RIP Mechanical Trading Systems?

[x4x]

Yes.

BOO!

Well, congratulations on discovering the Holy Grail :clap:

 

[Hotch]

A trader has to make lots of discretionary decisions on a day to day basis as news is released which can have a dramatic effect on the market.

You don't think programs can read the news?

 

[x4x]

You don't think programs can read the news?

If you say that a program can respond to things like 9/11 & 7/7 then I can't dispute it, I know very little about programming. I'm obviously the stupid one sitting at my screens 5 or 6 hours per day whilst you are partying, laying in bed, playing golf everyday or whatever else it is you do with all your spare time - If you've achieved that then I genuinely congratulate you....

...but to be honest I enjoy trading and will continue to do it (manually) for the rest of my life, or at least as long as there is a market to trade.

 

[Shanghai]

Do you know your annualised daily vol? Without that, or max drawdown, the percentage return is meaningless

I've mentioned my drawdown before. A few years (3-4) ago it was just under 19%. As I have gone along my equity curve has been getting a bit smoother and in the couple of years I have had a drawdown of 15.6% although my annual returns have been to the mid/lower end of my normal range. I don't have detailed stats as I don't backtest much due to the decretionary element of my trading. I do backtest by hand but that tends to be to compare different approaches and not look at the system as a whole.

 

[meanreversion]

I've mentioned my drawdown before. A few years (3-4) ago it was just under 19%. As I have gone along my equity curve has been getting a bit smoother and in the couple of years I have had a drawdown of 15.6% although my annual returns have been to the mid/lower end of my normal range. I don't have detailed stats as I don't backtest much due to the decretionary element of my trading. I do backtest by hand but that tends to be to compare different approaches and not look at the system as a whole.

Right, you're the only person who has put a figure on drawdown.

However, no-one else seems willing to put a figure on the kind of annual return they would like to see, or the drawdown they anticipate/could tolerate.

So, I wonder, maybe a good portion of traders simply have no idea what they are targeting, or what they expect on the downside? A bit like flying blind, isn't it?

 

[meanreversion]

I am a full time 'system trader' - so I have no problems with systems. However, I have to adjust my strategy according to market conditions. No point trying to trend trade when the market is flat or choppy, or range trade if it's trending.

What annual return would you like to see (% of equity)? What sort of drawdown would you expect? If you won't answer these questions, then saying EAs "don't work" means precisely nothing.

 

[Pazienza]

What annual return would you like to see (% of equity)? What sort of drawdown would you expect? If you won't answer these questions, then saying EAs "don't work" means precisely nothing.

I confess I'm not a fan. But if it was a fully mechanical system, a consistent low double digits would be very good indeed, anything more would be positively blissful.

So, 15% would be perfectly adequate, anything above say 20% and I'd be dancing.

At that you'd be beating the vast majority of fund managers, with very little ongoing effort.

 

[Shanghai]

Right, you're the only person who has put a figure on drawdown.

However, no-one else seems willing to put a figure on the kind of annual return they would like to see, or the drawdown they anticipate/could tolerate.

So, I wonder, maybe a good portion of traders simply have no idea what they are targeting, or what they expect on the downside? A bit like flying blind, isn't it?

I agree, drawdown is almost never mentioned. But as all the posters here are very successful traders with large accounts that shouldn't be an issue..... ehmm.

Annual returns for me are 30-60% (lower end of that for the past year as I have been too cautious). That sort of return tends to get scoffed at on here by those that wouldn't get out of bed unless they are doubling their money year on year but it does compound up nicely.

I also risk 1% per trade and I am trying to reduce that, not increase it.

 

[Pazienza]

I agree, drawdown is almost never mentioned. But as all the posters here are very successful traders with large accounts that shouldn't be an issue..... ehmm.

Annual returns for me are 30-60% (lower end of that for the past year as I have been too cautious). That sort of return tends to get scoffed at on here by those that wouldn't get out of bed unless they are doubling their money year on year but it does compound up nicely.

It certainly does.

I also risk 1% per trade and I am trying to reduce that, not increase it.

Me too, although I'm still over 2%, so I've a way to go. Ultimately my goal is 1%.

.

 

[DionysusToast]

I can only define what is a good system for me, but that is not a good system for everyone.

Go on then...

 

[meanreversion]

And there you have it.

This thread is utterly pointless. On one side, you have Shanghai who earns 30%+ per annum having seen a max drawdown of 20%. This ratio over 1.5 is very good, and enables us to do a little more analysis.

Max drawdown of 20% in 5 years is probably in the bottom 5th percentile of drawdowns as experienced by t2w punters, i.e. it's low. There's another mechanical trader out there (JRP) who has said he's prepared to tolerate 50% drawdowns, which would lead to an annual return of 75% using this ratio.

So, some analysis with numbers.

Then on the other side you have x4x and other people parrot repeating "EAs don't work" WITHOUT ACTUALLY SAYING WHAT THEY WOULD CONSIDER TO BE A GOOD SYSTEM.

If you can't define a "good" system, how the F*** can you say EAs are not "good"? It's simply idiotic.

Think about it. You are saying -

"I can't define a good system in any way, but without having done any analysis, I conclude that all EAs suck."

 

[meanreversion]

Annual returns for me are 30-60% (lower end of that for the past year as I have been too cautious). That sort of return tends to get scoffed at on here by those that wouldn't get out of bed unless they are doubling their money year on year but it does compound up nicely.

I also risk 1% per trade and I am trying to reduce that, not increase it.

Seriously, why do you give a crap what other people think is an acceptable return. Some people DO return 200% a year, but that's on a tiny account with ridiculous drawdown and occasional topping up from salary.

The vast majority of people on this site talk unadulterated sh1t. We can see this from x4x, who keeps repeating that EAs don't work without any analysis or research. His only "reason" given so far is - well, hey look what happened in 2008, very volatile, nothing could handle that. Most systematic trend funds had their best year in a decade in 2008, and they continued to make money in 2009 and 2010.

I know you don't know your vol of returns, Shanghai, but with a "Calmar" ratio (return/dd) of 1.5+, I reckon your Sharpe is higher than 1, which is exceptionally good.

 

[DionysusToast]

In all fairness MR - x4x is flat outright lying.

but he's saying what the newbs want to hear.

 

[Shanghai]

MR I'm going to show my ignorance now but I don't always understand your metrics. How to calculate annualised daily vol for example?

Out of interest what metrics do you use to assess a system, and how would you define "robust" for example?

 

[meanreversion]

In all fairness MR - x4x is flat outright lying.

but he's saying what the newbs want to hear.

It's like having a conversation with a four year old.

Parent - "Try some broccoli"
Child - "No, I don't like broccoli"
Parent - "Have you ever tried it?"
Child - "No, but I don't like it"
Parent - "What food do you like?"
Child - "I don't know, but I know I don't like broccoli"

 

[meanreversion]

MR I'm going to show my ignorance now but I don't always understand your metrics. How to calculate annualised daily vol for example?

Out of interest what metrics do you use to assess a system, and how would you define "robust" for example?

You need all your daily NAVs or closing account balances. Then take daily percentage returns, and find the standard deviation. Multiply by sqrt 260 to get annualised vol. I reckon yours is 30%.

The easier metric is return/drawdown, which you have already provided. Anything above 1 I think is good.

I would also look for a win rate above 35%, not that it really matters per se, but it makes the system easier to trade.

99% of punters define their trading purely by return, with no reference to drawdown. Totally and utterly pointless.

Which is better, system A with a return of 10% and DD of 5%, or system B with return of 40% and DD of 40%?

Well, all you need to do with system A is increase stake size 4 times and hey presto, it's return of 40% and DD of 20%. Clearly from that perspective it's better.

 

[x4x]

In all fairness MR - x4x is flat outright lying.

but he's saying what the newbs want to hear.

You just want to believe a dream - you need put no effort into something and get a fantastic lifestyle.

If you want to call me a liar, fine - you just carry on dreaming.

 

[x4x]

It's like having a conversation with a four year old.

Parent - "Try some broccoli"
Child - "No, I don't like broccoli"
Parent - "Have you ever tried it?"
Child - "No, but I don't like it"
Parent - "What food do you like?"
Child - "I don't know, but I know I don't like broccoli"

I think you are the one being childlike - I have expressed an opinion that I don't believe EA's will work in the long term - what's your problem with that - have I upset your dream too?

Secondly I have said that if Hotch has found the Holy Grail, then I congratulate him - I didn't start going off subject and making personal insults - that I left at the playground 30 odd years ago. I think you need to grow up and accept that people have different opinions to you.

 

[meanreversion]

I think you are the one being childlike - I have expressed an opinion that I don't believe EA's will work in the long term - what's your problem with that - have I upset your dream too?

Secondly I have said that if Hotch has found the Holy Grail, then I congratulate him - I didn't start going off subject and making personal insults - that I left at the playground 30 odd years ago. I think you need to grow up and accept that people have different opinions to you.

Put your testosterone to one side and consider my question.

What annual return expressed as percentage of trading pot would you consider a good return, and what drawdown do you think is acceptable to achieve this?

If you can't answer either of these questions, then you can't define what is a "good" system. If you can't define a good system, how can you say EAs are not good?

 

[meanreversion]

Secondly I have said that if Hotch has found the Holy Grail, then I congratulate him - I didn't start going off subject and making personal insults - that I left at the playground 30 odd years ago. I think you need to grow up and accept that people have different opinions to you.

Hotch did NOT say he had found the Holy Grail, because he knows it does not exist. Your response is a pure straw man argument, namely incorrectly exaggerating something said by the other person in order to disparage it. It's quite a common tactic in children.

Anyway.. let's look at what you said again. You do not believe EAs work in the long run. Why not? Have you researched them? Would you care to explain how systematic trend funds have made good coin for the last four years? What WOULD an ideal return be for you?