Why choose Fixed Fractional Money Management?

[Liquid validity]

I agree with you that Forwards testing is the testing that accounts for slippage and spread and shows you how you perform with the strategy (not just taking every trade like you would with back testing).

Agreed, that is why rigorous forwards testing is vital.

Also if your back testing for the last 5 years then you get a good sample test, unless your testing the 240 min charts and up, while covering all market conditions, because in the last 5 years the markets have been trending up, down and have gone sideways as well. So it wouldn't be curve fitting.

The battle between optimization and curve-fitting
You place more emphasis on sample span rather than sample size.
Its good to have both.
Yes you do mention sample size, but you gloss over the fact that
sample size is more important than sample timespan.

Also if your forward testing for 1 year how ca you know it's not curve fitting garbage as well... It's only 1 year worth of data... For example if you had a 60 min trend following system and you forward tested it from the 1/1/2013 till now on any AUD cross you would have made a killing...is that not curve fitting garbage?

You make the false assumption that forwards testing is separate from back testing.
A decent forwards test sample size and timespan is important yes.
What is more important is to see how a backtest of the forwards sample compares
to the forwards test results.
That will give an idea of how reliable your backtest is.
If the results are vastly different, the original backtest is worthless.
Backtesting is just a guide, confirmation comes from forwards testing.

I didn't say that everyone should risk 50% of their account, the percentage of the account risked if maximum drawdown happens depends on each person and should be chosen by each individual according to their risk tolerance.

You mention 50% drawdown as a maximum:

if you know roughly what your maximum drawdown will be, you can also see how much of your account you will lose if that happens and you will not make that % be more than 50% of your total account

So you say 50% (which is poor advice) yet don't offer a more realistic drawdown %...

i said "you will decrease the amount risked per trade as you increase in position size and implicitly decrease your risk of ruin"
and you respond "So you decrease size as risk of ruin decreases".
Just so I clarify this for you as you increase in position size, so from 1 lot to 2 lots and so on, your the percentage of your account risked per trade decreases, so at 1 lot you risk 2% per trade at 2 lots you risk 1.5% per trade and so on... This was only an example I'm not advocating any risk per trade that is mentioned above.

Risk of ruin is greatest at the start of an account.
Suggesting anyone start with greater risk at that point is lunacy.
Excessive size or risk per trade at the start of the account amounts
to the same thing.
Its the classic mistake of an over leveraged, undercapitalised account.
Simple fact is, if someone does not have sufficient account size,
they are better off not trading at all.

 

[Shakone]

Back testing, if done properly, is not just an execution test, when back testing you can also look for the optimal place to put, stops, targets and help you better understand the strategy itself without having to sit for 1 or 2 years in front of the computer and watch charts.

Which can lead to curve fitting

Also if your back testing for the last 5 years then you get a good sample test, unless your testing the 240 min charts and up, while covering all market conditions, because in the last 5 years the markets have been trending up, down and have gone sideways as well. So it wouldn't be curve fitting.

It can still be curve fitting.

I didn't say that everyone should risk 50% of their account, the percentage of the account risked if maximum drawdown happens depends on each person and should be chosen by each individual according to their risk tolerance.

The problems with a drawdown of 50% are not just personal risk tolerance. A 50% drawdown requires a 100% gain just to get back to your previous peak.

 

[Maza]

Agreed, that is why rigorous forwards testing is vital.

The battle between optimization and curve-fitting
You place more emphasis on sample span rather than sample size.
Its good to have both.
Yes you do mention sample size, but you gloss over the fact that
sample size is more important than sample timespan.

You make the false assumption that forwards testing is separate from back testing.
A decent forwards test sample size and timespan is important yes.
What is more important is to see how a backtest of the forwards sample compares
to the forwards test results.
That will give an idea of how reliable your backtest is.
If the results are vastly different, the original backtest is worthless.
Backtesting is just a guide, confirmation comes from forwards testing.

You mention 50% drawdown as a maximum:

So you say 50% (which is poor advice) yet don't offer a more realistic drawdown %...

Risk of ruin is greatest at the start of an account.
Suggesting anyone start with greater risk at that point is lunacy.
Excessive size or risk per trade at the start of the account amounts
to the same thing.
Its the classic mistake of an over leveraged, undercapitalised account.
Simple fact is, if someone does not have sufficient account size,
they are better off not trading at all.

I agree with you when you say sample size is the most important part of back testing and forward testing. I was talking about 5 years of back testing under the idea of having a sample size of at least 100+ trades per year. That was my mistake for not specifying that.

I also agree with you on the comparison of back testing with forwards testing. But the problem with the forward test is that if your not testing an automated system the results will vary quite a lot if your trading the 15 min time frame or lower. This of course depends on, if you took into account the exact time you actually will be trading in your back testing. Agreed with you also that you do need forward testing to confirm your back testing results.

I just threw a number out there. Me personally I risk, if my maximum draw down occurs, 25% - 30% of my total account equity. Plus when I do this calculation, I take my maximum drawdown, see the % risked of the account and add another 10% to that % and that number is my possible risk. I do this because the maximum drawdown will not necessarily be the same as in the results gotten from testing, so the extra 10% gives me a little more breathing room.

I also agree with you on the idea that if the account is too small then people shouldn't trade at all.

If you read the PDF document I have attached you can see that the idea of this money management is not to increase the risk to equity by much, it increases the total equity risked by a maximum of 5%.

 

[Maza]

Which can lead to curve fitting


It can still be curve fitting.




The problems with a drawdown of 50% are not just personal risk tolerance. A 50% drawdown requires a 100% gain just to get back to your previous peak.

If you have a big enough sample size it will not be curve fitting... big sample size is 500+ trades in my opinion. Actually that is still curve fitting, but the curve is much more accurate .

Agreed with you on the problem with a 50% drawdown. The idea was that it was the largest draw down, that is how you can calculate in the most pessimistic was how your strategy will work, because I would rather be pleasantly surprised when putting real money into the market than staring at the screen asking myself what I did wrong.

Also there are ways to reduce the loss to total account equity during the draw downs, I forget what it's called but it is quite effective.

There are a whole bunch of techniques that can be used to reduce draw downs and to augment results in a negative way, that the back testing has given you to mimic reality or close enough to.

 

[Liquid validity]

I just threw a number out there. Me personally I risk, if my maximum draw down occurs, 25% - 30% of my total account equity. Plus when I do this calculation, I take my maximum drawdown, see the % risked of the account and add another 10% to that % and that number is my possible risk.

If you read the PDF document I have attached you can see that the idea of this money management is not to increase the risk to equity by much, it increases the total equity risked by a maximum of 5%.

If that is unleveraged Drawdown, its still very high really.
Increase of risk per trade from 2% to 5% might not seem much,
but as you agree nobody should really trade a small account,
so this only applies to reasonable account size.

A sequence of losers or drawdown at 5% per trade on a reasonable sized
account is going to make anyones eyes water and hit the bottle...

 

[Maza]

If that is unleveraged Drawdown, its still very high really.
Increase of risk per trade from 2% to 5% might not seem much,
but as you agree nobody should really trade a small account,
so this only applies to reasonable account size.

A sequence of losers or drawdown at 5% per trade on a reasonable sized
account is going to make anyones eyes water and hit the bottle...

No no... I said a risk to max draw down of x+5%... for a risk per trade, if you read the PDF, you can see that I don't advise going any higher than a 3.5%. Optimal in my opinion is something like 3% at the beginning stages of position size growth.

And I agree with you on that... 5% risk per trade is huge in my opinion as well.

Read the PDF, I say this because you would be a good person to scrutinize what I wrote, I like your scrutiny because it is informed and detailed, you don't just say "That's not right!!!".

Maza.

 

[Liquid validity]

I'm well aware of the flaws and advantages of fixed fractional, fixed ratio, kelly and optimal F.

Particularly fixed ratio:
Fixed ratio flaws PDF

If you trade FX and can size down to microlot resolution, moving in 1 lot steps
is not an issue.
All MM systems have flaws and drawbacks.

For me, fixed fractional still wins, simply due to the fact that I trade
FX with brokers who offer microlot sizing resolution, so whole lot or contract
steps aren't an issue.
Another reason is Ed Seykota amongst others gives fixed fractional his vote:
Risk
Thats good enough for me.

 

[DionysusToast]

Maza - are you trading profitably right now?

 

[Maza]

Maza - are you trading profitably right now?

Yes I am. Why do you ask?

 

[Liquid validity]

Yes I am. Why do you ask?

Probably because of this:
https://groups.google.com/forum/#!topic/misc.invest.futures/xao2I6ZRwE4
Trading System Solutions - Publications - Money Management - The Basics of Money Management III

As to Ryan Jones’ attempt to break the trading record of Larry Williams in The Robbins 2001 Futures Trading Contest described in the previous article, he failed again... After his account grew by 600% from $15 000 to $107 000, he sent an offer to buy his method, proven by statements capable to bring such profits. Besides, he offered a $299 per month subscription to stay informed of all trades taken in the contest. As a result the drawdown on his capital reached 95%, just what had to be proven.

That was using fixed ratio MM as devised by Ryan Jones.
http://www.google.co.uk/url?sa=t&rc...hsWDBt6ur1YOErg&bvm=bv.50768961,d.d2k&cad=rja

Fixed ratio is simply too risky in the early stages of an account and has other flaws.

 

[DionysusToast]

Yes I am. Why do you ask?

Simply because you are on here giving advice.

And advice that is a tad 'iffy'.

Also - you speak of 'systems' and that's usually the domain of people still trying to figure out how the markets work.

 

[Maza]

Probably because of this:
https://groups.google.com/forum/#!topic/misc.invest.futures/xao2I6ZRwE4
Trading System Solutions - Publications - Money Management - The Basics of Money Management III


That was using fixed ratio MM as devised by Ryan Jones.
http://www.google.co.uk/url?sa=t&rc...hsWDBt6ur1YOErg&bvm=bv.50768961,d.d2k&cad=rja

Fixed ratio is simply too risky in the early stages of an account and has other flaws.

Lol ) that is funny as. I don't follow Ryan Jones, I just read his book and liked the concept of his money management strategy, plus I know quite a lot of people that use it profitably and have had no problems with it. The idea that I like is the one about fixed pip increments to go up in position size (when you talk about the Forex market) and how the risk decreases as your account equity grows.

As for Ryan Jones... that is probably the biggest fail out there lol... I love seeing these sort of things .

 

[Maza]

Simply because you are on here giving advice.

And advice that is a tad 'iffy'.

Also - you speak of 'systems' and that's usually the domain of people still trying to figure out how the markets work.

I am here giving advice because I want to help people not make big mistakes like throwing money into the market without having a clue what going long even means. This all started when I saw one of my acquaintances loose about $50,000 in the market because he thought the market was a get rich quick scheme... I'm trying to prevent anyone else from making the same mistakes as him.

"iffy" in what way?

I speak of "systems" because that is what I saw was spoken on the forum... personally I call them strategies for lack of a better word. I know how the market works, I studied a hell of a lot of things before I put real money into the market. It took me a year before I actually put my hard earned money into the market and since the I have only had 1 loosing month, my first one, Back then I was scared to put a live trade on the board lol. I'm not saying I'm an expert (because I'm not even close to being one) but I'm not a beginner either, or a guy that is still trying to figure out how the markets work.

Maza.

 

[barjon]

.. personally I call them strategies for lack of a better word. I know how the market works, I studied a hell of a lot of things before I put real money into the market. ............

Strategies - yes, and if you are successful I wouldn't mind betting that there is quite a discretionary element involved and you can't back test that. After well over thirty years I haven't yet seen a mechanical system that will give guaranteed profit over time, however good countless back-tests might look.

I think your opening post gives the game away since it concentrates on making money and making it quickly at the start. That's stars in the eyes stuff and the more sensible approach is to concentrate on limiting losses and have a money management strategy to suit.

One problem I have with these money management discussions is that I don't believe all trades are equal in terms of the risk you are prepared to bear. Some trades look better than others (rightly or wrongly ) and I'm prepared to take a higher risk with those.

 

[Maza]

Strategies - yes, and if you are successful I wouldn't mind betting that there is quite a discretionary element involved and you can't back test that. After well over thirty years I haven't yet seen a mechanical system that will give guaranteed profit over time, however good countless back-tests might look.

I think your opening post gives the game away since it concentrates on making money and making it quickly at the start. That's stars in the eyes stuff and the more sensible approach is to concentrate on limiting losses and have a money management strategy to suit.

One problem I have with these money management discussions is that I don't believe all trades are equal in terms of the risk you are prepared to bear. Some trades look better than others (rightly or wrongly ) and I'm prepared to take a higher risk with those.

Barjon,

Yes that is true, the discretionary element is key in any successful trading strategy.
Thirty years!!! wow that is... just wow...
That is also true, that mechanical systems will not give guarantied profits over time, because market conditions change and tweaking is necessary as time passes, what was valid 10 years ago might not have any relevance today.

As for the opening statement I like going for shock and awe right from the beginning... it gets everybody fired up and you get some good informed opinions on the subject at hand. The assumption that I was working under was that everyone had a strategy that they have tested till the cows came home and were quite set in all their ROE's and were looking to just trade that strategy as they have up to now, just with live money.

I also agree with the fact that people are different and if one risk % works for me (for example) might not work for "John" or "Maria". That is why in my opinion every money management system should be tailored to each individual person. This of course has to be done by each individual person for themselves not by someone else for them. This is the exact same idea as with developing a trading strategy or a trading plan. We both might have the same basic idea on how the strategy will work, but the rules and restrictions must be made by each individual separately depending on their character and risk tolerance.

Maza.