Optimized Risk vs Reward Equation

Correct me if I'm wrong, but the way I interpret the lessons learnt is that [1] means "Don't average up!" and [2] is a given.

From a quantitative perspective, there is a handful of parameters that determine whether or not money management can actually create an edge. Because identification of these parameters is quite advanced, I second fti's lesson [1].

Average down is for idiots who don't know math.

Search "Martingale" in google. If you want to average down, just give me your money, it's faster that way.

The book was good, and if anyone wants more on Soros, there's an unauthorized biography of him.

@twoblink: You've been watching 21 a few times in a row, haven't you?
Ah, the advantages of socialistic education, inequalities in first grade of primary and matrix calculus in second year of secondary school.

I think I've managed to grasp the whole fuzzy logic non linear constrained optimization problem (FLNLCOP in future ). I'm now trying to put my thoughts into excel and then on to demo account. It looks easy now.

I've taken Fuzzy Matrix Games Multi-Criteria Model for Decision-Making in Engineering as my next reading material.

This thread celebrated it's first birthday four days ago!

Has anyone found the equation yet?

And 4 years later, people are still averaging down.. so obviously I'm a failure in what I preach..

BigV.. Fuzzy logic is awesome.. You can apply it to a bunch of irrelevant items in your life!! ;-)

I'm not a great Blackjack player; but so far, my stats are spot on..

made me think of averaging up continuous ephemeral C:
http://blog.makezine.com/archive/201...ind_black.html

i still struggle imagining a generic formula for interplay of long and short, and without using market bias.

(just juggling ideas) instead of abrupt stops (direction bias), some smooth scaling progression of long inside short inside long inside short... that actually extracts profit instead of returning to 0 (-spread) on retrace to starting price.

Two Blink I am not sure if your still monitoring this

Assuming that one of us does solve this - how would we go about testing?

@almo,

I'm not sure I understand the question.

If you have an equation that is correct; then when you plug in numbers and graph the results; you will see a mini-max (saddlepoint, in calculus vernacular) when you plot risk vs reward. Is that what you are referring to..??

https://upload.wikimedia.org/wikiped...ddle_point.png



Since I haven't written on this thread for like 4 years.. Things that I have discovered and I almost would like to retract this thread..


1) Average down: Still a stupid idea. Given X to be the price, and n to be the the iteration; probability asymptotically approaches ZERO as n increases when you average down. Think of it like a slot machine. Payout is 98%. So what? 2% loss every pull on average; so after 100 pulls you will be broke. Same with average down.


2) Average up: It's still a better idea that average down; but I've discovered a few things about it. Given X to be the price, and n to iteration, probability asymptotically approaches ONE as n increases when you average up.

The problem with that is that it NEVER reaches 1.. And so your expectancy will never get to positive based purely on the mechanics of position manipulation. MM magnifies; that's all it does. If you have a positive expectancy, you will magnify profits. If you have a negative expectancy, you will magnify the losses.

Rereading my old posts is somewhat amazing to me..

I thought the original intent was to prove/disprove what you found by having someone else attempt to recreate the work.

I guess in my mind I was thinking if someone was able recreate it - the testing would go something like

"given A ( Starting lot size), what would be optimal position size at y distance, and minimal risk distance at z."


Assuming your formula is correct, the tester would need to supply you X(1), Z(1), and Y(1). - where Y(1) would be next buy/sell point.


Side note - I do want to thank you for starting this thread. If nothing else, you got me started on re-learning Calculus.

That was my thought as well; but thus far, nobody seems to have been able to reproduce my work; of if they did, they aren't posting it.

It's been 4 years so I can give a hint;

It contains phi.

I have something I want to submit via PM - but I also must profess my ignorance on one aspect of Calculus - that of Convergence/Divergence. Also still working on the Saddlepoint aspect of this.

No more update now? Like to read more. It is very helpful to me.

Twoblink ,

Is this the same thing http://forum.mql4.com/41087 you are talking about?

If I remember correctly Optimized Risk vs Reward Equation in a different form was stated by Michael Harris.
http://www.amazon.com/Profitability-.../dp/047022908X

Not sure what was shown in this thread since I didn't read it all....

I had meant to post on the anniversary; but such is life.

Looking back at my emails, July 19th, 2009 was when I formalized my mini-max proof; I had been working on it for 2 years, but it was on that day that I was able to give a proof. Actually, that was done on July 17th, 2012. I had sent it to someone to independent verification; and not only did he verify it; but sent back a summary equation for it; so I had an equation to plug into.

Well, it has been 3 years, and I have seen NO PROGRESS that's even close to it. I had promised someone that if nobody is even close in 3 years; then I might consider publishing it.

So...

I am now officially considering publishing the results;

I guess the questions are:

1) Should I publish it?

2) Would any of you guys buy it?

3) How much should I charge for it?

The first one is very philosophical; part of me thinks the publishing of it will greatly advance the FX world; and part of me feels like this is my FX Coral Castle and I'm more than happy going to my grave with it. So I'm still on the fence about it.

The second question I guess is just that, would you all buy it? I keep hearing people say they will buy it; but that does not equate to sales IMHO.. and I've had no less than 20 people tell me my ebooks are posted on a bunch of FX forums that I can't access or can't read the language to (for some reason, the arabs love my ebooks) and so I make very little money from the hard work I put in since most are just going to pirate it..

The third question however, is the one to ponder... Quite a few have said I should charge in the thousands.. or at least low four digits to weed out the people who would repost it on some forum... If you paid $29 for an ebook; you probably don't feel back reposting it or giving it to someone that asks you for it, if you paid $3000 for an ebook, then I'll bet you you don't even show it to your wife..

So those are the questions at hand...

keep it for your self because if you publish it and the people are starting to use it then it will NOT work again!! ... then you should find another formula which it works as long as the people don't use it ...

By the way if it works why the hell do you need it to sale and ask here how much should i charge for it? .... if it works you will be Billionaire in no time ...

Lol, I thought Friday Funnies are published on fridays.

Hi Two,
I've a good idea of what you're doing / have done. I don't think you're the first btw, but that's a different story. In terms of pricing, my own personal barometer is that if something is priced like other similar professional finance texts like Taleb's, OK. If it's priced like it's missing chapter in the book of revelations, I know what to expect and give it a wide birth.

Good luck whichever way you cut it.

Hello TwoBlink..all.

Interesting and fascinating topic in spite some of the math is way over my head knowledge.

Nevertheless I call myself an strategist for the forex market.

The math is exact but without taking in consideration certain fundamentals of the market itself the application of the math could be worthless.

For example... Reading your latest toughs regarding averaging down and averaging up and generalizing a slight advantage of one over the other.

If we do not take in consideration the most fundamental aspect of currency exchange ..which is ::: "at certain rate of exchange the crap from one country is appealing to merchant/investors/people of another country and a demand is created for that currency, that currency will raise in value.... to a point that the crap from that country is not longer appealing ... and from there, either come down by lack of demand or it is brought down like The Swiss and the Japs do (and China that is doing it in a continuous base) .
And that is a never ending history... even from before the currencies lost the gold value and became just a mere "legal tender".

I guess my point is that regardless of any equation, math model or whatever anyone place in front of me, proving that Average down is "mathematically" a stupid idea, I would still continue to do it when it suits and make money from it.

... Would I average up, just because of a math model tells me that is the best thing to do? for example when selling eur/chf at 1.2010? Obviously the question is why anyone would sell the eur/chf at 1.2010 rt? ...but just for the example sake. So, I would buy it at 1.2010... and would average down with another buy if the price drop to 1.2000 ... why? Because the expectancy (risk/reward, in this case, override the math model. ).

Anyway, thanks for great thread twoBlink ... I have been working in equation trading myself but not in the risk/reward area. But the high math is always a weak point for me... so I do everything the long and hard way... Back testing all my theory.

Would I buy your risk/reward equation from you??? Hard to say. The difficult part for me would be to know what to do with it. I think that would be for most of us. Would I need to create a strategy around the equation? Would I be able to apply the equation to any/all strategies? Would improve the strategies that I am already using? Maybe I am already using the equation without me knowing?... Those thing I would never know until I have it and try. In the other hand.... Would I buy a "strategy" in which you use the equation and profits are guarantee? Of course I would... regardless of the price.

In my own findings, to be honest I have use very simple math concepts in my strategy.. part of it, I believe that I am averaging down but from what I read here ,, I may be doing both... the bottom line... the strategy is profitable , and the math seems to be correct as the system works both ways... doing it one way... and doing the exact opposite, so in my real trading I am doing both simultaneously.

J.

Hi Twoblink,

The fact that you have been absent for years from this forum and reviving a basically dead thread with lots of typical (stupid) FF user comments after a few minutes of your reappearance says it all: We are interested in your optimal re-entry point and lot size formula.

With regards to myself paying for it: I have been paying for your ebooks before and found them somewhat useful, but not conclusive - i.e. I ended up understanding and coding what a trend is, playing around with it and testing various strategies and periods, but that's it. It didn't help me make money, because I couldn't apply it with repetitive success. The market changed direction too often and too far, to use the method consistently.

What you are offering now is basically another puzzle stone: a mathematically optimized averaging up method - no matter the direction. However, we still need to figure out when to enter for the first time and in what direction. We will end up with something half baked or semi-finished - or put differently: it's up to us to do with it whatever we make out of it. No How-to, simply some Food-for-further-Thought. Intrinsically motivated, I will be paying for it like I will be paying for an interesting book, but nothing more.

Regards, !iii!