My System vs. The Rest

[fiftyeight]

So I have continued reading, and have come across some good books mentioned on here.

I have also had a look at the ******************** as recommended by Tech/A.

With the amount of time I have available to dedicate to my trading education it will be some time until I have a system let alone a fully tested and profitable system.

My question is, until my system is more profitable (if ever) after accounting for all fees and charges should I not have my money invested in something like thechartist?

I am really enjoying my time reading and researching and will continue to do so, I just cant see a time where anything I can come up with could compete with something produced by somebody as educated, well funded and experienced as the thechartist.

Note
I have only used the thechartist as an example, it could be anyone with a proven track record.

 

[wayneL]

Re: My System Vs The Rest

On a scale of one to ten, one being the most subtle spam and ten being the most blatant, how would you folks score this OP?

 

[fiftyeight]

Re: My System Vs The Rest

Ahhhhh, I knew I should of removed the thechartist. Silly me.

Pretty sure Nick Radge has commented on many threads under his own name?

 

[CanOz]

Re: My System Vs The Rest

So I have continued reading, and have come across some good books mentioned on here.

I have also had a look at the ******************** as recommended by Tech/A.

With the amount of time I have available to dedicate to my trading education it will be some time until I have a system let alone a fully tested and profitable system.

My question is, until my system is more profitable (if ever) after accounting for all fees and charges should I not have my money invested in something like thechartist?

I am really enjoying my time reading and researching and will continue to do so, I just cant see a time where anything I can come up with could compete with something produced by somebody as educated, well funded and experienced as the thechartist.

Note
I have only used the thechartist as an example, it could be anyone with a proven track record.

I think it depends on a few things really:

-How much capital do you have, could you benefit with more capital?
-How is your coding and development skills?
-How much time do you have?

I think buying open code (or using managed trading systems) is a great idea if you are time or skill poor or both, but with ample capital to deploy immediately.

That was certainly my case.

CanOz

 

[fiftyeight]

Re: My System Vs The Rest

Yup, I currently have the trifecta. Capital, time and skill poor.

Just in education mode at the moment so once the capital becomes available I wont loose as much when I start.

It was more of a hypothetical problem. In general, do managed trading systems become less profitable as their size increases?

I guess on the flipside then, does that mean that I have an advantage in being small?

Just struggling to see how I will ever beat somebody/anybody who has vastly more experience, time and sometimes a team of people.

 

[CanOz]

The advantage in the "Boutique" version of managed systems in my view is that they are small, relative to the liquidity on offer. I guess the key to the whole system is really that its not too widely used. Most managed systems that i know of do not release the code and only so many clients or contracts are traded. This preserves the edge that system is taking advantage of to a degree.

Systems like Larry Williams "oops" no longer work because it is widely known, so i've been told but not actually tested.

You can with experience, even some luck stumble upon a really good trading system.

I was playing around with some reasonably common code the other day, then had the well know person mentioned above, polish it up a little...

This is what i got without adding anything complicated for an exit yet....see PDF

Then again sometimes you completely waste your time and effort to end up with something hopeless or having a post-dictive error...(another reason its good to have a pro check it).

Even all of this though is contributing to a greater understanding of how systems work, which will allow you to make better decisions on even what code to buy, or what managed systems to use...

I really enjoy spending some time with Amibroker on EOD data and coding. I'm going to use Monday's for Amibroker time now, get me away from messy market monday!

Now, just got to pick up a couple of Howard's books!





CanOz

 

[sinner]

Then again sometimes you completely waste your time and effort to end up with something hopeless or having a post-dictive error...(another reason its good to have a pro check it).

CanOz

My experience is that you can avoid this entirely by focusing only on exploiting actual market return factors and anomalies. If you try to discover a new form of edge, very high probability it doesn't exist or will eventually become arb'd out of existence. If you try to exploit the edge in being long volatility or short volatility, there shouldn't be any issues about timewasting, only degrees of efficiency in harvesting that alpha (assuming the market structure is appropriate for that edge - e.g. mean reversion in equity indices since the 90s or long term trend following in EURUSD since 1999).

Just looking at the equity curve you posted there CanOz, I'd guess it's long equity upside volatility and can perform as long as equity upside volatility exists.

 

[Gringotts Bank]

General question to anyone:

How much do the World's best system traders make?

The best I've heard is a consistent 100-150%pa with a futures system (ie. leveraged). But i suspect there are better out there....

 

[systematic]

General question to anyone:

How much do the World's best system traders make?

The best I've heard is a consistent 100-150%pa with a futures system (ie. leveraged). But i suspect there are better out there....

No one consistently, on any scale makes that return. That's the thing - at decent scale, those sorts of returns become an impossibility.

A fund with only $100m under management returning 150%pa would, after 13 years own the NYSE. The following year it would own the NASDAQ, Tokyo, London, Hong Kong & Shangai exchanges, TMX (Canada) and the German Bourse - with probably enough pocket change to pick up ASX, and the Bombay stock exchange to boot.

 

[fiftyeight]

100-150% sure beats the banks.

Sinner, I think you have to the core of what Im thinking. Any edge is already exploited or wont last long, so its about making the most of using already known strategies and knowing when they will actually work.

Working my way through "Encyclopedia Of Trading Strategies" at the moment Can, I think that was a book you recommended to someone else. Then ill start playing around with Amibroker.

General consensus seems to be, keep it simple.

 

[craft]

General question to anyone:

How much do the World's best system traders make?

The best I've heard is a consistent 100-150%pa with a futures system (ie. leveraged). But i suspect there are better out there....

Unless your system traders name is 'Carlos' or 'Bill' you can be pretty sure something does not compute on anything over 20% long term. [most likely not scalable or susceptible to outliers].

Over the last 48 years (that is, since present management took over), book value has grown from $19 to $114,214, a rate of 19.7% compounded annually

http://www.berkshirehathaway.com/letters/2012ltr.pdf

 

[sinner]

Unless your system traders name is 'Carlos' or 'Bill' you can be pretty sure something does not compute on anything over 20% long term. [most likely not scalable or susceptible to outliers].

I sort of agree but also sort of disagree. If we are talking commodity CTAs or value investors, then yep absolutely. These guys are doing one thing, exploiting one kind of edge over and over again. Most of these sorts actually tend to underperform their benchmarks because they mistake the "factor" driving their returns for the baseline when generally it is somewhere near the ceiling. Berkshire is obviously an exception, where the "low beta" and "quality" factors explain the return profiles almost entirely (i.e. they have strictly stuck to the factors, no tweaking and have been paid handsomely for it).

However there is no rule saying you can only exploit one factor over and over again. There are obviously cases of high probability, high frequency setups (by high frequency I mean 1-5d holding periods) which occur maybe only 2-3 times per year. Utilising a large basket of these systems, constantly adding new ones and turning old ones over as their efficiency is arb'd away is one way I have seen some systems traders (mostly ex engineers and compsci types) achieve very very high and surprisingly consistent returns. It can be done, but not by factor based trading. The tradeoff is there is 0 guarantee of future performance and you have to constantly be searching for new anomalies and inefficiencies, where as factors such as value which follow some intuitive principle "buying assets for cheaper than their fair value" will always have some guarantee of future performance and you don't have to go looking for a new one every 4 weeks.

Personally I like to track and follow the market factors I deem to be most robust, concentrating my efforts on harvesting those as efficiently as possible. However, here is a very good example of the "other way"

http://quantifiableedges.blogspot.com

 

[Gringotts Bank]

...is one way I have seen some systems traders (mostly ex engineers and compsci types) achieve very very high and surprisingly consistent returns.

...as in 100-150%? It's not that important I guess but I'm interested in what 'World class' would mean quantitatively.

For general interest, the guy who calls himself the Bell Ringer (Fin Review) made 800% during 2008. But that's only one year, and i don't know about scalability factors. It was also discretionary, I think, so not directly relevant to this thread.

 

[Country Lad]

General question to anyone:

How much do the World's best system traders make?

The best I've heard is a consistent 100-150%pa with a futures system (ie. leveraged). But i suspect there are better out there....

Quite possible indeed. Keep in mind that 100% in a year is about an average of 1.34% per week and 150% in a year is an average of 1.78% per week.

Cheers
Country Lad

 

[chops_a_must]

If you try to discover a new form of edge, very high probability it doesn't exist or will eventually become arb'd out of existence. If you try to exploit the edge in being long volatility or short volatility, there shouldn't be any issues about timewasting, only degrees of efficiency in harvesting that alpha (assuming the market structure is appropriate for that edge - e.g. mean reversion in equity indices since the 90s or long term trend following in EURUSD since 1999).

Just looking at the equity curve you posted there CanOz, I'd guess it's long equity upside volatility and can perform as long as equity upside volatility exists.

Indeed.

It seems that most effective systems are actually trading off the volatility, one way or another.

How you use elements that interact with that, appears to be the key.

And because these systems are forward looking in essence, they include an inbuilt edge that is a fundamental part of the market, and may not be arbed out of existence. You only have to compare performance on breakout strategies from mid and small caps, with big caps to see that.

 

[SuperGlue]

Quite possible indeed. Keep in mind that 100% in a year is about an average of 1.34% per week and 150% in a year is an average of 1.78% per week.

Cheers
Country Lad

Quite possible indeed if you think like the corner store owner, a few cents here and there. Your corner store owner is actually a sclaper

From memory, reading through Van Tharp or Howard Bandy books, most system trade profit are only averaging $100-$200 per trade. Correct me if I'm wrong.

Please note Theory only.
Entry must be right most of the time (eg. higher low bar with uptrend long term chart).If not Van Tharp position sizing has to come into play.
Brokerage must be low.

Capital = $100,000

5 trades per day X $100 (no more no less) = $500 (0.5% per day)
Per week = $500 X 5 days = $2,500 (2.5% per week)
$2,500 x 40 weeks = $100,000 (100% per year)

Percentage return is a little higher than mention by Country Lad

That's the theory anywhere, then the greed & stop loss steps in and the whole system falls apart.

I was discussing about this theory last week with my sister.

Quite possible indeed, ideally a robot sytem, that does it automatically, so greed doesn't come into play.
Just dreaming........, thats what makes the world go around.


Now, where is my "Introduction to Amibroker" book by Howard Bandy.

 

[craft]

I sort of agree but also sort of disagree.

My simple point – put another way.

Buffet is the third richest person in the world – after Carlos Slim and Bill Gates. He has made his money through investing and his results are recorded and verifiable - It is 19.7% over 48 years.

So in relation to this post.

General question to anyone:

How much do the World's best system traders make?

The best I've heard is a consistent 100-150%pa with a futures system (ie. leveraged). But i suspect there are better out there....

You will note the underlining of consistent!

I’m simply saying, something doesn’t compute either those high return rates are not scalable, transitory, or susceptible to outliers or the world’s rich list is incorrect.

Just trying to put some realism around consistent return expectations.

Sure you can do better with non-consistent systems – four spins on black is easily with-in random outcomes and gives you a 1600% return in a few minutes – filth spin?

Potentially high return systems are viable if they are protected by transitory opportunities or a non-scalable niche - is it worth perusing them though or is it better to get on with a realistic, robust, scalable approach?

 

[Gringotts Bank]

My simple point – put another way.

Buffet is the third richest person in the world – after Carlos Slim and Bill Gates. He has made his money through investing and his results are recorded and verifiable - It is 19.7% over 48 years.

So in relation to this post.




You will note the underlining of consistent!

I’m simply saying, something doesn’t compute either those high return rates are not scalable, transitory, or susceptible to outliers or the world’s rich list is incorrect.

Just trying to put some realism around consistent return expectations.

Sure you can do better with non-consistent systems – four spins on black is easily with-in random outcomes and gives you a 1600% return in a few minutes – filth spin?

Potentially high return systems are viable if they are protected by transitory opportunities or a non-scalable niche - is it worth perusing them though or is it better to get on with a realistic, robust, scalable approach?

The Stock Market Wizard series highlights fund managers and individuals who make these extremely high returns over prolonged periods, like say 10 years. Some of these guys would be big spenders on depreciating assets and entertainment, so their wealth might not accumulate like it has with Buffet. Others might fail to adapt to, or fail to recognize, new markets conditions. Or they might lose the plot in some other way (personal issues have a way of affecting concentration, for example). So after 10 years of spectacular performance, anything could happen. 10 years of high returns is consistent enough for me.

Also, Buffet isn't a systems trader.

The guys mentioned who make the high returns are often managers of a fund, so while they probably receive big pay checks, they don't necessarily have all their own money compounding in the fund they work for. They might have it in a balanced industry fund! heee! :