risking 1% vs 2%

I have a question regarding capital risked per trade. What is the difference between risking 1% vs 2% per trade. How much more conservative is 1%? Would I have to lose on 100 straight trades to lose all my capital vs. 50 trades risking 2%? Is it that simple? I am trying to decide how much to risk per trade?

The optimal amount to risk is highly dependant upon your win/loss ratio. No one can come in here and definitively tell you how much to risk. You should know that there is a number for every system and you have to find it. Generally speaking it's between 1 and 5.

Theoretically, you can lose more than 100 times risking 1% becuase you always risk 1% of your remianing balance which continues to decrease as you continue to lose. Of course, at some point a margin call comes into play and that is where theory jumps over to reality.

It's the same as the age old calculus question: If you walk half way to a wall, and then go half again (repeat many times) will you ever reach the wall?

Phil,


What criteria would you use to decide between 1% and 5%?

I actually use software that I can feed in my trades. It analayzes, profit, loss, risk, reward, and drawdown and then spits out an optimal value to risk. Of course this number is highly dependant on your system. Me telling you my numbers would be of no use to you. The number 3% does come up frequently though.

To give you an idea of how important it is to find your number consider the following experiment. This is from Ryan Jones' book The Trading Game.

Imagine a coin flip system which takes 100 flips. 50 are heads and 50 are tails. Let's say that for every heads you win $2 and for every tails you lose $1. Obviously if you bet $1 per trade you will end up with $50 (100 - 50).

Now 4 risk scenarios beginning with a $100 account

A. Bet 10% of your account on each flip
B. Bet 25% of your account on each flip
C. Bet 40% of your account on each flip
D. Bet 51% of your account on each flip

After 100 flips, the results would be as follows:

A. $100 turned into $4700
B. $100 turned into $36100
C. $100 turned into $4700
D. $100 turned into $31

As you can see, risking too little is as foolish as risking too much. The one thing this experiment ignores is drawdown. If you desire to limit drawdown to the 20-30% area the % risked comes down to around the 1-5% value that I mentioned previously.

To give you an idea of what this looks like visually, here is a system that I use. Remember this graph ignores drawdown but it does line up well with the results of the coin flip example from Jones' book.

Where can I obtain the software you use?

Hopefully it's OK to post this: Just google "Market System Analyzer" and you'll find it. I don't want this to turn into an advertisement.I have absolutely no affiliation with this guy but I do use his software. There is a free demo. Also, the author has written some amazing articles on money management. You should read them if you find the time.

I've seen that software before and was about ready to demo it before I gave up my Tradestation account.

It would work well with TS... as well as Rina models... I love different money management models... it's very interesting.

I'm going to play around with this as well. Seems very interesting, indeed. Good information Phil.

if you use kelly's criterion you can figure it out pretty quickly. Lots of people say to only risk 1% but if your system is good you are hurting yourself by risking less. That said, you really need to know what your system produces and it is better to make the mistake on the low end.

Kelly % = W – [(1 – W) / R]

W= win rate of your system. if you win trades 55% of the time then w=55
R = pay ratio. if every time you win you get $2 and every time you loose you loose $1 you have a 2:1 pay ratio or 2.00.

so a 55% winrate, 2.00 payrate would be

0.55-[(1-0.55)/2] = 0.325% = 32.5%

so 32.5% would be the optimal bet. This is a bit tricky. If you are betting full kelly then your risk of ruin is about 22%, meaning you have a 22% of loosing everything. so what you want to do is take risk down to the level you are a little bit more then comfortable with. If you risk to much the stress will steal all of your money

kelly/1 = 22% risk of ruin
kelly/2 = 7% risk of ruin
kelly/3 = 1.02% risk of ruin
kelly/4 = 0.46% risk of ruin
kelly/5 = 0.20% risk of ruin

so if you can handle a 1% chance of loosing everything do kelly/3. This system would be 32.5/3 = 10.83% which seems alot, and it is, but not if the system is really 55%/2.00. The 1st trick is knowing how good your system is and being honest about it. The 2nd trick is never betting so much that if you loose it you panic and start trading outside of your system rules.

a real life example would be 40% winrate / 1.80 payrate

<TABLE style="WIDTH: 109pt; BORDER-COLLAPSE: collapse" cellSpacing=0 cellPadding=0 width=146 border=0><COLGROUP><COL style="WIDTH: 56pt; mso-width-source: userset; mso-width-alt: 2742" width=75><COL style="WIDTH: 53pt; mso-width-source: userset; mso-width-alt: 2596" width=71><TBODY><TR style="HEIGHT: 15pt" height=20><TD class=xl65 style="WIDTH: 56pt; HEIGHT: 15pt" width=75 height=20>Kelly/1:</TD><TD class=xl66 style="WIDTH: 53pt" align=right width=71>6.7%</TD></TR><TR style="HEIGHT: 15pt" height=20><TD class=xl65 style="HEIGHT: 15pt" height=20>Kelly/2:</TD><TD class=xl66 align=right>3.3%</TD></TR><TR style="HEIGHT: 15pt" height=20><TD class=xl65 style="HEIGHT: 15pt" height=20>Kelly/3:</TD><TD class=xl66 align=right>2.2%</TD></TR><TR style="HEIGHT: 15pt" height=20><TD class=xl65 style="HEIGHT: 15pt" height=20>Kelly/4:</TD><TD class=xl66 align=right>1.7%</TD></TR><TR style="HEIGHT: 15pt" height=20><TD class=xl65 style="HEIGHT: 15pt" height=20>Kelly/5:</TD><TD class=xl66 align=right>1.3%</TD></TR></TBODY></TABLE>

if you went with kelly/5 you would have 1 in 500 chance of loosing all your money

Mr. Trend, do you know of any such software that would work on Macs?
thanks.