185 times

Eventhorizon
Roundtable Knight
Roundtable Knight
Posts: 229
Joined: Thu Jul 08, 2010 2:36 pm
Location: Boulder, CO
Contact:

Post by Eventhorizon » Tue Jan 31, 2012 12:02 pm

Dean,

There is something funky with those numbers.

For example, if the data is for 5 years of returns, there are several entries where the compounded return is positive but the Sharpe is negative. Since arithmetic return is always greater than geometric (i.e. if compounded return is positive then arithmetic return must be too) and variance is always positive, how can Sharpe be negative.

Unless the Sharpe values in the table are for the most recent 12 months, not the entire period in the title of the table.

Also, there might be aissues regarding confidence intervals on the correlations i.e. how significant are the relationships you are seeing?

Interesting stuff!

rajivm
Roundtable Knight
Roundtable Knight
Posts: 129
Joined: Fri Jul 25, 2003 12:50 am
Location: New Delhi, India

Post by rajivm » Tue Jan 31, 2012 1:19 pm

Chris67 wrote:rajivm

I dont understand your point - show me a strategy with a long track record of making better returns
I dont mean some little fancy short term strategy that has done well for 2 years as we all know where they end up ?
Agreed the Rentechs of this World may have found the holy grail - they may have not - but if you look back over 5 years the CAGR's of many TF's hasnt been too bad in comparison ? Some have been excellent and some have been dissapointing I agree - but on the whole it depends on if you think things have fundamentally changed ?
C
Well Chris,
How to show? But there definately exist short term systems :D
Can you show me one LTTF strategy where 50% return can be achieved without almost 50% drawdown..... Such a system from my point of view is not a very good system..

R

marriot
Roundtable Knight
Roundtable Knight
Posts: 350
Joined: Thu Nov 20, 2008 3:02 am

Post by marriot » Tue Jan 31, 2012 1:36 pm

And belive me Rajvim,
i am the ship's boy here.
Attachments
ravjim.JPG
ravjim.JPG (109.51 KiB) Viewed 5523 times

AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2040
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner » Tue Jan 31, 2012 1:39 pm

For many years I traded new issues of bonds, equities and convertibles, encompassing also secondary stock offerings. A US IPO opened up 30% last week. You very rarely have to run overnight positions, other than with some of the Far Eastern IPOs.

You can make a complete hash of it if you get the colour of the deal wrong or are too much of a pig. You have to pick your deal and be careful about liquidity and the quality of the underwriting syndicate. On balance though it is less nerve wracking than the agony of trend following and I'm not sure I ever remember too much of a draw down for any length of time.

Not a system perhaps but certainly a way of trading. There are periods of drought of course.

As for CAGR, well that's a difficult one. Given the right prime brokerage arrangement we used to make rather healthy returns on rather slim capital. Pick your figure.

There is more than one way to trade, that's for sure. And you can forget all about back testing and worrying whether your system has broken down.

Moto moto
Roundtable Knight
Roundtable Knight
Posts: 427
Joined: Mon Jun 01, 2009 4:12 am
Location: once again in the UK

Post by Moto moto » Tue Jan 31, 2012 3:04 pm

AFJ G - I agree with you...less nerve racking and more patience needed, however, you key point of avoiding the dogs is key.....plus you usually end up getting a small percentage of the good ones, and a large percentage of the dogs in allocations unless you pay enough brokerage.....its a very different game.
Been there done that, and IMHO - you may as well just wait until the market is running and get on board with extra leverage, unless of course you are talking about pre pre IPO where 10 X rather than 30% might be more in line with possible returns.
(by the way - welcome back)

DPH
Roundtable Knight
Roundtable Knight
Posts: 113
Joined: Tue Nov 17, 2009 1:07 am
Location: Central Pennsylvania
Contact:

Post by DPH » Wed Feb 01, 2012 2:15 pm

Eventhorizon wrote:Dean,

There is something funky with those numbers.

For example, if the data is for 5 years of returns, there are several entries where the compounded return is positive but the Sharpe is negative. Since arithmetic return is always greater than geometric (i.e. if compounded return is positive then arithmetic return must be too) and variance is always positive, how can Sharpe be negative.

Unless the Sharpe values in the table are for the most recent 12 months, not the entire period in the title of the table.

Also, there might be aissues regarding confidence intervals on the correlations i.e. how significant are the relationships you are seeing?

Interesting stuff!
The Sharpe Ratio can still be negative if the funds CAR is less than the Treasury Bill CAR....

Sharpe Ratio:
(Compound Annual ROR - risk free ROR (calculated from T-bills)) / Annualized Std. Dev. of Mo. ROR

As far as confidence intervals, I have looked at many different rolling windows of data with as large of a sample size as BarclayHedge offers (which to my understanding is the biggest in the industry for CTAs), and I have been able to thoroughly convince myself that the Sharpe Ratios predictive ability (by itself) is null (based on the time frames I look at anyway).......If anybody is interested in this additional work just private message me (or keep an eye on my blog, where I intend to keep publishing this kind of thing)

rajivm
Roundtable Knight
Roundtable Knight
Posts: 129
Joined: Fri Jul 25, 2003 12:50 am
Location: New Delhi, India

Post by rajivm » Wed Feb 01, 2012 10:17 pm

marriot wrote:And belive me Rajvim,
i am the ship's boy here.
Hello Marriot,
One question - 30300 trades in 10 years -- 3030 trades in one year....How do you manage such a system?
I try to keep number of trades per year in short term system below 500 per year..I feel it is too much work even at that level..::))
Raj

marriot
Roundtable Knight
Roundtable Knight
Posts: 350
Joined: Thu Nov 20, 2008 3:02 am

Post by marriot » Wed Feb 01, 2012 10:53 pm

Thanks to Corziine, i do not have problems with orders.
Simply i do not.

rajivm
Roundtable Knight
Roundtable Knight
Posts: 129
Joined: Fri Jul 25, 2003 12:50 am
Location: New Delhi, India

Post by rajivm » Fri Feb 03, 2012 1:04 pm

marriot wrote:Thanks to Corziine, i do not have problems with orders.
Simply i do not.
Marriot,
what you asked in private message I put here..because I cannot attach file in private message.. you asked about orders -- on open only...
Attachments
performace capture Tblox.JPG
performace capture Tblox.JPG (67.23 KiB) Viewed 5288 times

marriot
Roundtable Knight
Roundtable Knight
Posts: 350
Joined: Thu Nov 20, 2008 3:02 am

Post by marriot » Sun Feb 05, 2012 1:48 pm

Hi Ray,
your test is amazing and more if looking at starting equity.
Would you mind to say something more about Portfolio used?
Thank you.
Last edited by marriot on Sun Feb 05, 2012 3:34 pm, edited 1 time in total.

DPH
Roundtable Knight
Roundtable Knight
Posts: 113
Joined: Tue Nov 17, 2009 1:07 am
Location: Central Pennsylvania
Contact:

Post by DPH » Sun Feb 05, 2012 3:29 pm

Ahh yes, the beautiful art that can painted with the perspective of past performance…..Makes one wonder what the likes of a Picasso, Van Gogh or Dali could have created with such tools~~~ For those feeling enchanted by these museum worthy pieces START HERE ……..If it’s too late and you’ve already fallen to their irresistibly hypnotic spell GO HERE

Consider this blurb I wrote 9 years ago as a system vendor (prior to managing money as a CTA) : Trading Systems That Work

rhc
Roundtable Knight
Roundtable Knight
Posts: 464
Joined: Tue Nov 18, 2008 8:46 pm
Location: Oz

Post by rhc » Sun Feb 05, 2012 10:40 pm

The afore-mentioned trading system returns 295% CAGR
The unfunded US debt figure is approx. 60 Trillion dollars (i.e. $60,000,000,000,000)
A bionic man costs 6 million dollars (i.e. $6,000,000)

If the US government is willing to forgo the manufacture of only one such bionic man and use those funds as starting capital for this system then using the compound interest calculator as found here;
http://www.1728.org/compint.htm

We find that if will take 11.7 years to retire this debt. - - - >That's not half bad!

Now,
If we can just get Oscar Goldman to agree . . . . . ,
Attachments
CICalc.jpg
CICalc.jpg (57.75 KiB) Viewed 5068 times

Chris67
Roundtable Knight
Roundtable Knight
Posts: 1046
Joined: Tue Dec 16, 2003 2:12 pm
Location: London

Post by Chris67 » Mon Feb 06, 2012 3:08 am

was I the only one who used to really fancy Jamie Summers - the bionic woman !!! she was cute

rabidric
Roundtable Knight
Roundtable Knight
Posts: 243
Joined: Mon Jan 09, 2006 7:45 am

Post by rabidric » Mon Feb 06, 2012 3:19 am

DPH wrote:Ahh yes, the beautiful art that can painted with the perspective of past performance…..Makes one wonder what the likes of a Picasso, Van Gogh or Dali could have created with such tools~~~ For those feeling enchanted by these museum worthy pieces START HERE ……..If it’s too late and you’ve already fallen to their irresistibly hypnotic spell GO HERE

Consider this blurb I wrote 9 years ago as a system vendor (prior to managing money as a CTA) : Trading Systems That Work
whilst ravijm's graph may display a rate of return that would be unsustainable beyond a certain moderate capacity, I do believe he actually trades it for real( maybe 2008/9 is the "since live" bit). Futhermore, while the graph may show a run using a specific static set of rules on past data, that may not remain stable into the future; it does not mean that it is not possible to achieve that kind of graph in live trading if you evolve your system(s) as required.

It would seem to me that an easy trap to fall into, that catches many here, is that because it is difficult to achieve something very good within their paradigm, it must therefore be impossible under all paradigms. Far easier to trot out the usual received wisdom and sneer. :roll:

AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2040
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner » Mon Feb 06, 2012 5:20 am

The title of this thread can be justifiably applied to the art of back testing. However corny some of the old sayings may sound, they are nonetheless true. Your biggest draw down is always in the future. Markets are ruled by fear and greed and you need to battle against both of these emotions to survive. Back testing inevitably involves a degree of curve fitting, conscious or unconscious.

Most CTAs seem to have made a mess of things at some time or times in their career. Many CTAs have records including discarded systems and blow ups. Risk control and reduced leverage may be the only way to survive long term. Seductive back tests are almost inevitably mirages and are best treated with a wry smile and a knowing shaking of the head.

marriot
Roundtable Knight
Roundtable Knight
Posts: 350
Joined: Thu Nov 20, 2008 3:02 am

Post by marriot » Mon Feb 06, 2012 6:50 am

No complain Anthony,
you know i love you : )

> Your biggest draw down is always in the future.

Sluggo does not agree with you.
Toc, toc " hi Sluggo, have you change your mind in the last year?"

>reduced leverage may be the only way to survive long term.
Yes, but is not easy balance risk with differents starting equities.
Starting from two or three millions, seem easy.
But trying to start from 100 k and appling all the safest rule of money management i am coming to ask myself: "why should i do all this ?"


"the ship boy".

AFJ Garner
Roundtable Knight
Roundtable Knight
Posts: 2040
Joined: Fri Apr 25, 2003 3:33 pm
Location: London
Contact:

Post by AFJ Garner » Mon Feb 06, 2012 8:55 am

marriot wrote:No complain Anthony,
you know i love you : )

> Your biggest draw down is always in the future.

Sluggo does not agree with you
Maybe not but that does not influence me or change my mind. Take a look at the new lows reached by Tactical and Chesapeake. As for a low starting capital, take a look at Sluggo's oft posited and interesting idea to bet large to begin with (because you have to with limited capital) and reducing leverage as equity increases.

Trading is a dangerous game and none of us are immune to its perils. My accountant told me recently of a copper trader who built up his capital to £160m and lost 40% last year.

Post Reply