SafetyinNumbers

Another great set-and-forget investment is E-L Financial. Special company and an incredible price.

The other part of context that is missing is that shareholders really bought into this narrative. The company issued 1m shares at C$735 (~US$550) in March 2016 or ~1.35x book. They bought Allied World in Dec 2016 and issued shares at about the same multiple. Were the hedges bad. Yes. Are the people who use it as a reason not to buy the stock now exactly the same people that bid the stock up in 2016. Also yes. Prem gets criticized on the investment side but as Viking has pointed out in previous posts, the return on Allied World has been phenomenal. Just because it was an insurance acquisition doesn’t mean it shouldn’t count as brilliant asset allocation. Just like Digit or GIG, etc… Another piece of context is that if negative rates are a real possibility then float becomes really expensive and capital would around quickly. In that sense deflation hedges and shorts make sense. Especially if the risk is existential. If the trade off is to hedge or stop growing float, I can understand the decision. Fairfax arguably got positive Social Value for shorting the market because investors who shared the view bought the stock. It’s not that different from Bitcoin enthusiasts buying Microstrategy because it’s a Bitcoin proxy. Thankfully Prem was able to use that to issue equity and buy a world class insurance business. Maybe the hedges served their purpose after all but it still wasn’t the right thing to do. In context, it seems like the criticism is mainly resulting.

Do you know how the SOFR + spread compares to the cost of other borrowing or the returns on alternative investments including buying their own shares back?

How do you both think about ELF.TO?

That's the last thing I would do. It's a sizable buyback where we have already paid the premium. Might as well wait, buy stock on the NCIB and enjoy the higher weighting in the index. I think the closet indexers will be the big driver for the stock over the next five years as they are way underweight and Fairfax is now 80bps of the index which is about what CSU was when they started chasing it in 2018. There are a lot of similarities between Fairfax now and CSU in 2018 like market cap, share price, shares outstanding, price momentum i.e. stuff that matters to closet indexers. The big difference is CSU starting valuation looked optically expensive 5 years ago which is what active managers had to overcome to get back to market weight. Fairfax meanwhile, is exceptionally cheap on every measure so in theory it should be easier to convince PMs to go to market weight or even overweight but we also have more jaded PMs to overcome as many stayed overweight FFH long after the outperformance ended. On the same note we likely have more willing sellers as Mark Leonard has achieved god status over the past 5 years while mostly everyone is waiting for Prem to make a mistake so are much more likely to take profits as price goes up even if valuation and technicals are screaming buy.

With an expected duration of claims at 4 years, I wouldn’t follow that strategy.

That would be amazing for the stock. If the insurance market softens then billions comes back to the holdco every year which would be used for buybacks or new equity investments if there are better opportunities there. Float would print even more money at 5-7% plus the opportunity to do more deals like KW this morning where the float is earning 10%. My contention is that currently you are either getting the insurance business (including float) or the equity portfolio for free which makes execution risk less important at current prices.

The prefs and warrants almost replicated the common exactly. The same yield (6%) and the same strike for the warrants where the shares closed ($16.21). Plus the pref are cancelled when the warrants are issued. It just gives us priority in capital stack which is valuable and potentially better capital treatment which is also valuable.

If my theory is right on the chase by Canadian PMs who are underweight FFH then we could see its outperformance accelerate. Those managers will buy until valuation is “fair” and that range is big as we can see from this board. In fact, FFH has already outperformed the index in each of the past three years and the relative outperformance has increased each subsequent year but it might just be a coincidence. Obviously execution has to continue but I think the biggest risk is underestimating intrinsic value and selling into multiple expansion due to the index chase. My tweet has the relative performance charts vs the XIC which is the S&P/TSX Composite ETF for June 2020 - June 2023 and each year in there.

The way I measure the high end of intrinsic value for Fairfax is to look at the price that gives an expected long term return of 10%. Maybe it’s too simplistic and we are in a high inflation world and I should adjust to 12% i.e. moving my long term inflation estimate to 5%. Ultimately that just means a higher interest rate environment which means the float is more valuable so there is a natural hedge. So I call it a wash and stick to 10%. On that basis, if Fairfax has a long term expected ROE of 15%, then 1.5x book value seems reasonable as it’s a 10% return on price. Any discount to that is margin of safety including the potential for cat losses that results in a zero. There should be some discount even if the odds are low they are non-zero despite that at some point in the next 5 years, I think we’ll trade north of the above intrinsic value estimate (i.e. have positive Social Value). The next few years, we could reasonably earn north of a 15% ROE so 3 years hence, BV could be US$1250 including some embedded gains in Ambridge, GIG, Digit and Fairfax India (Anchorage IPO). That would put the IV estimate (1.5x BV) in 2026 at US$1875 discounted at 10% to present day is ~US$1425 including dividends. Another way to triangulate intrinsic value is to listen to Warren Buffett when asked in 1996 about BRK’s $7b insurance float that he wouldn’t sell that “liability” for $7b in cash unencumbered if it meant they couldn’t compete in insurance again. On that basis, a low end estimate of intrinsic value is book value (BV) - insurance goodwill (IG) + insurance float (IF). For Fairfax that is approximately: + US$18.7b BV - US$3.4b IG + US$31.2b IF = US46.5b IV Or ~US$2000/share These are IV estimates and I try to buy things well under these targets so I get a high margin of safety which I need as I’m wrong about a third of the time. Buffett’s method results in a lower expected return than 10% which is why the IV is so much higher but arguably it’s not unreasonable if the quality of Fairfax’s insurance businesses are as good as they appear. The high class hardship for me will be trying to hold on to the majority of my position until it moves above my estimate of intrinsic value and I don’t think it’s out of the question we get there reasonably soon for two related reasons. First, we have a reasonably strong shareholder base as the share turnover is low which is what happens when a company is trading well below intrinsic value. Even though the share turnover is low, it’s hard to understand for investors that look at the company the way Viking does, for example, why anyone is selling at these prices. I talk to a lot of investors and my conclusion is that most who are selling are doing so because they made a great decision a while ago and have too high a concentration and the rest are worried about what Prem’s next mistake is going to be and don’t want to be around when it happens. Congrats to those on this board in the former category. For those in the latter category the stock could appreciate a lot before the next mistake and there will be one. When they do make a mistake we might not know it for years afterwards and actually cheer it on when it happens like the hedges when they were initially put on. Ultimately, it’s not obvious to see what looms as a problem now which is why most of us here are so bullish I suppose. They might have already made the mistake that will cause the next drawdown. Essentially both groups are making their decisions around price and not intrinsic value. They want to avoid the next drawdown. We’ve all done it and we’ll probably all do it again. Maybe even on Fairfax. If we buy the premise that our shareholder base knows what it owns for the most part (this thread was a great idea Viking as always!) should mean institutional buying will move the shares up significantly. The source of the buying is the second reason we can get above intrinsic value. It could be the company via share buybacks but ideally its active Canadian managers trying to keep up with the S&P/TSX. I have a friend and former colleague who is a PM at a Canadian asset manager. He likens it to Constellation Software about 5 years ago when its weighting was around 80bps. Fairfax is around that weighting now and has outperformed the index by 100% in the past three years. CSU outperformed the S&P/TSX by ~85% cumulatively from June 2015-2018. The pre is that pain caused PMs to reconsider how they were valuing CSU and the 5 years since it has outperformed the S&P/TSX by a further ~125%. That’s not to say the performance or outperformance for FFH will be similar or if it will happen at all. CSU had to execute and so will FFH but at this starting valuation and set up the odds seem excellent. PMs had to get over the multiple they were paying for CSU to get to market weight or even better overweight. With Fairfax, the valuation is easy to get over but the hurdle is defending Prem. I have had a financials analyst at a small Canadian broker tell me he can’t send his brokers out to defend Prem. I don’t think he’s thinking about it in terms of market weight yet but his strategist might soon. A signal that it might be playing out is if the volume picks up. It’s been stubbornly low for a while so while some funds might be accumulating, there isn’t a lot of urgency or competition. Perhaps, I’m just being impatient however given FFH has already outperformed XIC by ~18% YTD.

I read the press release differently. I agree they owned ~4m shares coming into the transaction. They added 2.7m shares at €18.5 and the 2.5m €20 strike warrants. They must have originally owned ~700k shares before acquiring ~4.9m shares in 2013 at ~€5.13 based on the MYTILINEOS 2013 AR (reference attached). They also must have sold 1.6m shares at some point before this most recent transaction to get to 4m shares. I doubt the cut it all the way to zero but I’m not sure. But ultimately ~6x on the original shares in ~10 years. In the words of Larry David, pretty, pretty, pretty good.

Fairfax showing some good timing. They participated in a financing in December 2022 for €50M and are up over 100% since including the warrants.

Almost certainly the counterparties are just long the stock and short the TRS to Fairfax. I assume, they will just buy back the shares when they are ready to close it out making it “risk-free” to the counterparties.

I think lots of successful investors have heuristics they live by. Complicated derivatives on their own stock while brilliant and positive signalling for me, probably rules out a bunch of potential shareholders for Fairfax. That’s what makes a market!

I met a smart investor at the Markel AGM a few weeks ago who sold his Fairfax because he was scared away by the TRS, thinking it was too aggressive. I told him it wasn’t too late to get back in!

No listed options on FFH. I tried to get them listed last year so I could buy calls but was told by MX, it’s too illiquid for options.

I think what we are both saying is correct. If I understand your argument correctly, our risk of impairment from the insurance business is so big, our expected intrinsic value should be significantly lower than any of our peers and so we should trade meaningfully cheaper on common metrics based on expected returns. I use a 10% expected forward return as a benchmark for my intrinsic value estimate. That’s usually where I’m entirely out of a position since there are higher returns expected elsewhere. Right now, Fairfax seems like it’s close to 20% (based on forward earnings yield) and I think there might be upside optionality from the portfolio which is mostly ignored. I agree with you there is risk associated with owning a large insurance business but I’m not sure what odds the market is placing on such an event and if it’s consistent across the peers. Ultimately though, if intrinsic value is the price where one can expect to earn an expected 10% forward return, what is IV right now? I would expect a forward 15% ROE which equates to 1.5x P/B or about US$1200 IV estimate to earn 10% forward returns. We are over earning now based on most estimates and have some embedded gains to come into BV soon (Digit, Ambridge, GIG) so arguably IV is a touch higher providing some margin of safety. I might be underestimating the odds of an existential event for Fairfax and don’t think the discount should be very big vs MKL, BRK and others. Arguably the expected returns are lower than 10% for some of the peers. I’m curious if others think about valuation similarly.

Fairfax didn’t have “normal interest rates” or a float 2x the share price for the last decade. Just some pepper to add to the grain of salt

No one is forcing you to trade Parsad! Congratulations on your gains.

it’s ok if we stay perennially undervalued but with equity growing so fast, we’ll probably be rewarded with a P/BV above one despite your objections!

I don't think you like float enough Parsad! Or maybe I'm just more bullish on interest rates than you.

Buying more BIAL to sell it through Anchorage is effectively a pass through especially if they can get a higher valuation on the flip. My point is selling it through the IPO or as a PIPE is kind of irrelevant if it helps them raise money at valuations closer to intrinsic value.

Could they sell a stake in Anchorage to buy a stake in IDBI? I thought the plan was to IPO Anchorage to get a currency to buy other assets in India. Getting other outside investors to buy a private stake at a fair valuation does the same trick and maybe has the same impact as a catalyst on FIH shares.

E-L Financial is another one, I would consider putting 100% of my net worth in, especially at less than half of intrinsic value where it is now. With respect to FFH fair value, Buffett has said that book value plus float is a fair estimate of intrinsic value. For Fairfax that’s about a triple from here.

Management fees really aren’t that high with the 5% hurdle rate.