Fairfax 2023

[Viking]

7 hours ago, Thrifty3000 said:

Here is a historical chart showing average portfolio returns:

 

In terms of return on their total investment portfolio, I have Fairfax tracking at 8.2% for 2023. And 7.2% for 2024 and 2025.
 

For both 2024 and 2025 i am modelling nothing for large realized gains: items like selling pet insurance or Ambridge. Or revaluing of GIG (this should drop into 2023). I think it is highly likely Fairfax will do something meaningful to surface value in 2024 or 2025. So i think my 7.2% estimate is conservative. 
 

What about 2026? And further out? That is simply a bet on management. And their capital allocation skills. Fairfax has been putting on a clinic the past 5 years when it comes to capital allocation. Volatility is a tailwind for Fairfax - not a headwind. Why do people expect they are going to all of sudden get stupid? From my perspective that approach is not being conservative. 

Edited December 20, 2023 by Viking

[Haryana]

I am counting on at least one large realized gain or monetization or revaluation per year in guessing 200/year.

 

1. Digit IPO is in the pipeline

2. BIAL IPO is in the pipeline

3. New items will get added in the pipeline with their opportunism

4. They buy a small portion of an existing asset that revalues it upwards

5. They sell a small portion of an existing asset that revalues it upwards

6. They sell some unknown/insignificant position for an significant/outsized valuation

7. They sell some debt earning high interest for large gain when interest rates lower

8. Some of their distressed assets like BB or FDGE produce unexpected value

9. Fairfax India gets revalued to book value for whatever turn of events

 

[OCLMTL]

11 hours ago, Thrifty3000 said:

The investment portfolio will probably be closing in on $70 bil in 3 years. Bonds and cash will probably make up 60% of the portfolio. A blended 4.5% return, or $3.2 bil, should net around $100 per share.

 

If the bonds are only earning 2% then the rest of the portfolio has to earn 8% to maintain the $100 per share bogey. I don’t think that’s too big of a hurdle for the investment team.

 

I also think if interest rates are 2% that it’s not far-fetched to assume sub-100 combined ratios, which should help FFH stay in the neighborhood of $150 EPS through the cycle.  

To me that’s the biggest « hidden » upside. A lot of people are worried about rates heading lower and the potential impact to EPS/FCF in 3 years. It’s totally legitimate if you ask me. The problem though is that if we look at Viking’s earnings expectations, for example, and we see that $4B of earnings PER YEAR gets added to BV, and slap an incremental ROE on that $4B of 10%, you’re talking about a LOT of earnings power in years 4+ simply on the added equity we’ve built in the last few years. 

 

Am I completely wrong to think that incremental ROE of 10% is achievable on each tranche of $4B of earnings per year the company adds? Am I seeing this in a way that is overly simplistic? Thanks for all the feedback. Trying to humbly contribute to this dialogue. 

[Xerxes]

If that famous constant $150 per share per annum earning builds up against 2024 equity, as that equity goes up and accumulates then ROE would drop in year 3, as equity now has a larger base. 
 

So there is a case for FFH not trade at a premium multiple in the out year. 

[Viking]

52 minutes ago, OCLMTL said:

To me that’s the biggest « hidden » upside. A lot of people are worried about rates heading lower and the potential impact to EPS/FCF in 3 years. It’s totally legitimate if you ask me. The problem though is that if we look at Viking’s earnings expectations, for example, and we see that $4B of earnings PER YEAR gets added to BV, and slap an incremental ROE on that $4B of 10%, you’re talking about a LOT of earnings power in years 4+ simply on the added equity we’ve built in the last few years. 

 

Am I completely wrong to think that incremental ROE of 10% is achievable on each tranche of $4B of earnings per year the company adds? Am I seeing this in a way that is overly simplistic? Thanks for all the feedback. Trying to humbly contribute to this dialogue. 

@OCLMTL great add. I think what you are getting at is the power of compounding - earnings on earnings. It cracks me up that we have a bunch of value investors on this board who seem to be ignoring this ‘8th wonder of the world’ when they look at Fairfax and the significant earnings that is rolling in each quarter.
 

Yes, we don’t know exactly what Fairfax is going to do. But does this mean we assign it a value of zero? Expecting Fairfax to earn 10% on reinvested earnings (on average) makes sense to me.
 

Related, the size and timing of earnings matter. A lot. Well, at least that’s what Buffett says. Companies that get the cash flow up front are supposed to be valued higher. Especially if it outsized.
 

Fairfax has been generating record operating earnings starting back in 2022. New record in 2023. Likely a new record in 2024. Outlook for 2025 and 2026 is good. This is a massive amount of earnings coming over a 5 year period. 
 

This pile of gold has been/will be reinvested each year. Those investments are/will grow new earnings streams. Which, over time, will grow into rivers. So in 5 years time guess what? 

 

The idea that earnings at Fairfax will flatline at $150/share for the next two years and then drop from there makes no sense to me. At least based on what we know today. 

 

Edited December 20, 2023 by Viking

[Viking]

44 minutes ago, Xerxes said:

If that famous constant $150 per share per annum earning builds up against 2024 equity, as that equity goes up and accumulates then ROE would drop in year 3, as equity now has a larger base. 
 

So there is a case for FFH not trade at a premium multiple in the out year. 

That only makes sense if Fairfax puts the earnings under a mattress. 
 

Or if they allocate capital in a value destructive way. With each passing year, this is looking less and less likely. 

[Xerxes]

45 minutes ago, Viking said:

That only makes sense if Fairfax puts the earnings under a mattress. 
 

Or if they allocate capital in a value destructive way. With each passing year, this is looking less and less likely. 

Viking

while earning on earnings would be logical, they could simply choose to hold steady and build up capital and not go on risk curve with their annual “proceeds”.
 

Pre Covid, they have demonstrated their patience. 
 

Their investment “greatest hits” all revolve around accumulating drypowder for a big swing. I don’t think it is illogical to assume that the surplus won’t be re-invested with same market rate, either because (1) rate peaked or (2) credit spread didn’t blow out or (3) internal opportunities in insurance subs are less available. 
 

We just don’t know any of this. 
 

In absence of (1), (2) and (3), it would either accumulate dropping that ROE overtime or do a return of capital. 

 

[Viking]

1 hour ago, Xerxes said:

Viking

while earning on earnings would be logical, they could simply choose to hold steady and build up capital and not go on risk curve with their annual “proceeds”.
 

Pre Covid, they have demonstrated their patience. 
 

Their investment “greatest hits” all revolve around accumulating drypowder for a big swing. I don’t think it is illogical to assume that the surplus won’t be re-invested with same market rate, either because (1) rate peaked or (2) credit spread didn’t blow out or (3) internal opportunities in insurance subs are less available. 
 

We just don’t know any of this. 
 

In absence of (1), (2) and (3), it would either accumulate dropping that ROE overtime or do a return of capital. 

@Xerxes i really enjoy the back and forth. Writing is a great way to think and get ones thoughts straight. Here is a good thought exercise. Go back to Fairfax version 2019. Look at all the important metrics. Now look at Fairfax 2023 and look at those same metrics. Importantly, look at all the decisions made each year from 2019-2023. The company is unrecognizable today to the company that existed 4 years ago. But here is the key: Fairfax did all of this when they had much lower earnings!
 

Today Fairfax is earning gobs of money that is sustainable. Do people seriously think Fairfax in 2027 is going to resemble the Fairfax of 2023? Seriously? 

Edited December 20, 2023 by Viking

[Parsad]

3 hours ago, Viking said:

That only makes sense if Fairfax puts the earnings under a mattress. 
 

Or if they allocate capital in a value destructive way. With each passing year, this is looking less and less likely. 

 

Correct!  Cheers!

[steph]

You don't only pay a higher P/B multiple because of high expeced return on book.  You pay a higher multiple because of the notion of quality.  Quality company, quality management,... .  Fairfax got a low P/B because of distrust towards management and the idea that FFH was lower quality.  With what they have just done I sincerely do believe that the market will reward them again with a P/B more in line with quality competitors.  1.3 to 1.5 times would be reasonable.  So if after 2026 rates are lower and ROE will be somewhat lower I am convinced the P/B could be higher as long as they continue doing sensible things. 

[petec]

19 hours ago, MMM20 said:

 

You can argue that earnings in the out years would be lower (I'm skeptical) but don't forget that the fair multiple of those earnings would be higher - holding ERP constant at least.

 

 

In theory, but this is not what happened during the 2010s, in my view.

[petec]

19 hours ago, Tommm50 said:

 

I don't believe lower CR's are driven by higher interest rates, if fact, the opposite. If an insurance company is not making income on the investment side it is forced to try to make it on the underwriting side i.e. lower CR's. I'm not sure the relationship to growth but growth is not as important as profit.

 

Sure - but if investors can't make reasonable returns in fixed income they go looking elsewhere, driving capital into all sorts of things including insurance. lower rates did NOT drive a hard market in the 2010s - why would the extend this one?

[petec]

9 hours ago, Viking said:

Why do people expect they are going to all of sudden get stupid?

 

I happen to agree with this entirely, but don't forget this is the same board that KNEW that Prem was stupid from about 2011 to 2021, so sentiment can shift

[petec]

8 hours ago, Haryana said:

I am counting on at least one large realized gain or monetization or revaluation per year in guessing 200/year.

 

1. Digit IPO is in the pipeline

2. BIAL IPO is in the pipeline

3. New items will get added in the pipeline with their opportunism

4. They buy a small portion of an existing asset that revalues it upwards

5. They sell a small portion of an existing asset that revalues it upwards

6. They sell some unknown/insignificant position for an significant/outsized valuation

7. They sell some debt earning high interest for large gain when interest rates lower

8. Some of their distressed assets like BB or FDGE produce unexpected value

9. Fairfax India gets revalued to book value for whatever turn of events

 

 

I preferred this investment when one didn't have to count on such things - they were available for less than free.

 

FFH is still my biggest position by far (17%) but I am wondering how long I should keep it there.

[petec]

4 hours ago, Viking said:

Go back to Fairfax version 2019. Look at all the important metrics. Now look at Fairfax 2023 and look at those same metrics. Importantly, look at all the decisions made each year from 2019-2023. The company is unrecognizable today

 

I actually disagree with this. Most of the decisions that have led to the current metrics had been made in 2019, and some of them much earlier. They were short duration. Most of the big equity investments that have moved the needle were already there. They were limiting premiums in a soft market, waiting for the hard market. I think (can't remember the exact timing without checking) that they had already started to wind down the hedges.

 

Very few people on here were prepared to pay for the earnings potential then, despite a sometimes significant discount to book.

 

Today, all those decisions and that patience are paying off in droves, so the potential is obvious, and there's a huge love-in, even at 1x book (which remember is arguably inflated vs peers due to goodwill and the marking up of things like GIG).

 

Interesting debate - thanks all.

[steph]

Not so sure that FFH book value is more inflated than competitors.  At Markel the difference between book and tangible book is also quite big.  And with Berkshire you have a lot of hidden value, many companies that are worth much more than what book says, but on the other hand you have today, more than ever, 200 billion in Apple, 150 billion in cash plus all other equity holdings for which you pay 1.35 times book. 

[SafetyinNumbers]

2 hours ago, petec said:

 

I actually disagree with this. Most of the decisions that have led to the current metrics had been made in 2019, and some of them much earlier. They were short duration. Most of the big equity investments that have moved the needle were already there. They were limiting premiums in a soft market, waiting for the hard market. I think (can't remember the exact timing without checking) that they had already started to wind down the hedges.

 

Very few people on here were prepared to pay for the earnings potential then, despite a sometimes significant discount to book.

 

Today, all those decisions and that patience are paying off in droves, so the potential is obvious, and there's a huge love-in, even at 1x book (which remember is arguably inflated vs peers due to goodwill and the marking up of things like GIG).

 

Interesting debate - thanks all.

It’s hard to argue it’s a big love in when the multiple is 1x BV and IFC and TSU trade closer to 2.5x BV. I appreciate it’s because their earnings have less variability and estimates show consistent growth which quants like. It doesn’t mean they will create more value over the next 5 years and certainly doesn’t guarantee they have better stock performance although anything is possible. 
 

I think a lot of FFH investors are scarred from the last decade and are very afraid of drawdowns. There is no consideration for what could go right despite a very strong set up. I think it’s greater than 50-50 odds that ROE is north of 15% for the next 3 years based on the current balance sheet. It could be materially north or it could be lower but ultimately I think the odds of a 20% ROE are higher than a 10% ROE over the next 3 years and thankfully my hurdle is 10% so I’m still happy to own it even if my lofty expectations aren’t met.

[Xerxes]

10 hours ago, Viking said:

@Xerxes i really enjoy the back and forth. Writing is a great way to think and get ones thoughts straight. Here is a good thought exercise. Go back to Fairfax version 2019. Look at all the important metrics. Now look at Fairfax 2023 and look at those same metrics. Importantly, look at all the decisions made each year from 2019-2023. The company is unrecognizable today to the company that existed 4 years ago. But here is the key: Fairfax did all of this when they had much lower earnings!
 

Today Fairfax is earning gobs of money that is sustainable. Do people seriously think Fairfax in 2027 is going to resemble the Fairfax of 2023? Seriously? 

thanks. I am looking forward to watch this movie 4x a year (through quarterly releases) for the next several years, watching it adapt and change. This almost beats reading Tolkien’ work. 
 

One thing that can be a bull case for artificially high ROE (in absence of external and internal opportunities), is a major return of capital that would drop equity, giving that ROE a low base to be computed from. 

[Thrifty3000]

3 hours ago, Xerxes said:

thanks. I am looking forward to watch this movie 4x a year (through quarterly releases) for the next several years, watching it adapt and change. This almost beats reading Tolkien’ work. 
 

One thing that can be a bull case for artificially high ROE (in absence of external and internal opportunities), is a major return of capital that would drop equity, giving that ROE a low base to be computed from. 

 

YES. I was just about to say the same thing about return of capital. With so much cash flooding in the door it will be very interesting to watch where they park it. If the stock price continues hovering around book value then if they aren't using excess cash to buy back shares it means they must have REALLY juicy alternatives. I assume the only reason they would park excess cash into ultra-low yielding bonds rather than return capital would be if those bonds allowed them to underwrite additional insurance at extremely favorable rates.

Edited December 20, 2023 by Thrifty3000

[Hamburg Investor]

6 hours ago, SafetyinNumbers said:

It’s hard to argue it’s a big love in when the multiple is 1x BV and IFC and TSU trade closer to 2.5x BV. I appreciate it’s because their earnings have less variability and estimates show consistent growth which quants like. It doesn’t mean they will create more value over the next 5 years and certainly doesn’t guarantee they have better stock performance although anything is possible. 
 

I think a lot of FFH investors are scarred from the last decade and are very afraid of drawdowns. There is no consideration for what could go right despite a very strong set up. I think it’s greater than 50-50 odds that ROE is north of 15% for the next 3 years based on the current balance sheet. It could be materially north or it could be lower but ultimately I think the odds of a 20% ROE are higher than a 10% ROE over the next 3 years and thankfully my hurdle is 10% so I’m still happy to own it even if my lofty expectations aren’t met.

To me that seems a reasonable way to think about Fairfax. Heads you win a lot, tails it‘s ok. 
 

Again and again the discussion here touches interest rates. Where will they be in one year? In two year or three years? I don’t know. And I’m pretty sure the answer is: Nobody knows. Nobody has ever known. But the point is: it’s not important to know where interest will be in such a short timeframe, if you have lots of time to invest.
 

And what if we would think more about the question, if there is a “new normal” that is 0% or 2%? I think that is way more easy to answer for most of us. If you think there is a new normal, then Fairfax Financial (as well as every other insurance company, but Fairfax in particular) won’t prosper. If interest rates go back to zero and stay there for two decades, then there will be better buys than insurance companies. 
 

But I don’t think so. If anything history shows, that there has never been such a thing like “a normal interest rate.” There have been peaks and lows… but interest rates have never been staying at any level for long. 

 

Of course, others might disagree. There is always the possibility that “this time is different”. But I think there are a lot of cases in history, that show, that betting on “this time is different” normally isn’t a brilliant move. In fact going that way is really risky. 

 

So my thinking is more like: “If interest rates on average are at 2% or higher over longer timeframes then Fairfax has to do a lot of more dumb things then good things for not beating the S&P 500 buy a whole lot. 
 

And history. Again. How likely is it, that a company that has compounded within the top 1% (0.1%?) of stocks over 37 years is going to do so many more dumb things then good things? In fact you don’t find a lot of management teams, that have been working for one company for 37 years and being that good in the end. Buffett? Hasn’t reached 19% CAGR since 1986. Tom Gayner? I don’t think so, but I haven’t checked

 

Of course, people are free to believe that - again - “this time is different” with regards to the Fairfax management team. Maybe Prem and his team have lost all their skills in 2011 and 2021 to 2023/2025 is just a very short upswing episode in that downswing. Maybe.

 

But normally people that have done a very very good job over 25 years on average (1986 to 2011) don’t lose their skills overnight. 
 

A value investor having a CAGR of 19% of 37 years in his books at the end of a lost decade, which happened parallel to a zero interest rate episode, where growth has been beating value like never before, tells me another story.
 

The question is which perspective is right and which is wrong? 

[Hamburg Investor]

19 hours ago, Viking said:

In terms of return on their total investment portfolio, I have Fairfax tracking at 8.2% for 2023. And 7.2% for 2024 and 2025.
 

For both 2024 and 2025 i am modelling nothing for large realized gains: items like selling pet insurance or Ambridge. Or revaluing of GIG (this should drop into 2023). I think it is highly likely Fairfax will do something meaningful to surface value in 2024 or 2025. So i think my 7.2% estimate is conservative. 
 

What about 2026? And further out? That is simply a bet on management. And their capital allocation skills. Fairfax has been putting on a clinic the past 5 years when it comes to capital allocation. Volatility is a tailwind for Fairfax - not a headwind. Why do people expect they are going to all of sudden get stupid? From my perspective that approach is not being conservative. 

 
I totally agree. And I asked myself this question again, and again.
 

Maybe it’s because: Prems results were always bumpy. Not like Berkshire or Markel or most of the other insurance companies. Maybe it’s harder for people to trust management with bumpy results, even if there is a clear pattern over 37 years? A pattern with absolute returns that are way better than the smooth results of the competitors fishing in the same lake? Don’t get me wrong. Berkshire and Markel are opposition two and three in my investment portfolio. But Fairfax is on position one.
 

@Viking As you are discussing this historical chart, I look at again and again. But I don’t get the point of that chart.

 

All I can read is that the returns of the investment portfolio more or less define the end results in conpounding.
 

But even there are exceptions. And I can’t read anything out of the combined ratio. There’s just no relationship getting visable with regards to the compounding. Of course there is a relationship, I know. But this historical chart doesn’t help to understand that. and normally you show chart because you want the reader to understand an inner structure. 
 

So is the cr just included to point out to shareholders that Prem is aware that the combined ratio is important? Sometimes I ask myself if Prem thinks that in the future, the relationship will be getting more and more visible and I would agree on that point. 
 

What is he trying to show to us? What do you read in it?

[Viking]

2 hours ago, Hamburg Investor said:

 
I totally agree. And I asked myself this question again, and again.
 

Maybe it’s because: Prems results were always bumpy. Not like Berkshire or Markel or most of the other insurance companies. Maybe it’s harder for people to trust management with bumpy results, even if there is a clear pattern over 37 years? A pattern with absolute returns that are way better than the smooth results of the competitors fishing in the same lake? Don’t get me wrong. Berkshire and Markel are opposition two and three in my investment portfolio. But Fairfax is on position one.
 

@Viking As you are discussing this historical chart, I look at again and again. But I don’t get the point of that chart.

 

All I can read is that the returns of the investment portfolio more or less define the end results in conpounding.
 

But even there are exceptions. And I can’t read anything out of the combined ratio. There’s just no relationship getting visable with regards to the compounding. Of course there is a relationship, I know. But this historical chart doesn’t help to understand that. and normally you show chart because you want the reader to understand an inner structure. 
 

So is the cr just included to point out to shareholders that Prem is aware that the combined ratio is important? Sometimes I ask myself if Prem thinks that in the future, the relationship will be getting more and more visible and I would agree on that point. 
 

What is he trying to show to us? What do you read in it?

@Hamburg Investor if i understand your question correctly, i think Prem is trying to provide a shorthand way to value Fairfax (this is just a guess): CR + Total return on investment portfolio = possible ROE = way to value Fairfax.

 

Here is a link to an old writeup on Fairfax from 2019 that discusses this relationship. 
- https://www.woodlockhousefamilycapital.com/post/the-horse-story

 

“Moreover, I think the assets collectively could generate a ~10%-type ROE. Watsa has made a public goal of hitting 15%. (FFH’s ROE was 15% in the second quarter, thanks to investment gains). He says a 95% combined ratio and a 7% return on FFH’s investments gets to a 15% ROE.

 

“But in a low-interest rate environment, and given a large bond portfolio, a 7% return seems unlikely. But possible. Sustaining a double-digit ROE is key. (FFH can reach 10% by following a number of roads. For example, one road requires a ~95% combined ratio and ~5% return on its portfolio. That seems do-able.)

 

“Anyway, a consistent 10% would grow book value at a decent clip and then you’d likely get an additional lift from the valuation even if the stock moved just to 1.2x book. As RayJay reports, a comparable set of North American insurers with an 11% ROE trades for 1.7x book value per share.”

Edited December 20, 2023 by Viking

[gfp]

3 minutes ago, Viking said:

Here is a link to an old writeup on Fairfax from 1999 that discusses this relationship. 

Oh man I really thought that was going to be a 1999 research piece but I guess you meant 2019.  

[Hamburg Investor]

17 minutes ago, Viking said:

@Hamburg Investor if i understand your question correctly, i think Prem is trying to provide a shorthand way to value Fairfax (this is just a guess): CR + Total return on investment portfolio = possible ROE = way to value Fairfax.

 

Here is a link to an old writeup on Fairfax from 1999 that discusses this relationship. 
- https://www.woodlockhousefamilycapital.com/post/the-horse-story

 

“Moreover, I think the assets collectively could generate a ~10%-type ROE. Watsa has made a public goal of hitting 15%. (FFH’s ROE was 15% in the second quarter, thanks to investment gains). He says a 95% combined ratio and a 7% return on FFH’s investments gets to a 15% ROE.

 

“But in a low-interest rate environment, and given a large bond portfolio, a 7% return seems unlikely. But possible. Sustaining a double-digit ROE is key. (FFH can reach 10% by following a number of roads. For example, one road requires a ~95% combined ratio and ~5% return on its portfolio. That seems do-able.)

 

“Anyway, a consistent 10% would grow book value at a decent clip and then you’d likely get an additional lift from the valuation even if the stock moved just to 1.2x book. As RayJay reports, a comparable set of North American insurers with an 11% ROE trades for 1.7x book value per share.”

I know the discussion and the theoretical frameset. I like that persoective.
 

But how do you get those numbers together?

 

- 1986 to 1990: cagr of 57.7%, a combined ratio of 106.7 and an average total return on the investment portfolio of 10.4%.

- From 2001 to 2005 the investment returns were 8.6%, the cr was 105.4%, but the cagr was minus 0.9 per cent.

 

So both the insurance businesses as well as the investment returns are relatively similar in both timeframes; but the cagr couldn’t differ more. 58 per cent plus to minus 1 per cent. So the relationship between the numbers shown doesn‘t get visible. 

Or look at the timeframe 1996 to 2000. Investment returns were equal, but the cr was way more bad. Still the cagr was 31 per cent better then in the years with the worse cr.

 

At least we’d need to know how big the investment portfolio was and the premiums and the equity. But he hasn’t added that. Why not? Without that context and other contexts (like the macro bet) that are important in some of the timeframes, I can‘t read the relationship out of the numbers.  
 

That would be a great framework and one would understand a lot. What does the chart alone show to you intuitively?

 

As he hasn’t added those numbers, I wonder why? How does the chart help without that? The numbers shown alone to me seem totally random, one (at least me) isn’t able to get to any conclusion by looking at the numbers shown about any relationship between those. There’s just nothing to understand intuitively by looking at the chart for me - so I guess, there’s something I miss.  (There are other tables and charts like e. g. the table with the insurance companies results or the table with the tech stocks, that are easy to understand).

[Viking]

1 hour ago, gfp said:

Oh man I really thought that was going to be a 1999 research piece but I guess you meant 2019.  

 My mistake - edit has been made