dartmonkey

I have done this before, quickly selling something for a 8% gain, because the math looks so appealing, if you annualize it. And it would make sense if you thought the shares were worth $14.50 give or take 25c, and would trade quite closely around that number, and then you buy at $14 (at least 25c beneath intrinsic value) and get an opportunity to sell them at $15 (25c above the upper limit of intrinsic value.) If all this were true, you might very well get lots of opportunities to do this kind of profitable round-trip trade. My experience, however is that our assessment of intrinsic value has a much much wider error bar around it than 50c, and Mr Market is also very undecided about exactly what that value is. When the price starts moving up, it can go a long way up, without ever revisiting the previous range, and too often I have ended up sitting on my $1 gain when I could have had a $10 gain a few months later. In the case of FIH,the price might stay around $15 for a long time, or it could move quickly to $25 which would still not be a crazy price. So my investing strategy has moved to being patient with a company that is performing well and not selling shares just because they have moved up 10% or so. Of course, personal circumstances can differ - you might sell because you need the money, or because you see some better investment opportunity, or because you are a lot less optimistic about the long-term value of a company. But I do have trouble imagining a business that I would buy at $14 because I thought it was underpriced and then selling it at $15 because I thought it had reached full value or was even overpriced. I hope you get an opportunity to buy back in again at similar prices!

Check out this recent analysis: But for thoroughness, it's hard to beat Viking's spreadsheets, but I can't find it quickly. If memory serves, he had one posted here a few months ago. Maybe someone else can post a link. As for the fee structure, it is 1.5/20 on the book value as you say, but the 20% only kicks in after a 5% hurdle. It is paid in cash or in shares, every 3 years, and after 9 1/2 years of existence, has been paid 3 times already, the first 2 times in shares, and recently in cash. The hurdle is a fixed 5%, not annualized (i.e. it is $10, $10.50, $11.00, etc., not $10, $10.50, $11.025, ...), so it is already only about a 3.4% hurdle because of the doubling of FIH's book value in the last 9 years, and in 10 more years it will have drifted down to 2.6%.

Yes, it's nice to see a company go from 8 to 19 to 56 to 73b INR, although most of this is a rebound from COVID - it was 68b in 2019-2020, before dropping to 8b. And at 241 INR, the stock price is also just back to where it was 5 years ago. Odd that the share price drop seemed to slightly precede when you might have expected it to happen - it started in June 2019, before anyone had heard of SARS-CoV-2 dropping to 160 in December, and then went as low as 22 INR! Please remind me of this next time the world has a global anxiety attack!

Your link needs a slash instead of a period, at the end: https://finance.yahoo.com/quote/GMRINFRA.NS/ It looks like they had an IPO in July 2021. I don't know how they have a price chart prior to 2021, but the price since July 2021 is up from about 29 to 92 in the 3 years since the IPO. If the value of the Bangalore airport, has followed anything like the same trajectory, we are sitting on dynamite.

Walk me through your reasoning or tell me if you think this seems right: The preferreds are given a carrying value of about $1790 on March 31st, and the equity portion (49% of Digit's shares) is worth $477m (fair value), but is carried at $147m. That's presumably how you get a carrying value of $1.94b : $1.788b for the preferred plus the $.147b carrying value of the equity. Once the preferreds are converted, Fairfax expected to own 68% of Digit. Digit has a market cap of $3.74b at today's price (340.80 INR). If Fairfax still owned 68%, that would be worth $2.54b. But there are a couple of things I am not sure about. First, why are the preferreds, if they can be converted to 29% of the company, worth $1790, whereas the shares, worth 49% of the company, only have a fair value of $477m (end of 2023) or $476m (end of March)? Second, Digit issued 41.4m new shares at the IPO, so the total market cap of Digit is higher now than it was pre-IPO. Before the IPO, they had 145.6m shares, so unless Fairfax ponied up more money to buy a proportion of those new shares, that 68% share would now only be .68*(145.6/187.0)= 52.9% of the $3.74b market cap, or $1.98b. If my reasoning is right, then Fairfax's $1790+476 = $2.27b stake is now worth $1.98b, a drop of about 13%. It's a bit higher than the carrying value of $1.79b + $0.147b = $1.937b, but that may not be adjusted, given Fairfax's controlliing stake. But I am pretty sure that parts of this analysis may be wrong.

See Watsa’s mea culpa, p.11 of the annual letter. And he left out the terrible losses from the shorts.

I don't know, but since they quote a fair value of $2265m for Digit in the annual report, they may be including the value of the full 68% stake once converted - I don't think they are ascribing a value of $4.622b to the whole 100% of Digit, it makes more sense for 68%% to be 2265 and the whole company to have a fair value of 2265/.68=3.331b. Fairfax investment Date of Initial Inv. Ownership Cost Fair Value at Dec.31, 2023 Compounded Annualized Return ... Digit Feb-17 49.0% 154 2,265 61.9% IF this is correct, they may continue assigning fair value and not consolidate, as long as they have not gone from 49% to 68%. I guess they will probably talk about this investment in the Q2 report, now that the IPO has happened, and we will get a better idea some time in late July or early August.

Brief recap of recent Digit valuation: Fairfax has this at a fair value of $2.265b on its 2023 annual report. Although Fairfax notes that it owns 49%, it also has 'securities' that, when converted, would give it a 68% stake. My understanding (someone please correct if this is not true) is that the $2.265b number is for the whole 68% stake, not just the 49% they own outright, so this would put the total fair value at $3.331b. In February 2024, Muddy Waters contended that "Since 2021, the valuations for "InsureTech" stocks and Indian unicorns have collapsed. Digit's prospects for an IPO in the near future seem minimal." They also said that "it is incontrovertible that Digit is worth far less today than where Fairfax ultimately marked it in 2021, ... we see Digit as generously being valued at ~$1.5 billion presently. We therefore adjust Fairfax's book value downward by -$1.1 billion to align Digit's carrying value with a more reasonable value." Four months later, these predictions are not doing well. The IPO, with minimal prospects of happening in the near future, is done. The market capitalization, at today's share price (337.6 INR) is $3.78b, higher than the $3.331 fair value on Fairfax's books, and considerably higher than Muddy Waters 'generous' valuation of $1.5b. I know it takes a while to process this new information, but 3 weeks after the IPO, I am eagerly awaiting Muddy Waters acknowledgement that at least this one adjustment, representing $1.1b out of their "~-$4.5 billion "conservative adjustment to book value," turns out to be unneeded. I don't know the details of the IFRS accounting for majority stakes, so it may well be that there will be no fair value adjustment in Q2, given the controlling stake, but the collapse in valuations sure hasn't played out.

I agree with Thrifty and think that the remaining 55% of earnings will either be reinvested in buybacks or else something even better. So a reasonable approach might be to plug in 75% in buybacks instead of 30%, even if we are not really predicting that level of buybacks. Equivalently, that extra 45% of earnings could be in 'Unknown Investment', with a 15% return; that unknown investment could end up being either more FFH shares or else something better. Implicitly assuming a 0% return on that 45% of earnings is unrealistically low. In my opinion, we don't want to just throw in extra 'conservatism', we want to get our best estimate of value. It would be a shame to sell or reduce the size of a good investment like Fairfax because we've made assumptions on returns that are significantly lower than what we really expect.

OK, thanks. 6.405m shares is exactly 1/9 of their previous holdings of 57.641m shares, so they took full advantage of the share offering, meaning there was no dilution for FIH shareholders. IIFL shares outstanding went up from 392.6m sh to 424.5m, so it looks like they issued 31.9m new shares altogether, FIH accounting for 20.1% of the new shares. FIH previously owned 14.7% of the shares, so given the fact that not all shareholders took advantage of the rights offer, FIH will have slightly increased their ownership from 14.7% to 15.1%. More important, they will have confirmed to the Indian investment community (including the government) that they are a reliable financial backer.

At first governments build walls to keep people out, then eventually they build walls to keep people in. That's pretty pessimistic - I don't think most countries have ever prevent people from leaving. But I'm curious, what governments are currently forcing people to stay? I can't think of a lot of examples. I don't think China prevents people from immigrating. And obviously the emigration ban that used to apply in the USSR and the Eastern Bloc has now disappeared. Even Cuba now allows people to leave, although passports are expensive. Are there any others?

Exactly, this does seem irrational. It makes sense that IIFL Finance (IIFL) should be up, but it is not clear how this has any bearing on IIFL Securities which, since the split of IIFL into 3 independent companies, all publicly traded: IIFL Finance (IIFL), IIFL Securities (IIFLSEC), and another company with the mysterious name 360 One Wam Limited (360ONE.NS) (a wealth management firm.) (A previous spinoff, 5Paisa, is also a small publicly traded FIH holding.) Could it just be confusion between the 2 tickers with IIFL in them? And isn't it odd that the one that is seemingly NOT concerned by the news of the end of the government audit (IIFLSEC) is up 20%, while the one that IS concerned (IIFL) is up 14%? Here's why it makes sense for IIFL Finance: From April: BENGALURU, April 23 (Reuters) - IIFL Finance said a special audit directed by the Reserve Bank of India (RBI) started on Tuesday, about one-and-a half month after the country's central bank barred the non-bank finance company from disbursing gold loans. In early March, the RBI ordered IIFL Finance to stop sanctioning, disbursing and selling gold loans, citing "material supervisory concerns" in its gold loan portfolio, raising liquidity concerns among its investors and lenders. Gold loans accounted for nearly a third of the company's total loan assets as of 2023-end. Since the order, IIFL has lost nearly 30% in market value. The RBI said in March that the restrictions will be reviewed after completion of a special audit and the company's rectification of the audit findings. "We are committed to extending full cooperation to the special audit team to ensure a comprehensive and thorough audit," IIFL said on Tuesday. Meanwhile, the company has received liquidity support from its top shareholder Fairfax India and decided to raise funds to assuage some of the concerns. This is actually quite good news for Fairfax India (at the end of a week of disappointing election news) - IIFL Finance represented about 7% of FIH's holdings on March 31st, worth $234m. It had lost $177 in Q1 on news of the 'supervisory concerns'. FIH stepped in with liquidity support. I don't know if the company announced what terms this support came with, all I can find is this from March 6: "Fairfax India has agreed to invest up to $200 million of liquidity support on terms to be mutually agreed and subject to applicable laws, including regulatory approvals (if any),” IIFL Finance said in a press release. Given that IIFL Fiinance was trading at about 620 INR prior to the concerns, and dropped to about 380 before the FIH support was announced, and after a further dip have recovered INR56 today and are up to 470, shares have now recovered about 3/8 of the drop. Since shares were at 345 on March 31st when the quarter closed, half of the Q1 loss has now been reversed, for an unrealized gain of almost $100m. And FIH may make a little extra on whatever money they provided as support. I suppose we should wait to find WHAT the concluded audit found before we get too excited, but at least this Q1 problem looks like it is on its way to being resolved. Meanwhile, IIFL Securities, 4% of FIH's holdings at the end of Q1, is up from 123 INR to 219, including today's (confused?) gain of 36. This part, I doubt we get to hold on to...

It looks like you wanted to add the effective shares outstanding for the 5 years 2016 to 2020. 2016: 23.1m 2017: 27.8m 2018: 27.2m 2019: 26.8m 2020: 26.2m Average, 2016-2020: 26.2m 2024e: 22.2m Change, 2016-2020 to end 2024: -15% By the way, looking at annual reports to compile these numbers, I noticed how often the word 'outstanding' is used in the annual letter, apart from referring to shares outstanding. A lot!! 2016: 17 2017: 13 2018: 14 2019: 9 2020: 23 Average, 2016-2020: 15 2023: 29 Increase: almost 100% In retrospect, 2020 would have been a pretty nice time to invest, with a sharp uptick in the outstandings, and the trend is still favourable. Maybe this technical indicator performs as well or better than the exclamation point indicator, what do you think?

Ouch, what a shitshow. Railway stocks tank 20% as election trends diverge from exit polls Definitely a disappointment that the BJP looks like it will not get an outright majority. True, with 272 seats necessary to pass laws, it looks like his NDA coalition is heading towards about 292 seats. This is substantially less than the NDA hoped for (355-370), and less than the outgoing parliament (353), but still 20 seats over a majority. But Modi's BJP party, while it is by far the biggest party in the NDA coalition, will only get about 240 seats, so they will need about 32 additional votes from one of the parties they campaigned with (there are about 30 of them). That will obviously mean compromises to Modi's political agenda. https://indianexpress.com/article/explained/explained-politics/lok-sabha-results-bjp-nda-partners-9372070/

Well, in this case, Bangalore is 50% of book value, and that number might end up being a fair bit higher. Fairfax India pays Fairfax Holdings 20% of book gains over 5% a year, so if the Bangalore Airport ends up generating 20%/year returns, you would lose 3% of that to Fairfax (i.e. 3 out of 20), so you would only get 17%. I could live with a flattening like that. Long and tedious addendum; skip to the last paragraph if this is TL;DR: Here's the company's description of the benchmark: You will recall that under the investment advisory agreement with Fairfax Financial, Fairfax Financial is entitled to a performance fee, calculated at the end of each three-year period, of 20% of any increase in Fairfax India’s BVPS (including distributions) above a non-compounded 5% increase each year from the BVPS at inception in 2015. If book per share was $10 on January 30th, 2015 (see last year's annual report, p. 70), then the benchmark is presumably $10.50 after a year, $11.00 after 2 years, and $11.50 after 3 years, and so on, with the benchmark higher by 50c every year. This is, I think, what is meant by 'non-compounded' in the above quote. I think this is borne out by my calculation for the fee paid after 3 years, when book value was $15.24 on Dec 31st, 2017, before fees. They probably adjusted for the fact that there were only 11 months in 2015, but roughly, we would expect them to have paid out 20% of book value growth beyond $11.50, which would be 0.2*(15.24-11.50)=0.2*$3.74=$0.75. $0.75 would represent 7.5% of those 52.4% in gains, meaning the gain after fees would be 52.4-7.5%=44.9%, and this corresponds quite closely with the $4.46 book value gain reported. FIH shareholders kept $4.46 out of the $5.24 in book value gains, or 85%. So I think this calculation is probably correct. But what it means is that, every year, the 5% benchmark means 5% of the original $10 per share book value, not a 5% return on the previous year's book value. This makes no difference in the first year, and only a tiny difference in the next few years, but presuming that FIH's book value continues increasing, it will eventually mean that the benchmark becomes a very small percentage of book. For instance, at 2023 year end, book value was $21.85 (after fees). At the end of 2024, if the book value has increased by 20% (let's be optimistic), it would be $26.22 before fees, a gain of $4.37. Fairfax Holdings would take its 20% fee on the gain minus the benchmark, i.e. $4.37-0.50 = $3.87, 20% of which is $0.77, leaving $4.37-$0.77 = $3.60 for shareholders like us. In other words, we would keep $3.60 our of the $4.37 in book value gains, i.e. 82% of gains, with Fairfax getting 18%. Compare this to 2017, when we paid 15% of gains and paid . As the 50c annual benchmark becomes a smaller and smaller proportion of the book value per share, the performance fee will get closer and closer to 20%. For instance if book value grows by 15% a year over 20 years (it was 14.3% for the first 9 years), the book value per share would go from $164, a 15% annual gain would $24.55, and we would be paying out 20% of $24.55-$0.50, i.e. essentially 20% of the whole book market gain, with the 50c benchmark, initially 5% of book value, now representing only 0.3% of book value. TL;DR: The performance fee will get a bit worse, because compounding is making the non-compounded benchmark disappear. Eventually we will just pay 20% of all book value gains. It's not a deal-breaker for me, but it's a little worse than paying 20% of the annual gains beyond 5%.

This is just generous Mr Market giving us one last chance (maybe) to buy FIH at 72% of book value ($1.9b market cap, $2.66b book value). And that is book value calculated with a fair value of $1.6b for FIH's 64% holding of the Bangalore Airport, giving the whole airport a fair value of $2.5b. If FIH is able to float this at anything like the $3.7b valuation they were hoping a year and a half ago, Going from $2.5b to $3.7b would take FIH's book up from $2.66 to $3.42, and at the current price, that would put them at 56% of book (or maybe a little more, since they would have to pay some of this out to Fairfax , since book gains would be substantially over the benchmark of 5% annual. If that happens, who knows when, something spectacular is going to happen to FIH, either to its price:book or, perhaps, to its share price.

Yes. Ultimately, it is alpha that matters; but whatever Greek letter it might be, some volatility around price:value is a bonus, ultimately, for anyone who is prepared to take advantage of dips (this one has been a long one!) and lighten up on climbs. And even for long-term holders who do neither, if the company does it for them. But that would work better if there were a bit more high points…

I'm with TwoCities about the persistent discount - the important thing is that the investments are doing well, and if so, the share price will eventually follow, providing for some additional oomph in the meantime via repurchases. So the private investments have to be strong, but the translation into market price doesn't concern me. Anchorage is the same sort of deal - is the airport doing well, or not? It seems that it is. At over 50% of assets, an IPO would be nice, just to take some cash off the table and to allow for other investments (IDBI?), but there are likely to be opportunities to do the same thing with private investors, so maybe we don't have to worry about whether an IPO happens or not.

Interesting. Do you care to expound, or say where one could read more anlaysis about this company?

That IS impressive, I love it! I don't at all mind the ceremony, on the contrary, but the business end of it, it is the setting the price of the IPO that seems archaic to me. Does anyone think such a process wil be around when companies issue shares in a hundred years? It seems like a lot of work for nothing, not even getting the best price for the issuing company and not properly insuring that everyone has an equal access to buying the shares they might want to buy. As woodman points out, what is important is how the company does in the next 3-5-10-20 years, not whether it came out at 250 rupees or 300, and whether it popped 10% the first day or the first month. Anyway, file this away as a pointless gripe, it is what it is, and the company seems to have navigated through this complex landscape in a satisfactory way.

The whole process seems downright mediaeval to me, with its price bands, its GMP (grey market premium), its alotment dates, its public and private subscriptions, the 55 share block requirement (why on earth 55??), the concern about whether the IPO is 'successful' or not because it pops on day 1, etc. I know this is not so different from North American IPOs, but why so much drama? Why not just offer blocks of shares on the market, with a bid and an ask, and do it at the price that settles, like an auction? Does it have to be this complicated?

think I can settle with the thought that prem really was just overweight and wants to have some eggs at the side lines which is fair. Yes, according to his statement, he bought "482,600 subordinate voting shares of Fairfax at a price of US$308 per share, or approximately US$150 million in total." Now he has sold 275,000 of these shares to the company (so they can be retired), at $US1,106.48. He continues to control 1,548,000 outstanding multiple voting shares and 519,828 subordinate voting shares of Fairfax, for a total of 2,067,828 shares, meaning that he had 2,342,828 shares before the sale, and has sold 11.7% of these. If the new stake is worth 90% of his net worth, that would mean that he previously had 101.9% of his net worth in Fairfax; if his current stake is higher than 90%, then his previous stake was even higher than 101.9%. That probably means he took out a loan to make the purchase back in 2020, when short term interest rates were still less than 1%. So there's overweight and then there's really spectacularly overweight. We think we are overweight at 50%, but he was at over 100%. Taking that down to 90% can not reasonably be taken as evidence that he think the shares are fully valued, it's just elementary prudence. If he had reduced from a smaller percentage, like going from 10% to 9%, that would have been a different story. And the increased capital gains tax in July, while not mentioned, just adds another reason why it makes sense to pay off the loan now, even if he still think Fairfax's prospects are great.

I have the same problem, 50% in Fairfax after the big share price increase in the last few years, but nowhere too tempting to go if I sell my stake down to a more reasonable size. And 10% in coal/gas/oil which seem like the best alternative for the moment - Yancoal, Suncor, CNQ, Petrobras...

Why wait? Here's hoping RoaringKitty was just waiting for Prem to get off the Blackberry board to give Fairfax one last chance to unload this, and maybe reap some tax losses to book against the Pet Insurance business.

if once digit starts writing insurance at < 100% CR if it will start to invest its profits in equities What about Odyssey or Brit or Allied? Should they also start to invest their profits in equities (both public and private) and slowly morph into conglomerates? Seems like a waste to replicate the Fairfax structure at the subsidiary level.