Liberty

There's some discussion of the letter here: http://brooklyninvestor.blogspot.ca/2014/04/jpm-annual-report-2013.html

Can you elaborate on why you think it "unfairly penalizes Fairfax"? You don't feel 15-20 years is multiple cycles? If they had not hedged and had done great for the past 5 years, I don't think people would say that right now is not representative, just like they weren't saying that after the CDS win, and just like nobody is saying that now is not a good time to evaluate the performance of BRK and MKL, whether we're at a top or not. Another question we could ask is how fast book value would have to grow over the next few years to bring up the average BV growth when you include the past 15 years. They say they're aiming for 15%, and I know they say results will be lumpy, but you'd need many years of 30% growth to compensate for those who bought 15 years ago..

Maybe. It depends if I think NAV understates IV. You don't think Berkshire was worth more than book value a few decades ago? I never disagreed with that. My main point is: "it's incorrect to automatically assume that issuing stock above book is not dilutive for shareholders."

My understanding is that they believe that IV parallels book more or less, so it's the most useful metric to use as a benchmark, but it doesn't mean that they think that Fairfax's IV is literally exactly 1.0x book (ie. if they really can compound at 15% or more while protecting the downside, it would be crazy to say that IV was 1.0x, especially in a low interest environment -- but these days, maybe it is...). Before Berkshire had a significant portion of its earnings from operating businesses, its IV was significantly higher than book because of the quality of its insurance and investing. Like Markel now, despite Ventures still being small, Tom Gaynor says he thinks MKL is worth 1.5-2.0 book. It's also incorrect to say that Fairfax doesn't have "any such businesses" (operating), though they are still small. I'm not saying that FFH's IV is much higher than book - I said I didn't know - just that it's incorrect to automatically assume that issuing stock above book is not dilutive for shareholders. For example, if you estimate that IV is 1.3 book, then issuing at 1.25 book reduces per share value slightly.

This does not change the fact that paying dividends leads to sub-optimal capital allocation at Fairfax. Given his net worth, I figure Prem can sell a few shares every year to pay for his lifestyle and this should not change his control of the company. Buffett has been giving away 5% of his stock (he converts them to B shares first) to charities for several years now, without any effect on his ability to control Berkshire. I am sure Prem can do something similar w/o forcing a penalty of additional share issuance on all shareholders. Doesn't seem to be about control since he owns tons of multiple-vote shares iirc. I agree it seems very sub-optimal. I wouldn't want to see Berkshire or Markel pay such a high percentage of book in dividends.

I believe Prem said that he does it instead of paying himself a higher salary, so that all shareholders are treated equally (and he doesn't have to sell shares).

How do you know that, though? Someone could have written the same thing a couple years ago. What if the market goes up another 30% in the next few years and the real economy picks up a lot? That's the problem with trying to time macro. Over the past 100 years, there were always reasons to think the sky was about to fall, yet our economy mostly keeps driving forward. How many points was the dow jones 100 years ago? Markel and Berkshire will participate fully in a good economy, but they'll also do well if things go to hell. If things go well, FFH will create value more slowly than they would otherwise because the hedges are like a huge weight tied to their ankles, but if things go to hell, they'll just rewind the tape a bit and get back some of the money they lost in the past few years. On the long-term net, I don't think they'll come out ahead unless there's something just as bad or worse than 2008 that happens soon, in which case their other equity holdings would probably suffer more than the average company since they tend to be somewhat distressed situations (how would Blackberry do in a collapse? probably worse than Wells Fargo and Johnson & Johnson, I'd guess). I would rather have seen them reduce their leverage or raise minimum cash to 2b at the holdco level or something like that. That reduces risk, but it's less directional, so that if things don't go the way you think, you still move in the right direction. Gio, what do you think of FFH's performance if you remove the first few years when they were really small? It looks like both BRK and MKL have done better for quite a long time despite not having successfully bet on CDS in the GFC and such. Again, I think FFH could just reduce its leverage a bit and it would make them have to be a lot less paranoid about macro, which is never a position you want to be in anywyay...

They did, though, they issued 1 million shares in November 2013 at Cdn$431/sh. I had totally missed this until AZ pointed it out. This dilutes shareholders and raises cash, but actually increases book value since the stock issuance was done at a significant premium to book. I have mixed feelings about this, on one hand if they really are cash strapped because their "hedges" are moving against them I think they need to be be a little more clear on what risks these hedges pose to shareholders. On the other hand quite frankly I am surprised that Fairfax stock is continuing to trade so high, it's at around 1.3x book now. When was the last time you remember it trading at such a premium? From that perspective Watsa may as well issue as many shares as he can since he's raising cash with overvalued stock and increasing book value. What matters is not book value, though, it's IV. If Fairfax considers that they are worth more than BV (in the same way that Buffett considers that BRK is worth substantially more than BV), then issuing new stock above BV could still be dilutive if it is below IV. I don't know if that's the case now, though. I find Fairfax's IV very hard to estimate these days.

That was in 2011, though. http://investor.bankofamerica.com/phoenix.zhtml?c=71595&p=irol-newsArticle&ID=1600359#fbid=S7hVyiigUpF

Of course. That hasn't stopped hundreds of millions of them from moving to cities in the past decade, and the government has recently started to loosen the hukou registration rules iirc.

Thanks 50cent and AZ Value.

Thanks!

http://www.economist.com/news/china/21599806-our-asia-economics-editor-takes-his-leave-less-worried-many-his-peers-about-frailties I thought this article by the economist's China correspondent (his last article from there) made a few points that we see too rarely in all this talk of empty cities and bubbles. I'm not saying he's right or not (who really knows about macro, right? I certainly don't use it to make investing decisions, but it can still be fun to learn about), but I think it's an interesting point of view. china-The_economy__On_cloud_nine_trillion___The_Economist.pdf

That's what I'm thinking. This kind of stuff hasn't been around that long (there was a legal change around 2006 that made a lot of this possible), and probably won't be around that long now that there's more of a spotlight on it. Note that this is different from the legit kind of HFT, which even IEX wants to encourage. You're a market maker or trying to react really fast to information? That's fine for the long term. But exchanges and banks (dark pools, etc) and HFT working hand-in-hand to create a rigged game, that's probably going to see the pendulum swing in the other direction sooner than later.

That is funny. Looking forward to reading the book. It is worth it. By about 2/3 through it, it becomes very clear what the problem with HFT is, where they describe the HFT strategies but also how exchanges works behind the scenes, will all kinds of hidden staircases and trapdoors that most people don't know exist (brokers selling their own customers' info to HFT firms so they can be front-run; 150 different order types, some described by 20 pages of inscrutable legalese, with no obvious purpose except to allow a HFTer to not do the trade that he publicly appears to want to do or to get a kickback without providing the liquidity that the kickback is supposed to incentivize? yeah, that's just random luck that all that stuff is there and that the exchanges make most of their profit from HFT..).

I thought this anecdote from the book about IEX was pretty funny. They called their exchange Investors Exchange. They shortened it to IEX because when they registered the website, they realized there was a problem.. See for yourself: www.investorsexchange.com ;D

I've looked at it briefly and it was interesting. Seemed like a solid conglomerate with good capital allocation operating in decent niches, but I haven't yet dug deep enough to really have a solid opinion.

I'm enjoying it too so far. Now looking back on the media coverage of it, it's pretty obvious that none of the media people had actually an advance copy to read, they just had an intern do a quick search through the book for certain keywords and asked questions based on the paragraph surrounding whatever that was. Well, they probably always do that...

Thanks for pointing them out, I've started reading their public letters and they're interesting so far.

http://i.imgur.com/dmev4A4.jpg If true that order info is sold before execution, that's pretty slimy.

Thanks, I snagged it!

If I could I would go the Steve Jobs route and wear the same thing every day :) I've always liked the quality of American Apparel basic t-shirt (2001 model, iirc). They are usually $21, but they went on sale for $8 a few days ago so I ordered 5 black ones. Not sure I ever would pay $50+ for a t-shirt... http://store.americanapparel.ca/2001.html?cid=198-313

I think it depends on what your product is. Both methods work best with different things. Releasing early and often and correcting bugs and adding features on the go works great for software and services, even better if ad-supported and people. But when you're selling hardware, it better be great because you can't easily go back to fix it and people paid you good money, so once you've delivered crap to your customers, they stuck with it and won't forget it for a while... Since Apple builds the whole thing (hardware-software-services), they can't afford to be in perpetual beta like Google.

Thanks for the list Sanjeev!

I wish you the best of luck. I know that I couldn't invest like that, but if it works for you, that's cool.