Ryan Jones' Money Management

[SL]

Mike & Co,

I got side tracked a couple of weekends ago working on something and was playing around with the money management code on my system wasting time as you do. It sounds similar to what is described here in this thread so I will add my bit.

What I did was start out with an account of 50K, at a risk of about 3% or so and when it got to 100K I wound it back to 2% and when it got to 200K I wound it back again. Interesting things started to happen. With the right combination of increments the account balance increased significantly in some cases for no or little change in the drawdown. I was not using FF but volatility based algorithm. I am at work other wise I would cut and past what I did. I use TR by the way. Let me have a stab at it here in free hand.

'******resources page***********
col1= ATR[1]
Manager[1] = EMA[col1,10] * pointvalue

'*****Trade entry page******
memory[1] = (50000 * .03) / manager [1] * 3 ' 3* times volitility
if equity > 100000 then memory[1] = (100000 * .02) * 3
if equity > 200000 then memory[1] = (200000 * .01) * 3
newcontracts = memory[1]

Bugs and omissions are supplied free of charge with this. By playing around with the * 3 and making it * 4 and alterations like that you can get it to boost earnings while maintaining the same or almost the same drawdowns (another 'strange but true' piece of weird maths going on). You can change the trigger points, and the bet size fractions, or even the volatility EMA, it all does something.

If nothing else it illustrates again just what a waste of time playing with trade entries all night really are when you can make such a dramatic difference just by manipulating the bet size algorithm to a system. I think the standalone system made around 80K. By playing around with it I had it up around the 500k mark and some pretty high ROI figures but the DrawDowns started to catch up with me at that point.

It kind of looks like curve fitting the MM to the trade history. I can't remember if I walked it forward or not I may have done, but I pushed it aside to look at later on when I have more time. A walk forward test like that used on TR would surely destroy this in a second if it has no credibility. It may inspire some other variations. It sort of has that anti-martingale feel to it somehow. I tested it I believe on a regular breakout trend-following system on 21 markets (a very plain vanilla system).

RE: iitm and others-- These outfits trying to nickel and dime everyone because they claim they have discovered some new science is just a lot of leg pulling and grandstanding. You can have this for nothing. Anyone that spends enough time playing around with the MM can discover all sorts of interesting things, this is the tip of the ice berg. You are only limited by your imagination. This is not new science when it's this simple some people like to make out it is. Somebody would have thought of it before me, they just aren't saying so, probably card players

Cheers

Stephen

[SL]

FYI,

See, already spotted a couple of mistakes so I will get in first. Just testing!

'******resources page***********
col1= ATR[1]
Manager[1] = EMA[col1,10] * pointvalue

'*****Trade entry page******
memory[1] = (50000 * .03) / manager [1] * 3 ' 3* times volitility
if equity > 100000 then memory[1] = (100000 * .02) / manager[1] * 3
if equity > 200000 then memory[1] = (200000 * .01) / manager[1] * 3
newcontracts = memory[1]

Stephen

[smodato]

If you can trade you trade. If not .....

Toby Crable told me that his book got his trading career launched. He recommended writing a book as a way to get noticed.

Sorry Forum Mgmnt, I don't understand now what book Toby is speaking about, Jones' book or some book of his own?
Thanks, bye
Smodato

[Forum Mgmnt]

I was referring to Toby's Book: "Day Trading With Short Term Price Patterns and Opening Range Breakout".

He said that several industry veterans read his booked, liked it, and thought he would be a good person to place money with.

- Forum Mgmnt

[smodato]

Thanks Forum Mgmnt, now I'm a little bit full of doubts, I'm reading Ryan's book about money management and the reading is convincing me about the advantages of fixed ratio vs fixed fractional, I see not everybody here "accepts" this position and even though I fully understand everybody is presenting his own work with the right propaganda ( so certainly does mr Jones) there must be a truth somewhere in everything.
FFM sounds very dangerous in the examples mr. jones presented in his book , FRM seem to have left some pain in the very last periods so I have one simple question:
different methods show how to increase significantly the account by proper MM, it is shown however how these methods may also create problems in heavy drawdowns scenario, so what's left? a less agressive technique maybe? I'm first of all aiming at keeping my money, if it goes up 1000% or 10% it does matter but not that much to risk the end of the game, is Kiwi perhaps saying better techniques are a little bit more conservative?
Thanks
Smodato

[Roscoe]

smodato, the only way to resolve this to your own satisfaction is to test the heck out of it! My preference for such evaluation is to look first at the equity curve to get a feel for the general nature of the MM strategy being tested, then review the statistics in detail.

FWIW my findings to date indicate that, for me, the preferable MM strategy is one that shows a decreasing risk profile as the account size grows, so I am still a FR fan, although FF with a decrease in risk % based on account size is still a possibility.

Put aside concerns about the character of Ryan Jones or whoever and instead concentrate of proving whether or not their idea has any merit. It is the idea that you are evaluating, not the author of the idea.

So, getting back to my intial point, test everything of interest to you, and then test it some more. I know of no other way to find the answers that we all seek.

[smodato]

IMO and many others (there is much discussion 2 yrs back on omega group threads with good testing to justify the opinions) Ryan is just selling books. If you can trade you trade. If not .....

The best money management for normal players is to bet a fixed fraction of your equity on every trade. You can add to variations that may pay off:
- the ryan jones bet heavy when new approach (try that if u started 2 years ago just for fun).
- the "bet harder on the markets money" until the end of the trading period.

I suggest keeping it simple.

Then the next variation is how you calculate your equity. Do you use your closed trade equity, your open trade equity or a cleverer version which is the equity protected by stops. I think the last one is intellectually most attractive but like most people I end up using my Open Trade Equity because that is what I get on my report each night so its easiest.

There are also other discussions about all of these aspects on this site with more depth and some opposing opinions esp wrt the value of RJ. Tests and his inaccuracies in his book suggest to me that you should keep it simple.

John

I'm just curious, it seems many oubts are left on Ryan's work, but in the referrals Kiwi is giving I can't identify precisly the fixed ratio method. I tested some of my strategies with fixed fractional method (keeping the % from 2% to 5% maximum), fixed ratio with some twice a margin delta and Larry Williams method dividing by the largest loss the % of tha capital we are willing to risk. as my systems exit on close on the same day position is entered LW method loses part of importance in my opinion as largest loss should be of less impact (I suppose the real "largest loss" occurs when maret opens against you on the net morning).
I liked fixed ratio scenario and I'd like to know what particular limits you can see in it, c.f. explained quite well his doubts about optimal f but with fixed ratio I see quite less risk.
Thanks
Smodato

[smodato]

You've got test results but you don't trust them. Why not?

Thanks for your answer, I go immediately to the point, my results look very good, maybe too good, I can imagine worse scenarioes and live with them with no problm as there is still room for good profits; what I fear is the disaster, I can't see the hidden trap if there is one in the method, so I ask to those who showed some concern in order to get some practical experience maybe.
I'm happy for my results, I'm happy for the money management applications results, I would avoid being happy now and desperate later on.

Smodato

[bolter]

I borrowed a copy of the notes from Jones' course (circa late 90's). After working through his math and applying a little quadratic algebra ...... his fixed ratio approach can be expressed with the following formula.

l = (1+sqrt(1+8*e/d))/2;

where:
e = equity
d = delta
l = number of lots

After testing with it I would conclude:
1. It works well with a small account size.
2. But you can blow your account if your system goes into an immediate drawdown.
3. The real danger period is around the first increment in lot size.
4. If you manage to get a good start your risk of ruin rapidly approaches zero.

In order to reduce some of the inherent risk:
1. Use TRUNC() rather than ROUND() on the above formula.
2. Experiment with various methods for determining delta.

My interest is purely academic (ie: I don't use Fixed Rato) - it is nonsensical for managing OPM.

Cheers

[sluggo]

I think your equation is incorrect. It has you increasing from one lot to two lots when the account equity is greater than delta. But that is clearly opposite to every F.R. example in Jones's book and articles.

I think you have totally omitted a variable which I will call "B", the Beginning equity. At the beginning of a trading campaign, equity equals "B" and the F.R. method trades one-lots. It doesn't increase position size until equity >= (B + delta). This is certainly how all of Mr Jones's examples work.

But your equation effectively sets B=0 which I think isn't correct.

sluggo

[bolter]

hi sluggo,

Thanks for the heads up. I am working from his course notes dating back to 1997 - which may be problematic. I found them quite difficult to follow but my formula seemed to give the correct results for various worked examples he provided.

Having said that, the introduction of initial equity into the equation as you suggest seems perfectly logical to me, although it would alter the underlying concept. For instance, my version of the FR formula does not mandate an initial size of one lot, and delta would be viewed differently.

My intention here was to reduce Fixed Ratio to a single formula, in accordance with what Ryan Jones advocates. If I've got it wrong because my material is dated or I have misinterpreted it then please help me get it right.

Any chance you post the correct formula from more recent material sluggo?

Thanks - I appreciate the exchange.

[Roscoe]

I am using this:

Code: Select all

    if(CurrentProfit > 0)
        NumContracts = max(1,(int)(sqrt(2*CurrentProfit/Delta+.25)+.5));
    else
        NumContracts = 1;

[bolter]

Thanks Roscoe.
Your formula is less agressive initially but the results converge as profit grows, which is as you would expect.

Did you source this from his book/recent material?

Cheers.

[Roscoe]

Did you source this from his book/recent material?

No, I got it from a mathematically clever friend and the results closely match my understanding/calculation of Ryan's book.

I would like to work out ways to incorporate starting equity into the equation. I have tried the following variations:

Code: Select all

	if(CurrentProfit > 0) {
		if( SelectVersion == 1 ) {
			NumC = max(1,(int)(sqrt(2*CurrentProfit/Delta+.25)+.5));
		}
		else if( SelectVersion == 2 ) {
			int DefinedProfit = InitialProfit + CurrentProfit;
			NumC = max(1,(int)(sqrt(2*DefinedProfit/Delta+.25)+.5));
	    }
		else if( SelectVersion == 3 ) {
			NumC = max(1,(int)(sqrt(2*openEquity/Delta+.25)+.5));
		}

but with no clear winner (so far anyway). Any ideas?