dartmonkey

That's the 5 second version! I think Eurobank has to be in the one-minute version. It is 10% of Fairfax's market cap, bigger than Apple as a proportion of Berkshire's market cap (about 7%, now.) Fairfax India and the other Indian holdings (Digit, Thomas Cook, ICICI Bank, etc.) are a similar size, and Atlas (shipping) is a bit smaller, but close to another 10%. A variety of fully-owned private companies, mostly Canadian, is also maybe roughly 10%, if you include Recipe, Sleep Country, AGT, Bauer, ... what else am I missing? And then there are the minority public equity stakes in the USA, worth about 5% of Fairfax's market cap: Orla, Occidental Petroleum, CVS, the dreaded BB, Cleveland Cliffs, Kennedy Wilson and Kraft-Heinz being the big ones. So, a stab at the 60-second version: A $40b insurance conglomerate with some similarities to Berkshire, growing BV at 18% for the last 40 years, trading at 1.6 times understated book value and 10 times earnings, with diversified income streams from insurance underwriting, a $39b bond portfolio, $4b stakes in Eurobank (Greek banker), $2b in Poseidon (shipping), $4b (?) in Indian investments (Fairfax India, Digit) and several billion in a portfolio of fully owned Canadian companies (restaurants, mattress retailing, agricultural products, etc.). Investment returns heavily levered by ability to invest $37b (end of 2024) of other people's money (insurance float) in addition to its own $23b in equity.

The high share price probably does serve as a barrier preventing some people from considering Fairfax as an investment. The fact that there is no listing in the USA or Europe is probably the other major barrier. The first would be easier to address, hopefully with a large split like 20:1 or 50:1. The second does involve costs and regulatory requirements which a frugal company might want to avoid. I guess the counterargument is that maintaining these 2 fairly trivial barriers improves the investor base, in that they tend to exclude people who know very little about investing principles or Americans or Europeans who would not consider investing in a foreign country. Is the company better off not to have such investors? Perhaps, and if it means the share price is a bit lower, this only makes share repurchases more profitable. But in the long run, a high share price also has advantages, for a company that is willing to not only repurchase shares when they are cheap, but also to issue new shares when they are dear, à la Henry Singleton.

It’s hard to summarize why Fairfax is a good investment when you’ve only got one minute. $40b market cap, 1.6 times book and 11 times earnings, excellent investment record with investments in India and in shipping as well as insurance… Not a bad summary, when there’s not enough time to talk about float (on a program where a guest has to explain that Kinder Morgan doesn’t make candy eggs), or Greek banking, or Canadian restaurants and mattress companies or hockey equipment maker. At least there was no mention of Blackberry. A good challenge for the board here: to write the ideal elevator pitch, one minute or less, touching on the most important points.

Great news! Anyone have any estimate of how much this might lower the interest cost on their new corporate debt, going forward?

Maybe this was not quite the final act, since $10/sh of the $70/sh price of the sale to Cleveland Cliffs was in CLF shares, for a total of 5.9m shares, and Fairfax bought another 9m shares of CLF last quarter. The current price of CLF shares is $7.35, down from the $9.26 price when the sale went through, so in your table, I think you could say that that 282% return was the return as of July 2024. The total return is surely lower now, as almost another year has gone by, the 5.9m CLF shares are down from $9.26 to $7.35, and the 9m additional shares were acquired last quarter when prices were consistently higher than today's price.

Good point - this is both why Fairfax is so much smaller than Berkshire was at the same age, and also why it is reasonable to hope that Fairfax will not grow its market cap as quickly.

Thanks for posting that, I hadn't been able to see it with the expired link. I can see that the G&M reporter doesn't quite get what float is, but that's not SafetyinNumber's fault. There is one statement about Fairfax's size, relative to Berkshire's, that is interesting to think about - it says that Fairfax is now about the size that Berkshire was in 1995. By my calculation, Berkshire had a market cap of $43.8B (USD) at the end of 1995, while Fairfax's market cap is US$37.9, so that seems about right. But if you adjust for CPI inflation, Berkshire would have been worth about $91.5b. If Fairfax grows its share price 20%/year, it would need almost 5 years to be Berkshire's 1995 equivalent size, putting Fairfax 35 years behind Berkshire. In my opinion, Berkshire has been pretty limited in terms of what it can acquire to move the needle, for about the last 10-15 years. That might suggest that Fairfax has a good runway for another 20-25 years, even if it keeps up its recent torrid growth rate.

Sure are. According to the 2019 annual report, Fairfax bought 40% of the company in 2016 at €0.94/share. Today's interim price of €2.85/share is a nice 200% return in 9 years, but most of that has been in the last 3 years, as the price was still onlyin the low €0.30s in May 2020 (not quite as bad with dividends included), and after they were still under €0.90 as recently as October 2022.

What does this even mean? If Eurobank repurchases its own shares, that would mean that all investors' percentage holdings of the company would increase, as long as the don't sell their shares. Fairfax has recently had to sell shares on the market to remain below their allowed 33.3% stake. So it would make no sense for Fairfax to also repurchase shares, alongside Eurobank. If anything, it might mean that Fairfax would be tendering its shares to Eurobank, so as not to break through the 33.3% cap. (Edit: While I was typing this, villainx and gfp said essentially the same thing...)

This would be a reasonable addition to the index’s current list of categories, and companies like Brookfield and Fairfax, currently called ‘financials’, might be called ‘multi-sector conglomerates ‘, in contradistinction to the pure financials like the banks and the pure insurers like Manulife and Intact. But as long as they keep the current categorization, they are very unlikely to drop big BCE from the overweighted telecom sector just to add even bigger Fairfax to the even more overweighted financial sector. Algonquin would be a better bet, because it is small, very close to the traditional 20 b.p. cutoff. But it has stubbornly stayed at 20 b.p. and the index committee hasn’t pulled the trigger when it was even smaller. True, Fairfax keeps getting bigger and is I think by far the biggest company not yet in the index. But a more likely scenario is that one of the current 61 members gets taken over and leaves a spot that would probably go to Fairfax, despite the worsening this would cause to sector weightings.

LULU didn't trade today...that's Friday's close. LULU's only listed on the Nasdaq as far as I know. Cheers! True enough. And LULU jumped back in front again today, with a 3% rise instead of FFH's 1%, so we're back to #24. Consolation: all-time intra-day and closing price highs again today, $2385.66 (Canadian) and $2350.61, respectively. In unrelated news, Coca-Cola Bottling (COKE) last week decided to go ahead with a plan to do a 10:1 split of its $1150 shares, to "make its stock affordable to a broad range of investors".

According to the list, those values are with FFH up 1.17% and Lulu down 3.03% today, meaning yesterday it would have been Lulu at $52.9b and FFH at $51.4b, so the switcheroo would have happened today. Looks like we have an all-time high today (2330) and likelyan all-time high close if it holds the current 2322.85 price another 10 minutes...

I too voted for other, and while I love the Irish and Greek bank investments, Allied and the other insurance expansions, including Digit and Ki, the Bangalore airport, and the aggressive total return swaps, the thing I voted for is their willingness to acknowledge mistakes and make major course corrections, particularly with the shorts. It can’t have been easy to close a losing short trade when they were most behind, with huge unrecoverable losses. I think they made the decision after Trump was elected, end of 2016, and the last 10 years would have been far worse if they hadn’t. And mine would have been far better if I had followed their lead. (I have now, mostly, except for a 12% index short on the S&P 50; I’m still clinging to my guns and religion, I guess.)

Yeah, the way this has often been framed is even worse: Buffett chose Apple vs Watsa who chose Blackberry. Admittedly not Watsa's finest hour. But I think Apple, BAC and KraftHeinz belong in the same group: they may seem have seemed unassailable once, but a lot can change, and if I were a Berkshire shareholder, I would hate the BAC and KHC holdings and I would worry a lot about Apple. Where Watsa and Buffett probably agree, is buying Apple at 10 times earnings, with a great brand and opportunities for growth, was a great idea. Holding it now at 40 times earnings, with less opportunities for significant growth, is another thing, but that is part of the Buffett favourite holding period schtick. (Forever)

I guess it's good that Watsa is not the Buffett of the North, after all. I think I'll take Meadow Foods, AGT and Recipe over Kraft, Heinz and Dairy Queen. Times change, and some of the old things we used to think were reliable may not be so reliable any more.

I think it is confusing to ask the question about whether they should be in the denominator or not - what is clearer is, what is the book value of the firm, i.e. assets less liabilities ? The treasury shares, which the company owns and has paid for with cash, are clearly assets, as you say, and should be coutned in the numerator. And anything that is counted in the numerator should probably also be counted in the denominator. One way of thinking about it is to suppose that these treasury shares were all held by a company that belonged 100% to Fairfax. Let's say Fairfax has 20m shares outstanding, and there are 1 million shares in this wholly owned company that belongs to Fairfax, earmarked for distributing to employees but not 'owned' by employees yet, i.e. not vested. Let's say Fairfax has a book value of $21b, altogether, including these treasury shares in the assets column. Now who do these assets belong to? Well, let's say Fairfax changed its policies, and sold those 1 million shares on the open market, at 1.5x book, so the company they own now has $21.5b in assets, and clearly, they will now have 21 million shares outstanding. I would say their book value is not 21.5b/21m= $1024/share, since they sold 5% of their shares were sold at a 50% premium to book value. The fact that Fairfax will be using these assets as future remuneration is irrelevant - they are an asset on the balance sheet, and the full assets of the company, including these ones, will be eventually owned by 21 million shares. So I would put them in both the numerator and the denominator.

Hmmm, it's making my brain hurt. But how's this: The cash is already gone from the balance sheet, and they have shares instead. If you don't adjust for the future shares, it seems you would have a company which appears to have less assets than it really has. So I would have naively said you should count those new shares, even if they haven't vested, or it would be 'undercounting', no? It would give you a too low book value, and that would make the P:B seem too high.

I (and probably many others) would be curious to hear your reasoning. For most companies, it DOES seem like the right approach, doesn't it? These are shares which will essentially be divvying up the earnings that are coming, not just the ones that are already issued. In Fairfax's case, there are shares in treasury that are destined to be used as share compensation, but which have not yet been assigned, so these should not be used. Is it the general principle of using a diluted count that you oppose, or something specific about the Fairfax calculation of the diluted count? TIA,

The WEF 3 is for $1.4b CAD, right? I notice in Fairfax's annual report they already had $218m (USD) marked as being with that fund, but they may well have added more. Have the details of the subscription been publicly announced?

Oh, right you are. $75m from Fairfax, not $28m, which is from a different investor. But your conclusion is still right: they previously owned 23% of the company, so they will be (slightly) reducing their ownership percentage.

Cliffs was worth about $7.2b when they bought Stelco for $2.5b, with very bad timing, as about a third of Cliff's assets are now on the wrong side of the border wrt tariffs. Going from 6m shares of Cliffs (after selling the Stelco business to Cliffs for about 90% cash and 10% shares) to 15m shares now is a small bet, $122m compared to the $2.5b they got for selling Stelco, but the timing is nice. And maybe it also means that Fairfax doesn't just stand for fair and friendly acquisitions, but also fair and friendly divestitures ?

Fairfax is subscribing for $28m of the total $350m, i.e. 8% of the new shares. They had a 23% ownership position at the end of 2024, so yes, their stake will now increase but their % ownership will decrease.

Oy vey, that must be it. Actually, while riding my bike, I thought of one more major similarity and one more major difference. Similarities: 6) Centralized capital allocation, massively decentralized operations Differences: 6) Capital return. Both Buffett and Watsa have expressed admiration for Henry Singleton, but Buffett has practically never paid a dividend, and only got on the buyback bus in the last 10 years or so, so if Fairfax is at all aggressive with repurchases, and maintains or increases the current 1% dividend, it can significantly postpone the size problem that Berkshire is now stuck with. Also, apart from starting 21 years later than Berkshire, Fairfax started smaller and gives back capital, so it is still much smaller than Berkshire was 21 years ago. In May 2004, Berkshire had a market cap of $142.8b, or $243b adjusted for inflation, compared to Fairfax's $35b today. So Berkshire still has another 25-30 years before it reaches Berkshire's current size, if it maintains 15% annual increases, and potentially much longer if it keeps returning significant amounts of capital.

What are the major differences and similarities? For me, the biggest difference is the importance of float: crucial for Fairfax, useful but not necessary for Berkshire. My take is that the 5 biggest similarities are: 1 ) Run by people with a long-term focus and who have been with the company for a long time 2) Insurance company focus with float (from the beginning with Fairfax, only gradually with Berkshire as it got bigger) 3) Value investors, capital and float invested in companies with good long-term return prospects, both have about 19-20% annual returns, Berkshire over 60 years (October 1964-2024) Fairfax over about 39 years (September 1985-2024). 4) CEO and top managers: Low pay, no stock options, honesty and integrity, good communicators, no guidance 5) Circle of competence, not interested in tech (exceptions being the Apple homerun and the Blackberry strikeout) The 5 biggest differences: 1) Fairfax is mainly an insurer, with some big equity investments; Berkshire is mainly an investor, with some big insurers. So Fairfax has much more float leverage: float represents 160% of Fairfax's book value, 26% of Berkshire's book value 2) Fairfax buys and sells, Berkshire only buys, and holds forever; I think this is an advantage for Fairfax; it may have been an advantage for Buffett acquiring private assets, but doesn't seem to be any more. 3) Fairfax makes more macro calls, has shorted, in general swings at pitches that Berkshire would not swing at (Resolute, Blackberry, Farmer's Edge...). On the other hand, Fairfax has a much larger percentage of its assets invested in bonds, because of the size of its float and the need to meet insurance regulation requirements. 4) Fairfax is much smaller, $35b instead of $1106b, i.e. Berkshire's size about 20 years ago. 5) Berkshire is transitioning to a new CEO, Blackberry may have another 10-20 years with Watsa I would be curious to see how others feel about the Fairfax/Berkshire comparison, and whether their lists are different.

It would only be a bug if they had somehow changed the degree to which their reserving was conservative, and then current years's CR would be reflecting past virtues and not necessarily good underwriting going forward. I'm not at all saying that this is the case with Fairfax (I have no idea), but it would seem helpful to also have a revised version of combined ratio history, where they DID go back and revise history, in the sense of better reflecting what really happened rather than what they thought was true when they wrote the first draft. As I write this, the word 'loss triangle' comes to mind and I vaguely remember they do do this, in their annual report. On p.136, they have a table that indicates that prior year reserve development has been favourable for 2020-2024, with gains of 3.3%, 2.2%, 0.9%, 1.4% and 2.4%, for 2020 to 2024 respectively, although obviously, the most recent years haven't had long to mature. But I can't find the full loss triangle for previous years. I guess the question I would really like to answer is, to what extent do recent favourable CR figures just represent previous over-reserving, as opposed to good current underwriting.