How do you weight your holdings?

[Packer16]

Was wondering what the boards approach to security weights?

 

Packer

[argonaut]

I take into account the strength of my conviction in the upside vs downside risk against my goal of a pretty focused portfolio of at the moment 13-15 positions..some eg are large WFC and CHK of over 10% size at the time of purchase when bought a couple of years ago with lots of safety and upside. Vs newer smaller positions where I am not as sure of so the sizes are smaller eg. Perhaps 4-6% each. I'd like to find a few more 10% positions but as the portfolio grows I find I need a certain level of comfort to make this size purchase...Is this the sort of info you are looking for?

 

[LC]

I weight my portfolio by the relative percentage discount from intrinsic value.

[bz1516]

I also weight my holdings by conviction.  Over time as I gather more information and increase conviction I can increase the weighting, but never buy more than 10% of nav.

[adesigar]

My weighting is based on 

1. What is the downside protection.

2. What I feel is the upside.

 

I stay highly concentrated currently top 5 holdings are 92% of portfolio. Cant keep track of too many things.

Current weighing is 27%, 26%, 20%, 11% & 9%.

 

[Yours Truly]

a combination of conviction + gut feeling

[BG2008]

1. More concentrated on highest conviction and knowledge of the underling

2. Familiarity with the country, product, industry, also plays a factor.  Tend to tip toe into a new situation rather than make concentrated bets.

3. Market Neutral and market dependents plays a role.  Generally size investments where I am expected to receive a distribution larger (if less than 12 months)

4. Tail risk - Take a Black Swan approach.  Even if everything makes sense, if there is some sort of tail risk that will render the investment to go to zero, then size it smaller than I would like.  Examples would include highly levered companies, regulatory risk, blow up risk (rigs), warrants, options, lawsuit outcomes that are binary, etc.

5. More concentrated in PA than fund

6. Sleep well at night test - If a position sizing causes me to lose sleep at night.  I tend to trim it down.  There's a threshold where even the best ideas can make good investors nervous.  Have to admit we are not all Buffet and Munger.  #2 and #4 should be considered in situations like this. 

 

   

 

[ERICOPOLY]

I think weightings is more complicated than it looks on the surface.

 

You could have a 10% weighting in Sears, or a 26% weighting in BAC.  For the same amount of capital at risk.

 

Reason?  The at-the-money $35 strike SHLD put trades at 26% of strike, and the $17 strike at-the-money BAC put trades at roughly 10% of strike.

 

I think without looking at the options, it can be a bit misleading to announce what your limitations are for weightings.  It is much more nuanced than it looks.

 

Saying you have a 10% weighting limitation doesn't take into account reality.  It doesn't really make sense.

 

I don't understand fund managers who claim they have this limit of 5% to holdings, or 10% to holdings.  You can see that 10% in one holding equates to 26% in another.  Get it?

 

[bz1516]

I think there is a lot more art, or maybe a better word is feel, than science to weightings.  Certainly taking implied volatility into account is one way to risk adjust weightings to make limits more representative of actual risk. 

 

However though I might be very comfortable with a 10% weighting in SHLD I would not be at all comfortable with a 26% weighting in BAC.  While implied volatility is one measure of risk it is not a measure of all risk, especially longer term tail risk.  Implied volatility is a function of day to day market movements not what may happen 4 months from now.

[ERICOPOLY]

I think there is a lot more art, or maybe a better word is feel, than science to weightings.  Certainly taking implied volatility into account is one way to risk adjust weightings to make limits more representative of actual risk. 

 

However though I might be very comfortable with a 10% weighting in SHLD I would not be at all comfortable with a 26% weighting in BAC.  While implied volatility is one measure of risk it is not a measure of all risk, especially longer term tail risk.  Implied volatility is a function of day to day market movements not what may happen 4 months from now.

 

You can write the at-the-money SHLD put and use the premium/proceeds to purchase the at-the-money BAC put (thus hedging your BAC common).

 

You now have 10% downside risk in SHLD, 0% downside risk in BAC, and 26% upside concentration in BAC.

 

That's what I meant.

 

So if you can accept a 10% downside in BAC, and a 10% downside in SHLD, then you can go 36% into the BAC common and only have a 10% downside in BAC (plus a 10% downside in SHLD).

 

You are really only 20% exposed to the downside (10% BAC + 10% SHLD = 20%), but you have 36% allocated to BAC. 

 

So do you call this 10% position weighting in BAC, or 36%?  Are you 80% in "cash", or 64%?

 

So quite literally, volatility and risk (even tail risk) are in the real world market closely connected.  There is quite literally a market for volatility, and you can use that market to swap risk, so you can't say volatility and risk are strictly disconnected.  Only if you refuse to use the tools readily available is that true.

[ERICOPOLY]

So you can be levered 2.6x on the upside, while having no leverage on the downside.

 

Or you can be invested on the downside only 38.46%, while being invested on the upside at 100%.

 

So this is why I'm saying it makes no sense to say "I'm 50% invested, or I'm 30% invested, or I'm 100% invested" without giving a lot more information to what you are doing and what companies you are invested in.

 

After all, being 38.46% invested in high volatility companies like SHLD is exactly the same very similar to being 100% invested in companies with the implied volatility of BAC.  You can swap the risk back and forth until they are logically equivalent position sizes.

 

So implied volatility quantifies a price for risk -- then you can use that quantity of risk to "raise cash" if you use it to purchase puts on stocks that have lower implied volatility.

 

Then you are effectively in cash (partially) while perhaps not being at all in cash.

 

[constructive]

Deja vu:

 

http://www.cornerofberkshireandfairfax.ca/forum/general-discussion/weighting-of-holdings-in-portfolio/

[RichardGibbons]

You can write the at-the-money SHLD put and use the premium/proceeds to purchase the at-the-money BAC put (thus hedging your BAC common).

 

You now have 10% downside risk in SHLD, 0% downside risk in BAC, and 26% upside concentration in BAC.

 

In general, this is a bad way to think about options, since it obfuscates the situation and thereby encourages cognitive biases.  Typically, decomposing the position is much more useful than mentally co-joining positions.

 

If selling a put on SHLD seems like a good idea, it should be a good idea regardless of whether BAC options exist.  If buying puts on BAC seems like a good idea, it should be a good idea regardless of SHLD.  If one of these two legs doesn't seem like a good idea on its own, then the aggregate position could be improved by removing the leg that doesn't seem like a good idea on its own.

 

Obfuscating the individual trades is a bad idea because it can make you think, for instance, "that option was free because I got the purchase price from selling SHLD options".  The option wasn't free, because you're still spending cash that you could have kept as cash.  Decomposing the position lets you better determine whether each leg is good, without the unnecessary additional complexity of the combined position. 

 

(Unless, for instance, you believe that the two positions should be relatively correlated, like BAC and JPM.  But I'm assuming that you're using SHLD and BAC for examples because you believe that they're relatively uncorrelated.)

 

[ERICOPOLY]

You can write the at-the-money SHLD put and use the premium/proceeds to purchase the at-the-money BAC put (thus hedging your BAC common).

 

You now have 10% downside risk in SHLD, 0% downside risk in BAC, and 26% upside concentration in BAC.

 

In general, this is a bad way to think about options, since it obfuscates the situation and thereby encourages cognitive biases.  Typically, decomposing the position is much more useful than mentally co-joining positions.

 

If selling a put on SHLD seems like a good idea, it should be a good idea regardless of whether BAC options exist.  If buying puts on BAC seems like a good idea, it should be a good idea regardless of SHLD.  If one of these two legs doesn't seem like a good idea on its own, then the aggregate position could be improved by removing the leg that doesn't seem like a good idea on its own.

 

Obfuscating the individual trades is a bad idea because it can make you think, for instance, "that option was free because I got the purchase price from selling SHLD options".  The option wasn't free, because you're still spending cash that you could have kept as cash.  Decomposing the position lets you better determine whether each leg is good, without the unnecessary additional complexity of the combined position. 

 

(Unless, for instance, you believe that the two positions should be relatively correlated, like BAC and JPM.  But I'm assuming that you're using SHLD and BAC for examples because you believe that they're relatively uncorrelated.)

 

 

I don't see how we get from my post to the things you write in yours. 

 

You went off topic -- I never talked about a free option, nor did I hint at one (in fact I explicitly explained taking the option premium from SHLD puts to pay for BAC puts).  However, you did, and then you offered the helpful suggestion of decomposition, which would be a nice tip if we were talking about free options, which we are not (you might be talking about that, but not me).

 

See, you can have a rule that you'll never go over 10% for a position size.  So you might be envisioning a portfolio of 10 equally weighted stocks.  Naturally, this is in part to prevent a blowup from single-company risk.  Yet you can "cheat" your way into much larger positions by doing what I said -- taking a 36% position in BAC common and hedging it down to a 10% maximum downside using puts (and spending one of your 10% positions on writing SHLD puts).  Therefore, you can have a portfolio with much larger than 10% positions in names you have conviction on, without taking the accompanying single-company risk.  And your portfolio may wind up with larger than 100% notional upside exposure without suffering any drag from net put options decay. 

 

Anyone can just buy a bunch of calls, but there is net decay.  So if the stocks go nowhere, you can suffer permanent capital loss.  This is not the case in what I describe above because the puts you write merely net out against the puts you purchase.

 

[ERICOPOLY]

I guess it's harder to grasp than it looks to me.  So I'll make it really short: 

 

You can have more upside notional exposure than downside exposure, without any net frictional decay from options premium.  You can have upside-only positioning in some names (like BAC), which is paid for by taking downside-only positioning in other names (like SHLD).  Just swapping risk back and forth.

 

That's it!  Nothing fancy, nothing insightful, it just is what it is.

 

Yet how do you phrase your "weighting" of holdings?  Are you talking only about net downside exposure from a given name, or are you counting upside exposure? 

 

That's all.

[ERICOPOLY]

See, if you count only net downside exposure, then in December 2011 I had a 0% position in BAC.

 

And if you count only net upside exposure, then my exposure was 100%.

 

0% notional downside, and 100% notional upside.

 

What do you call that?  What is the position weighting?

 

Now, if you count the options premium as the actual position size, then it becomes misleading.  A 10% position for one person (if entirely in options) could be another investors 50% position (in the common).  Then the two people aren't communicating with each other on equal ground if one guy says he's in a 50% position and the other guy says he won't ever put more than 10% into it.  They just aren't speaking the same language.

 

So, basically, I think notional value (upside and downside) is an essential thing to mention when talking about weighting of holdings. 

 

 

 

[RichardGibbons]

Anyone can just buy a bunch of calls, but there is net decay.  So if the stocks go nowhere, you can suffer permanent capital loss.  This is not the case in what I describe above because the puts you write merely net out against the puts you purchase.

 

Yeah, this is precisely what I meant when I said that it leads to cognitive biases.  (In this case, it's called the Zero-Sum Bias.)  Basically, you're illustrating exactly what I was saying, trying to rationalize that the decay loss in one is matched by the decay profits in the other.

 

If you think the decay in the long option is a big problem, then don't buy the long option.  You'll get to keep the profits that you got from the decaying short option.

 

That said, it is reasonable to have a variety of options positions with different bullish and bearish bents in such a way so as to reduce risk in the portfolio.  It's just a mistake to group them together in your head as a single position, since that's what introduces the cognitive biases.

 

(Also I think one of the biggest novice mistakes with options is thinking that long options are bad because they decay, and short options are good because you profit from the decay.  I don't think you suffer from that issue, but it is worthwhile noting it when talking about trying to "offset option decay".)

 

Richard

 

[biaggio]

Ericopoly's level of thinking is like Ph.d quantum physics vs my level of comprehension (grade 3 math).

 

Found this good write up for the avg. folk - which will probably not give you the outstanding gains like Eric

 

http://www.beyondproxy.com/thoughts-portfolio-construction/

[ERICOPOLY]

 

Yeah, this is precisely what I meant when I said that it leads to cognitive biases.  (In this case, it's called the Zero-Sum Bias.)  Basically, you're illustrating exactly what I was saying, trying to rationalize that the decay loss in one is matched by the decay profits in the other.

 

If you think the decay in the long option is a big problem, then don't buy the long option.  You'll get to keep the profits that you got from the decaying short option.

 

 

Continuing to misrepresent what I am saying does not make it true.  You did the same thing the last time you argued with me.

 

You are effectively arguing something that implies a common stock holder is getting a "free option".  This however is not what I am arguing.  So if the assertion is wrong, then I'm not the one who is wrong (because it's not my assertion).  You are just claiming it to be mine -- you are just pulling the assertion out of thin air.

 

Example:

A person with two accounts buys straight BAC common in one account (the first of two accounts). 

He then writes a covered call in the first account where he holds the BAC common.  Then he takes the proceeds from writing the covered call and moves this cash to the second account.

 

In that second account, he has the choice of investing that cash in absolutely anything.  He can buy SHLD calls, JPM calls... anything.  But he instead purchases BAC calls.

 

 

Does this imply that he is getting a free option on BAC?  Is this what you call "Zero Sum" bias, because the decay from one will offset the decay from the other?  So do all common stock holders (who don't trade options) then implicitly suffer from what you call "Zero Sum" bias?

 

Logically, his position is no different from just being in the straight common at this point.  So does this mean that anyone with straight vanilla common is getting a "free option"?  After all, you preached about decoupling.  Once you start down that decoupling path, the cash in the second account has been decoupled from the first account.  So your argument is such that the upside in BAC is claimed to be "free".  This is a false assertion because anyone can clearly see that it's not "free" at all.

 

You are just barking up the wrong tree.

 

[gary17]

 

In my portfolio I count leaps as their true cost... ie I add strike and premium... because that's my commitment when I bought the option. if one does not do this then it distorts the exposure / weighting.

 

See, if you count only net downside exposure, then in December 2011 I had a 0% position in BAC.

 

And if you count only net upside exposure, then my exposure was 100%.

 

0% notional downside, and 100% notional upside.

 

What do you call that?  What is the position weighting?

 

Now, if you count the options premium as the actual position size, then it becomes misleading.  A 10% position for one person (if entirely in options) could be another investors 50% position (in the common).  Then the two people aren't communicating with each other on equal ground if one guy says he's in a 50% position and the other guy says he won't ever put more than 10% into it.  They just aren't speaking the same language.

 

So, basically, I think notional value (upside and downside) is an essential thing to mention when talking about weighting of holdings.

[frommi]

ERIC, correct me when i am wrong but you are saying that you can lever the upside without risk. And that is simply not possible, you can only move the tail risk to an end where you want to have it. Its not possible to remove the tail risk completly from the equotation. If that was possible, every investment company in the world would do this because this would mean free money. And Richard is right when the BAC idea stand alone is a good one, it doesn`t need additional legs. What you are doing is a pairs trade where you move your blow up risk to another end.

[Straddle]

a combination of conviction + gut feeling

 

Me too. I have done some expirements with the Kelly formula and volatility adjusting my position size, but this is hard to combine with value investing in my opinion.

[ERICOPOLY]

you can only move the tail risk to an end where you want to have it. Its not possible to remove the tail risk completly from the equotation. If that was possible, every investment company in the world would do this because this would mean free money.

 

Yep, but let me ask you something.  Where in the hell are you getting the notion that I don't understand any part of that?

 

 

And Richard is right when the BAC idea stand alone is a good one, it doesn`t need additional legs.

 

The BAC idea "standalone" is the same as one where you hold the BAC common, then write a BAC call, then buy the BAC call back.  The legs are there, implicitly.  You have that cash from writing the covered call, and from there you can choose what upside you want to hold.  Perhaps you choose to hold the BAC upside -- fair enough.  It doesn't add any new "legs" if you choose to invest it in the calls of a different company instead.  There were already legs, you've just swapped them for different legs.  There is still risk, you've merely traded your risk for some other trader's risk.  This is why I call it "swapping risk back and forth".

 

This is merely a method of diversifying the risk of a concentrated position.  You can initiate a 10% position in the common of 10 different companies.  You are familiar with that of course as a tactic of reducing single-issue risk -- it's called diversification of course (your downside in any single issue is only 10%).  So I can utilize this time-tested approach of diversifying the downside -- meanwhile selling the upside (covered calls) in all those names and using the proceeds to purchase the upside just on one name.

 

So I can achieve 100% concentration without ANY downside concentration.  It's diversified across all the names I swapped risk with.  There is no concentration on the downside.  So given that this is a topic about position sizing, I am claiming that you don't have to worry about your position sizing on the upside -- you just have to find enough diversification on the downside. 

 

You can start first by:

1)  building the diversified portfolio of common

2)  writing covered calls on each

3)  using the proceeds to purchase a concentrated position in your top pick

 

Or do it the tax-preferred way that I mentioned -- buying the concentrated position in the common, and then writing puts on other names to pay for the puts on the concentrated name.

 

 

What you are doing is a pairs trade where you move your blow up risk to another end.

 

Well, excuse me if I say... and no personal offense to you... "no SHIT Sherlock!".  That's just a colorful phrase that I find humorous and I don't mean any disrespect.

 

Didn't I make myself clear enough when I said I'm taking the upside of BAC and the downside of others (C, JPM, SHLD)?  Or perhaps it wasn't clear when I talked about the COST of non-recourse leverage?  Or that I'm deducting an expense (the depreciating puts).  And the margin interest?

 

[ERICOPOLY]

I believe if I talked about the great, inexpensive campout I had with the kids, roasting marshmallows over an open fire, you guys would first state that I don't understand the risks of a campfire, and from there expound on the cognitive bias I am suffering from.

 

Of course, the fact that I mentioned "FIRE" in the story would be overlooked and soon you would be warning me that campfires are hot, that they contain a fire, that a fire can spread, that my bias is causing me to overlook this danger, etc.. etc.. etc.. etc..

 

I'm then sitting here wondering... do these guys really believe I am totally unaware of this?  When did I tell them that I thought fires could not spread, that they were not hot, etc.. etc...  After all, I roasted something in the fire.

 

[frommi]

 

Yep, but let me ask you something.  Where in the hell are you getting the notion that I don't understand any part of that?

 

 

Nowhere. You are a great mind and it is a pleasure to understand everything you do. Thanks for being so kind to lay down this whole thing, because now i fully understand what you are doing and it starts to make sense for me.