Position Sizing & Longs/Shorts

[Rise-T 10]

Hi everyone,

 

I used to consider myself something like a 'living position sizing encylopdia' over the years, but enhancing my position sizing / real-time risk management spreadsheet to handle longs & shorts (used to be longs only), I've noticed that I am not 100% sure about the answer to the following rather simple question:

 

To make things easier, let's just talk about stocks, so there's a pretty high correlation between positions.

 

If I have a 100,000 USD portfolio and I have long positions worth 100,000 USD and short positions worth 100,000 USD.

 

Do I consider myself 0% invested then or 200%...?

 

I mean, 0% doesn't really make sense, but so doesn't 200%.

 

Any informed thoughts?

 

Thank you in advance!

[diablo272 10]

Unless the long and short positions are both on the same instrument, you would be 2:1 invested in your example. You have $100k in cash, and you're using full (I think that's full in equities retail) buying power to short the extra $100k. You've used 2x the capital you actually own, so 2:1.

 

For reference, if you were long and short the SAME instrument the same amount, you would have a net neutral position. This is because as the value of one position goes down, the value of the other would go up the same amount.

[DugDug 10]

Hi, I can tell you from direct experience in talking to many, many people over 15-16 years that the answer to your question depends on who you are talking to....unfortunately.

They measure it differently, they talk about it differently.....i dont think there is a universal language.eg; is 100% leverage = 1:1 or 2:1...

HOWEVER.....when talking to someone the best thing to do is differentiate between exposure and leverage, and talk about GROSS verses NET exposure and or leverage.

eg; long 100, short 100 gives gross exposure of 200, net exposure =0.

 

Or something along those lines.

We encountered this problem when talking to clients especially when it came time to talk about options and delta exposures, leverage and risk....(but thats a whole other story)

[Kojak 10]

If it is the risk that you are trying to manage then by going long and short you have hedged away the exposure and the risk to the extent that they are correlated. So if the scrips have a correlation of 95% by hedging your next exposure has reduced to 5%. Hedging is a strategy of portfolio managers - not speculators.

 

But if you are a speculator and leveraging one position taking directional bets across different time frames then these positions are in reality not correlated. Your exposure and risk is twofold as you could profit / lose on both legs.

 

Jose

[Cruiser 10]

  Rise-T said:

...To make things easier, let's just talk about stocks, so there's a pretty high correlation between positions....

 

If you are talking about stocks where there is a high correlation between positions then why are you both long and short at the same time?

 

Unless there is very little correlation between the positions then in all likelihood they will both be moved to a great extent by the same forces that are effecting the broad market.

 

If there isn't much correlation and one could expect different movements then you would be invested in the gross amount, or $200K in your example. If they are correlated then one would infer that your strategy of going long and short would only make sense under some sort of hedging strategy in which case your strategy would dictate the risk. Of course if you have the same luck I do (or did) there would likely be some sort of market movement that would whipsaw things quickly and stop both out in opposite directions.

[plungee 10]

What's most important is that your technically have 0% risk, not counting the cost of trading.

 

As for your 0% invested or 200%, neither is correct. There is never more than 100% of anything. !00% is all there can be. Look at it as if you were invested in stocks, long 5000 shares and short 5000 shares. You are 100% invested with 0% risk, but, in my case, I must pay 5 cents a share commission ($500) plus slippage just to go nowhere.

[sr100m 10]

Assuming you have long and short positions in different instruments - one way to think about it is - assume all your positions (both long and short) could go against you at the same time (unlikely but not beyond the realms of possibility - if someone had told you 3 years ago that Lehman and AIG would go bankrupt within weeks of each other ...... ) - how would you determine things then.

 

I am more comfortable with the portfolio heat type approach. You need to have a system with an initial stop loss to do this (although you can substitute your average historical loss or use atr to calculate volatility stop losses). The difference between your entry and the stop loss along with the size of your position allows you to estimate what percent of your equity you would loose if stopped out for each position you have. The total percent at risk among all instruments is the portfolio heat. 20% is a figure you see banded around as a max heat although depending on the performance of the system it may need to be a lot lower.

 

If you adopt this approach it does not matter whether a position is long or short - all you are concerned about is how much you could loose if it goes against you. Van Tharp is the best I have come across on the topic of position sizing (Trade your way to financial freedom book). I have no affiliation with his company - just a grateful reader.

 

Just to muddy the waters a bit further Curtis Faith in his book about the turtles says that one of the turtle "rules" addressed this - basically the rules meant that turtles were allowed to offset long and short positions against each other to a certain extent (and effectively have a higher portfolio heat than if all positions were in the same direction).

[Rise-T 10]

Hi all,

 

Sorry for replying sort of late, but I have been pretty busy the last couple of days.

 

And many thanks to all for taking the time to post!

 

1.) I guess the concept of Net & Gross exposure is what I was looking for. I now realize I was somehow trying to get those two stats into one figure - which I guess is next to impossible :crap: So realizing this was of great help for me. Also, I am a great fan of clear & systematic wording (not sure where I got that from...) - so this perfectly fits the bill. Thank you.

 

2.) Regarding being long & short in highly correlated instruments:

Firstly, my intention was to create my risk management spreadsheet as versatile as possible.

Secondly, when there are single stocks I've been watching breaking out in a constructive manner and (I want to own them in a continued market uptrend) while the general market seems stills suspect to me, what I sometimes do is initiating corresponding shorts with weak action at low risk points (i.e. rallies) to somewhat hedge the longs. When the general market finally decides its way, I slowly fade out of the shorts (or the longs) in order to get net long or short. Sure, sometimes I get faked out on both sides, but other times it works.

 

3.) Portfolio Heat etc.

I guess I've spend the better part of 5 years or so studying Van's material (after a huge gain followed by a big loss - gold stocks in 2002... In fact, that is what got me into trading vs investing in the first place - controlling risk) and I am familiar with pretty much every concept he's written about. I was also reading avidly (and sometimes participating in) his old MasterMind forums (which sadly don't exist anymore, I suspect partly because of his recent strange affinity to Service Marks... ).

On a sidenote, if you haven't read it yet, I would recommend his Definitive Guide to Position Sizing since it summarizes almost all position sizing concepts pretty well. It is a bit on the expensive side since he published it himself, but I definitely would have bought it (I don't have any affliliation with Van's company as well - besides being a client - but I kindly got the book for free, though, since I've been helping him in a - really very tiny - way with the book).

 

The measuring of exposure is just one out of many stats I've integrated into my risk management - I've tried to integrate as many as possible (at least the ones that make sense to me & you don't have to have a PHD in rocket science to calculate - as my math skills are rather questionable & I don't think it's really necessary - e. g. the material of Ralph Vince). Some stats that I use to measure risk: Open Initial Risk, Open Risk, Exposure (all of them for individual positions, for correlated groups and on a porfolio level (e. g. Open Position Risk, Open Group Risk, Open Porfolio Risk (= Portfolio Heat - see my preference for systematic wording above... ). All of them just for longs and just for shorts. All net & gross long/short. Adjusting position size for volatility. R-Multiples. Market's Money concept. Keeping positions & groups & the portfolio within defined max risk limits by peeling off units when necessary.

 

Regarding Curtis, I've followed his writings early on when he was still participating at his former forum at tradingblox.com. I bought his first book when it came out - and I agree, it is an excellent one. I try to incorporate that mentioned Turtle rule by measuring the risk of longs vs shorts.

 

Again, thank you all for your kind comments!

[Rivest 10]

I have to go with Plungee on this, you are 100% in game. Your all in.

[dangermouseb 10]

My 2p.

 

I think if the problem is framed as either 0% or 200% then it gets oversimplified. You trade your beliefs about position sizing and correlation. If you are very technical then you'd want to do some Monte-Carlo - a boot strap similar to van's approach (figure out your non-correlated R multiples and your correlated R multiples from backtest results), a permutation analysis (e.g. from Evidence Based Technical Analysis). You should also look at the sharpe ratio as well as Van's System Quality Number. The point is to be familiar to how much volatility is in your system and tuning the risk amount in relation to your capital to meet your objectives (e.g. 50% chance of making 50% per annum with a 10% chance of a 25% peak to trough drawdown).

 

Happy to provide further pointers.

 

DM