Rule of thumb for position sizing

[intradaybill]

Any arbitrary percentage you consider will be some fraction of %Kelly. So what %Kelly is good for?

To be honest, I think whoever trades with %Kelly is an idiot and deserves it.

 

[smallStops]

I was wondering, has anybody modified Kelly's approach to include Martingale in it. One could use Martingale on a given trade, based solely on the profits.

 

[optionnew]

Quote from Ghost of Cutten:

If the win rate is low, risk 0.5-1%.
If the win rate is roughly evens, risk 1-2%
If the win rate is high, risk 2-4%

Any improvements on this, critiques etc?

Would it make a difference what kind of trade you are doing? Short term or long term? Stock or option? Leverage or unleveraged? As for myself, I wouldn't have a problem investing 40-50% of a long term portfolio in SPY "with out leverage", although SPY could drop 50-60% in value and bring my total drawdown to 25-30%. But I would NOT be comfartable using 25% of my trading portfolio to buy an option that has a 90% win ratio, but would loose the full 25% some times. Now both might have the exect % portfolio drawdown, but their structure differs. Does it make sense to paint everything with the same brush?

 

[optionnew]

Quote from optionnew:

Would it make a difference what kind of trade you are doing? Short term or long term? Stock or option? Leverage or unleveraged? As for myself, I wouldn't have a problem investing 40-50% of a long term portfolio in SPY "with out leverage", although SPY could drop 50-60% in value and bring my total drawdown to 25-30%. But I would NOT be comfartable using 25% of my trading portfolio to buy an option that has a 90% win ratio, but would loose the full 25% some times. Now both might have the exect % portfolio drawdown, but their structure differs. Does it make sense to paint everything with the same brush?

Anyone here has an opinion on the question above?

 

[baro-san]

Quote from optionnew:

--------------------------------------------------------------------------------
Quote from optionnew:

Would it make a difference what kind of trade you are doing? Short term or long term? Stock or option? Leverage or unleveraged? As for myself, I wouldn't have a problem investing 40-50% of a long term portfolio in SPY "with out leverage", although SPY could drop 50-60% in value and bring my total drawdown to 25-30%. But I would NOT be comfartable using 25% of my trading portfolio to buy an option that has a 90% win ratio, but would loose the full 25% some times. Now both might have the exect % portfolio drawdown, but their structure differs. Does it make sense to paint everything with the same brush?
--------------------------------------------------------------------------------


Anyone here has an opinion on the question above?

... sorry couldn't resist; we all make typing errors.

 

[optionnew]

Quote from baro-san:

... sorry couldn't resist; we all make typing errors.

That's all you have to add to discussion..?

 

[Ghost of Cutten]

Quote from optionnew:

Would it make a difference what kind of trade you are doing? Short term or long term? Stock or option? Leverage or unleveraged? As for myself, I wouldn't have a problem investing 40-50% of a long term portfolio in SPY "with out leverage", although SPY could drop 50-60% in value and bring my total drawdown to 25-30%. But I would NOT be comfartable using 25% of my trading portfolio to buy an option that has a 90% win ratio, but would loose the full 25% some times. Now both might have the exect % portfolio drawdown, but their structure differs. Does it make sense to paint everything with the same brush?

The thread is about a "Rule of thumb for position sizing", hence the title. It is not, and never purported to be, a precise calculation of optimal position size under all circumstances, timeframes, leverage conditions etc.

Some of you seem to misunderstand what a rule of thumb is, or what one is for, so I will clarify - a rule of thumb (aka a heuristic) is just a fast, simple, but useful and robust rule to aid in quick decision-making. It's something you can work out in a few seconds when time or simplicity is of the essence. Why use them? Because it is much more useful in a pressure situation to be able to get an approximately correct answer very quickly, than to get a more precise answer over a longer period of time. That is why rules of thumb exist. Once you understand this, you will realise how pointless and misguided it is to criticise one just because it isn't always precisely correct in all circumstances. All that matters is whether it boosts performance. That is why Daal's suggestion of fractional Kelly is inferior - it takes longer to calculate, and provides no performance benefits.

As for your stock market vs option scenario, at no point did I suggest risking 25%. If you read my post, you will see that the maximum risk I proposed was 2-4%. With a 90% win rate, and positive expectation, you are unlikely to run into problems risking 2-4% per option trade. As for your willingness to risk 25-30% of capital on any given position, that's merely evidence that you have a very high willingness to risk career-endangering losses. The Dow once fell 89% over 3 years, so clearly SPY can fall at least that much in future. The fact is that if you had risked this in the Nikkei in 1990, then 22 years later you would have lost 70%+ of your trading capital. So, I think your risk tolerance is insane, and you have no risk management to speak of.

 

[Ghost of Cutten]

Quote from Daal:

I mentioned 1/20 of kelly as a example to counter the flawled argument 'but you are going to go broke with kelly' I don't think there is anything special about 1/20. Could be 1/10 1/16

Kelly is not 1/20 (or 1/10 or 1/16) Kelly, therefore suggesting fractional Kelly doesn't counter the argument that you will suffer gigantic drawdowns with Kelly. In fact, the very proposal to use small fractions of Kelly, is itself an implicit condemnation of fully Kelly's absurdly high levels of risk.

Anyway, you have totally missed the point of this thread, which is to have a quick position sizing calculation that is useful, easy to calculate, and robust - not to engage in pointless intellectual masturbation in order to score points with 2 or 3 strangers on an internet message board.

 

[tenthousandmen]

Quote from abattia:

By "win rate", do you mean "% winners" (i.e. winning trades as a percentage of all trades)?

If so, then (IMO) the rule misses a consideration of average winner from each system (and profit factor), which ought to be part of your ranking of alternative systems. No? You could have a very high win rate system with a dismal profit factor...

See http://www.istockanalyst.com/financ...t-the-win-rate-profit-factor-and-payoff-ratio

Starting with equation (3) of the above link, and taking r = "avg winner"/"avg loser", you could re-arrange to get ...

avg loser (i.e. what to risk each trade?) = (win rate x avg winner)/[ProfitFactor x (1-win rate)]

[Sorry .... had to re-edit this post, as made several algebraic errors first time .... ]

How does average loser figure into how much to risk? Are you taking the inverse of average loser or something?

Also there is little talk about the Sharpe ratio formula... this takes into equation all of the factors being mentioned (win rate, profit factor, drawdown, win rate)

 

[tenthousandmen]

Quote from Ghost of Cutten:

Thanks for the link & comments.

I agree that risk tolerance is subjective, but I don't agree that treating unrealised and realised gains radically differently is at all rational. There is no difference between 'initial' capital and 'profit', at least not on closed positions. Dollar for dollar they are exactly the same thing, and treating them differently for psychological reasons is no more rational than -EV loss aversion, gambling on -EV games like roulette, or other behavioural finance oddities. So, I will call a spade a spade and say that this makes no sense.

If the goal of trading is to maximise risk-adjusted return for a given drawdown tolerance, then anything other than treating all dollars the same is going to be an inferior approach. Being reckless with 'open profit', and very cautious with 'initial capital' will result in excessive drawdowns when ahead, and inferior profitability when 'behind', or both. The optimal approach is to take the right amount of risk at all times. Not to mention, after a certain point in a profitable traders career, almost all their capital will be 'open profit'. Or, for people who use calendar years to reset their 'open'/'initial' distinction - why is a trade worth risking 3% on 31st December, and only 1% on Jan 1st? It makes no sense.

If a trader's emotions are such that they are tempted to inferior or irrational behaviour, then the trader needs to master their emotions, just like they would do when buying a market panic, or selling a super-profitable stock that everyone is bullish on.

You're right, but I think the idea of adjusting risk based on initial capital versus profit is a psychological measure, not so much a quantitative one. It also relates to how one should consider adjusting risk as they grow. One will have to do this anyways because trading becomes more difficult with more capital, even if it's hitting that $100k marker your broker says requires double the margin per contract. Thanks for the great thread btw...


Edit: Read some more of the thread (great material) - I agree, not that I no anything , that being reckless with open profit and cautious with initial is silly. It's similar to how making long term stock investments should not be based on where the price is now, but whether your initial or current hypothesis has been broken.

Also, in regards to Kelly - the formula was first used for gambling, not trading, to the best of my knowledge, so even though it fits in very well with trading since the differences in edges with card games and trading are not much, most who use the formula take the "half kelly", which is half of the calculated risk exposure. Furthermore, Kelly criterion says you should reduce risk exposure as you enter a draw down (recalculate). This of course leads to the center of all of this - never crossing below your initial capital.

Quote from ronblack:

The author of that paper is probably confused. Kelly is about bet sizing, not leverage sizing.

Yes! More info on the Kelly http://en.wikipedia.org/wiki/Kelly_criterion

http://www.jimgeary.com/poker/letters/KELLY.HTM