Fairfax 2025

[glider3834]

5 hours ago, wondering said:

Is Fairfax too complex to be valued at the same metrics as its insurance peers?

 

I raise this question because I have read this criticism twice from analysts.  Once from the RBC analyst mentioned above (see attached full report) and another from an analyst on Seeking Alpha https://seekingalpha.com/article/4790903-fairfax-holdings-misunderstood-conglomerate-with-some-upside

 

What is the complex part of Fairfax?

• TRS investment 
• private equity holdings
• other

I can understand how the TRS might be view as complex (I have to admit re-read the explanation of the TRS 5 times before I begin to understand it).  Are the private holdings too difficult to model for income projections?

 

I know Safetybynumbers has argued that Fairfax has the low value metrics because it doesn't screen well for the quants.  Maybe screening well and complexity is the same thing?

 

Could this problem be fixed by greater disclosure or explanation in the annual report?

 

 

1654 (1).pdf 1.38 MB · 17 downloads

I actually feel like they have lifted the hood a bit on the value of their private holdings with the annual report showing multiples to support their carrying values - that allows us test for reasonableness  

 

I also see complexity in what Fairfax does a function of their own creativity in deal making eg  preserving optionality (retaining a small equity stake in what they have sold eg Sigma, or taking equity warrants with convertible notes, Orla) or providing downside protection (leading an investment with mostly debt instruments eg Vacatia Blizzard & just a small equity commitment) 

 

 

Edited August 6, 2025 by glider3834

[glider3834]

Below I have taken the average NYMEX futures pricing (USD per MMbtu) (source https://www.nrg.com/resources/energy-tools/nymex-settlement-history.html )& compared to Exco's results which tend to lag as they are likely hedging to lock in price for future production

1H 2025 share of loss (20.1) M (NYMEX futures avg 2.27MMbtu in 2024)

2024 share of profit 39.6M (NYMEX futures avg 2.73 in 2023)

2023 share of profit 129.1M (NYMEX futures avg 6.64 in 2022)

2022 share of profit 81.9M (NYMEX futures avg 3.84 in 2021)

2021 share of loss (41.2) M (NYMEX futures avg 2.07 in 2020)

 

For 2025, avg NYMEX futures pricing (USD per MMbtu) is 3.45  so would expect Exco's results would improve in 2026. It also looks like their profitability break even might be around mid 2 area.

[SafetyinNumbers]

1 hour ago, glider3834 said:

I actually feel like they have lifted the hood a bit on the value of their private holdings with the annual report showing multiples to support their carrying values - that allows us test for reasonableness  - it doesn't make strategic sense for Fairfax to tell us what they think those private holdings are worth IMHO - they would want to play those cards close to their chest with any future potential sale negotiation

 

I also see complexity in what Fairfax does a function of their own creativity in deal making eg  preserving optionality (retaining a small equity stake in what they have sold eg Sigma, or taking equity warrants with convertible notes, Orla) or providing downside protection (leading an investment with mostly debt instruments eg Vacatia Blizzard & just a small equity commitment) 

 

 

Perhaps a function of the short report which argued they inflated numbers. A better report might be  all the ways carrying values understate fair value!

 

[Hsmpanl]

1 hour ago, glider3834 said:

Below I have taken the average NYMEX futures pricing (USD per MMbtu) (source https://www.nrg.com/resources/energy-tools/nymex-settlement-history.html )& compared to Exco's results which tend to lag as they are likely hedging to lock in price for future production

1H 2025 share of loss (20.1) M (NYMEX futures avg 2.27MMbtu in 2024)

2024 share of profit 39.6M (NYMEX futures avg 2.73 in 2023)

2023 share of profit 129.1M (NYMEX futures avg 6.64 in 2022)

2022 share of profit 81.9M (NYMEX futures avg 3.84 in 2021)

2021 share of loss (41.2) M (NYMEX futures avg 2.07 in 2020)

 

For 2025, avg NYMEX futures pricing (USD per MMbtu) is 3.45  so would expect Exco's results would improve in 2026. It also looks like their profitability break even might be around mid 2 area.

Nat gas expectations were high for ‘25 and beyond but have been declining. A lot of sales processes for gas assets were cancelled. Don’t see a lot to be excited about with Exco.

[glider3834]

49 minutes ago, SafetyinNumbers said:

Perhaps a function of the short report which argued they inflated numbers. A better report might be  all the ways carrying values understate fair value!

 

I was wondering about Poseidon where the fair value has not moved - it was 2046 at Jun'25 and was also 2046 at Jun'23 - does IFRS require them to use fair value in reporting at which the buyout took place around 2023?

 

 

Edited August 6, 2025 by glider3834

[SafetyinNumbers]

58 minutes ago, glider3834 said:

I was wondering about Poseidon where the fair value has not moved - it was 2046 at Jun'25 and was also 2046 at Jun'23 - does IFRS require them to use fair value in reporting at which the buyout took place around 2023?

 

 

I think that’s probably an example of conservatism. IFRS would look for an objective value first for a fair value assessment and Fairfax could be aggressive and use Sokol’s estimate but instead they just use last trade. 

Edited August 6, 2025 by SafetyinNumbers

[Txvestor]

15 hours ago, gfp said:

I think you guys are wrong about QE and QT.  They do almost nothing and often times have at least as much effect in the opposite direction of their intended outcome.  One thing QE accomplished was tightening mortgage spreads to their treasury benchmark but I think most of that effect was the effective duration of a 30 year mortgage plummeting during an extended refinancing boom and the appropriate treasury benchmark moving closer to the 2 year bond than the 10 year.

 

QE removes useful treasury securities from the private sector and replaces them with bank reserves, a neutered form of token that only the largest Fed member banks can use.  Since lending by the largest banks is not in any way constrained by the level of bank reserves in the system (lending is constrained by demand for loans / balance sheet / regulation / risk adjusted capital requirements), excess bank reserves parked at the fed is not the same as the "cash" most people imagine QE raining down on the economy.

 

A treasury security is the bedrock collateral of the world financial system and can be leveraged, pledged, transformed, swapped, repo'd - it is useful collateral to the private sector.

 

A bank reserve is next to useless unless the level of reserves in the system is too low (it's huge relative to history).

 

At the time they were big on QE, the treasury coupon securities they were removing from the private sector paid interest and the bank reserves paid nothing or next to nothing - removing interest income from the private sector.  That is why the Fed's balance sheet was earning a big "profit" during those years - remitting those profits to the treasury.  That is the same as a tax - interest income that would have been earned by the private sector is earned by the Fed instead and remitted to the treasury to help reduce the deficit.  Tax.  (opposite of stimulus)

 

Presently, the excess reserves pay a higher rate than the treasury coupon securities on their books so the Fed's balance sheet is producing losses.  This is stimulus.  When the Fed pays more net interest into the private sector than it earns it is additive to the fiscal stimulus.

Thank you for that clarification. I wasn't aware that the banking reserves were involved in the treasury sales process. 
irrespective it's a force function on LT interest rates and often to sub inflation levels. 
with the federal govt debt where it is, and with this president it's not clear to me that LT rates won't be jawboned lower. 
 

All of that said, it will be interesting to see how they position over the next quarter. Bradstreet may indeed be pivoting to something or see something coming which we don't. 

[villainx]

6 hours ago, djokovic1 said:

That perception will inevitably change over time as long as the stellar execution continues.

 

I don't know, that Blackberry thing is still there.

 

 

6 hours ago, SafetyinNumbers said:

Fairfax also invests in securities most quality investors (growing majority since the GFC) wouldn’t touch like commodities and emerging markets.

 

For me, definitely a feature!  

 

 

[TwoCitiesCapital]

I don't think it's a 'complexity' issue as much as it is laziness. 

 

Fairfax has different investments accounted for in different ways - market to market, consolidated, and equity accounted. One has to follow each and know how each flows through to understand to the true value vs what is reported. And also have to follow the larger ones, at minimum, casually to know how they may be expected to flow through in an upcoming quarter. 

 

None of it is complicated - but it is significantly more work than what the analysts would have to do for an insurer that simply owns corporate bonds. 

[gfp]

3 minutes ago, TwoCitiesCapital said:

I don't think it's a 'complexity' issue as much as it is laziness. 

 

Fairfax has different investments accounted for in different ways - market to market, consolidated, and equity accounted. One has to follow each and know how each flows through to understand to the true value vs what is reported. And also have to follow the larger ones, at minimum, casually to know how they may be expected to flow through in an upcoming quarter. 

 

None of it is complicated - but it is significantly more work than what the analysts would have to do for an insurer that simply owns corporate bonds. 

 

 

Yeah, RBC's recent report was like - 'earnings were way ahead of our estimates because of the gains on the TRS.' 

 

I mean, that was like the most predictable figure in the entire earnings release.

[TwoCitiesCapital]

2 minutes ago, gfp said:

 

 

Yeah, RBC's recent report was like - 'earnings were way ahead of our estimates because of the gains on the TRS.' 

 

I mean, that was like the most predictable figure in the entire earnings release.

 

Yup

 

It was pretty clear just from tallying obvious items that $90-100 in economic value was on the table. But then adjusting for how that gets accounted for, taxes, surprises on insurance (positive or negative) etc was clearly going to bring the number down -  not difficult, but is just more work. 

 

I'd argue the top number is what matters and is easier to get to - who cares what is reported unless if you're the analyst guessing. 

Edited August 6, 2025 by TwoCitiesCapital

[Munger_Disciple]

On 8/5/2025 at 6:09 AM, gfp said:

I think you guys are wrong about QE and QT.  They do almost nothing and often times have at least as much effect in the opposite direction of their intended outcome.  One thing QE accomplished was tightening mortgage spreads to their treasury benchmark but I think most of that effect was the effective duration of a 30 year mortgage plummeting during an extended refinancing boom and the appropriate treasury benchmark moving closer to the 2 year bond than the 10 year.

 

QE removes useful treasury securities from the private sector and replaces them with bank reserves, a neutered form of token that only the largest Fed member banks can use.  Since lending by the largest banks is not in any way constrained by the level of bank reserves in the system (lending is constrained by demand for loans / balance sheet / regulation / risk adjusted capital requirements), excess bank reserves parked at the fed is not the same as the "cash" most people imagine QE raining down on the economy.

 

A treasury security is the bedrock collateral of the world financial system and can be leveraged, pledged, transformed, swapped, repo'd - it is useful collateral to the private sector.

 

A bank reserve is next to useless unless the level of reserves in the system is too low (it's huge relative to history).

 

At the time they were big on QE, the treasury coupon securities they were removing from the private sector paid interest and the bank reserves paid nothing or next to nothing - removing interest income from the private sector.  That is why the Fed's balance sheet was earning a big "profit" during those years - remitting those profits to the treasury.  That is the same as a tax - interest income that would have been earned by the private sector is earned by the Fed instead and remitted to the treasury to help reduce the deficit.  Tax.  (opposite of stimulus)

 

Presently, the excess reserves pay a higher rate than the treasury coupon securities on their books so the Fed's balance sheet is producing losses.  This is stimulus.  When the Fed pays more net interest into the private sector than it earns it is additive to the fiscal stimulus.

 

Excellent post @gfp! 

 

QE did appear to have created a large number of asset bubbles due to deprivation of safe income producing securities from the hands of private sector. These include VC, PE, PD, Crypto, real estate & publicly traded stock valuations, especially in MAG-7 & other tech stocks. 

Edited August 6, 2025 by Munger_Disciple

[Pellom]

1 hour ago, TwoCitiesCapital said:

I don't think it's a 'complexity' issue as much as it is laziness. 

 

Fairfax has different investments accounted for in different ways - market to market, consolidated, and equity accounted. One has to follow each and know how each flows through to understand to the true value vs what is reported. And also have to follow the larger ones, at minimum, casually to know how they may be expected to flow through in an upcoming quarter. 

 

None of it is complicated - but it is significantly more work than what the analysts would have to do for an insurer that simply owns corporate bonds. 

 

I also think there is a significant set of analysts who won't look at it because it is Canadian. 

 

[vinod1]

I don't think complexity is the reason. There are a lot more complex companies for which analysts have detailed reports.

 

I think it is underfollowed for a couple of reasons:

 

1. It is unconventional. It does not align with how the other P&C companies operate. And unconventionality is a significant source of value generation.

 

2. It does not generate much investment banking fees I think. They are the ones coming up with creative ideas, not the investment banks!

 

This is similar to Berkshire before the year 2000.

 

Vinod

 

 

[Viking]

It is very interesting to read the updated analyst reports on Fairfax after results are released each quarter. With each report we learn as much (if not more) about the analyst/firm than we learn about Fairfax.

 

The biggest single driver of economic results at Fairfax in Q2 was the significant increase in the value of Eurobank. This was captured in excess of FV over CV for associate the consolidated equity holdings, which increased $1 billion in Q2 to $2.4 billion. This was an increase in the quarter = $42/diluted share (pre-tax) or about $34/share (after tax). This puts 'economic EPS' in Q2 at about $96/diluted share ($62 + $34).

 

What is interesting is some analysts don't even discuss the increase in 'excess of FV over CV' in their Q2 reports on Fairfax. Does that mean it didn't happen? Or that it is not important?

 

What is interesting is if Fairfax owned less than 20% of Eurobank (they own 32%), the significant gain would have been reflected in Q2 earnings. 

 

What is puzzling is Fairfax's communication is very clear and transparent when it comes to 'excess of FV over CV.' So 'I didn't know' is not a good excuse. 

 

As updated Q2 reports come out we are learning which analysts/firms understand Fairfax’s business model. And which clearly do not.  

Edited August 6, 2025 by Viking

[gfp]

And Eurobank keeps on trucking.  At least FFH's sales into Eurobank's repurchases will book the types of gains these analysts can count.

 

 

Edited August 7, 2025 by gfp

[Parsad]

2 hours ago, vinod1 said:

I don't think complexity is the reason. There are a lot more complex companies for which analysts have detailed reports.

 

I think it is underfollowed for a couple of reasons:

 

1. It is unconventional. It does not align with how the other P&C companies operate. And unconventionality is a significant source of value generation.

 

2. It does not generate much investment banking fees I think. They are the ones coming up with creative ideas, not the investment banks!

 

This is similar to Berkshire before the year 2000.

 

Vinod

 

 

 

+1!  I agree.  I remember back in 1998, not that many business people, let alone analysts understood Berkshire.  We only had a couple of books on Buffett then...Of Permanent Value, The Money Masters...maybe that was it back then!  My only source of information was the letters themselves which went online in late 1998 and the Motley Fool BRK Board.  Today, every Tom, Dick and Harry is quoting Buffett.

 

Back in 2002, maybe a couple of thousand people really knew about Fairfax.  It's taken almost 20 years just to get the analysts that do cover Fairfax now.  The Fairfax meetings today are a quarter the size of the first Berkshire meeting I went to in 2001.  Probably why at the right price, it may be a generational opportunity to find a worthy company growing capital for the next 30-40 years!  Cheers!

[nwoodman]

On 7/6/2025 at 6:37 AM, Viking said:

@Maverick47, my view is a number of factors have come together for Fairfax over the past 5 years that have resulted in a unique business model in P/C insurance. 
 

I don’t think it is the Berkshire Hathaway business model (conglomerate). 
 

Fairfax has built a unique platform. Their focus the past 10 years has been to aggressively grow their insurance business. Much more so that BRK ever has. As a result, Fairfax has been increasing the amount of leverage it has (to float) on a per share basis.
 

At the same time, the quality of their insurance business has improved dramatically. I just finished reading Mark Adee’s book, Once and Future Crum and Forster, and it provided more insight into the many improvements. 

• https://www.cfins.com/the-once-and-future-cf-landing/

At the same time, Fairfax has built out a wonderful investment management platform. It has spent decades building extensive capabilities. Venture capital (start-up) investor. Private equity investor (LBO light). International investor (India, Greece). Value investor. Sometimes they are planting acorns that are growing to oaks (to steal a metaphor from Mark Adee). Other times that are more tactical. 

They have a wonderful breadth of capabilities. 
 

At the same time, Fairfax has been building out its business/relationships with outstanding external capital allocators. This is having a big impact on deal flow - they are likely now getting many more juicy opportunities than they have the money for - a first class problem to have. 
 

Bottom line, Fairfax’s investment management business is much more diversified than Berkshire Hathaway has ever been. That also bodes well for longevity of its model - it is increasingly not reliant on any one person. 
 

The external environment has also changed in recent years. We have moved away from a zero interest rate world (with suppressed volatility). And we appear to moving into a higher inflation/interest rate regime (with higher volatility). The current external environment is ideally suited to Fairfax and their business model. 
 

When I put it all together, I think Fairfax is poised to perform exceptionally well over the next 5 years. I think a 15% ROE is a good baseline number to use (on average). I think this estimate has a margin of safety built into it. 

Marc Adee has been doing the rounds on the book.  A couple of recent interviews:

 

https://cpcusociety.libsyn.com/unpacking-200-years-of-insurance-history-with-marc-adee

 

 

 

[Viking]

Earnings Estimate For Fairfax For 2025 and 2026

 

Below is my updated earnings estimates for Fairfax for 2025 and initial estimate for 2026. This forecast includes learnings from Fairfax’s AGM (in April) and Q1 and Q2 earnings reports. And general developments since my last update (which was on March 15). My old estimate for 2025 was diluted EPS of $161/share. Yes, I know... I am such a Fairfax fanboy!

 

Let me know your thoughts. Too optimistic? If so, why? Or too pessimistic? If so, Why?

 

Summary

 

For 2025, my current estimate is diluted EPS at Fairfax will be about $183/share. This does not include the increase in excess of fair value over carrying value for Fairfax’s associate and consolidated holdings. I expect ‘excess of FV over CV’ to deliver another $38/share (after tax) in value in 2025. So, my forecast for ‘economic EPS’ for Fairfax in 2025 is about $221/diluted share.

 

For 2026, my initial estimate is diluted EPS at Fairfax will be about $182/share. This does not include the increase in excess of fair value over carrying value for Fairfax’s associate and consolidated holdings. I expect ‘excess of FV over CV’ to deliver another $10/share (after tax) in value in 2026. So, my forecast for ‘economic EPS’ for Fairfax in 2026 is about $192/diluted share.

 

Over the 4-year period, from 2023 to 2026E, diluted EPS at Fairfax is poised to average about $175/share and the increase in ‘excess of FV over CV’ is poised to average about $22/share (after tax). This would put the average increase in ‘economic EPS’ each year from 2023 to 2026E at about $197/diluted share. This amount is likely a good number to use as a baseline (starting point) when trying to estimate ‘economic EPS’ for Fairfax in future years (2027 and further out). With shares trading today at around $1,750, this gives it a PE of about 9x 'normalized economic EPS.' That is very cheap for a company of Fairfax's quality, earnings and prospects.  

 

 

Will retained earnings be re-invested in a way that builds value for shareholders?

 

Perhaps the hardest piece to forecast with Fairfax today is what they will be doing with the substantial amount of earnings that they are currently generating (about $4.5 billion per year). And the impact the re-investment of current earnings will have on future earnings. Both the size - how much. And the speed - how fast. 

 

When it comes to re-investing earnings, Fairfax has lots of very good options:

• Grow insurance - Continuation of the hard market? Bolt-on acquisitions?
  • Buy out minority partners in insurance (Allied World/Odyssey)?
• Buy fixed income securities?
• Buy equities?
• Buy back a meaningful amount of Fairfax stock? 

What Fairfax does will determine which of Fairfax’s 5 income streams will grow the most. Because we don’t know what Fairfax will actually do when we build our forecast, we have to guess which income streams will benefit and by how much. Please keep this in mind when you review our forecast – it is a guess at a point in time. As results come in each quarter, we will update our forecast to reflect the new news.   

 

Looking at the last 5 years, the management team at Fairfax has done an outstanding job with capital allocation. My guess is they will continue to make good decisions (on balance) and this will benefit shareholders in the coming years – likely providing a tailwind to my forecasts for 2025 and beyond.

 

What are the key assumptions we have used to build our forecast?

 

To estimate future EPS, BVPS and ROE for Fairfax, an investor needs to think about three things:

1. Combined ratio – How good is the P/C insurance business?
2. Total return on the investment portfolio – How good is the team at Hamblin Watsa?
3. Capital allocation – How good is the senior management team?

 

 

Note, when calculating the total return for the investment portfolio, I am including the change each year for ‘excess of FV over CV’ for associate and consolidated holdings. As stated earlier, this is value that is being created by Fairfax and it needs to be incorporated into models.

 

Interest rates: I am assuming interest rates remain roughly at current levels (at August 2, 2025). Of course, this will not be the case. The average duration of Fairfax’s fixed income portfolio (about 2.4 years) is less than the average duration of their insurance liabilities (a little under 4 years?). Changes in interest rates will offset (in ‘net gains/losses on investments’ and ‘effects of discounting and risk adjustment- IFRS 17’) but not perfectly. Bottom line, with IFRS 17, changes in interest rates will result in less volatility in Fairfax’s reported final results moving forward – but there will be some volatility. The investment community should like that (the less volatility part).

 

Below is a 6-year snapshot of earnings for Fairfax. It communicates in a concise manner the dramatic transformation that has happened at the company, beginning in 2021. There has been a spike in operating income per share – from an average of $39/share from 2016-2020, to $235/share in 2024. This much higher amount has become the new baseline for the company. For 2025, my estimate has operating income coming in at $229/share, which is a 487% increase from the average from 2016-2020. ‘Normalized earnings’ at Fairfax have moved to a much higher level – and, importantly, this higher level looks durable/sustainable.

 

 

What are analysts’ current earnings estimates for Fairfax?

 

At Aug 6, 2025, analysts’ average diluted EPS estimates for Fairfax are:

• 2025 = US$185
• 2026 = US$187

These estimates do not include changes in the value of ‘excess of FV over CV’ for associate and consolidated holdings. Yes, analysts today understand Fairfax much better than they did a year or two ago. As a result, their average diluted EPS estimates are much more accurate than they were. 

 

Here are the most important assumptions that went into each line item in our forecast:

 

1.) Underwriting profit: Estimate = $1.55 billion in 2025.

• Net premiums written growth of 5% in 2025. 
  • Yes, the hard market is slowing. 
• Combined ratio (CR) of 94% in 2025.
  • Catastrophe losses: 2025 will finish the year higher than 2024.
  • Reserve releases: continuation of the positive trend observed in 2024.

2.) Interest and dividend income: Estimate = $2.6 billion in 2025.

• Tailwinds:
  • The size of the fixed income portfolio continues to increase (from $47 to $50 billion.
  • Interest income from $810 million investment in Blizzard Vacatia Equity Partners.
  • Continued growth of mortgage loan portfolio (managed by Kennedy Wilson).
• Headwinds:
  • Lower short-term interest rates.
• The average yield of the fixed income portfolio was about 5.1% in 2024. For 2025 we estimate the average yield will be about 5.1%.

3.) Share of profit of associates: Estimate = $770 million in 2025.

• Earnings at Eurobank and Poseidon/Atlas should continue to chug along.
• Headwind in 1H not expected to continue in 2H:
  • Unrealized loss of $157.7 million on Waterous III investment in Greenfire.
• Two more headwinds for 2025:
  • Shift of Peak Achievement to a consolidated holding ($57m). 
  • Sale of Stelco ($18m).

4.) Effects of discounting and risk adjustment (IFRS 17):

• The two key drivers for this bucket are the trend in net written premiums of the insurance business and changes in interest rates.
• Net written premiums growth of 5% in 2025 should be a small tailwind. 
• We assume Treasury yields remain constant at June 30 levels.

This bucket is difficult to model – therefore, my confidence level in my estimates is low. 

 

Important. Plug number. For actuals, we are combining Bucket 4 (impacts from IFRS 17) and Bucket 5 (life insurance and run-off) together. We do not get either of these numbers from Fairfax. However, we can back in to the total for these two buckets (because they give us all the other numbers we use in our earnings model).  

 

5.) Life insurance and runoff: Estimate = a loss of $150 million in 2025.

• Adverse reserve development at runoff should be offset by earnings from the life insurance business in Greece.

6.) Non-insurance consolidated operations (Other revenue – expenses): 

• Estimate = $270 million in 2025. 
• This income stream has a number of significant tailwinds for 2025:
  • Acquisitions: Sleep Country (closed Oct 1, 2024).
  • Shift from associates: Peak Achievements and Meadow Foods (Q4, 2024).

This bucket is poised to grow nicely in the coming years. 

 

7.) Interest expense: Estimate = $828 million in 2025.

• Run rate: Q2 2025 = $208 million 

8.) Corporate overhead and other: Estimate = $470 million in 2025.

• An increase from 2024, which was $450 million.

 9.) Net gains on investments: Estimate = $2.4 billion in 2025

• Key drivers in 1H-2025:
  • FFH-total return swaps
  • Increase in value of remaining mark to market holdings.
  • Benefit from decline in bond yields.
  • Realized gain on sale of Sigma in Q1 = $178.7 million

10.) Gain on sale/deconsol of insurance sub: Estimate = $0 in 2025

 

This is where I put the large asset sales/revaluations. These items are very lumpy and therefore difficult to forecast precisely for any one year. Sometimes these gains show up as a separate line item and other times they show up in investment gains. I like to break them out at the start of the year as a separate line item. Over the past 5 years, large one-time gains from asset sales/revaluations have averaged about $500 million per year – so using an estimate of $400 million for 2026 seems like a reasonable and conservative estimate.

 

Bottom line, this bucket is a wild card. But Fairfax has a long history of surfacing the significant value that is residing/hidden on its balance sheet. When they do, we see significant realized gains (from both insurance and non-insurance holdings).

 

11.) Income taxes: Estimate = 21% for 2025

 

Fairfax’s tax rate was 24.4% in 2024. We expect investment gains to be a significant driver of earnings in 2025 – and these are taxed at a lower rate.

 

12.) Non-controlling interests: Estimate = 7% for 2025 

 

As Fairfax continues to take out its minority P/C insurance partners this number should shrink. In 2024, Fairfax took out its minority partner in Brit. Later in 2025 it is likely they will increase their ownership in Allied World. As minority P/C insurance partners are taken out, non-controlling interest should continue to shrink in size.

 

13.) Effective Shares Outstanding (year-end): Estimate = 21.4 million for 2025

 

We focus on effective shares outstanding as this is what Fairfax highlights in their reporting. Fairfax finished 2024 with effective shares outstanding = 21.7 million. This was down 1.3 million in 2024 (from 23.0 million at 2023YE). In 2025, we estimate Fairfax will reduce effective shares outstanding by 300,000 which is a slower pace compared to 2024.

 

Additional notes:

• ‘Underwriting profit’: Includes insurance and reinsurance; does not include life insurance and run-off.
• ‘Interest and dividends’ and ‘share of profit of associates’: Includes insurance, reinsurance and life insurance and run-off.

Edited August 7, 2025 by Viking

[Santayana]

@VikingI think I would bump up that 2025E CR a bit.  It's at 95.9% after Q2 and historically Q3 is the highest number of the year.  I realize the CA fires in Q1 are skewing things and that might not be the case this year, but just thinking conservatively.

[Viking]

14 minutes ago, Santayana said:

@VikingI think I would bump up that 2025E CR a bit.  It's at 95.9% after Q2 and historically Q3 is the highest number of the year.  I realize the CA fires in Q1 are skewing things and that might not be the case this year, but just thinking conservatively.

 

@Santayana, I have done as you suggest each of the past three years and I have been proven wrong each time. The reason is not that Q3 losses are not elevated but that Q4 tends to be quite favourable (the level of reserve releases will be key). 

 

With my 2025 forecast I tried to nail the final number (we are less than 5 months away). 50% chance it is a little higher. And a 50% chance it is a little lower. 

 

With my 2026 forecast I think there is a little more conservatism. 55% it is higher. 45% chance it is lower.  

 

I appreciate the question. Please keep them coming.

Edited August 7, 2025 by Viking

[Maverick47]

1 hour ago, Viking said:

The reason is not that Q3 losses are not elevated but that Q4 tends to be quite favourable (the level of reserve releases will be key). 

In my past life as a North American insurance employee, I also noticed quarterly seasonality in loss experience, with Q4 generally being the most favorable of the four during the year.  While reserve releases from prior accident years tends to be the largest favorable item during the quarter,  there may be an additional weather related claims reason for the quarter to be more favorable than the other three for the current accident year as well.

 

Weather/catastrophe events tend to be lowest in occurrence during this quarter.  When these large events happen earlier in a calendar year, the ultimate losses (such as for California wildfire losses in Q1 this year) necessarily have to be estimated based on limited actual claims data.  Management has to estimate both the ultimate number and average size of the claims that will be paid.  They are always going to be wrong, but human tendency is to add a bit of conservatism into the estimates of both.  By Q4, they have more actual experience and have probably settled a good number of the property claims, so they can fine tune their ultimate estimates (often downwards) which can benefit the ultimate losses for weather for the current accident year as estimated during Q4.  Sort of like a current accident year reserve release for weather related claims that occurred during the year.

 

The first two quarters of the year tend to have more severe storm losses than average, while Q3 can also have some hurricane/tropical storm weather.  In a typical Q4, winter has not generally set in so severely as to drive significant winter storm losses.

 

The obvious exception would be if a huge event happened during (perhaps even near the end) of a fourth quarter.  Some catastrophes, such as earthquakes or tsunamis, do not depend upon weather patterns.  If one of those happened near the end of a fourth quarter, one could not expect a beneficial Q4 catastrophe adjustment….

[Hoodlum]

Thanks @Viking for this great summary.  I am in general agreement with your conservative approach to EPS especially as it pertains to any Mark to Market equity, due to possible overall market headwinds in the coming months. 

 

I do think your ‘excess of FV over CV’ increasing $10/share is a bit low considering how much Eurobank is already up since Q2.   I think we could also see a sizeable increase in FV for Poseidon when 2024 YE results are released.

[djokovic1]

@Viking thanks as usual for your detailed analysis. I have ~$200/ share (diluted) for 2025 and ~ ~$195/share for 2026.

 

There are 2 main differences in our assumptions:

 

i) I have higher gain on investments for 2H 2025 and 2026. It looks like your assumptions for 2H 2025 and 2026 are a bit below the long term average of their aggregate investment returns of 7.7% which all else equal should be achievable with the locked in 5.1% fixed income yields and FFH undervaluation (TRS). Of course in the short term it is impossible to forecast with precision and is a fools errand and is dependent on what FFH stock price does (amongst a lot of other variables).

 

ii) What do you assume they do with significant ROE (~15%+ so far) / net income they generate in 2025. In general that additional cash should simplistically either lead to a significant (~10%) reduction in share count (if all deployed in buybacks at <10x earnings) or lead to a 15% larger investment portfolio (ie higher interest income, profit from associates and all else equal higher investment gains) or deploy into M&A which again means higher underwriting profit and investment income. Of course in reality it will be a mix of all and likely organic premium growth will be muted with the softer market but it seems like you dont have that assumption in there going from 2025 to 2026.

Edited August 7, 2025 by djokovic1

[Viking]

52 minutes ago, djokovic1 said:

@Viking thanks as usual for your detailed analysis. I have ~$200/ share (diluted) for 2025 and ~ ~$195/share for 2026.

 

There are 2 main differences in our assumptions:

 

i) I have higher gain on investments for 2H 2025 and 2026. It looks like your assumptions for 2H 2025 and 2026 are a bit below the long term average of their aggregate investment returns of 7.7% which all else equal should be achievable with the locked in 5.1% fixed income yields and FFH undervaluation (TRS). Of course in the short term it is impossible to forecast with precision and is a fools errand and is dependent on what FFH stock price does (amongst a lot of other variables).

 

ii) What do you assume they do with significant ROE (~15%+ so far) / net income they generate in 2025. In general that additional cash should simplistically either lead to a significant (~10%) reduction in share count (if all deployed in buybacks at <10x earnings) or lead to a 15% larger investment portfolio (ie higher interest income, profit from associates and all else equal higher investment gains) or deploy into M&A which again means higher underwriting profit and investment income. Of course in reality it will be a mix of all and likely organic premium growth will be muted with the softer market but it seems like you dont have that assumption in there going from 2025 to 2026.

@djokovic1, I think we are very similar in how we are looking at things. 
 

1.) For 2025, I think my estimate for investment gains might be light. But I did lean out a little with underwriting, interest and dividend income and share of profit of associates. Bottom line, investment gains could surprise to the upside each year moving forward (Fairfax is sitting on so many material assets they could monetize). 
 

A big part of the returns from the equity portfolio are showing up in the excess of FV over CV bucket. My estimate for investment returns are 10.1% for 2025 and 8.0% for 2026 (and I include excess of FV over CV). 
 

2.) What do they do with significant earnings from 2025?

• Blizzard Vacatia investment in January ($835 million).
• Grow size of fixed income portfolio (in addition to Blizzard Vacatia).
• Equity investments: Buy out minority owner in Recipe. Buy out minority holders in KRIF. Materially increase size of investment with Waterous. Increase size of investment in Metlen, Cleveland Cliffs and Foran.
• buy back stock. 
• Move to 100% ownership in Allied World - Q4?

The big ‘investment’ is taking out the minority partner in Allied World (if it happens). This will reduce ‘non-controlling interests’ and increase ‘net earnings attributable to Fairfax shareholders.’ I took down ‘non-controlling interest’ for 2026. 
 

Recipe is getting a bunch of attention. As a result, their earnings should increase nicely in 2026. The large Waterous investment is long term in nature and it is a commodity investment (oil) so it will be volatile and will likely take a few years to come together. 
 

Anyways there are some examples of things I built into my 2026 forecast. I am likely light on the net impact - I have been every year for the past 4.

 

Edited August 7, 2025 by Viking