How Much Are You Risking On Each of Your Bets?

[thefairarbiter]

I'm getting into option strategies and trying to develop a risk model for the capital i have (about 4k). My platform is IB TWS.

The risk/reward of options strategies can be complicated when you start factoring in probabilities and integrating risk over a series of trades as opposed to just a single trade.

How much, in relative terms to your account, do you typically risk on option contracts at a given time? (In other words, how much of your account is at risk at once, integrated over all your options positions)

What is the best way of calculating the risk of, say, a short butterfly call, considering maximum loss is unlikely, but probability of any loss is significantly likely? Is their a more effective way of calculating this risk than just simply integrating over the loss zone, assuming a normal distribution of underlying price at the expiration date?

How do you maintain a risk measurement of the trade after you've entered it, and the ongoing movements of the underlying change the probabilities of PnL?

 

[SunTrader]

Risk if of an event happening or not. Not a dollar amount.

Capital employed is a better term ... that nobody uses. 2% or less in my case.

Same as R/R is said to be "risk"/reward at 2 or 3 to 1 typically, even though they don't mean "risking" 2 or 3 to gain 1 but the other way around. Which of course makes it reward/risk.

But once something sticks ... it sticks.

 

[murray t turtle]

Risk if of an event happening or not. Not a dollar amount.

Capital employed is a better term ... that nobody uses. 2% or less in my case.

Same as R/R is said to be "risk"/reward at 2 or 3 to 1 typically, even though they don't mean "risking" 2 or 3 to gain 1 but the other way around. Which of course makes it reward/risk.

But once something sticks ... it sticks.

%%
On options?? 00.00%
I did some of that years earlier.................................................................More cash + 2x, some 3x TFs....................................................................Less = more.
7 or 8 scale positions on average for what i want to risk
Less on inverse/bear trends.

 

[yana hansen]

i don't recommend risking more than 2%

 

[Specterx]

Risk per trade for me is 0.5% to 1.5% (for rare fat pitches). 1.5% to 3% max on correlated positions.

 

[Magic]

FWIW I risk over 10% max theoretical loss on some trades for sure. Anyone who is even unleveraged long on the market is risking more than 10% over a single time-step.

If you are trying to compound your capital; the variance of your PnL distribution is going to create a drag on your geometric mean. Also the skewness of the PnL distribution will have a lesser effect as well. (Negative skewness drags the return.) There's math for all this.

If you trade limited risk with options you are at least capping your variance and your skewness at some point, and stopping an outlier scenario where your PnL volatility overpowers the edge of your strategy and causes your wealth creation itself to have negative expectancy.

Optimal risk is always going to be a function of your edge and your personal utility curve. If you haven't read about the Kelly criterion yet, start there. Many have further developed the concept of maximizing mean return from there. If your only goal is maximizing mean terminal wealth, there are bets you should be risking 50% of your capital on. For example 60% chance of making 3:1 and 40% chance of losing the whole bet scores above .5.

However, that feels pretty insane when you are trading with significant capital, and also the odds of a trade aren't known; they're estimated. Overbetting has harsh consequences for capital growth. So many people try to calculate Kelly and then back it down by some multiple. Your chance of having a positive capital path over shorter timeframes also goes up as you back down variance, at the cost of lowering long-term geometric mean. Personal utility and circumstance comes in a lot here.

This is a pretty complex topic but a very important aspect of trading and investing in my opinion. Having some positive expectancy is necessary but risk control is the mechanism you ultimately use to leverage it into wealth generation whether you know it or not.

 

[zghorner]

The problem i have found with options are the huge price swings...I usually trade 6 month out contracts and still if one was to put a SL @ 5% even 10% i think you would get stopped out of nearly every trade. For that reason i think your position sizing should be much smaller than if you were trading the underlying. I personally have started pyramiding into my positions, starting with one contract to prove my bet as right or wrong then going from there...I also only go long contracts with no hedging strategies and you really have to program for a lot of failure this way.

also full disclosure, up to this point i have really sucked at trading and have made every common mistake out there...this is my first attempt at developing a system and actually employing discipline to follow the rules. Before this I was only gambling with options hoping for a WSB style dream payout. So take my recommendations lightly as there are some really smart guys on here that give better advice.

also, never sell options unless you know exactly WTF you are doing...i dont give a shit what tasty trade tells you.

 

[thefairarbiter]

@Magic thank you

If you are trying to compound your capital; the variance of your PnL distribution is going to create a drag on your geometric mean. Also the skewness of the PnL distribution will have a lesser effect as well. (Negative skewness drags the return.) There's math for all this.

Do you have a recommendation for reading about these maths? Obviously I could just grab a statistics textbook and start reading about stochastic processes, but is there a book out there that discusses these subjects from a financial engineering perspective?

This is a pretty complex topic but a very important aspect of trading and investing in my opinion. Having some positive expectancy is necessary but risk control is the mechanism you ultimately use to leverage it into wealth generation whether you know it or not.

What are your favorite books for this topic? I already read Aaron Brown's Red Blooded Risk and I could use some more recommendations.

 

[thefairarbiter]

also full disclosure, up to this point i have really sucked at trading and have made every common mistake out there...

My opinion is this is for the better, all things considered. Losing money is the best teacher. But you have to size your bets such that one loss, or even a string of losses, won't wreck you. If you're still in the game, you've probably done that.

this is my first attempt at developing a system and actually employing discipline to follow the rules

Pretty much in the same boat myself. I was once way up with options.. no longer. I want to develop a strategy that has a positive expectation to improve my consistency.

also, never sell options unless you know exactly WTF you are doing...i dont give a shit what tasty trade tells you.

IMHO, if these are uncovered writes, yes. If you are selling options within a combo of options, you really just need to know a few extra details about how to properly close the trade.

 

[Magic]

I think "The Kelly Capital Growth Investment Criterion" is a great resource. One of the editors is Ed Thorpe who was one of the pioneers of this research and also a very successful practitioner for many years. It's just a huge compilation of research papers on bet sizing and capital growth all the way from Bernoulli to recent times.

Otherwise you can probably source a lot of the concepts from various miscellaneous blog posts, and I've heard a few podcasts going over the concept from time to time. Euan Sinclair goes over some Kelly math in his books, as did Aaron Brown as you mentioned. Ralph Vince wrote a few books over the years and kind of developed his own way of thinking about the problem. I've only read one of them. You can look them up and peruse a few reviews to see if that might be helpful for you.