racemize

Well, again, here is my rationale and comparison: If I'm putting $18 in and I'm comparing whether I buy the common or the warrant, I'm comparing the expected future total return of those two vehicles. I do lose 3%, because I would have bought the common, which would have paid me 3% dividend in addition to the appreciation (which is already accounted for). I do not lose 6%, as I could never have gotten it with that $18. The math above shows this, and it does not show a cost of 6% (or the full $1 dividend). I'm looking for the annualized rate that makes the TR of the common the same as the TR of the warrant. After going back and forth on this, I do not think we are measuring the same thing. Your example of exercising before the option requires $36--I don't have $36, so I don't see how it applies. The other example should correspond to my math above.

Well, I still think the extra cost is 3% and not 7%, but that is based on my comparison between common and option versus Eric's comparison to margin possibility. So, I'd say, it depends on what exactly you are calculating. See my previous example comparing the total return of the common versus the total return of the warrant.

I didn't engage because it was getting off track. What interest rate did you use? There is $18 being "borrowed" in the case of the call, but not in the case of the put. Did you omit that expense from the call? You're right, I'm more interested in the other half of the discussion. I was just thinking that since calls don't get the dividend on one side and the puts + stock do, then the puts would necessarily be more expensive than the calls. It would depend on the bond yield, so that's probably why the dividend yield didn't matter that much. Still, though, if one can get double the dividend by margining + puts, it would seem like the puts price should reflect that difference in possibility. I'll focus back on the other conversation though.

FCF isn't by definition a metric that applies to only the equity, you can also calculate FCF to the whole firm (so without subtracting interest payments). I see, so EV/FCF could be calculated as: EV / (EBITDA - maintenance capex) I guess the issue with EV/FCF (and the above) if you don't include interest payments is that the taxes will necessarily change. So do you also pull out the taxes, right?

I agree here, and it makes sense. However, to be equivalent to the call version, you would have to buy puts at $18 right? Wouldn't the puts be more expensive than the calls because of the implied dividend yield? I tried to verify that earlier, but failed. Is that just not right? I was expecting it to work out due to put/call parity. There is no cost to the put. The example, if you recall, is a $36 stock with $18 call strike price. The call is priced at $18. Thus, the embedded put in the call is free. No, this isn't terribly realistic, but I wanted to keep the discussion on just the missing dividend. So I reduced the complexity. It is assumed (apples to apples) that the put is also free on the market (if it's free for the call). Okay, not real-world, but I didn't want to get the discussion off track. Too late! :D Sure. If we did go ahead and assume a dividend, wouldn't the put cost more by the dividend yield amount than the call though? That was what I was trying to calculate with the JPM and WFC examples, but it didn't work out that way, or at least didn't seem to.

I would suspect that many people use P/FCF or P/OE (owner earnings) on the board. Or some kind of normalized amount of FCF/OE as Packer just said. Packer, with respect to EV/FCF - that seems somewhat strange since the debt is "paid for", but you are still subtracting interest payments. What is the rationale behind it?

I agree here, and it makes sense. However, to be equivalent to the call version, you would have to buy puts at $18 right? Wouldn't the puts be more expensive than the calls because of the implied dividend yield? I tried to verify that earlier, but failed. Is that just not right? I was expecting it to work out due to put/call parity. I see what you are saying, but it just doesn't make much sense to me. There was nothing that I could invest in, besides borrowing the margin and buying puts, that would have yielded 6% off of the dividends, and in that case, I still have the question above. Let's put it another way. I think I've calculated "cost of leverage" in the same way you do when there is no dividends. Doing this allows me to know how much the common has to return for the warrant to break-even, or in other words, it allows me to directly compare at what point the warrant will return the same amount as the common. In the ridiculous case of $0 cost, the break even is right when I buy it, everything else is levered upside. Let's say the price goes absolutely nowhere for a year, and a dividend of $1 is instituted. Ok, now I need to make up for the $1 in total returns in having bought the warrant instead of buying the common, right? So, the total return for the common is now 1/36 = 2.7%. In order to break even with that, I need the warrant to go up to 18*1.027 = 18.5, or just $0.50 and not the $1. So, it seems like if I "punish" the warrant by requiring a 5.4% yield or the full $1 dividend, it no longer provides the total return comparison I've been calculating. Perhaps that means we just aren't calculating the same thing?

We are thinking the same way, I just wasn't careful in how I described my example. yes, two shares of upside. two dividends missed. twice we "borrowed" $18. Thus, 1 missed dividend per $18 borrowed. Therefore, the missed dividend of 1.08 (per call) is 6% of the "borrowed" amount of $18 per call. I get that 1.08 / 18 is 6%. I just don't understand why we would think we should get the 1.08 in the first place. If we had bought $18 of shares, we would have got a 3% yield, or .54. If my choice was to buy the common unlevered or to buy the call, or some combination, nothing was going to give me a 6% yield, so why would I be missing that unattainable 6%?

Ok, I'm still on the dividend issue. Since a call misses dividends and a put doesn't (when paired with appropriate counterparts), one would expect there to be an implied dividend yield difference in cost. So I looked at JPM Jan 15 calls/puts. JPM has a 2.7% yield right now and is at 56.06. The JPM Jan 15 57.5 call is 4.5. Using (price + strike - stock price) / strike I get (4.5+57.5-56.06)/57.5 = 10.3% (using short method out of laziness comparing to JPM Jan 15 57.5 put: price is 7.24. So that cost is (7.24 - (57.5 - 56.06)) / 57.5 = 10.1%. Ok so that failed. Maybe I did something wrong. I'll try out WFC, which is currently 44.11 with a 2.7% yield. The WFC Jan 15 $45 call is 2.89. So (2.89+45-44.11)/45 = 8.4% The WFC Jan 15 $45 put is 4.85. So (4.85 - (45 - 44.11))/45 = 8.8% failed again. Oh well.

This example makes more sense to me, but may be simply because the SP and the stock is similar. It makes the most sense to me to use the dividend yield (annual dividend / stock price) as the additional cost for the dividend rather than dividing the dividend from the strike price. e.g., going back to Eric's example, there was no ability to invest $18 and get 1.08, it could only be gotten by investing $36.

The thesis again appears to be based on guarantor/non-guarantor in bankruptcy. I haven't seen anything that addresses Kraven's points earlier in this thread.

Well, I guess where I get confused is this: I could have bought the 36 dollar stock and gotten the $1.08 dividend. However, I could elect to buy two 18 dollar calls. Thus, it would seem I'm missing the $1.08 from 36, not from 18. Or, in other words, I had no other investment choice where I could get 1.08 out of 18, so I'm not sure why I would divide by 18 to get 6% instead of 36 to get 3%. I do understand you could have gotten 6% if you had margined 100%, but presumably the puts would cost 3% more than the calls, which would again be a net loss of 3% for the calls, not 6%, right?

I see--I think I'm just doing the comparison all in one go.

I think your math is off once you hit 1.51. From that number do the following: 1.52-1 = 52% raw gain needed to hit 74.95 from this point. Annualzing that number is 5.98%. Then, you can model dividends however you like. The current dividend is 0.1, so all of those missed dividends adds from 74.95 to 76.07 (this uses yield and not raw missed dividends), which is the new break-even point, giving an annualized rate of 6.2%. Alternatively, you could push all the way up to .17 (the threshold) since it will likely get to that point in the future, pushing 74.95 to 76.82 or 6.35% annualized.

Right, that makes sense if you compare your alternative to cash--I guess when I'm looking at leverage, I'm comparing it as an alternative of buying the common, so the cost of leverage has to go up, as the common necessarily goes up some amount in order for you to make money in the levered vehicle. With regard to the dividends, I only account for "missed" dividends. In the case of BAC-A warrants, you will only miss the 0.01 per quarter, since that is the threshold. Any amount of dividends above 0.01 is accounted for by the adjustment in strike/# of warrants. Thus, the lower strike price on the warrants is offset by actually receiving the dividends as a common holder. That being said, that thinking only applies when comparing warrant to common, and not warrant to cash, which is what you appear to be doing.

I suppose you'll just come back and say that with a margin + put strategy, you would have bought $20 of stock with $10 of capital, giving you $0.6 dividends on $10 of cash. Since before the dividend, you were already paying the margin cost + puts at $10, the additional 0.6 is all "free", so a warrant holder, having a similar cost of leverage as the margin cost + puts, would just be losing that 600 bp. I'm not sure how to apply that to the call strategy, since $10 of calls only misses $10 * 3% dividends...

They have almost everything the TARP warrants have (I think), except for ordinary dividend thresholds, so ordinary dividends will always cost extra. There is protection for extraordinary dividends, stock splits, etc. With regard to the dividends and accounting for missing them in the cost of leverage calculation, I'm not sure of the best way to do that. In your quote there, you indicated that a 3% dividend equates to a 6% raise in cost of leverage if the warrant strike were 50% of stock; I'm a little confused there. For example, if I put $10 in the common, I would get $0.30 cents in dividend for the 3% yield. If I put $10 in the warrants instead, I would lose only the $0.3, not $0.6. Alternatively, if I put $5 in the warrants (equivalent notional exposure to $10 common), I would only miss $0.15. How is it that the cost of leverage goes up to 6%? Edited As noted above, what I've been doing is calculating the break-even price for the warrants/common, and then simply adding the dividend yield of all of the missed dividends to that, and then annualizing that number / current stock price. Thus, you have to get to the break even point, and then get the sum of missed dividends (using yield values, not raw) on top in order to truly break-even. I've been feeling that there is something wrong with this approach, but I'm not sure what it is.

I see--I calculate it by figuring out where owning the warrants are better than owning the common and annualizing that number. For example, the point where the warrants make the same as the common is currently $22.94. Thus, my rate is $22.94/$15.56, annualized, which is currently 7.9%, or since this is at parity, it would also be ($22.94-$13.30)/$6.54, or 7.9% annualized. I also usually adjust for missed dividends, which adds be 0.01*remaining number of quarters to the $22.94 number, adjusting to 7.95% per year. I'm not totally sure if that dividend adjustment is strictly the right way to do it, but it makes some sense (those are the raw dividends that were missed). In your calculation, you are calculating how much extra your are paying now for the leverage (strike + warrant price - stock price), which makes sense. Why then divide that by the strike? Seems like you would want to divide by your cost (warrant price)? In any event, it seems like the first approach gives a slightly better indication of cost, since you could have bought the common instead of the option/warrant. Is there an advantage to the second approach I'm missing?

Hi Packer, I've got it at 0.6% for GM, but 7.95% for BAC--how do you get to 5.6%?

my wife has made a few of those types of presentations--there's some special software for it I think. I'll ask her for some names.

I liked: 32. The best investors in the world have more of an edge in psychology than in finance.

Why does that chart say "Household debt in Asia" and then list a bunch of non Asian countries? Does it mean something other than debt in those countries?

Horizon Kinetics commentary on Canadian REITs. No real mention of the danger of housing prices falling--I presume that is an issue for REITs? (I've always ignored them as they have seemed risky to me, but don't know much of the mechanics). http://www.horizonkinetics.com/docs/December_Commentary_Canadian_REIT.pdf

I can vouch for b & c. That was very, very common. I have even heard of excess of 100%. Not sure about point a. The one thing that concerns me about housing in Canada is the prevalence of 5 year mortgages. In the US, it seems that they are fully amortized over 15 or 30 years, so no interest rate risk. If interest rates ever rise in Canada, watch out. It is not just new buyers who are going to be hurting. well that certainly sounds frightening.

I like the Nature Conservancy. It's a pretty straight forward charity and I think the overhead is pretty low (or at least I've been told that by others): http://www.nature.org/ Here the charity navigator review: http://www.charitynavigator.org/index.cfm?bay=search.summary&orgid=4208#.Up5w7mSxNF8