wknecht

I was actually going to ask you if you've bought more in this recent dip! Any idea what's causing the move? Is it just people (wrongly) reacting to how rising rates will impact their returns? I think it's a combination of the disclosure of the New York AG subpoena, rates, spreads widening in the ABS market (raises cost of capital), slowly declining origination metrics (spread, mix), and additional headlines around subprime auto and auto loans being a bubble. Not necessarily in that order. (a) I think I've been conservative in my accounting for the first three items, (b) the loan metrics are cyclical and not too concerning yet, and © I think a bubble popping (if indeed there is one) is best case scenario for CACC because they have liquidity, are disciplined, and capital would leave the market. The legal matters should obviously be closely monitored and if they start looking like Ocwen or unexpected significant skeletons arise, folks should sell. I think the shareholder lawsuits are typical ambulance chasing so I'm not really worried about that. They may actually have some merit, but I doubt any outcome would be material. I think they may have a little merit because CACC received the subpoena 9/18 and didn't disclose it until the 10Q last Friday. I was a little surprised they didn't disclose it sooner. They must have thought it was not material. So folks that purchased between that period may be angry, given the market disagreed. I think these are the boxes a court may check. For long term shareholders it shouldn't really have made a difference. Because New York is always active and 10% of CACC's originations are in New York, so I don't think it should have been a surprise. If you bought and that caused you to sell (or refrain from buying), you may not have the right expectations going in and should refrain from buying anyway (possibly folks have Ocwen in mind). Hence, I was a little surprised by the market reaction. I doubt that's how the courts think though. I purchased during that period myself, would I have liked to purchase 8% lower than I did? Yes absolutely. Obviously do your own research.

I added a little CACC.

I'm speculating, but I wouldn't be surprised if their selling of WFC was part of the plan to protect from deflation. Deflation is generally thought to be bad for credit.

Not sure if you're referring to apps or web based, but as far as apps, I started using one called MyStocks that I like quite a bit. It's good for portfolios with multiple currencies which is a complaint I've had with other options.

Insurance float results from a cash inflow whereas the PP&E DTL is only possible after a huge cash outflow (the capex). If the Capex earns a poor return then the associated float cannot somehow redeem that. Berkshire's profitable insurance float is far superior.

I think an argument could be made that adding the float back is conservative because it doesn't go far enough. The float has produced income uninterrupted for a long time (12 years if my memory serves), absent any contribution from the assets, so are arguably assets.

In your post on the other thread your argument seems to be that the float should not be added back because the assets are invested in low yielding securities. Can you clarify this? Why are the assets at all relevant to the valuation of the liabilities? Seems to me you should value them separately, and if you think the assets earn below market returns, then subtract from the assets, not the float.

I do agree but: Of course FFH’s assets are not 100% in the US. Yet, cwericb’s example shows an exact correlation between the difference in FFH USD denominated share price and FFH CAD denominated share price, and the USDCAD exchange rate. How is it so? Thank you! Gio The two shares represent the same asset. The same asset should have the same price. Hence the only difference is the exchange rate the two are quoted in.

CACC (Brett Roberts; a subprime auto lender): Fantastic track record: Have compounded EPS 20.4% since 1992 (multiple cycles), and have done even better since Brett Roberts took over (26.6% since 2001). This is in a cyclical business where a lot of people have gone broke. Very disciplined underwriting, and generally very rational and focused Strong capital allocation: have bought back 26%/43% of the stock the last 5y/10y, while maintaining a pretty conservative balance sheet Shareholder friendly with clear and transparent communication USB (Richard Davis): extremely well run bank with a fantastic track record through cycles Possibly PSG Group in South Africa (Jannie Mouton; an investment holding company): I don't own this yet, so am going out on a bit of a limb. But based on what I've seen so far, they seem like a strong candidate. They've grown their intrinsic value (management's sum of parts calculation based mainly on listed equity prices of subs) around 50% annually for 20 years. Very friendly from a shareholder interest and communication standpoint. Edit: I own CACC and USB, though I have trimmed my CACC quite a bit recently as it has risen a bit lately. I do not own PSG.

My guess is some insurance companies with long dated liabilities (neutralizing the rate risk of the product). But Petrobras is probably a B credit if they didn't have sovereign support, so tough to immunize liabilities with that.

I heard they hired this guy to take his place during his absence: http://firewallnova.com/2015/05/patriots-announce-dom-grady-will-start-during-brady-absence/

I thought Loeb was supposedly smart?

I kinda agree. If you are valuing things from a BV perspective then obviously you need to adjust the DTLs b/c you readily admit they aren't worth face. Float from insurance/DTL/etc is kinda of irrelevant when you look at project planning. Almost everyone uses some sort of cash on cash RoE as a metric for project planning and the benefits of float would be captured there. There is an opportunity cost for insurance float though. You need to have X amount of equity capital per Y amount of float in order to write business. So you might want to consider that as well. Yes agreed.

Yes I understand mechanically how DTLs work. And like I said, I get that it's not a true dollar for dollar liability from a valuation standpoint. My point was that in the case of PP&E DTL, the economics are totally different from insurance float. But articles I've read seem to cheer them as if they were the same, which is what confuses me. In you're example in order to get the 1.2mm DTL, you have to pay 10mm for the equipment. But to get 1.2mm of insurance float you don't have to pay anything. So insurance float can generate net profit, absent any underwriting profit, whereas DTLs are a negative carry (opportunity cost on the 10mm outflow to buy the asset minus interest earned on the DTL is a negative number). Obviously the equipment earns money, but this seems irrelevant from the standpoint of the economics of the float. Again, one form (insurance) produces net economic benefit all by itself, the other does not (DTL). I just see good tax planning which everyone uses and anyone would expect from Buffett.

Thanks for the 10-k analysis by the way. I found it interesting.

Could you elaborate on the BNSF DTL comment? I've read articles on this and it doesn't really make sense to me. I of course get that when valuing BRK, you must understand that the DTLs reduce book value dollar for dollar, whereas economically they are worth way less. What I don't get is the sentiment in some articles that the DTLs related to PP&E are some fabulous insurance-float-like equivalent. Suppose the PP&E earns a 0% return. This would be analogous to a 100% combined ratio on the insurance float. In the PP&R case, the cost of carry is hugely negative because we have to lay out a ton of money upfront to buy it, only to receive a small consolation from uncle Sam. In numbers, suppose the machine costs 100, and the opportunity cost is 5%. If the machine earns nothing, we're losing 5 in year one with a minor offsetting benefit for DTLs (5% on some number way less than 100). So we still have risk that the government subsidizes slightly. In the insurance case, we don't have to lay out any money upfront (though we'll need to hold a smaller amount of capital), so we still come out ahead at 100% combined ratios. Long winded way of saying these two sources of float seem totally different to me. The insurance kind produces net economic benefits all by itself, whereas the PP&E kind are just standard tax minimization methods that don't add net economic value by themselves.

No I don't think they do. The conversation stemmed from the widening difference between the price of FFH shares (priced in CAD) and FRFHF shares (priced in USD).

CAD has fallen relative to USD. Around 10% since 2014, and more than 20% since 2011.

Why not? When a company pays dividend of $1, its cash account as well as its book value immediately get reduced by $1. Yes agreed, but that doesn't mean it's the same as selling the stock at book. For that to be true, excess cash would need to be valued at the same book multiple as the rest of the equity. But cash doesn't deserve the same multiple as the rest of the capital deployed in operations. By default cash is valued at 1x, maybe higher in Watsa's case but not necessarily the same multiple as the rest of the capital that has been deployed. Hence the stock falls by the same $1, and the P/B rises.

I used to think this (a dividend is the same as selling at book), but I'm not so sure this is true. I think cash is valued at cash. I don't know that I ever really hear folks talk about price to book minus cash though. That being said, a lot of folks (including me) think cash in the hand of Buffett, Watsa, Klarman etc is worth more than the cash in their own hands, and a dividend does not get this premium.

How do you rationalize the avoidable capital destruction? If anything, I feel like a stock dividend would make more sense. Then the people that want dividends can sell in a more tax efficient manner while maintaining the same number of shares, and the folks that don't want the cash and related tax inefficiencies forced on them can do nothing. I can see multiple reasons this wouldn't happen though.

How did you get to 1.1x? Reported BV was $395 and then I added $20 from roughly $450m of unrealised gains on equity accounted stakes. Prem gave that figure and it got me to 1.3x BV when I did the maths at CAD $670 per share. But I don't think they gave a figure for unrealised gains on consolidated stakes. Did I miss something? You've got to find another $91 in book value per share to get to 1.1x! Which would make me very happy ;) I still don't really get why people would rather a rights issue than just buying more, if it is so attractive! They also paid a 10 USD dividend (went ex in mid January I believe).

Currency futures are by far the cheapest way to hedge currency risk. The imbedded interest rates are much better than you can get as a retail investor. When the contracts come due and you don't want to take delivery of the underlying currency you simply roll the contracts to a future expiration date, i.e. you buy/sell the contract due and sell/buy the contract with a later expiration date. The contract with the next expiration date is usually the most liquid. At IB it's very cheap to trade futures. Are there futures that allow you to trade smaller dollar amounts? For JPY I see contracts for 12,500,000 JPY. For EUR I see contracts for 125,000 EUR. Are there any for smaller amounts or is selling cash short at the retail rates the only option? There are mini futures available for half the full contract amounts but they are slightly less liquid (which should be irrelevant unless you are day-trading them). Still way too large for my purposes. I guess my only option is to sell cash. I hedged my JPY. So folks in the Japanese threads that are unhedged can thank me later as I've no doubt done it at the worst possible time. EUR is next. I didn't adequately assess my ability to withstand currency fluctuations before buying international stocks. I was okay with moderate fluctuations, but watching them plummet hurts a lot because I don't understand them enough to know where they're going next. I have stocks that are extremely volatile and I couldn't care less generally speaking. Doesn't impact my decision, but the currency wars seem to be escalating.

Currency futures are by far the cheapest way to hedge currency risk. The imbedded interest rates are much better than you can get as a retail investor. When the contracts come due and you don't want to take delivery of the underlying currency you simply roll the contracts to a future expiration date, i.e. you buy/sell the contract due and sell/buy the contract with a later expiration date. The contract with the next expiration date is usually the most liquid. At IB it's very cheap to trade futures. Are there futures that allow you to trade smaller dollar amounts? For JPY I see contracts for 12,500,000 JPY. For EUR I see contracts for 125,000 EUR. Are there any for smaller amounts or is selling cash short at the retail rates the only option?

Up 14.8% this year, with cash in the 20-25% range. Not terrific, but am reasonably pleased to keep up with the market.