wknecht

I think on BOI they have a gain around 170%, and the EUR has weakened around 5% against USD. They bought at 0.10 EUR, and currently it's at 0.27 EUR. So I think the initial investment of $387mm is worth around $1b for a gain of around $30 per Fairfax share.

I feel like this type of policy would tend to be incredibly sticky, and reversing it would be even more difficult/painful and political than QE.

Bizarre, another 0.1% investment. I don't really understand all of these tiny positions. Trying to get the lay of the land / make connections in Greece in this case?

Random coincidence, but I ran into McCartney in times square today. Guess he was performing. Missed that part, but got a shot of him pulling away.

How much of Prem's time do you suppose this situation is consuming?

Yes I think that is pretty much right. Though I'm not sure what you mean by "our investments which contain the collateral." The $156mm is the gross value of the contracts in an asset position, not the net value of the swaps. I'm not sure if these are broker quotes or if there is a counterparty credit risk adjustment on top of the trivial part of marking the P&L. Minor point, but I guess I would note also that the $1b in collateral is mostly, but not soley attributable to the equity swaps.

My understanding: These are OTC swaps. To enter into the swaps, Fairfax was required to post "initial margin" of $1b. This is collateral to cover P&L between reset dates of the swaps in the case that Fairfax defaults and can't pay. The swaps reset/settle on some periodic basis. Meaning, periodically (a) Fairfax sends the P&L to the bank, and (b) the bank sends LIBOR plus some spread on a notional amount to Fairfax. After this settlement, the contract has 0 balance sheet value. It's similar to an exchange traded futures contract that settles daily after which a "new" contract is effectively created at 0 value. Except the settlement is less frequent. Fairfax has since lost $2b. So assuming this $2b is net (including interest accrual to Fairfax), yes Fairfax would have sent the $2b to the counterparty (+/- a relatively small amount of additional collateral to cover P&L since the last reset). If they unwind the contracts in a somewhat orderly fashion they should get most of that $1b back. Some would probably go to cover the P&L since the last reset. I believe the $132mm is what they have lost since the last reset date on certain contracts (emphasis on certain because other contracts are in asset positions). They have not paid this amount to counterparties yet, but appear to have posted a small amount of additional collateral.

Thanks, this is interesting. I'm curious though, has anyone seen specifics of how these principles and things are translated into their firms business processes? Meaning how their decisions are reached, how they divide roles/responsibilities, is capital centrally deployed or is this delegated selectively, what is the process they use to periodically monitor their investments etc. For example, I remember in reading in Snowball that at Graham-Newman, Buffett and other analysts filled out forms of some kind which they presumably passed to Graham and/or the Newmans who made the decision based on unanimous (or majority?) agreement.

Think these are the correct figures at June 30: 12% of the equity portfolio (541mm/4.7b), and 2% of the overall portfolio (541mm/24b). Still about 7% of BVPS, so will sting a bit if it heads towards zero. That could all change soon though if a deal of some kind is made.

Has anyone come across descriptions or interviews where some of the greats discuss the specifics of their firm's investment process that they could point me to? I'm thinking along the lines of Graham-Newman, Baupost, Hamblin-Watsa, Oaktree, Tweedy Browne etc. I know there's a little about Graham-Newman in Snowball, and HW on this board, but was curious what else is out there.

Currently in Brooklyn, NY. Mainly grew up in northern Virginia.

I would argue that some of the bond losses are indeed only a paper losses. Reason being, they're offset by insurance liabilities whose balance sheet value is not a function of interest rates (i.e., Fairfax's reserves are established on an undiscounted basis). What that portion is I'm not sure. It would be interesting to see how much of their bond portfolio is strictly part of an asset-liability management program and what portion is for investment purposes. Unless I'm mistaken, I don't think this information is disclosed. There is though an interesting table on page 69 of the interim report with an estimated payment schedule. $6.0b of the $19.0b of the loss and LAE provision are not expected to be paid for 5 years or more. This compares to $5.3b of the $9.6b in bonds with maturities of 5 years or greater. On the other hand, some of the gains were paper gains. Which is basically your original point.

I would argue that some of the bond losses are indeed only a paper losses. Reason being, they're offset by insurance liabilities whose balance sheet value is not a function of interest rates (i.e., Fairfax's reserves are established on an undiscounted basis). What that portion is I'm not sure. It would be interesting to see how much of their bond portfolio is strictly part of an asset-liability management program and what portion is for investment purposes. Unless I'm mistaken, I don't think this information is disclosed. There is though an interesting table on page 69 of the interim report with an estimated payment schedule. $6.0b of the $19.0b of the loss and LAE provision are not expected to be paid for 5 years or more. This compares to $5.3b of the $9.6b in bonds with maturities of 5 years or greater.

It's 360, the dividend was paid in Q1 and is now reflected in BV.

Aside from the potential vindication, what are the implications of all this from Fairfax's perspective? With the negative reputational impacts, and possibility for additional information to come to light, I would imagine a slight uptick in the probability that Fairfax reaches a settlement or favorable court decision. I'm not very familiar with the details or legal process, so am not sure what to realistically expect, if anything.

Consider a 1 year pure discount note, par value 100. The risk free rate is 10%. Assume that note pays you on time. The IV is 90.90- and that doesn't change from Day 1 to Day 365. And The Second Coming happening on Day 366 doesn't change the fact that IV was 90.90. I must give credit to Eric. He is explaining this concept so well. I see nothing special about day 1. Why wouldn't the IV change on day 2? Given an appropriate discount rate (separate discussion), present value is a function of time.

Intrinsic value is the number that if you were all knowing about the future, and could predict all the cash that a business would give you between now and judgement day, discounted at the proper discount rate- that number is what intrinsic value is. - Buffett speaking to UGA students Intrinsic value is the number that if you were all knowing about the future, and could predict all the cash that a crap table wager would give you between now and judgement day, discounted at the proper discount rate- that number is what intrinsic value is. +1 The key word: omniscient. Advance knowledge of all future events and risk free interest rates will produce a singular real IV. The idea that an omniscient being is going to engage in a range of potential outcomes or probabilities is ludicrous. Why do you think the omniscient being will spend any time predicting what he already knows? Maybe this concept is interesting to discuss from a philosophical point of view, but I'm afraid I don't see any practical use for decision making. Do you? Perhaps to look back for measurement, but even then you run the high risk fooling yourself. You could easily convince yourself that you made a good decision when in fact you made a bad one (paying $0.75 for the coin flip), or vice versa.

Yeah me too (that's why I thought of it ;D). Then I realized it was much more nuanced than I intended. When we think of a business selling at IV, wouldn't we still expect to see a profit by purchasing the company at IV? Why would you/what investor would pay the expected cash flow? You would always require a profit. Therefore, the IV needs to exclude $.50. Now, what is the right amount to pay? There is a large variance in the outcomes, so what price is appropriate to subject your capital to the variance? The Kelly Criterion comes into play here. The larger your bankroll, the closer to $.50 you should pay. The smaller your bankroll, you should pay closer to $.00. So here the value depends. I am not sure how this applies to share structures though so that's why I said we might be getting off topic. wknecht, The real IV of the coin toss is either 0 or it's $1.00. There is no other possible outcome. (hey, you asked how IV could be anything other than 50 cents ;)) You either get $0 if it's heads, or $1.00 if it's tails. There is absolutely no chance in hell that IV could be 50 cents. ironically though, estimated IV is just like you say, .50c because we have no other tool than predictive powers (and we have a lot of confidence in the equal heads/tails weightings). So we have no hope but to assign an estimate that we know for certain will not be correct in one iteration of the coin toss. In estimation efforts, we do our best. . I'm quite sure we're saying the same thing. Just quibbling about arbitrary words we use to define things. But just in case, my bid is a generous $0.25 and ask is $0.75.

Yeah me too (that's why I thought of it ;D). Then I realized it was much more nuanced than I intended. When we think of a business selling at IV, wouldn't we still expect to see a profit by purchasing the company at IV? Why would you/what investor would pay the expected cash flow? You would always require a profit. Therefore, the IV needs to exclude $.50. Now, what is the right amount to pay? There is a large variance in the outcomes, so what price is appropriate to subject your capital to the variance? The Kelly Criterion comes into play here. The larger your bankroll, the closer to $.50 you should pay. The smaller your bankroll, you should pay closer to $.00. So here the value depends. I am not sure how this applies to share structures though so that's why I said we might be getting off topic. Was trying to get to an agreement on value first. As you mentioned price and value are two separate questions (or in Buffett's words, "Price is what you pay, value is what you get"). Suppose we didn't pay, but were given the right to participate. Prior to the toss, what did we "get"?

IV estimates can and do change. Unless you are all knowing about the future, you have only estimates (guesses). Surely this we can agree on. How have you witnessed the movement of IV itself? You do not have perfect enough knowledge of the future to know what IV actually is, let alone any illogical change in it's value. Going back to the coin flipping example ($1 if we win, $0 if we lose). To buy this asset (ability to participate) we would pay no more than $0.50. If it turns out that we win, we now have a $1 asset. Had we paid $0.75 before the toss, we cannot claim this was a good decision on the basis of being less than the ex-post $1 "intrinsic value" (instead of $0.50 or less). So in my view the intrinsic value changed from $0.50 before the toss to $1 after the toss. I agree with racemize. I think folks are saying the same thing, just some are looking through the rear view mirror, others the windshield. This doesn't really have much to do with the main argument, but as it relates to IV: Are you sure the IV was $.50? Sorry if I got us off topic. I thought this simple example got to the heart of the point. In any case, we will lose money (on average) if we pay more than $0.50 and make money (on average) if we pay less than $0.50. (Note I'm obviously ignoring time value.) What should we base such decisions on if not mathematical expectation? As a corollary, how could IV be anything other than $0.50?

IV estimates can and do change. Unless you are all knowing about the future, you have only estimates (guesses). Surely this we can agree on. How have you witnessed the movement of IV itself? You do not have perfect enough knowledge of the future to know what IV actually is, let alone any illogical change in it's value. Going back to the coin flipping example ($1 if we win, $0 if we lose). To buy this asset (ability to participate) we would pay no more than $0.50. If it turns out that we win, we now have a $1 asset. Had we paid $0.75 before the toss, we cannot claim this was a good decision on the basis of being less than the ex-post $1 "intrinsic value" (instead of $0.50 or less). So in my view the intrinsic value changed from $0.50 before the toss to $1 after the toss. I agree with racemize. I think folks are saying the same thing, just some are looking through the rear view mirror, others the windshield.

I think Uccmal may have been referring to several events in the aggregate. A billion for the floods alone is extremely speculative. Based on Odyssey's annual report, 3% of GPW are concentrated in Canada ($82 mm). Northbridge will take a hit, but they write about a billion total. Recall that Sandy is estimated to cost Fairfax $261 mm (and Odyssey wrote at 88.5% despite this) and the industry $25-30 b. Uccmal, I'm curious how you come up with your estimate?

Some libraries have access to historical annual report databases like ProQuest (http://www.proquest.com/en-US/catalogs/databases/detail/pq_hist_annual_repts.shtml) or something similar. You could check if your local library has such a database. Or maybe could try a free trial from the link above.

It's Principle #9 in the owner's manual: http://www.berkshirehathaway.com/ownman.pdf

It's still early in the year, but with the market continuing to rise, it seems probable that Berkshire will not pass the five-year rolling capital retention test. I'm wondering folks specific thoughts on the implications of this until the time Berkshire begins to pass the test again. My thoughts have always been that once the test is not met they will: pay out 100% of whatever earnings are not used for bolt-ons (or other use at the subs), working capital, and capital to keep rating agencies/regulators happy and Buffett sleeping well at night; and continue to retain prior year's earnings not yet deployed (with money for acquisitions coming from this pool). I don't recall Buffett getting into specifics in the past, is this consistent with others expectations? It's is a little frustrating because while the market seems to be outpacing fundamentals, Berkshire's intrinsic value per share is most likely outpacing book value (with the accounting oddities he has recently discussed, repurchases above book etc). So if it were put to a vote, I would vote for them to still keep the money. Another reason to root for the market to fall (as if anyone needed one).