10M USD Reference Portfolio

[poyet]

As I had proposed sometime ago, I'll post on a specific blog http://cosmosportfolios.blogspot.com/ the evolution of a 10M USD reference portfolio which started on the 20th Dec. 2004.

I'll publish every Saturday a snapshot, the current portfolio, closed trades, and all actions taken, in an incremental manner. The system is a demonstration of what trading automation can deliver and I'd like to acknowledge the many ideas that we got from traditional trend followers as the turtle traders http://www.turtletrader.com/ and Jack Schwager in particular in his many books, some with excellent sections dealing with the design of trading systems.

One must understand that the system takes into account many specific parameter to each account / fund , including NAV (with size line in % of NAV having an impact on liquidity criterions), drawdown (DD) and hardcashDD, etc. all impacting the selection of the stocks and position of stops among other things. So one cannot just "adapt" the positions to its own capital to copy / paste positions (would one wish so...) but nevertheless this will give an idea of what an entirely automated trend following system does.

Looking forward to your comments,
Cheers.

PS: as some are sensitive to the issue, I'm not selling anything here except the wish to share some ideas with you around the general issues of investing / trading and software automation of these activities.

 

[justin]

Interested to see how things turn out.

I guess you know that posting how well you've done over the last 12 months isn't as impressive as posting your trades as you do them, and then looking at the end result 12 months from now. It sounds like you'll post weekly trade reports on Saturdays?

 

[poyet]

Yes Justin, I'll post weekly on Saturday and there will be no difference in the future as to what was done in the past as this account was set up for a client who got on a weekly basis the progress made by the system. As you can image I did not cook a posteriori all these data...

Of course, posting on Saturday does not disturb positions implemented, if any, as they're already opened during the week. At some stage I might have to remove information about where the stops are...

Anyway, 14% is not that impressive though it is not that bad either given the whipsawing market we've had. Hope we'll do much better next year. You'll see things unfolding before your eyes.

Hopefully we've made better on the French market this year.

Just to remind for some of you who might not know that the method and system implemented are (partially) documented at http://tradingautomation.blogspot.com/

Cheers,

 

[justin]

I'm not saying you're cooking the books. I assume you have more than one "portfolio" you are tracking. If I were in your shoes, what is to stop me from picking one of my model portfolios that had done well, and presenting it. What I'm saying is I'm more interested in the results going forward, since we've identified the portfolio under review, instead of seeing how our portfolios do for a year, then identify which portfolio we want to present after we know their 1 year performance.

 

[poyet]

You're right justin. I've got several running of different sizes both in the US and in France. For a similar market say the US or France they all perform on average in the same way though we have small differences due to the account size, starting date (and other parameters depending on the risk level the account owner is happy with, etc.). Normally the smaller the account the better the performances as we can trade "smaller" stocks with stronger moves, but this year it has not been the case and my own account his behind the one I presented ! but we talk about small differences...

whereas

on the other hand French portfolios have delivered +65% since Sept 04 but I did not post them as we can only run small accounts (the FR market is not big enough to find the liquidity for large accounts with our method and system) and furthermore these unknown stocks tou you guys would be of limited interest to US traders and investors.

So the account presented was kind of "representative" for US observers and is also "downward scalable" (trading the same stocks with smaller lots) whereas a smaller one would not be "upward scalable" as you would not necessary find the liquidity.

 

[Nords]

I'm not saying you're cooking the books.  I assume you have more than one "portfolio" you are tracking.  If I were in your shoes, what is to stop me from picking one of my model portfolios that had done well, and presenting it.  What I'm saying is I'm more interested in the results going forward, since we've identified the portfolio under review, instead of seeing how our portfolios do for a year, then identify which portfolio we want to present after we know their 1 year performance. 

It's like asking a stockbroker to post all their trades as well as online images of their tax return schedules D. Wouldn't it be great if that was an SEC requirement?

 

[poyet]

Nords,

I quite dislike quality through requirements and regulations, though if it were kind of a common practice for more money managers to disclose what they actually really do and how they really perform (to their clients !) that would increase transparency and ensure credibility, to the very least.

I've read lately unbelievable stories relating to funds & hedge funds collapses, not as big as LTCM but in a sense more amazing as the managers were simply going to Vegas with their clients' monies !

Well, I'll add a bit of a market snapshot as well on each Saturday based on automated market analysis tools. embedded in the system

Hope it will interest some of you.

 

[poyet]

Portfolio and actions as of 17th have been updated.
Best,

 

[poyet]

Portfolio, world markets, actions, curves as of 23rd Dec. have been updated.
Comments welcome.

 

[poyet]

Portfolio and actions updated as of 30th of Dec. World markets have been quiet and little change occured.

 

[Spanky]

I thought the world market was mostly down by 0.5% on December 30.

 

[poyet]

What I meant Spanky is that apart from the SSEC for which we passed from SHORT to Neutral all other markets remain LONG. What I summarize in the "world market snapshot" table are the general characteristics of a bunch of indexes offering a picture on the world markets. For the system, a change from Bull to Neutral or further to Short requires a change in configuration, not just a change in % be it significant (say more than 10-15%).

You're right saying that we've had a small retracement and the reference portfolio is behind where we were in mid Dec., but it does not change anything to the general stance the system has with respect to the market(s) so far.

The change to the SSEC is otherwise interesting as it might point out to a future Bull in the Chinese markets (mainland) which have remained noticeable laggards in asia. HSI is a bit of a different story.

 

[poyet]

Updated as of Jan. 7th 2006.
Portfolio makes a new highwatermark @ 11 685 986 USD.
Market expo is now leveraged as the market is accelerating.
Cheers.

 

[poyet]

Updated as of Jan. 13th 2006.
Portfolio makes a new highwatermark @ 12 015 378 USD.
Market expo is leveraged @ 17 546 876 as the system is on margin.
AMD was downgraded by Deutsche Bank analyst.
Comments welcome.
http://tradingautomation.blogspot.com/
http://cosmosportfolios.blogspot.com/

 

[poyet]

Portfolio has been updated. as of 20th of January COB.
Take note of a split on TALX - 3 for 2 and adjustements.
Not a good week as one would have guessed...

 

[BigMoneyJim]

Why $10Mil? I can think of few who couldn't happily put it all in fixed, relatively secure investments and live lavishly off the dividends alone.

Also, I've skimmed through quickly but couldn't see if there were any new contributions assumed or withdrawals assumed. Is this an accumulation exercise, a withdrawal phase exercise or just a lone untouched portfolio exercise?

 

[poyet]

Thanks for looking BigMoneyJim.

1) There are various reasons why I chose 10M. The first is that downscaling positions is always possible whereas upscaling is not necessarily (or seldom), due to liquidity constraints. If you have a 100k portfolio and wish to get some of these positions (or implement the portfolio) you'll have no problem getting 1/100th of the shares. Whereas if the systems manages a 100k reference portfolio and goes for rather illiquid securities or simply less liquid like say Bookham Technology plc A (BKHM), or GigaMedia Limited (GIGM) which I own in smaller portfolios at the moment and one wishes to place a large order, he'll make a big mistake as slippage (also on stops if executed) will be dramatic.

2) the system can accomodate adding or withdrawing money, but the objective is maximum capital gains with a given X investment, using trading automation as the backbone of the method. 10M is a large trading portfolio for individuals but a small one for organisations. The system operates well till 75/100 M, then performances decrease due to restrictions on the trading universe. This is annoying as organisations do not start a new strategy for less than 100.

3) people with more than 100M do not wish to let 10 sleep. It is just 1/10th of their net worth and they do not plan to live of fixed income from that. They are also a commercial target for the deployment of the system.

It's not just an exercise, though I like the term. Portfolios are run daily with this system.

Best luck with your trading.

 

[justin]

It's not just an exercise, though I like the term. Portfolios are run daily with this system.

Just curious, what is the ballpark dollar amount that outside investors are investing using your system?

 

[poyet]

Justin,

To run an account they must be accredited investors in US terms (1M min). Hope it answers your question. Was a bit curious about what "ballpark" meant. BTW there are large discrepancies as you deal with people with very different situations. Recommandation is always that such trading accounts should NOT be more than 1/10th of net worth or 1/10th of the sum of other strategies / investments. Hope it helps.

Nevertheless, I also run small accounts for a couple of friends (say 100k) and normally over statistically significant time horizons (say 5 years) the smaller the account the stronger the performances as the system goes for more illiquid stocks having stronger moves (more upside / downside potential).

 

[justin]

"Ballpark" - apparently it is an americanism. Just means a very rough estimate of something. As in a "ballpark guess" about something. Ie - "sir, can you give me a ballpark estimate as to how much those car repairs will cost".

I was actually wondering what the total value of all assets being managed using your trading system (total value of ALL accounts).

 

[poyet]

Several millions at stake. But can't say more as it would hint to clients / partners / prospects private information.

BTW, running the reference portfolio is also an exercise to help demonstrate people that trading is not throwing darts, there is knowledge and it can be modelled, and even programmed into software and run and therefore lead to make consistent money.

Any bunny (sampled out of a large chunk of bunnies under trial) can make money (different ways including darts) but making consistent money over long periods of time with mastered risks and different markets configurations is a different story. It was at first an intellectual challenge to prove ourselves that trading knowledge was consistent enough to program systems that would prove on all grounds (monte-carlo, etc.) that consistent money can be made that way. I love the answer of Warren Buffet to Burton Malkiel (though he's a pure fundamentalist as everybody knows, but demonstrates that he has knowledge picking stocks), kind of "then if you were righ, why is it that I'm so rich " ))

 

[poyet]

Portfolio has been updated as of 27th Jan.

We made 1,18M this week benefiting from leverage and strong positions overperforming the overall market. We make a new high water mark.

AMD has nicely resumed its run since downgrade by Deutsche Bank analyst which indeed made a good day as an entry point.

Do nor hesitate to comment, I'm here for that.
All the best for your trading.

http://cosmosportfolios.blogspot.com/

 

[Eagle43]

"Ballpark" - apparently it is an americanism.  Just means a very rough estimate of something.  As in a "ballpark guess" about something.  Ie - "sir, can you give me a ballpark estimate as to how much those car repairs will cost". 

I was actually wondering what the total value of all assets being managed using your trading system (total value of ALL accounts). 

You could also use WAG: "Wild-A$$ Guess".

 

[poyet]

Eagle, funny as WAG sounds like vague(ness) and we also use the same word in French, i.e. vague. But the truth is that there's nothing vague in the amount managed, it's known to the cents

 

[poyet]

Portfolio and blog will be updated on Monday as I'll be skiing until then.