Fairfax 2024

[SafetyinNumbers]

There seems to be a relationship between P/B and ROE which makes intuitive sense. I interpreted Buffett’s view on book value to mean, don’t use it to estimate intrinsic value. That is if the stock is trading above book it doesn’t mean the stock is expensive and one should not buy the shares. It might indicate the business has high future expected ROE. It follows that a P/B multiple is appropriate to estimate IV because it’s based on future expected ROE.

 

For BRK and FFH, their capital allocation decisions eventually show up in earnings. I haven’t studied BRK closely myself but it seems like investors have a return expectation of 8-10% but are willing to pay ~1.8x BV for that. IFC has forward return expectations of ~15% and it trades at 2.8x BV. Meanwhile, FFH has forward return expectations of ~15-20% and it trades at ~1.2x BV.

 

My conclusion is that BRK shareholders have much lower return expectations as the margin of safety is extremely high. IFC has low variability in their earnings estimates which quants like which explains its premium multiple. Finally, FFH doesn’t screen well so it’s underowned by quants and most of the remaining actively managed money as they mimic quants. 

[Viking]

4 hours ago, nwoodman said:

@@Viking Good post and guilty as charged, the P/B heuristic is just so tempting. I think it has caused me to underestimate the true earning potential—and hence value—of both Fairfax and Berkshire over the years.  Although P/B of 0.6x’s was a decent marker of a margin of dafety.

@nwoodman , what i find so fascinating about Fairfax is how much has been changing over the past 4 years. And with all the cash they are generating today (and the coming years) my guess is more good ‘surprises’ are coming. Because the historical information doesn’t help us very much - and this is what most investors rely on the most to guide them in their decisions. 
 

With my posts i am often trying to figure out and get ahead of what is perhaps coming next - stuff that is not on investors radar today that will be in another year or two.

 

Eventually, the pace of change will slow and Fairfax will likely become a more normal, boring, simpler company to value.

Edited August 19 by Viking

[nwoodman]

10 minutes ago, Viking said:

@nwoodman , what i find so fascinating about Fairfax is how much has been changing over the past 4 years. And with all the cash they are generating today (and the coming years) my guess is more good ‘surprises’ are coming. Because the historical information doesn’t help us very much - and this is what most investors rely on the most to guide them in their decisions. 

True, picking up on @SafetyinNumbers point above, I think that is why we follow their capital allocation so closely.  There is a bit of scar tissue that has built up with Fairfax over the years from capital allocation misadventures.  I just don’t have those concerns with Berkshire and they make mistakes too.  Even when Buffett is doing nothing it feels like optionality.  To this end Wade’s recent CC comments resonated when he said he wasn’t finding much value in equities at the moment, I can empathise there.  Happy for them to keep the allocation to equities small until there is some really interesting or just take out an entire company e.g. Sleep Country.
 

I have probably touched on this before, but I have often thought that Berkshire has access to more timely information than the Fed.  It strikes me that Fairfax is entering a similar position in terms of data from their various subs which are now far and wide.  This has to help in general but in terms of managing a bond portfolio surely it must provide some semblance of a competitive advantage.
 

This ‘access to information’ also extends to key relationships and deal flow.  For me deal flow is key and I think that is something that has definitely improved in the last 5 years or so.  Berkshire doesn’t seem to get the opportunities they did when I first started to follow them 20 years ago.  It might be the law of large numbers but I do feel they got crowded out by the Fed during Covid so there is that.  
 

The good thing for Fairfax is that a $US1-2bn deal can make a material difference, as long as they are rational.

 

[petec]

5 hours ago, SafetyinNumbers said:

There seems to be a relationship between P/B and ROE which makes intuitive sense.

 

Not just intuitive - there's a direct mathematical relationship where R is the market's required return: ROE/R = P/BV.

 

For example: 15%/10% = 1.5x

 

This is perhaps more obvious when we remember that the E in ROE is the same as BV.

[petec]

On 8/17/2024 at 6:31 PM, Xerxes said:

This may or may not have been posted.

 

QUARTERLY REPORT #5: PERIOD TO 31 MARCH 2024 - Google Search

 

East 72 Dynasty Trust

QUARTERLY REPORT #5: PERIOD TO 31 MARCH 2024

 

We added new holdings in Fairfax Financial Holdings, a Canadian reinsurance company after an appallingly argued short-sale thesis saw the shares sold off sharply for less than a week in February. We also added Occidental Petroleum, the oil business boasting Berkshire Hathaway as its largest and effective “controlling” shareholder, but more pointedly exercising a capital management philosophy of some note.

 

////

 

Monopolistic but regulated high margin volume growth. A new expanding airport. Bangalore If you sat down to conceive of the perfect investment, how might it look? One interpretation would be that it has monopolistic attributes – the famous “only bridge into town.17 ” But further imagine your “unregulated” toll bridge was located in an area of rapid population growth of highly qualified people - an area of high tech and IT manufacturing. So your toll bridge is going to get ever more use - so you can build a second one next to it. Moreover, you could essentially build your monopoly “bridges” from the ground up with few (if any) legacy issues, using the latest technology to service an industry which cannot function effectively without the latest know-how.
Unfortunately, it’s only a partial reality, but one which is priced at a discount to legacy assets in the same sector. Because this is not an unregulated toll bridge. It’s rather that the toll is actually regulated but you can keep your customers on the bridge for very lengthy periods and they have to spend money at your shops.

BIAL is Bangalore International Airport Limited, operator of Kempegowda18 International Airport located 40km from the centre of Bengaluru, India’s third largest – but fastest growing city – with 14million people in its metro area, up from 6.5million twenty years ago. BIAL has a monopoly within a 150km radius until 2033 and has a 30 year concession with a further 30year option to operate Kempegowda. The airport has an estimated catchment area of 250million people (it’s worth noting when you read the next section, how long any competing facility may take to be planned, let alone built when the monopoly expires).

The dynamics and analysis of Kempegowda are identical to most other “greenfield” airports – gradual expansion and use of adjoining vacant land to add non-airport revenues over time. It’s a model seen elsewhere around the world in recent years, but for one exception: the numbers here are far bigger than anywhere outside the Middle East – and it is predominantly a domestic operation for the time being. Dare to dream of the word “hub” sometime down the track.

BIAL is now 13% owned by each of Karnataka State Industrial & Infrastructure Development Corporation and Airports Authority of India 13% - two Government instrumentalities. The remaining 74% is 20% owned by Siemens Project Ventures – part of the €134billion market cap Siemens AG – and through two entities by Fairfax India Holdings19 (FIH-U.TO20).

 

////

 

I'm never sure what to think of East 72. They seem to do deep work, but taking this report as an example

• The first idea, Belron, looks super risky to me. Margins have tripled in a few short years and leverage has ballooned against this profitability. This happened after PE got involved. I don't see any real argument for pricing power so there's a distinct possibility that the business can't bear the debt long term.
• The second idea, FIH and specifically BIAL, looks fairly valued vs. most comps on 10-12x 2026 ebitda, which may be a near term peak. The only argument for undervaluation on the data E72 have provided is that a private buyer might pay 24x ebitda. I own FIH but didn't find this report particularly well-argued.
• The third idea is more obviously cheap, but looks very like an ego project to me, with sub-par capital allocation.

 

I'm just not convinced by their thinking.

Edited August 21 by petec

[SafetyinNumbers]

5 hours ago, petec said:

 

Not just intuitive - there's a direct mathematical relationship where R is the market's required return: ROE/R = P/BV.

 

For example: 15%/10% = 1.5x

 

This is perhaps more obvious when we remember that the E in ROE is the same as BV.

It seems the relationship is actually exponential which also makes sense. 15% ROE businesses seem to trade > 2x book value and in IFC’s case 2.8x BV. Kinsale which has a 25%+ ROE trades at ~9x BV. 

I would caution FFH shareholders (including myself) from selling at 1.5x BV thinking that’s fair value when there is still significant growth and multiple expansion ahead. 

[dartmonkey]

5 hours ago, petec said:

 

Not just intuitive - there's a direct mathematical relationship where R is the market's required return: ROE/R = P/BV.

 

For example: 15%/10% = 1.5x

 

This is perhaps more obvious when we remember that the E in ROE is the same as BV.

This is a good point, and it emphasizes that ROE is the missing link, when we are thinking about P/B and P/E and value. It would be more obvious if we didn't use E for 2 different things (equity and earnings), like if we said ROB instead of ROE, and even better if we said E/B instead of ROE. But we have to live with the conventions and everyone talks about ROE, not E/B.

 

At the end of the day, we shareholders shouldn't give a damn about what B is - as Buffett has said, value is the discounted sum of future earnings. Fairfax has a very high ROE, and this is partly because it is heavily levered by its massive insurance float, much more so than Berkshire. This is a great model to have, but it is true that leverage cuts both ways, and so Fairfax arguably deserves a lower multiple on its levered earnings, in addition to worries about possible misallocations of capital like Blackberry and shorts on tech stocks and deflation bets that may have made some investors mistrustful of Fairfax.

 

I think, and I suppose most of us probably think, that this discount for leverage/mistrust is a bit excessive. At current levels, P/B is about 1.1, ROE=E/B is about .20, so P/E is about 1.1/.20=5.5, whereas in Berkshire's case, P/B = 1.6, ROE=E/B is about .12, so P/E is about 1.6/.12 = 13.3. 

 

Thoughts?

[SafetyinNumbers]

24 minutes ago, dartmonkey said:

 

At the end of the day, we shareholders shouldn't give a damn about what B is - as Buffett has said, value is the discounted sum of future earnings. Fairfax has a very high ROE, and this is partly because it is heavily levered by its massive insurance float, much more so than Berkshire. This is a great model to have, but it is true that leverage cuts both ways, and so Fairfax arguably deserves a lower multiple on its levered earnings, in addition to worries about possible misallocations of capital like Blackberry and shorts on tech stocks and deflation bets that may have made some investors mistrustful of Fairfax.. 

 

Thoughts?

I don’t think it makes sense to punish FFH for having a higher insurance float to book value ratio. Buffett says insurance float at a well run insurer is worth more than book value. I don’t think the relative size of it to equity should restrain that valuation as it could be solved simply by adding equity (which is what is happening as earnings flow in). 
 

The reference to past mistakes contributes to negative Social Value as opposed to any impact to Intrinsic Value. The numbers speak for themselves and the theoretical valuation exercise should remove bias. 

[StubbleJumper]

15 minutes ago, dartmonkey said:

At current levels, P/B is about 1.1, ROE=E/B is about .20, so P/E is about 1.1/.20=5.5, whereas in Berkshire's case, P/B = 1.6, ROE=E/B is about .12, so P/E is about 1.6/.12 = 13.3. 

 

Certainly that's been true for the three years 2021 to 2023 that BV has grown by 20%+.  But, it's worth taking a long and hard look at the table that Prem publishes every year in his annual letter which depicts growth in BV (it appears on page 22 of the most recent annual letter).  Those annual growth numbers result in the much vaunted 18% compounded annualized growth in BV over the 38-year life of the company. 

 

One can hold the belief that the entire 38-year historical series is representative of the sort of results that one might expect over the next 38 years (ie, 18% growth in BV is the long-term norm and what we'll get in the future).  Or one might hold the belief that the first 10 or 15 years of FFH's results reflect a rapid growth start-up phase that cannot be replicated in the future because FFH is a much, much larger company (eg, BV grew 180% in 1986 and 48% in 1987 but that would be unthinkable with the current asset base), implying that you need to give a substantial haircut to that 18% historical result.  Or, one can simply take Prem's publicly stated objective of 15% growth in BV as an "expert opinion" of the sort of annual BV growth that one might expect. 

 

When I look at that table that Prem publishes every year, and I squint a little, I see a company that operates primarily in a highly competitive, cyclical industry that produces a commodity (insurance).  The value offered by FFH comes principally from shrewd investment decisions, but there remains an anchor on those investment returns driven by the requirement of keeping a large portion of its investment portfolio in sovereign debt.  Personally, I do not hold the view that the 38-year historical BV growth number of 18% can be replicated on a going-forward basis and I would be absolutely delighted if FFH were actually able to achieve Prem's 15% goal over the long-term.  My guess is that in 2039, we will look back on the previous 15 years and see a 12% annualized growth in BV, but time will tell.

 

One's view on this matters a great deal.  As you noted, if you hold the belief that 20-ish percent annualized growth in BV is the new normal over the long-term, FFH is screaming cheap priced at today's 1.1x.  If you simply accept Prem's 15% objective as representative of the long-term future, FFH is still very cheap.  But, if you look at Prem's BV growth table and you see 12% in the future, you likely wouldn't want to pay more than 1.5x for the shares, implying that the stock is cheap at 1.1x, but probably not outrageously cheap.

 

We shareholders have witness three really good years for FFH, and there's likely another couple of solid good years coming down the pipe.  Enjoy them while they are here!

 

 

SJ

[MMM20]

1 hour ago, StubbleJumper said:

if you look at Prem's BV growth table and you see 12% in the future, you likely wouldn't want to pay more than 1.5x for the shares

 

This probably comes out to something like ~11-13% total returns if the stock derates to ~1x BV over those ~15 years, depending on how much capital they return to shareholders (and how). My expectation (ok, wild guess) is that sort of return would still end up at least a few percentage points ahead of global indexes. Maybe ~2x BV or mid-high teens on earnings is fairer right now across scenarios. Someone tell the robots!

 

Edited August 19 by MMM20

[73 Reds]

13 minutes ago, StubbleJumper said:

 

Certainly that's been true for the three years 2021 to 2023 that BV has grown by 20%+.  But, it's worth taking a long and hard look at the table that Prem publishes every year in his annual letter which depicts growth in BV (it appears on page 22 of the most recent annual letter).  Those annual growth numbers result in the much vaunted 18% compounded annualized growth in BV over the 38-year life of the company. 

 

One can hold the belief that the entire 38-year historical series is representative of the sort of results that one might expect over the next 38 years (ie, 18% growth in BV is the long-term norm and what we'll get in the future).  Or one might hold the belief that the first 10 or 15 years of FFH's results reflect a rapid growth start-up phase that cannot be replicated in the future because FFH is a much, much larger company (eg, BV grew 180% in 1986 and 48% in 1987 but that would be unthinkable with the current asset base), implying that you need to give a substantial haircut to that 18% historical result.  Or, one can simply take Prem's publicly stated objective of 15% growth in BV as an "expert opinion" of the sort of annual BV growth that one might expect. 

 

When I look at that table that Prem publishes every year, and I squint a little, I see a company that operates primarily in a highly competitive, cyclical industry that produces a commodity (insurance).  The value offered by FFH comes principally from shrewd investment decisions, but there remains an anchor on those investment returns driven by the requirement of keeping a large portion of its investment portfolio in sovereign debt.  Personally, I do not hold the view that the 38-year historical BV growth number of 18% can be replicated on a going-forward basis and I would be absolutely delighted if FFH were actually able to achieve Prem's 15% goal over the long-term.  My guess is that in 2039, we will look back on the previous 15 years and see a 12% annualized growth in BV, but time will tell.

 

One's view on this matters a great deal.  As you noted, if you hold the belief that 20-ish percent annualized growth in BV is the new normal over the long-term, FFH is screaming cheap priced at today's 1.1x.  If you simply accept Prem's 15% objective as representative of the long-term future, FFH is still very cheap.  But, if you look at Prem's BV growth table and you see 12% in the future, you likely wouldn't want to pay more than 1.5x for the shares, implying that the stock is cheap at 1.1x, but probably not outrageously cheap.

 

We shareholders have witness three really good years for FFH, and there's likely another couple of solid good years coming down the pipe.  Enjoy them while they are here!

 

 

SJ

Makes sense, particularly if you believe that growth is the objective.  OTOH, may we conclude by the already proliferation of share buybacks that continued BV growth is perhaps more important and there could well be a cap - implicit or otherwise - on the growth of the company?  In any event, even 12% BV growth for the next 15 years would hardly be a disappointing result.

[StubbleJumper]

12 minutes ago, 73 Reds said:

Makes sense, particularly if you believe that growth is the objective.  OTOH, may we conclude by the already proliferation of share buybacks that continued BV growth is perhaps more important and there could well be a cap - implicit or otherwise - on the growth of the company?  In any event, even 12% BV growth for the next 15 years would hardly be a disappointing result.

 

 

Well, Prem's table depicts growth in BV because FFH's earnings have been overwhelmingly retained and reinvested over its history.  But the growth in BV in a particular year is really just another way of saying ROE for that year if you haven't paid a dividend or repurchased a bunch of shares.  So for most purposes, it's possible to take the numbers in that BV growth table roughly as ROE numbers, and as you noted, it's quite possible that reinvestment will not be the way that ROE gets used in the future.

 

 

SJ

[SafetyinNumbers]

2 hours ago, StubbleJumper said:

 

Certainly that's been true for the three years 2021 to 2023 that BV has grown by 20%+.  But, it's worth taking a long and hard look at the table that Prem publishes every year in his annual letter which depicts growth in BV (it appears on page 22 of the most recent annual letter).  Those annual growth numbers result in the much vaunted 18% compounded annualized growth in BV over the 38-year life of the company. 

 

One can hold the belief that the entire 38-year historical series is representative of the sort of results that one might expect over the next 38 years (ie, 18% growth in BV is the long-term norm and what we'll get in the future).  Or one might hold the belief that the first 10 or 15 years of FFH's results reflect a rapid growth start-up phase that cannot be replicated in the future because FFH is a much, much larger company (eg, BV grew 180% in 1986 and 48% in 1987 but that would be unthinkable with the current asset base), implying that you need to give a substantial haircut to that 18% historical result.  Or, one can simply take Prem's publicly stated objective of 15% growth in BV as an "expert opinion" of the sort of annual BV growth that one might expect. 

 

When I look at that table that Prem publishes every year, and I squint a little, I see a company that operates primarily in a highly competitive, cyclical industry that produces a commodity (insurance).  The value offered by FFH comes principally from shrewd investment decisions, but there remains an anchor on those investment returns driven by the requirement of keeping a large portion of its investment portfolio in sovereign debt.  Personally, I do not hold the view that the 38-year historical BV growth number of 18% can be replicated on a going-forward basis and I would be absolutely delighted if FFH were actually able to achieve Prem's 15% goal over the long-term.  My guess is that in 2039, we will look back on the previous 15 years and see a 12% annualized growth in BV, but time will tell.

 

One's view on this matters a great deal.  As you noted, if you hold the belief that 20-ish percent annualized growth in BV is the new normal over the long-term, FFH is screaming cheap priced at today's 1.1x.  If you simply accept Prem's 15% objective as representative of the long-term future, FFH is still very cheap.  But, if you look at Prem's BV growth table and you see 12% in the future, you likely wouldn't want to pay more than 1.5x for the shares, implying that the stock is cheap at 1.1x, but probably not outrageously cheap.

 

We shareholders have witness three really good years for FFH, and there's likely another couple of solid good years coming down the pipe.  Enjoy them while they are here!

 

 

SJ

I think the assumption that returns should mean revert is a mistake because it doesn’t consider the actual balance sheet leverage, available returns on new investments and what’s already in the portfolio.

 

I think it’s really hard to model ROE averaging less than 15% over the next 5 years based on the current outlook. Maybe you are correct, ROE will fall precipitously after that but my intention is not to sell unless forward ROE is expected to be < 10%. That seems far away for now.

[hasilp89]

8 hours ago, petec said:

I'm just not convinced by their thinking.

*his.

And agree.

(i got blocked by him on twitter for questioning his thinking...)

[Viking]

21 minutes ago, SafetyinNumbers said:

I think the assumption that returns should mean revert is a mistake because it doesn’t consider the actual balance sheet leverage, available returns on new investments and what’s already in the portfolio.

 

I think it’s really hard to model ROE averaging less than 15% over the next 5 years based on the current outlook. Maybe you are correct, ROE will fall precipitously after that but my intention is not to sell unless forward ROE is expected to be < 10%. That seems far away for now.

 

The problem with mean reversion is that logic assumes the past business model is largely the same as the current business model.

 

Let's go back to 2010 when Fairfax put on the equity hedge/short trade. 'Reversion to the mean' logic would have lead a 'rational' investor to expect a 15% ROE type of performance from Fairfax over the next decade (2010 to 2020). 

 

Of course that did not happen. Fairfax significantly underperformed. Why? Mostly because of decisions made by the management team. Yes, zero interest rates was also a headwind.

 

The Fairfax of today is very different than the Fairfax that existed 10 years ago and even more different than the Fairfax that existed 20 years ago. My view is the current version of Fairfax is the best yet (in the 20 years that I have followed the company). That suggests to me that future returns (the next 5 years) should be good, and there is a solid chance that they could be very good.

 

Will Fairfax's results 'revert to the mean'. I have no idea. What matters to me is management and the decisions they are making (capital allocation etc). And what that is doing to the fundamentals of the business.

[StubbleJumper]

18 minutes ago, Viking said:

The problem with mean reversion is that logic assumes the past business model is largely the same as the current business model.

 

Yep.  We could be in a brave new world where it will be possible to simultaneously write a 94 CR every year and to routinely stuff your float with 4%+ sovereign debt, for a net financing differential of +10%.  But, I wouldn't count on it.

 

 

SJ

[SafetyinNumbers]

12 minutes ago, StubbleJumper said:

 

Yep.  We could be in a brave new world where it will be possible to simultaneously write a 94 CR every year and to routinely stuff your float with 4%+ sovereign debt, for a net financing differential of +10%.  But, I wouldn't count on it.

 

 

SJ

That’s only part of the return although I do think insurance is a growth market because of climate change. I also think a lot of institutional investors that previously would have shown up to take advantage of the opportunity prefer less volatile returns so instead are in private equity and private credit.
 

Do you think the recent large premium growth has added any permanent operating leverage for Fairfax or will it all be given back in price?
 

What do you model for forward returns on the equity portfolio? 

[Munger_Disciple]

3 hours ago, StubbleJumper said:

 

Certainly that's been true for the three years 2021 to 2023 that BV has grown by 20%+.  But, it's worth taking a long and hard look at the table that Prem publishes every year in his annual letter which depicts growth in BV (it appears on page 22 of the most recent annual letter).  Those annual growth numbers result in the much vaunted 18% compounded annualized growth in BV over the 38-year life of the company. 

 

One can hold the belief that the entire 38-year historical series is representative of the sort of results that one might expect over the next 38 years (ie, 18% growth in BV is the long-term norm and what we'll get in the future).  Or one might hold the belief that the first 10 or 15 years of FFH's results reflect a rapid growth start-up phase that cannot be replicated in the future because FFH is a much, much larger company (eg, BV grew 180% in 1986 and 48% in 1987 but that would be unthinkable with the current asset base), implying that you need to give a substantial haircut to that 18% historical result.  Or, one can simply take Prem's publicly stated objective of 15% growth in BV as an "expert opinion" of the sort of annual BV growth that one might expect. 

 

When I look at that table that Prem publishes every year, and I squint a little, I see a company that operates primarily in a highly competitive, cyclical industry that produces a commodity (insurance).  The value offered by FFH comes principally from shrewd investment decisions, but there remains an anchor on those investment returns driven by the requirement of keeping a large portion of its investment portfolio in sovereign debt.  Personally, I do not hold the view that the 38-year historical BV growth number of 18% can be replicated on a going-forward basis and I would be absolutely delighted if FFH were actually able to achieve Prem's 15% goal over the long-term.  My guess is that in 2039, we will look back on the previous 15 years and see a 12% annualized growth in BV, but time will tell.

 

One's view on this matters a great deal.  As you noted, if you hold the belief that 20-ish percent annualized growth in BV is the new normal over the long-term, FFH is screaming cheap priced at today's 1.1x.  If you simply accept Prem's 15% objective as representative of the long-term future, FFH is still very cheap.  But, if you look at Prem's BV growth table and you see 12% in the future, you likely wouldn't want to pay more than 1.5x for the shares, implying that the stock is cheap at 1.1x, but probably not outrageously cheap.

 

We shareholders have witness three really good years for FFH, and there's likely another couple of solid good years coming down the pipe.  Enjoy them while they are here!

 

 

SJ

 

Excellent post! It's a folly to assume long term ROE of 20% (well above Prem's own internal target) for Fairfax. I would be happy if Fairfax's BV compounds at 10%-12% for the next decade without major screwups in investments or underwriting. 

 

In a commodity business like insurance, it is crazy to expect a long term CR of 94. Buffett repeatedly said his goal for insurance business is 100CR; meaning he just wants it generate cost-free float. Keep in mind Berkshire is a lot less levered in their insurance business with a ton more capital than Fairfax!

Edited August 19 by Munger_Disciple

[Viking]

23 minutes ago, StubbleJumper said:

 

Yep.  We could be in a brave new world where it will be possible to simultaneously write a 94 CR every year and to routinely stuff your float with 4%+ sovereign debt, for a net financing differential of +10%.  But, I wouldn't count on it.

 

SJ

 

@StubbleJumper I am not trying to put words in your mouth - so please correct me if I am off base.  

 

Fairfax's future is like the multiverse in a Marvel movie. Fairfax has an infinite number of futures. 94CR is one. 4% interest rates is another. But there are an unlimited number of variables - many of which are very important. What your analysis largely ignores is management. It assumes Fairfax is a leaf getting blown by the wind - and where Fairfax goes will be determined by the wind. 

 

I completely disagree. I think the biggest factor that will determine Fairfax's future is not 'fate' (the wind) - it is management and the decisions they make. 

 

Why has Fairfax and Berkshire Hathaway performed so well over the past 38 years? Was it because of the P/C insurance cycle? Or the interest rates available on sovereign debt?

 

Both companies excelled because of the P/C insurance model and the excellent decisions made by the management team over time.

-----------

IMHO, what your mental model is completely missing is what we don't know - what the management team at Fairfax will do, especially when adversity (= opportunity) strikes. That is the same line of thinking that caused my to completely miss out on Berkshire Hathaway as an investment for decades - I couldn't see with certainty what Buffett was going to do so I way underestimated the value he was going to create in the subsequent years. 

[Viking]

54 minutes ago, Munger_Disciple said:

 

Excellent post! It's a folly to assume long term ROE of 20% (well above Prem's own internal target) for Fairfax. I would be happy if Fairfax's BV compounds at 10%-12% for the next decade without major screwups in investments or underwriting. 

 

In a commodity business like insurance, it is crazy to expect a long term CR of 94. Buffett repeatedly said his goal for insurance business is 100CR; meaning he just wants it generate cost-free float. Keep in mind Berkshire is a lot less levered in their insurance business with a ton more capital than Fairfax!

 

@Munger_Disciple , I don't think anyone is assuming that Fairfax will deliver a long term ROE of 20%. Today, I think there is a pretty good chance that Fairfax can deliver a 15% ROE over the next 3 years. Important: in my ROE 'calculation' I include excess of fair value over carrying value for the equity holdings. That does not show up in earnings or book value, but I think it is value creation that needs to be captured. 

 

Year 4 and further out? My guess is Fairfax will continue to do well, but I have no idea how well. There are too many unknowns. But I don't need to know what Fairfax will do in 4 or 5 years. Because in 3 years time (August 2027) I will have a very good handle on what I think Fairfax can do from August 2027 to 2030. What I do with my Fairfax position will be driven primarily by what I think I KNOW. Not what I don't know. 

 

What your analysis is missing is the reflexivity thing that George Soros is so famous for (I hope I don't get this wrong). But let's assume the CR for all P/C insurance companies go to 100 for years - not a one year blip caused by record catastrophes. If this happens, the share prices for many P/C insurers will get killed. This will likely present the perfect opportunity for a flush with cash player (like Fairfax) to make a big P/C acquisition at an attractive price. Soft insurance markets are a great time to grow via acquisition. 

 

That is exactly what Fairfax did from 2015 to 2107 - they used the then soft P/C insurance market to build out their global P/C insurance footprint (the fact they were able to do this when they were cash poor is amazing). 

 

In a soft insurance market Fairfax will also have the ability to shift capital to more productive uses. 

 

Like 2020, maybe Fairfax's share price gets taken out behind the woodshed. This will simply give Fairfax the opportunity to take out a meaningful amount of shares at a very attractive price (perhaps below book value).

 

Whatever boogeyman we can think of for Fairfax there is a flip side to it - something Fairfax can use to its advantage. The worse the boogeyman the better the opportunity. 

 

Now could we see high volatility over the short term (say 12 to 24 months) in the share price? Yes. But that would simply create the opportunity for significant long term value creation (via significant buybacks).

 

I am not assuming Fairfax will always make the most optimal decision moving forward and always come out smelling like roses. But I am also not doing the opposite - I am not assuming the worst because that is what Buffett says will happen. Fairfax tends to make their best investments (capital allocation) when adversity strikes. It is very counter-intuitive. 

 

Active management, volatility, unconstrained capital allocation and the power of compounding is a very potent combination.  Especially when you are flush with cash - this is the new variable for Fairfax.

Edited August 19 by Viking

[Parsad]

5 hours ago, SafetyinNumbers said:

It seems the relationship is actually exponential which also makes sense. 15% ROE businesses seem to trade > 2x book value and in IFC’s case 2.8x BV. Kinsale which has a 25%+ ROE trades at ~9x BV. 

I would caution FFH shareholders (including myself) from selling at 1.5x BV thinking that’s fair value when there is still significant growth and multiple expansion ahead. 

 

Shareholders need to consider their after-tax position in taxable accounts as well. 

 

I have so much in capital gains on FFH in my personal holding company, that at the new capital gains tax rate in Canada, it would set me back a long ways if I sold Fairfax only to buy something else.  Unless BRK gets cut by 65%, it wouldn't make sense for me to sell Fairfax and buy Berkshire if both grew at 15%! 

 

That's why in that account, I will just annually buy VOO going forward and average in as cash comes in.  The other core holdings bought like FFH, META, etc...will just let them keep compounding all of that capital gains. 

 

Now in non-taxable accounts...that's totally different!  Cheers!

[StubbleJumper]

39 minutes ago, Viking said:

MHO, what your mental model is completely missing is what we don't know - what the management team at Fairfax will do, especially when adversity (= opportunity) strikes. That is the same line of thinking that caused my to completely miss out on Berkshire Hathaway as an investment for decades - I couldn't see with certainty what Buffett was going to do so I way underestimated the value he was going to create in the subsequent years. 

 

 

Oh no, I'm not missing out on Fairfax at all (at this stage I'd be embarrassed to reveal the percentage of my portfolio in FFH because it's ridiculous, bordering on irresponsible)!  I'm just investing with my eyes wide open.  When you reach the top of the insurance cycle and the money starts coming in hand-over-fist, it is a mistake to project it too far into the future.  To your credit, you only tend to look forward a couple or three years at a time, which I think is probably about the limit of what one can do with any degree of precision.  But, when we slap a valuation multiple on the short term results, we are implicitly forecasting the long-term, and for that caution is preferable.

 

43 minutes ago, Viking said:

Fairfax's future is like the multiverse in a Marvel movie. Fairfax has an infinite number of futures. 94CR is one. 4% interest rates is another. But there are an unlimited number of variables - many of which are very important. What your analysis largely ignores is management. It assumes Fairfax is a leaf getting blown by the wind - and where Fairfax goes will be determined by the wind. 

 

Like it or not, the operating subs are subject to the industry competitive dynamics.  That involves highly profitable periods where it's relatively easy to grow premium, and it also involves difficult periods when FFH discontinues policies because they are not adequately profitable.  CRs go up, and CRs go down.  If that's what you mean by being blown around by the wind, well, that's what it is.

 

Management is quite another thing.  The reason why anyone would ever invest in FFH is because of the company's investment record.  Over long periods they hit just enough home runs to achieve a superior return.  Nobody is ignoring that.  But, there's an appropriate price for everything.

 

 

SJ

[Viking]

12 hours ago, nwoodman said:

True, picking up on @SafetyinNumbers point above, I think that is why we follow their capital allocation so closely.  There is a bit of scar tissue that has built up with Fairfax over the years from capital allocation misadventures.  I just don’t have those concerns with Berkshire and they make mistakes too.  Even when Buffett is doing nothing it feels like optionality.  To this end Wade’s recent CC comments resonated when he said he wasn’t finding much value in equities at the moment, I can empathise there.  Happy for them to keep the allocation to equities small until there is some really interesting or just take out an entire company e.g. Sleep Country.
 

I have probably touched on this before, but I have often thought that Berkshire has access to more timely information than the Fed.  It strikes me that Fairfax is entering a similar position in terms of data from their various subs which are now far and wide.  This has to help in general but in terms of managing a bond portfolio surely it must provide some semblance of a competitive advantage.
 

This ‘access to information’ also extends to key relationships and deal flow.  For me deal flow is key and I think that is something that has definitely improved in the last 5 years or so.  Berkshire doesn’t seem to get the opportunities they did when I first started to follow them 20 years ago.  It might be the law of large numbers but I do feel they got crowded out by the Fed during Covid so there is that.  
 

The good thing for Fairfax is that a $US1-2bn deal can make a material difference, as long as they are rational.

 

 

@nwoodman you make some great points:

1.) Fairfax's past mis-adventures and scar tissue. Making mistakes can be a wonderful educator. My guess is past mistakes have actually made Fairfax a stronger company today. And they are still recent enough that management likely still is feeling their sting. And I like that. 

2.) Great point about 'access to information'. Given how Fairfax thinks and operates this (they think about the economy and macro) getting real time information from the various businesses they own should help.

3.) You have spoken many times in the past about relationships/deal flow. What Fairfax did with Kennedy Wilson ($4 in real estate mortgages) last year provides perhaps the best recent example. Fairfax has spend decades building out their partnership/relationship networks. This should be another tailwind moving forward.

[Viking]

9 minutes ago, StubbleJumper said:

 

 

Oh no, I'm not missing out on Fairfax at all (at this stage I'd be embarrassed to reveal the percentage of my portfolio in FFH because it's ridiculous, bordering on irresponsible)!  I'm just investing with my eyes wide open.  When you reach the top of the insurance cycle and the money starts coming in hand-over-fist, it is a mistake to project it too far into the future.  To your credit, you only tend to look forward a couple or three years at a time, which I think is probably about the limit of what one can do with any degree of precision.  But, when we slap a valuation multiple on the short term results, we are implicitly forecasting the long-term, and for that caution is preferable.

 

 

Like it or not, the operating subs are subject to the industry competitive dynamics.  That involves highly profitable periods where it's relatively easy to grow premium, and it also involves difficult periods when FFH discontinues policies because they are not adequately profitable.  CRs go up, and CRs go down.  If that's what you mean by being blown around by the wind, well, that's what it is.

 

Management is quite another thing.  The reason why anyone would ever invest in FFH is because of the company's investment record.  Over long periods they hit just enough home runs to achieve a superior return.  Nobody is ignoring that.  But, there's an appropriate price for everything.

 

SJ

 

+1

 

I really do appreciate the opportunity to debate with other board members. We have a wonderful community of smart, thoughtful investors   

Edited August 19 by Viking

[StubbleJumper]

1 hour ago, SafetyinNumbers said:

That’s only part of the return although I do think insurance is a growth market because of climate change. I also think a lot of institutional investors that previously would have shown up to take advantage of the opportunity prefer less volatile returns so instead are in private equity and private credit.
 

Do you think the recent large premium growth has added any permanent operating leverage for Fairfax or will it all be given back in price?
 

What do you model for forward returns on the equity portfolio? 

 

Well, you might end up being right.  It's possible that over long periods FFH's insurance subs will be able to write a 94 CR, buy treasuries at 4%+, and run a premiums to surplus ratio of 2:1, resulting in a ~20% ROE for the sub.  Usually in the past, those sorts of returns were competed away as companies retained earnings and capital flowed into the industry (who doesn't want to throw more capital at a 20% ROE?!!).  Maybe the institutional investors will steer clear this time, but I wouldn't count on it.

 

The question of premium growth and operating leverage is an interesting one.  Operating leverage is highest in industries with high fixed costs.  That's not really the insurance industry.  Let us hope that most of the premium growth of recent years will endure even if the market softens, but that hasn't always been the case.  In softer markets in the past, FFH has allowed written less business.  The reason why the operating leverage question is interesting is that FFH's subs had a no lay-off policy during previous soft markets, which sort of converts labour from a variable cost into a semi-fixed cost.  You can see the effects of that when you decompose the CRs into their elements and see how the relative importance of the elements changes during soft and hard markets.

 

 

SJ