Fairfax 2025

[Parsad]

2 hours ago, Viking said:

 

@Marco Van Basten, it looks to me like Fairfax has been pretty aggressive on the stock buyback front over the past 7 years. At the same time, they doubled the size of their insurance business and they have dramatically increased the size of their equity portfolio. They now have 5 income streams gushing cash. 

 

My guess is if the stock stays around current levels, we will see Fairfax continue to take out a meaningful number of shares in the coming months/year. Bottom line, it appears Fairfax has learned to walk (buy back stock) and chew gum (do everything else) at the same time. I like their balanced, shareholder friendly approach. 

 

Chubb is a fine, well run company. With the hard market slowing and their stock cheap, buybacks is a no brainer for them. Same with other P/C insurance companies. This is great news... a significant amount of capital (that being spent on buybacks) is being removed from the insurance market. 

 

PS: It is interesting to note that for Fairfax the average price paid in 2024 ($1,179/share) is already under the current BV ($1,204). Fairfax took out a significant number of shares in 2024. That is exceptional capital allocation. At the time, I am not sure everyone thought Fairfax was getting a good price (taking out shares at a price above BV). My guess is buybacks being done at current prices will also be viewed favourably looking out a couple of years.  

 

 

+1!  Cheers!

[Marco Van Basten]

5 hours ago, Thrifty3000 said:

Because they’re facing a very real existential threat playing out over the next two decades,

 

where the cost of repairing/replacing all things insurable will decline precipitously (along with insurance profits),

 

so they are having to diversify into more durable cash streams that will sustain as long as humans; think real estate, food, banking, shipping, mattresses/sleeping. This is why it pays to invest in intelligent/talented capital allocators with well-aligned interests. (And to size this investment accordingly, the way Prem and Wade have demonstrated this year. Be a lot in, but not all in.)

I hope not.  Banking tends to be a bad business, as Warren stated, there are more banks than good bankers.  Historically, shipping has not been a great business either.  Food is not growing.  

[SafetyinNumbers]

1 minute ago, Marco Van Basten said:

I hope not.  Banking tends to be a bad business, as Warren stated, there are more banks than good bankers.  Historically, shipping has not been a great business either.  Food is not growing.  

It’s more about returns than growth.

[Txvestor]

3 hours ago, Viking said:

 

@Marco Van Basten, it looks to me like Fairfax has been pretty aggressive on the stock buyback front over the past 7 years. At the same time, they doubled the size of their insurance business and they have dramatically increased the size of their equity portfolio. They now have 5 income streams gushing cash. 

 

My guess is if the stock stays around current levels, we will see Fairfax continue to take out a meaningful number of shares in the coming months/year. Bottom line, it appears Fairfax has learned to walk (buy back stock) and chew gum (do everything else) at the same time. I like their balanced, shareholder friendly approach. 

 

Chubb is a fine, well run company. With the hard market slowing and their stock cheap, buybacks is a no brainer for them. Same with other P/C insurance companies. This is great news... a significant amount of capital (that being spent on buybacks) is being removed from the insurance market. 

 

PS: It is interesting to note that for Fairfax the average price paid in 2024 ($1,179/share) is already under the current BV ($1,204). Fairfax took out a significant number of shares in 2024. That is exceptional capital allocation. At the time, I am not sure everyone thought Fairfax was getting a good price (taking out shares at a price above BV). My guess is buybacks being done at current prices will also be viewed favourably looking out a couple of years.  

 

@VikingGreat chart for giving a longer duration patient perspective on capital allocation. I wish it went just one year further back because that would have illustrated the impact on shares outstanding of the Allied world acquisition. I remember I considered the shares undervalued then and it was painful to have to issue shares for such a large acquisition. 
as I recollect, they were 21M or so shares outstanding before the acquisition. Initially, the $54 deal was structured for $10 in cash and the rest as Fairfax shares. Then Fairfax went to their old pension fund friends and managed to find an extra $18 per share in cash by the time of the deal closure. And by so doing managed to keep the dilution to about 5 million shares. If memory serves me correct they acquired 67% and the Olmers etc had 33% for their cash infusion. 
I remember Prem saying back then about the dilution and that they hope to buy those shares back in time. Now that we are back to the 21 million share count or so, we have come full circle.

I know there is still a piece of allied world and other subs outstanding, but I think we have a clear path to full ownership.

Prem is an admirer of Henry Singleton and has referenced his brilliance in capital allocation over his career at the helm of Teledyne. 

And with that and other acquisitions during this last cycle a wonderful insurance franchise is formed, one that is diversified, both geographically, by insurance lines, and with sufficient scale and quality that it provides Fairfax with a phenomenal insurance engine on which they are building a much larger multi engine investment vehicle. 

Edited November 8, 2025 by Txvestor

[gfp]

5 hours ago, Thrifty3000 said:

Because they’re facing a very real existential threat playing out over the next two decades,

 

where the cost of repairing/replacing all things insurable will decline precipitously (along with insurance profits),

 

 

 

What is this?  The cost of repairing/replacing all things insurable will decline precipitously?

 

What is this based on?

[Thrifty3000]

8 hours ago, gfp said:

 

 

What is this?  The cost of repairing/replacing all things insurable will decline precipitously?

 

What is this based on?

Among other things it’s based on:

 

- an awareness that roughly half the cost of all things purchased today is labor cost

 

- a basic understanding of the vector mathematics and technology driving AI and robotic advancement (proofs of concept by Waymo, Tesla, etc.)

 

- understanding there are 150 million jobs in the US and that Tesla alone is scaling production to 11 million bots annually.

 

- an expectation that many human-related risks will be rapidly removed from the insurance equation leaving insurers with a higher proportion of non-human risks like cat risk. (Ie. What will happen to the insurance cost of a trucking company when you remove those pesky driver-related risks from the equation - bots don’t fall asleep at the wheel. What happens to medical malpractice cost - bots don’t mind 24 hour shifts? Workers comp insurance - see ya. You get the point.)

 

^ this is happening. The internet materially impaired staples in our lifetime like newspapers and indoor malls (not destroyed, impaired). Robotics materially impairs the business of insuring human-related risk. Fairfax has time to respond, and they’ve already started diversifying.

[gfp]

32 minutes ago, Thrifty3000 said:

Among other things it’s based on:

 

- an awareness that roughly half the cost of all things purchased today is labor cost

 

- a basic understanding of the vector mathematics and technology driving AI and robotic advancement (proofs of concept by Waymo, Tesla, etc.)

 

- understanding there are 150 million jobs in the US and that Tesla alone is scaling production to 11 million bots annually.

 

- an expectation that many human-related risks will be rapidly removed from the insurance equation leaving insurers with a higher proportion of non-human risks like cat risk. (Ie. What will happen to the insurance cost of a trucking company when you remove those pesky driver-related risks from the equation - bots don’t fall asleep at the wheel. What happens to medical malpractice cost - bots don’t mind 24 hour shifts? Workers comp insurance - see ya. You get the point.)

 

^ this is happening. The internet materially impaired staples in our lifetime like newspapers and indoor malls (not destroyed, impaired). Robotics materially impairs the business of insuring human-related risk. Fairfax has time to respond, and they’ve already started diversifying.

 

Thanks for your response.  I'm watching the tsla shareholder meeting video as I read this.  I will still take the "under" on how rapidly things change but I get what you are saying.

[Phoenix01]

10 hours ago, Txvestor said:

@VikingGreat chart for giving a longer duration patient perspective on capital allocation. I wish it went just one year further back because that would have illustrated the impact on shares outstanding of the Allied world acquisition. I remember I considered the shares undervalued then and it was painful to have to issue shares for such a large acquisition. 
as I recollect, they were 21M or so shares outstanding before the acquisition. Initially, the $54 deal was structured for $10 in cash and the rest as Fairfax shares. Then Fairfax went to their old pension fund friends and managed to find an extra $18 per share in cash by the time of the deal closure. And by so doing managed to keep the dilution to about 5 million shares. If memory serves me correct they acquired 67% and the Olmers etc had 33% for their cash infusion. 
I remember Prem saying back then about the dilution and that they hope to buy those shares back in time. Now that we are back to the 21 million share count or so, we have come full circle.

I know there is still a piece of allied world and other subs outstanding, but I think we have a clear path to full ownership.

Prem is an admirer of Henry Singleton and has referenced his brilliance in capital allocation over his career at the helm of Teledyne. 

And with that and other acquisitions during this last cycle a wonderful insurance franchise is formed, one that is diversified, both geographically, by insurance lines, and with sufficient scale and quality that it provides Fairfax with a phenomenal insurance engine on which they are building a much larger multi engine investment vehicle. 

[mengan]

11 hours ago, Txvestor said:

@VikingGreat chart for giving a longer duration patient perspective on capital allocation. I wish it went just one year further back because that would have illustrated the impact on shares outstanding of the Allied world acquisition. I remember I considered the shares undervalued then and it was painful to have to issue shares for such a large acquisition. 
as I recollect, they were 21M or so shares outstanding before the acquisition. Initially, the $54 deal was structured for $10 in cash and the rest as Fairfax shares. Then Fairfax went to their old pension fund friends and managed to find an extra $18 per share in cash by the time of the deal closure. And by so doing managed to keep the dilution to about 5 million shares. If memory serves me correct they acquired 67% and the Olmers etc had 33% for their cash infusion. 
I remember Prem saying back then about the dilution and that they hope to buy those shares back in time. Now that we are back to the 21 million share count or so, we have come full circle.

I know there is still a piece of allied world and other subs outstanding, but I think we have a clear path to full ownership.

Prem is an admirer of Henry Singleton and has referenced his brilliance in capital allocation over his career at the helm of Teledyne. 

And with that and other acquisitions during this last cycle a wonderful insurance franchise is formed, one that is diversified, both geographically, by insurance lines, and with sufficient scale and quality that it provides Fairfax with a phenomenal insurance engine on which they are building a much larger multi engine investment vehicle. 

Back to 2016

 

[mengan]

Also combined ratio is 91.59% for Q3, not 92%. Not sure why it was rounded up to 92% instead of 91.6%

[MMM20]

1 hour ago, Thrifty3000 said:

Among other things it’s based on:

 

- an awareness that roughly half the cost of all things purchased today is labor cost

 

- a basic understanding of the vector mathematics and technology driving AI and robotic advancement (proofs of concept by Waymo, Tesla, etc.)

 

- understanding there are 150 million jobs in the US and that Tesla alone is scaling production to 11 million bots annually.

 

- an expectation that many human-related risks will be rapidly removed from the insurance equation leaving insurers with a higher proportion of non-human risks like cat risk. (Ie. What will happen to the insurance cost of a trucking company when you remove those pesky driver-related risks from the equation - bots don’t fall asleep at the wheel. What happens to medical malpractice cost - bots don’t mind 24 hour shifts? Workers comp insurance - see ya. You get the point.)

 

^ this is happening. The internet materially impaired staples in our lifetime like newspapers and indoor malls (not destroyed, impaired). Robotics materially impairs the business of insuring human-related risk. Fairfax has time to respond, and they’ve already started diversifying.

Maybe a discussion for another thread, but this feels very Malthusian. Your crystal ball is clearer than mine!

[mengan]

1 hour ago, Thrifty3000 said:

Among other things it’s based on:

 

- an awareness that roughly half the cost of all things purchased today is labor cost

 

- a basic understanding of the vector mathematics and technology driving AI and robotic advancement (proofs of concept by Waymo, Tesla, etc.)

 

- understanding there are 150 million jobs in the US and that Tesla alone is scaling production to 11 million bots annually.

 

- an expectation that many human-related risks will be rapidly removed from the insurance equation leaving insurers with a higher proportion of non-human risks like cat risk. (Ie. What will happen to the insurance cost of a trucking company when you remove those pesky driver-related risks from the equation - bots don’t fall asleep at the wheel. What happens to medical malpractice cost - bots don’t mind 24 hour shifts? Workers comp insurance - see ya. You get the point.)

 

^ this is happening. The internet materially impaired staples in our lifetime like newspapers and indoor malls (not destroyed, impaired). Robotics materially impairs the business of insuring human-related risk. Fairfax has time to respond, and they’ve already started diversifying.

This is an oversimplification.

 

You cannot "remove" risk, only shift it through different parts of the value chain. If robots are driving instead of humans, the risk doesn't disappear, it falls to either the manufacturer, owner or operators of the robots. Change in insurance pricing will depend on the difference in probability for an accident to occur between a human driver and a robot driver. Not all scenarios will have robot drivers be drastically safer than humans.

[Tommm50]

12 hours ago, Parsad said:

 

Cannot naked short in Canada.  Cheers!

I understand it's banned but like the U.S. a ban is only as good as the enforcement of the ban. I understand the Canadian Exchange has a pretty poor record in that area. 

[Tommm50]

Apologies, trouble with the submit button...

[Thrifty3000]

20 minutes ago, MMM20 said:

Maybe a discussion for another thread, but this feels very Malthusian. Your crystal ball is clearer than mine!

For me it’s worth a discussion until I can model the key variables/assumptions needed to handicap the risk (or identify any glaring red flags). Variables like:

 

- percentage of business derived from insuring human-related risks

- human-related risks that likely will be materially reduced by AI/Robotics

- a reasonable timeframe

 

The AI cliff is similar to the interest rate cliff we used to talk about on this forum.

 

Yes, it’s impossible to forecast interest rates and AI adoption. But, it’s not impossible to estimate how a business will be impacted by some reasonable negative case scenarios.

 

I sleep better having a model estimating FFH’s revenue in a zero bound interest rate environment. And, of course, I’m thrilled we’re not in one. However, I don’t yet have any such model related to AI impact, which I find irksome. 

[Thrifty3000]

51 minutes ago, Thrifty3000 said:

For me it’s worth a discussion until I can model the key variables/assumptions needed to handicap the risk (or identify any glaring red flags). Variables like:

 

- percentage of business derived from insuring human-related risks

- human-related risks that likely will be materially reduced by AI/Robotics

- a reasonable timeframe

 

The AI cliff is similar to the interest rate cliff we used to talk about on this forum.

 

Yes, it’s impossible to forecast interest rates and AI adoption. But, it’s not impossible to estimate how a business will be impacted by some reasonable negative case scenarios.

 

I sleep better having a model estimating FFH’s revenue in a zero bound interest rate environment. And, of course, I’m thrilled we’re not in one. However, I don’t yet have any such model related to AI impact, which I find irksome. 

Well, this is going to be easier to solve than I expected, thanks to AI. LOL

 

According to Grok’s dozens of sources, like McKinsey & Co, etc:

 

Only about 20% of P&C insurance volume is due to non-human related risk events.

 

The 80% that’s human-related is concentrated in three areas:

 

- Auto ($300 billion)

- General/Professional liability ($100 billion)

- Workers comp ($35 billion)

 

Current forecasts are projecting 30% to 50% DECLINES in claims volumes in those lines of business by 2040 (including some offset for emerging AI related risks like systemic hacking). Decline will continue beyond 2040, with certain lines - like auto accident policies - being all but extinguished.

 

Ergo, I’m starting to think it might be sensible to start forecasting no more than maybe 3% annual premium volume growth through 2029, and then to start REDUCING premiums by a negative case of, say, 3% annually from there on out!

 

^ And that, ladies and gentlemen, helps explain a lot of what we’ve been seeing from FFH management recently. They’re repositioning for a P&C Industry in runoff!

 

(Ps. I have a penchant for drama.)

[MMM20]

2 hours ago, Thrifty3000 said:

For me it’s worth a discussion until I can model the key variables/assumptions needed to handicap the risk (or identify any glaring red flags). Variables like:

 

- percentage of business derived from insuring human-related risks

- human-related risks that likely will be materially reduced by AI/Robotics

- a reasonable timeframe

 

The AI cliff is similar to the interest rate cliff we used to talk about on this forum.

 

Yes, it’s impossible to forecast interest rates and AI adoption. But, it’s not impossible to estimate how a business will be impacted by some reasonable negative case scenarios.

 

I sleep better having a model estimating FFH’s revenue in a zero bound interest rate environment. And, of course, I’m thrilled we’re not in one. However, I don’t yet have any such model related to AI impact, which I find irksome. 

Maybe I’m wrong but I think the stock trades at 8x P/E on mid-ish cycle globally. I’d worry much more about these risks, prediction markets, etc at mid teens P/E!

 

Edited November 8, 2025 by MMM20

[mengan]

3 hours ago, Thrifty3000 said:

Ergo, I’m starting to think it might be sensible to start forecasting no more than maybe 3% annual premium volume growth through 2029, and then to start REDUCING premiums by a negative case of, say, 3% annually from there on out!

 

 

I'm sorry, but I wouldn't trust Grok with a 10000 foot pole, especially on its predictive capabilities. 

 

 

[xboojum]

Why would Fairfax care about an enormous decline in auto insurance claims? (Separately, why would large language models reduce workers comp claims?)

[TwoCitiesCapital]

1 hour ago, mengan said:

 

I'm sorry, but I wouldn't trust Grok with a 10000 foot pole, especially on its predictive capabilities. 

 

+1 

 

I asked it which of my three goalies I should start for fantasy hockey - it's reasoning was because Sorokin wasn't even scheduled to leave up. And yet, when I went to go check myself, Sorokin was confirmed as the starting goalie. Grok later confirmed with additional prodding/prompting, but was clearly wrong the first time on something easily verifiable. 

 

Data centers will need to be insured. System-wide FSD will need to be insured. Catastrophic failure of the system resulting in 100s of accidents at once will need to be insured. Etc etc etc. I don't see the premiums falling as much as changing hands. 

[Junior R]

7 hours ago, Tommm50 said:

I understand it's banned but like the U.S. a ban is only as good as the enforcement of the ban. I understand the Canadian Exchange has a pretty poor record in that area.

to short the stock you just need to have a brokerage that will lend you the shares...Its pretty easy to short stocks in Canada...

[Viking]

On 11/7/2025 at 10:26 AM, Viking said:

“What possible assurance do you have that (a stock you own) will go up in price? And if you are buying, how much should you pay? What you’re asking here is what makes a company valuable, and why it will be more valuable tomorrow than it is today. There are many theories, but to me, it always comes down to earnings and assets. Especially earnings.” Peter Lynch - One Up on Wall Street

 

How is Fairfax performing? Very well. Why do I think this? Follow earnings.

• Q3 net earnings = $1.15b or $52 per diluted share
• YTD (9 month) net earnings = $3.53b or $156 per diluted share
• Book value = $1,204/share, up 15.1% in 9 months (including $15 dividend)

Excess of FV over CV for associate and consolidated holdings = $2.5b

• This item is not included in reported earnings (it would be if Fairfax's ownership stake in these equities was less than 20%). 
• YTD increase = $1.0b, or $43/share pre-tax = $35/share after-tax

How has Fairfax really been doing (economic value creation)?

• YTD (9 month) economic EPS = $190/share ($156 + $35)
• This is a conservative estimate (it does not include everything).

Bottom line, 2025 is shaping up to be Fairfax’s best year ever from an earnings perspective. The company’s two business units (P/C insurance and investment management) are performing at a very high level. Capital allocation at the company continues to be best-in-class. Fairfax’s future has never been brighter.

I am going to do my review of Fairfax’s Q3-2025 results in a series of posts. Today, we are going to review the P/C insurance results. Our review is going to be at a very high level - we are going to zoom out and try and look at the big picture. Getting the the big picture right is a critical part of getting an investment thesis right.
 

—————

 

Fairfax is an interesting and unique company. It has two complementary businesses: P/C insurance and investment management. P/C insurance is the core engine (float) that powers the whole company. As I will explain below, P/C insurance is becoming an even more important part of the company. 

 

Over the past 10 years, Fairfax has been completely transformed as a company. One important example of this transformation has been the changes that have been happening in its P/C insurance business. 

 

Historically, investors bought Fairfax because of its investment management business. Not because of its P/C insurance business, which was considered to be low quality. 

 

Fairfax got started on transforming its P/C insurance business back in 2011, when Andy Barnard was appointed to oversee it. Fairfax’s P/C insurance companies have been on a 14 year journey of continuous improvement. Importantly, it takes years to transform a P/C insurance business (as Markel has been learning in recent years). And it takes years for the improvements to finally show up in business results (better underwriting profitability). And that is what has been slowly playing out at Fairfax. 

 

This is a really important development for long term investors. Because it means Fairfax now has two high quality businesses: P/C insurance and investment management. Of course, this bodes well for the future returns that the company will be able to deliver.

 

This should result in higher earnings in the future. This should result in the company generating a higher ROE. Over time this should result in the stock receiving a higher multiple from investors and analysts. And as the multiple expands, the narrative around the company should also improve.

 

Information advantage

 

The interesting thing is, despite the growing evidence in recent years, investors and analysts have been slow to recognize or reward the improving quality of Fairfax’s P/C insurance business. This means it is likely not yet built into the price of the stock.

 

And as we all know, one of the best ways to do well with a stock is to have an information advantage - to know something important that is (temporarily) being missed by most investors and analysts. This often happens with small companies like Fairfax (that tend to be underfollowed and misunderstood). 

 

Investors on this board have been documenting the transformation that has been slowing playing out at Fairfax over the past 5 years. This has given them a big information advantage over other investors and analysts. It has been like having a crystal ball - being able to see the future 6 to 12 months before it is understood and appreciated by the larger investment community.

 

What is interesting is the story at Fairfax has been continually changing (getting better) for each of the 5 years. This is not normal - most companies are not that dynamic. Importantly, the story at Fairfax is still changing (getting better) in important ways that are not yet understood or appreciated. Like the improved quality of their P/C insurance business. In important respects, the Fairfax story (this iteration of the company) is just getting started - of course, that is a wonderful set-up for a long term investor.

 

Let’s get back to our story - Fairfax’s Q3 results. 

 

Each quarter when Fairfax reports results we get lots of new and important information. When analyzing the P/C insurance business, the key question I have been asking for the past couple of years is: ‘Does Fairfax’s have a high quality P/C insurance business?’ Getting the answer to that question right is a critical input in how to properly value the company. 

 

What did we learn in from Fairfax’s Q3-2025 results?

 

1.) The top line is slowing. 

• Q3 net premiums written = $6.55 billion, up 2.1%. Excluding a short term impact at GIG, they were up 3.1%. Interestingly, this is lower growth than P/C insurance peers.

2.) Underwriting profit came in ahead of expectations.

• Q3 underwriting profit = $540.3 million (up from $389.7 million in 2024). 
• Q3 CR = 92.0% (was 93.9% in 2024), from decreased catastrophe losses of $150.0m (2024 - $434.5m) and growth in business volumes, partially offset by a decrease in net favourable PY reserve development to $111.2m (2024 - $130.5m).

Ok. What does this mean?

 

Is a slowing top line a bad thing for Fairfax? 

 

No, a slowing top line is not a bad thing. It is the opposite. It suggests Fairfax’s many P/C insurance companies are exercising great discipline with their underwriting. And more discipline than peers. 

Managing the insurance cycle

 

“Be fearful when others are greedy, and greedy when others are fearful” Warren Buffett

 

Over the past decade, Fairfax has provided a textbook example of how to manage the P/C insurance cycle. Grow aggressively by acquisition in a soft market (2015-2017). Grow aggressively organically in a hard market (2019 to 2025). Fairfax grew faster than peers at the beginning of the hard market (the ‘be greedy’ stage). And now it is growing at a slower rate at the end of the hard market (the ‘be fearful’ stage). What Fairfax is doing is exactly what Warren Buffet has been preaching for decades should be done. (Are investors and analysts paying attention?)

 

Now investors and analysts will not like that Fairfax is growing more slowly than peers. And of course, this is completely wrongheaded. 

 

What are we learning about Fairfax’s P/C insurance business from Q3 results?

 

It looks like Fairfax is managing the P/C insurance cycle very well. Top line growth is slowing. And their underwriting results are very good and improving (getting better). This is further evidence that Fairfax has a high quality P/C insurance business - and it is higher quality than most analysts and investors think. With each passing quarter, this is becoming more apparent. 

 

—————-

 

Another thing many investors and analysts are missing

 

Despite a slowing top line number (net premiums written), Fairfax will still be able to grow its insurance rate at a pretty good rate. 

 

How will it do this?

 

1.) Per share metrics are what really matter to investors. 

 

Net premiums written (NPW) matter. What matters much more is NPW/share.  This is a really simple and fundamental concept. And it is ignored by most investors and analysts. 

 

My guess is Fairfax might buy back about 4% of shares outstanding in 2025. Guess what that does to NPW/share? Yes, it boosts it by 4%. 

 

So if NPW increases by 3% and the share count falls by 4% then NPW/share increases by about 7%. 

 

Of course, one of the keys is the price that is paid for the shares. Trading today at 1.3 x book value, Fairfax’s stock is cheap. So share buybacks are a great use of capital. And a great way to grow NPW/share and float/share. 

 

Share buybacks are not new to Fairfax. From 2017 to 2024, the company has reduced effective shares outstanding by 22.9% at an average cost of $637/share. The value creation for long term shareholders has been enormous, further spiking the important per share measures (like NPW and float). Fortunately, Fairfax gets it. 

 

2.) Minority partners in Allied World and Odyssey

 

Fairfax does not own 100% of its two largest insurance companies: Allied World and Odyssey. Minority shareholders own 16.6% of Allied World and 9.99% of Odyssey. This means that today  Fairfax shareholders are not getting 100% of the earnings from these two companies. 

 

When Fairfax buys out its minority partners in the coming years it will be kind of like making two large P/C insurance acquisitions. Importantly, Fairfax is able to buy out their partners at a very favourable price (due to the call option feature put in place when the deals were initially put in place).

Summary

 

As a result, Fairfax has a number of very good options at its disposal to continue to grow its insurance on a per share basis in the coming years. 

 

What Fairfax has done here is unique in P/C insurance. But because it is different it is not appreciated or valued by investors or analysts.

 

Bottom line, Fairfax has many levers it can pull to continue to grow its P/C insurance business in the coming years - even as the hard market slows. 

Edited November 9, 2025 by Viking

[gfp]

1 hour ago, xboojum said:

Why would Fairfax care about an enormous decline in auto insurance claims? (Separately, why would large language models reduce workers comp claims?)

 

Fairfax specifically might not have much exposure to the auto insurance business but other insurers (Berkshire for example) do, and this area is likely to be disrupted relatively soon compared to other areas of insurance.  I still think the messy middle period where some drivers are AI and some are human is a lot messier and longer lasting than most people are penciling in.  The same AI-led disruption that we are talking about is going to impoverish a ton of people.  It isn't an age of abundance for all individuals.

 

For workers comp claims it sounds like they are referencing humanoid robots and other specialized AI agents that will replace human workers in the labor force.  People who think AI is just large language models stringing convincing sentences together in some kind of chat-bot parlor game are missing where the frontier of AI is today.  Forget about 18 months from now or 5 years from now.

 

I think things like humanoid robots replacing humans in the workforce is going to progress slower than, say, Elon Musk - who described humanoid surgeons being better than human surgeons as some kind of near term certainty.  All this stuff takes longer than people think.  But completely autonomous robo-taxis are very close and those will be the first big bang that wakes a bunch of people up to vision based AI robots.  That's what a self driving tesla is after all.  A robot that uses cameras to see.

Edited November 8, 2025 by gfp

[Marco Van Basten]

Does anyone know what is the long-term say 10, 15, 20 and 25 year track record of Fairfax's equity investments?  (Them not blowing up on the fixed income side in 2022 like most insurance and bank companies did is clearly incredible.  Similarly underwriting is great.)

[Parsad]

1 hour ago, Junior R said:

to short the stock you just need to have a brokerage that will lend you the shares...Its pretty easy to short stocks in Canada...

 

It's easy to short...not easy to naked short.  Fail to delivers is not high in Canada...never has been high.  Not like the U.S. back between 2003 and 2010.  Cheers!