Money Management

This is one of the amazing things about the human brain: it interprets the data based on its predisposition. When it has an opinion, all further data confirms it. So it happens that OddTrader, nononsense and I, all agreed with Bernards's thoughts. And we have no clue what he actually meant when he formulated them.
Quote from OddTrader:

Q

Claude Bernard quotes
http://chatna.com/author/bernard.htm

Mediocre men often have the most acquired knowledge. Claude Bernard

It is what we know already that often prevents us from learning. Claude Bernard

Man can learn nothing unless he proceeds from the known to the unknown. Claude Bernard

Men who have excessive faith in their theories or ideas are not only ill-prepared for making discoveries; they also make poor observations. Claude Bernard

The experimenter who does not know what he is looking for will not understand what he finds. Claude Bernard

UQ

:D
More...
 
http://www.elitetrader.com/vb/showthread.php?s=&postid=632764&highlight=buddha#post632764

Quote from OddTrader:

Q
Do not believe in anything simply because you have heard it.

Do not believe in anything simply because it is spoken and rumored by many.

Do not believe in anything simply because it is found written in your religious books.

Do not believe in anything merely on the authority of your teachers and elders.

Do not believe in traditions because they have been handed down for many generations.

But after observation and analysis, when you find that anything agrees with reason and is conducive to the good and benefit of one and all, then accept it and live to it.

--- Buddha (Trend Follwoing by Michael Covel, p.191)
Q
:confused:
More...

:)
 
Quote from cnms2:

This is one of the amazing things about the human brain: it interprets the data based on its predisposition. When it has an opinion, all further data confirms it. So it happens that OddTrader, nononsense and I, all agreed with Bernards's thoughts. And we have no clue what he actually meant when he formulated them.
More...
Yeah, but that's not our point of common interest. The question is as you put it yourself: "The theory", i.e. "How to make money speculating?"
No kidding around with that one.

nononsense
 
Regards

Money Management - What is it and why do we need it?

Every trader needs an edge in order to make money from the markets in the long run. That edge differs from trader to trader and it is depending on his trading style and on what he trades. The edge for example is different for a market maker in stocks than that for a long term trend follower in futures markets. Without it, he will be an easy target and he will gradually lose his capital. But the edge by itself alone is not a passport to success.

Overexploit a small edge and you are going to end in the poorhouse.

What do I mean by this sentence? Trading is a tough job and whatever edge a trader might have is usually small. If there were big opportunities for easy profits then those would be easy spotted and exploited by many and they would cease to exist. Having this small edge doesn't mean luck steps out of the equation. Luck will still be the major short term's determining factor. Luck will tell you if your bank account's balance will head north or south in the short term. In the long term however your edge is the only thing that it will matter. If it is positive you will earn money. If it is very positive you will earn big money. The bursts of good luck and bad luck will cancel its other in the long run and all that will be left is the edge. BUT, there is one big but :-). If you played big, in other words if you were overexposed then a bad luck's burst will catch you in the short term and it will not let you exploit your long term advantage.

Think about it. It is very simple!!! If you risk too much then you will not 'live' long enough to exploit your edge. You will have a winning system but eventually you will have manage to end up broke. That's were Money Management enters the equation. Money management will simply tell you how much you can afford to risk for exploiting an edge. It will assure that you will take the cream from an edge without in the meantime ending up broke from a portion of bad luck.

Every successful trader uses a version of the money management concept to control his risks. You can hear someone say. I never risk more than 2% of my capital in every trade. What he means is that the size of his position ensures that even if his stop loss order is triggered he will not lose more than 2% of his total capital. This is a simple way for ensuring that a losing streak won't hurt you much. You will survive to trade the next day.

A concept that you will hear about in money management is the famous 'Kelly Fraction'. Often you will hear criticism about Kelly and in most cases it will be something like this: "Kelly fraction is very dangerous to trade. You will end up broke using it.". Let's sort things out by first stating what Kelly fraction is. Every trader following a trading system has a summary of that system's statistics. That statistics describe the profitability, in other words the edge of this system and are easily available if you test your system in any back testing platform. From that statistics you can extract a number, 'the Kelly Fraction'. The Kelly fraction is the absolute maximum percentage of your capital that you can risk in every trade. Read this statement carefully. The concept of Kelly fraction in trading is not advising you to risk that percentage of your capital in every trade but is advising you never to exceed that percentage because you have nothing to gain by doing so.

Having calculated this 'Kelly fraction' for your system and assuming that your system will keep working in the future, there is only one question to answer. How aggressive you want to be? Can you sustain a big drawdown in your equity curve in exchange for a chance to achieve your goal sooner? Then by all means allocate a percentage close to the Kelly Fraction in every trade. But be prepared mentally for a very tough ride!!! A rule of thumb may look something like this.

Suppose you have calculated you Kelly fraction to be 10%.

Very Aggressive trader : Risk 7% of your trading capital in every trade (in other words 0.7 * Kelly).

Aggressive trader : Risk 5% of your trading capital in every trade (in other words 0.5 * Kelly).

Conservative trader : Risk 1% - 2% of your trading capital in every trade (in other words 0.1 - 0.2 * Kelly).

Very Conservative trader : Put your money in the bank :-).
More...
 
Quote from nononsense:

You're really spoiling me! I love your kind of guys.
The main vocation of nononsense is to bring the more outrageous nonsense to the attention of the more subtile minded speculating public. (It's the only thing that can help people learn how to make money.)

nononsense
:cool:
More...

Awwwww, how cute and nice of Mr Nononsense to be the savior of the subtle minded speculators out there. He scavenges the threads of ET to rid us of nonsense.

Now listen up you freak, no one on this thread is claiming to have anything to sell or claiming to have the holy grail. You mentioned before that risk of ruin theory was well known for tens of years.... SO WHAT? The fact that markets are complex and can challenge even the most theoretically informed of traders does not mean that one should not fully understand all tools to control risk, including their limitations! To understand, one has to investigate and discuss, which is what we were doing on this thread until it got trolled by idiots like yourself.

If your only advice on money management is "Don't lose any", please, realize that you have nothing to offer. The Kelly formula is one way to begin to grope for an answer to the question of how much to risk. Cnms mentioned he uses 20% kelly to trade and the accompanying numbers for drawdown probabilities. They are not pixie dust, or holy grail, they are a candle in the dark you fool.
Of course one has to try not to lose money to stay alive. Using a fraction of your fraction of the Kelly fraction is probably a decent way to derive a systematic way to derive position sizing, vs having to rely on useless platitudes/heuristics the kind you like to trumpet on this site, hoping for a gold medal of recognition.

You haven't given one practical suggestion on this thread, not one as to how to go about not losing money. Now if you are looking for 'the theory' or a trading system, piss off because no one is going to hand you one.

Here's a suggestion, why don't you email Seykota and call him a money management clown for posting some info on the Kelly ratio, expectancy and other things of the sort on his website. Tell him, "These things were know tens of years ago! clown!"

Nononsense, you have as much common sense as a rock.
 
I suppose "% of your trading capital in every trade" assumes in this context daytrading, where you have only one position opened at any given time. The 10% Kelly fraction refers to the optimum risk for your whole account ...
Quote from gbos:
...

Suppose you have calculated you Kelly fraction to be 10%.

Very Aggressive trader : Risk 7% of your trading capital in every trade (in other words 0.7 * Kelly).

Aggressive trader : Risk 5% of your trading capital in every trade (in other words 0.5 * Kelly).

Conservative trader : Risk 1% - 2% of your trading capital in every trade (in other words 0.1 - 0.2 * Kelly).

Very Conservative trader : Put your money in the bank :-).
More...
 
Quote from cnms2:

I suppose "% of your trading capital in every trade" assumes in this context daytrading, where you have only one position opened at any given time. The 10% Kelly fraction refers to the optimum risk for your whole account ...
More...

Yes, the example is for one open position because otherwise I had to make a long heavy on math post to explain the idea. For example for opening two positions the analysis is something like this.

Correlated trading systems with non-normal returns

Analytical solutions for Kelly-f capital allocation among trading systems are easy to obtain as long as they exhibit normal like payoffs. The well known solution in this case is that the matrix describing capital allocation between systems equals:

F* = C^(-1) x M

C the variance - covariance matrix between available systems
M the column-matrix containing the expectancy of each system
F* the row-matrix containing optimal betting fraction that we must allocate in each system

The geometric growth rate for our wealth will be:

g(f1*,...,fn*) = ½ x transpose(F*) x C x F*

When the systems don't have normal-like payoffs analytical solutions are perplexed and not easy to obtain. We will study the simple case of allocating capital between 2 correlated systems with binary payoffs. A binary system has a payoff x1 (success) with probability p and a payoff x2 (failure) with probability (1-p).

- The mean of this system is p*x1 + (1-p)*x2
- The variance is p*x1^2 + (1-p)*x2^2 - mean^2
- If this system was traded alone then the optimal Kelly fraction is mean/abs(x1*x2) were abs denotes the absolute value.

In the two system case we will use the notation p1 for the probability of success for the first system and p2 for the probability of success of the second system. We will also use the notation xij for the payoffs where the first subscript denotes the system and the second subscript denotes the payoff of the system. For example x12 denotes that the first system gave us the payoff x2 (failure).

So, the problem we have is: If probabilities p1 and p2 are known and payoffs xij are also known then what is the optimal Kelly capital allocation in these systems given that they have a correlation r?

This is equivalent of finding the f1 and f2 that maximizes the expression

g(f1,f2) = Sumi,j pij*ln(1+f1*x1i+f2*x2j)

Where f1 and f2 are the simultaneous capital allocations in systems 1 and 2 respectively, and pij denote the probability that system 1 gives the payoff i at the same time the system 2 gives the payoff j.

For finding f1 and f2 in the maximization expression, the only unknowns are the pij. For finding them we will use the fact that when system 1 gives payoff #1 then there is a probability that system 2 will give payoff #1, and (1-a) probability that system 2 will give payoff #2. Similarly when system 1 gives payoff #2 then there is b probability that system 2 will give payoff #1, and (1-b) probability that system 2 will give payoff #2. We will find those unknowns a and b.

We will need 2 equations to solve for the two unknown parameters a and b. If mi is the mean of system i and si the standard deviation (i.e. sqrt(variance)) of system i then:

(1) Covariance = r*s1*s2 = p1*a*(x11-m1)*(x21-m2) + p1*(1-a)*(x11-m1)*(x22-m2) + (1-p1)*b*(x12-m1)*(x21-m2) + (1-p1)*(1-b)*(x12-m1)*(x22-m2)
(2) p2 = p1*a + (1-p1)*b

Solving the two equations we obtain the two unknowns a and b. The unknown probabilities pij are now easily obtained.

p11 = p1 * a
p12 = p1 * (1-a)
p21 = (1-p1)*b
p22 = (1-p1)*(1-b)

Now the last step is to use excel solver and maximize the expression:

g(f1,f2) = Sumi,j pij*ln(1+f1*x1i+f2*x2j)

thus obtaining the Kelly-allocation of our capital at each system f1 and f2.

Example: We have two similar systems with the following characteristics:

p1 = 0.8, x11 = 0.2, x12 = -0.6
p2 = 0.8, x21 = 0.2, x22 = -0.6
Correlation = 0.5

We can easy calculate that each system has mean = 0.04 and variance = 0.1024. Solving the equations (1) and (2) we obtain a = 0.9 and b = 0.4. Thus we have:

p11 = p1 * a = 0.72
p12 = p1 * (1-a) = 0.08
p21 = (1-p1)*b = 0.08
p22 = (1-p1)*(1-b) = 0.12

The maximization of the expression g(f1,f2) = Sumi,j pij*ln(1+f1*x1i+f2*x2j) leads to the result:

f1 = 21.75%
f2 = 21.75%

So we allocate simultaneously 21.75% of our capital at each system (total exposure 2 * 21.75% = 43.5%).

If only one of the systems were available to us then the Kelly fraction would be 33.3%.

Note: If we use the matrix notation (applicable only for normal-like payoffs) then we would obtain the solution:

f1 = 26.04%
f2 = 26.04%

which is obviously far enough from the correct one.
More...

The conservative approach will lead us to allocate 0.1 Kelly in the above example so we would risk 2% of our capital in each of the two open positions (total 4%). If only one position was available then 0.1 * Kelly would be close to 3%.

If you want to trade more open positions simultaneously then either you use the math for normal like payoffs quoted above or you use a simplistic math treatment for calculating rough Kelly allocation fractions for your expected open positions.
 
Quote from ES335:

Awwwww, how cute and nice of Mr Nononsense to be the savior of the subtle minded speculators out there. He scavenges the threads of ET to rid us of nonsense.

Now listen up you freak, no one on this thread is claiming to have anything to sell or claiming to have the holy grail. You mentioned before that risk of ruin theory was well known for tens of years.... SO WHAT? The fact that markets are complex and can challenge even the most theoretically informed of traders does not mean that one should not fully understand all tools to control risk, including their limitations! To understand, one has to investigate and discuss, which is what we were doing on this thread until it got trolled by idiots like yourself.

If your only advice on money management is "Don't lose any", please, realize that you have nothing to offer. The Kelly formula is one way to begin to grope for an answer to the question of how much to risk. Cnms mentioned he uses 20% kelly to trade and the accompanying numbers for drawdown probabilities. They are not pixie dust, or holy grail, they are a candle in the dark you fool.
Of course one has to try not to lose money to stay alive. Using a fraction of your fraction of the Kelly fraction is probably a decent way to derive a systematic way to derive position sizing, vs having to rely on useless platitudes/heuristics the kind you like to trumpet on this site, hoping for a gold medal of recognition.

You haven't given one practical suggestion on this thread, not one as to how to go about not losing money. Now if you are looking for 'the theory' or a trading system, piss off because no one is going to hand you one.

Here's a suggestion, why don't you email Seykota and call him a money management clown for posting some info on the Kelly ratio, expectancy and other things of the sort on his website. Tell him, "These things were know tens of years ago! clown!"

Nononsense, you have as much common sense as a rock.
More...
Keep on reading your lill Seykota, you visibly need it. It can only strengthen your dreams of shining in the 95% heap. Lots of your buddies love that kind of stuff. MM is good for you, but don't forget about the trend - lots of possibilities for you to jerk off on that one too.
:D
nononsense

PS "He scavenges the threads of ET to rid us of nonsense." If you had any "common sense" at all, you would know that with bums like you, that would be an impossible task.
 
You must log in or register to reply here.
Back
Top