Fairfax 2023

[ValueMaven]

@treasurehunt big issue there has been GEICO - so not really comparing apples to apples imho

[treasurehunt]

2 hours ago, ValueMaven said:

@treasurehunt big issue there has been GEICO - so not really comparing apples to apples imho

 

It's true that the comparison is not apples to apples, but GEICO by itself does not explain the difference. 2022 was the only bad year for GEICO with an underwriting loss of -$1.9 billion. In the four years prior, GEICO's underwriting profit totaled over $8.6 billion, or almost $2.2 billion per year.

[DM1]

[DM1]

On 8/15/2023 at 12:48 PM, DM1 said:

 

[DM1]

Just now, DM1 said:

 

[Viking]

Fairfax and the Transition from Good to Great: The Flywheel Effect

 

Warning: Mr Market might be right. And I might be completely wrong. This post is not intended to be financial advice. It is intended to educate and entertain. Please contact your financial advisor before making any stock purchases.   

 

Fairfax’s stock trades today at a PE of 5.2 (to my estimate of 2023 earnings).

 

 

Of course, it is not normal for a stock to trade at a PE of 5.2. The PE multiple for the S&P500 is currently 20. So Fairfax’s stock could double in price and it would still be trading at a 50% discount to the S&P500.

 

Fairfax’s PE of 5.2 screams that one of two things is clearly wrong:

1.) the price of the stock is way too low.

2.) the estimated earnings are way too high (and not ‘durable’)

 

Let’s take a look the stock price first.

 

Fairfax has been one of the best performing stocks over the past 31 months (since Dec 31, 2020). Over this time period, Fairfax is up 143% while the S&P500 is up 16%. Fairfax has outperformed the S&P500 by 127%. That is stellar outperformance.

 

 

After a run like that, Fairfax’s stock price must now be fairly valued - in fact, it might even be overvalued. Looks like we might have our answer to our question above. If the stock is fairly valued then that means the earnings estimate must be way too high.

 

Let’s take a look at earnings estimates.

 

After Fairfax released Q2 earnings, I updated my three year earnings estimate for Fairfax and came up with the following:

• 2023 = $160/share
• 2024 = $166/share
• 2025 = $174/share

My forecast is for earnings to go up each of the next 3 years. Clearly, my estimates must be too high. Right? I actually think they might prove to be conservative. Why? Because every forecast I have done for Fairfax for the past 30 months has proven, in hindsight, to be too conservative and usually by a lot.

 

Why have my estimates been too low? Because i have been consistently underestimating the management team at Fairfax and  the earnings power of the collection of assets they have today. So i trust my earnings estimate looking out three years. A lot.

 

So what explains Fairfax’s current PE of 5.2?

 

Despite a 143% gain over the past 31 months, the stock price of Fairfax is still dirt cheap. Yes, that probably sounds like crazy talk. 

 

How can a ‘still dirt cheap’ stock price be explained?

 

Operating earnings are the holy grail for insurance companies because it is made up primarily of predictable items. And these items tend to be durable. Let’s focus on this bucket of earnings at Fairfax and see what we can learn.

 

The average for total operating earnings at Fairfax from 2016 to 2020 was $1 billion per year ($39/share). But this dramatically changed beginning in 2021.

• in 2021, operating earnings doubled to $1.8 billion or $77/share (from 2016-2020 average)
• in 2022, operating earnings tripled to $3.1 billion or $132/share (from 2016-2020 average)
• in 2024, operating earnings are forecasted to quadruple to $4.3 billion or $185/share (from 2016-2020 average)

The increase in operating earnings at Fairfax has been like a goat climbing straight up the steep side of a mountain.

 

 

Let’s now do some historical comparisons to see what we can learn.

 

From 2016 to 2020, Fairfax’s stock price averaged about $500/share (I am ignoring the covid drop in 2020). Over this same 5-year period, operating earnings at Fairfax averaged about $1 billion per year ($39/share). So investors over this 5 year period thought $1 billion in operating earnings (let’s call that baseline earnings) at Fairfax was worth a stock price of about $500. Back then, Fairfax’s stock was considered to be fairly valued.

 

In 2023, operating earnings at Fairfax will be about $4.3 billion ($185/share). Operating earnings for 2023 are up 330% compared to the old 5-year baseline trend from 2016-2020, or 374% on a per share basis (the share count has come down over the past 5 years). Fairfax’s stock price closed Friday at $828. Fairfax’s stock price is up only 66% compared to the 5-year trend from 2016-2020.

 

So operating earnings per share at Fairfax have increased a staggering 374% over the 2016-2020 trend while the share price has increased a modest 66%. I think we just learned something useful. The increase in Fairfax’s stock price has not kept up with the increase in operating earnings. And ‘not kept up’ is a big understatement.

 

What is Mr Market missing?

 

Mr Market clearly is not understanding the new trajectory for operating earnings at Fairfax.

 

This is likely because Mr Market is still looking at Fairfax’s financial performance through the rear view mirror - focussing primarily on past reported results. That approach makes sense for most companies. But it makes no sense for Fairfax today. Because it completely misses (ignores) all the significant positive changes that have been happening at Fairfax over the past 5 or 6 years - the benefits of which are only just now fully flowing through to reported results.

 

The good news is Mr Market will eventually figure things out at Fairfax. Earnings are the key. And as Fairfax keeps reporting stellar results quarter after quarter, Mr Market will price Fairfax’s shares appropriately. 

 

What is causing the massive increase in operating earnings?

 

What we are seeing today, with record operating earnings at Fairfax, is the cumulative effect of slow, organic (internal) change that has been happening at Fairfax for many years - a process of continuous improvement. It is the result of the conscious choices and actions being taken at all levels of the organization - senior management, the insurance operating companies, the investment team at Hamblin Watsa and the CEO’s of the various equity holdings. All parts of the organization are working in a disciplined way towards the same end purpose - the consistent delivery of solid results leading to the improvement of the long term performance of the company. It is the slow methodical process of doing what needs to be done.

 

For Fairfax the process also involved some soul searching - there were lessons that needed to be learned. Fairfax stopped doing the things that were not working (like the equity hedges and short positions). It got better with its new equity investments.

 

The improving operating earnings are also not due primarily to circumstance. But active management (taking advantage of circumstance) is an important part of Fairfax’s business model.

 

The record operating earnings we are seeing at Fairfax today is simply the end result of years of good decisions and hard work.

 

What is the new baseline for operating earnings at Fairfax today?

 

The level of operating earnings at Fairfax have likely reached an inflection point - a breakthrough of sorts - given their size. Significant sums are now being reinvested every year (billions). The seeds that are being planted will grow new streams of operating earnings for Fairfax in both insurance and investments in the coming years. Compounding will work its magic. Fairfax looks like it is now in that virtuous circle where success begets more success.

 

My estimate for operating earnings for 2023 is $4.3 billion and I think that is a reasonable number to use a new baseline for Fairfax moving forward. Why? Because all the inputs I use are reasonable and mildly conservative.

 

How durable is $4.3 billion in operating earnings?

 

My guess is it is quite durable. At least as durable as operating earnings at other insurance companies like WR Berkley, Markel or Chubb. Why wouldn’t they be? In fact, the management team at Fairfax has been best-in-class in terms of overall management of the business in recent years - this suggests that we should have more confidence in Fairfax’s future results than that for peers. I know, that is a very non-consensus view. But it is where logic takes me. 

 

As I like to say, the once ugly caterpillar called Fairfax has magically transformed itself into a beautiful butterfly. What thing happened to cause the transformation? There was no one thing. It was a bunch of things. From the butterfly’s point of view, what happened was perfectly natural. Only to the outsider does it look like magic.

—————

 

Jim Collins, in his book Good to Great, has a concept called the ‘flywheel effect’ that describes very well what has been happening ‘under the hood’ at Fairfax for the past 5 or 6 years that has got the company to where it is today.

 

The flywheel effect: “The Flywheel effect is a concept developed in the book Good to Great. No matter how dramatic the end result, good-to-great transformations never happen in one fell swoop. In building a great company, there is no single defining action, no grand program, no one killer innovation, no solitary lucky break, no miracle moment. Rather, the process resembles relentlessly pushing a giant, heavy flywheel, turn upon turn, building momentum until a point of breakthrough, and beyond.”

• https://www.jimcollins.com/concepts/the-flywheel.html

 

—————

What are some of the decisions/actions made by Fairfax in recent years that have caused the Fairfax 'flywheel' to pick up more and more speed? To provide some context, we are going to separate the decisions/actions into Fairfax’s three economic engines:

• insurance
• investments - fixed income
• investments - equities/derivatives

Each on its own is driving earnings for Fairfax. Together, they help illustrate why Fairfax is delivering record operating earnings - and why the Fairfax flywheel has now likely reached ‘breakthrough’ speed.

 

Economic engine 1: insurance

• Turn 1: 2015-2017: rapid growth - driven by international expansion by acquisition
• Turn 2: 2017: strategic pivot in India - sold ICICI Lombard for significant gain ($950 million) and seeded Digit with an investment of $154 million that is now worth $2.3 billion..
• Turn 3: 2019-today: rapid organic growth - driven by hard market.
• Turn 4: 2022: increased ownership in Allied World from 70.9 to 82.9%
• Turn 5: 2023: increasing ownership in Gulf Insurance Group from 44% to 90%. Strategic; secures Fairfax’s position in MENA.
• Turn 6: ongoing: methodically improving quality of the insurance businesses. Resulting in improving CR.

Net written premiums have increased from $8.1 million in 2016 to an estimated $24.1 billion in 2023, an increase of 198%. At the same time, the combined ratio has improved from an average of 98 from 2016-2020 to an average of 95 the past three years. Much higher net written premiums and a lower CR has resulted in much higher (record) underwriting profit. Underwriting profit averaged $191 million per year from 2016-2020. It was $801 million in 2021, $1.1 billion in 2022 and is forecasted to be $1.3 billion in 2023. This increase is sustainable (with some volatility in both directions).

 

Economic engine 2: investments - fixed income

• Turn 7: Dec 2021: average duration of fixed income portfolio was reduced to 1.2 years. In 2021, sold $5.2bn in corporate bonds at a yield of 1% for a realized gain of $253 million (most were purchased in March/April 2020). Avoided billion in unrealized losses on $40 billion fixed income portfolio as interest rates spiked higher in 2022 and 2023 (protected the balance sheet).
• Turn 8: 1H 2023: average duration of fixed income portfolio extended to 2.4 years. This locks in more than $1.5 billion in interest income for each of the next three years (this estimate is low).
• Turn 9: 2020 and 2023: real estate debt platform partnership established with Kennedy Wilson. $4 billion portfolio is delivering an average return of about 9% total = $360 million, mostly in interest income.

Driven by the significant increase in the insurance business, the size of the fixed income portfolio at Fairfax has doubled in size from $20.3 billion in 2016 to $40 billion today. From 2016-2022, the average yield of the fixed income portfolio was 2.4% and today the average yield is 4.8%. As a result of the two doubles (portfolio size and rate of return), interest income is spiking higher. Interest income averaged $650 million per year from 2016-2021. It was $874 million in 2022 and is forecast to come in at $1.8 billion in 2023 and $2.1 billion in 2024. This increase is sustainable (with some volatility in both directions).

 

Economic engine 3: investments - equities / derivatives

• Turn 10: 2016: ending the ‘equity hedge’ in late 2016.
• Turn 11: 2020: closing out the final short position in late 2020

These two programs cost Fairfax an average of $494 million per year on average from 2010-2020. Ending these two programs eliminated what was essentially a $494 million annual expense for the company (meaning Fairfax became $494 million more profitable). Fairfax has also said multiple times that they have learned their lesson and that they will no longer short indices or individual stocks.

• Turn 12: 2014-2017: poor equity purchases - Fairfax made a string of poor equity purchases from 2014-2017.
• Turn 13: 2018-today: very good new equity purchases - Fairfax has been hitting the ball out of the park with their more recent new equity purchases.
• Turn 14: 2020- present: Fairfax also have been taking advantage of recent bear market low stock prices by adding significantly to many of the equity holdings they already own.

Fairfax has done a great job over the last 5 years fixing their poor equity purchases from 2014-2017. These holding were burning  about $200 million per year in cash (losses/write downs/restructuring etc) and now they are all largely fixed and delivering solid returns for Fairfax shareholders. Eurobank is the shining star in this group. The equity purchases from 2018-today have been performing well. And Fairfax has been aggressively adding to positions in equities they already own - buying at bear market low prices.

 

Share of profit of associates at Fairfax averaged $151 million per year from 2016-2021. It was $1 billion in 2022 and it is forecast to come in at $1.1 billion in 2023. This number should grow nicely in the coming years (with some volatility in both directions).

 

The quality of Fairfax’s total portfolio of equity holdings has likely never been better than it is today.

• Turn 15: late 2020/early 2021: purchase of total return swap giving Fairfax exposure to 1.96 million FFH shares at an average cost of $372/share. This one investment has delivered to Fairfax an unrealized gain of more than $900 million since inception.

Asset sales:

• Turn 16: 2020/2021: sold Riverstone UK (runoff business) for $1.3 billion (plus $230 million contingent value instrument).
• Turn 17: 2022: sale of pet insurance business delivered a $1 billion after tax gift to Fairfax shareholders.
• Turn 18: 2022: sale of Resolute Forest Products for $626 million (plus $183 million CVR) at top of lumber cycle.

Asset sales (insurance and investments) have always been an important part of the capital allocation framework at Fairfax and have delivered significant value to shareholders over the years.

 

Stock buybacks:

• Turn 19: 2021: dutch auction - Fairfax purchased 2 million shares at $500/share.
• Turn 20: 2018 to present - via NCIB, Fairfax has bought back about 2.5 million shares via the NCIB at about $490/share.

Share count at Fairfax peaked at 27.75 million in 2017. Since that time Fairfax has reduced effective shares outstanding by 4.55 million or 16.2%. Share count has returned to about where it was in 2016 fully offsetting the dilution caused by the Allied World acquisition. As a result, shareholders today enjoy the full benefit of the significant growth Fairfax has achieved since 2016. As well, all 4.5 million shares were repurchased at a very attractive average price of around $490. Fairfax shares closed today at $828 (BV is $834).

 

Conclusion:

As you can see from the list above there is no one thing (action, event, luck) that is driving record operating earnings at Fairfax. Rather, it is the cumulation of many, many things that have happened over the past 5 or 6 years. But it is only now that the impact of these many actions are becoming fully visible to outsiders - because now they are all together ‘all of a sudden’ showing up in record underwriting profit, record interest and dividend income and record share of profit of associates. The flywheel has achieved breakthrough. And Fairfax as a company has made the leap from good to great.

 

Is Fairfax’s stock fairly valued at a PE of 5.2 (to 2023 estimated earnings)? You decide.

—————

The Benefits of active management:

Above is a list of 20 actions taken by the management team at Fairfax in recent years. A significant number of the actions mentioned have on their own delivered $1 billion or more in value to Fairfax shareholders. Each on its own is an impressive accomplishment. But when you put them all together… well that is simply an amazing collection of accomplishments. And a big reason why i am so confident Fairfax’s new baseline for operating profit is likely around $4.3 billion. It clearly demonstrates the huge impact active management, when done well, can have on a company like Fairfax.

 

Edited August 20, 2023 by Viking

[petec]

57 minutes ago, Viking said:

The record operating earnings we are seeing at Fairfax today is simply the end result of years of good decisions and hard work.

 

Viking your work is usually excellent but this statement is wrong.

 

Fairfax have absolutely done excellent work - I have argued this for a long time, and actually think it goes back decades, notwithstanding clear mistakes regarding the big short.

 

However, what's really driven the increase in operating earnings is the big shift in the macro backdrop - covid stimulus and then rising rates helped (in rough order) Atlas, then the broader economy, float income, and underwriting profits (because CRs are directly linked to interest rates, which control the amount of capital flowing into the industry). Eurobank's recovery is not entirely unrelated to this either.

 

Don't get me wrong: Fairfax's management made some great decisions in the down market to grow by acquisition (not writing more policy) and put themselves in a position to write record amounts of business and generate record amounts of float in the up market. They got roundly criticised for some of those decisions on this board (the Brit and Allied deals were evidence of Prem's towering ego, not his ability to make different decisions at different points in the cycle). I agree with almost every example you've given of good decisions on their part. But the fact is that if rates were still at 2019 levels, which they might well have been without covid, Fairfax's operating earnings wouldn't be at this level.

 

That said, I do think Fairfax's next 2-3 years look strong, and by then they will be so large in float terms that operating earnings will be far higher than historic levels regardless of rates. And I think the stock looks fairly cheap on that basis. It's my largest holding, but I trim on spikes these days.

 

 

[Viking]

41 minutes ago, petec said:

 

Viking your work is usually excellent but this statement is wrong.

 

Fairfax have absolutely done excellent work - I have argued this for a long time, and actually think it goes back decades, notwithstanding clear mistakes regarding the big short.

 

However, what's really driven the increase in operating earnings is the big shift in the macro backdrop - covid stimulus and then rising rates helped (in rough order) Atlas, then the broader economy, float income, and underwriting profits (because CRs are directly linked to interest rates, which control the amount of capital flowing into the industry). Eurobank's recovery is not entirely unrelated to this either.

 

Don't get me wrong: Fairfax's management made some great decisions in the down market to grow by acquisition (not writing more policy) and put themselves in a position to write record amounts of business and generate record amounts of float in the up market. They got roundly criticised for some of those decisions on this board (the Brit and Allied deals were evidence of Prem's towering ego, not his ability to make different decisions at different points in the cycle). I agree with almost every example you've given of good decisions on their part. But the fact is that if rates were still at 2019 levels, which they might well have been without covid, Fairfax's operating earnings wouldn't be at this level.

 

That said, I do think Fairfax's next 2-3 years look strong, and by then they will be so large in float terms that operating earnings will be far higher than historic levels regardless of rates. And I think the stock looks fairly cheap on that basis. It's my largest holding, but I trim on spikes these days.

@petec you are a night owl! 
 

My view is the true anomaly was the period 2010-2020 and zero interest rates. Interest rates appear to be normalizing. This is causing the investment world to return to a more normalized environment… one where active management, when done well, matters (can deliver serious outperformance). Something Fairfax has historically been very good at. So i give the management team the benefit of the doubt for the very good decisions they have made in recent years. 
 

The part of your comment i do not understand is: “And I think the stock looks fairly cheap on that basis.”

 

My estimate is the stock is trading at 5.2 x 2023 earnings. That is not ‘fairly cheap’… that is crazy cheap. Do you not think $4.3 billion is a reasonable estimate for operating earning for 2023? 
 

Or is it more a weighting issue… where Fairfax is getting too big and you want to lighten up to rebalance your overall portfolio? Regardless of fundamentals or what the stock might actually be actually worth?

Edited August 20, 2023 by Viking

[SafetyinNumbers]

3 hours ago, petec said:

 

 

 

Don't get me wrong: Fairfax's management made some great decisions in the down market to grow by acquisition (not writing more policy) and put themselves in a position to write record amounts of business and generate record amounts of float in the up market. They got roundly criticised for some of those decisions on this board (the Brit and Allied deals were evidence of Prem's towering ego, not his ability to make different decisions at different points in the cycle). I agree with almost every example you've given of good decisions on their part. But the fact is that if rates were still at 2019 levels, which they might well have been without covid, Fairfax's operating earnings wouldn't be at this level.

 

 

I think this take on Brit and Allied ignores that Fairfax issued stock at 1.3x book plus and issued preferred at tight spreads to partially fund these acquisitions. Singleton is a legend not only for the buybacks below book but for issuing stock early on well above book to grow the earnings power of the business. Call it towering ego if you like, I call it accretive capital allocation.

 

[StubbleJumper]

2 hours ago, Viking said:

@petec you are a night owl! 
 

My view is the true anomaly was the period 2010-2020 and zero interest rates. Interest rates appear to be normalizing. This is causing the investment world to return to a more normalized environment… one where active management, when done well, matters (can deliver serious outperformance). Something Fairfax has historically been very good at. So i give the management team the benefit of the doubt for the very good decisions they have made in recent years. 
 

The part of your comment i do not understand is: “And I think the stock looks fairly cheap on that basis.”

 

My estimate is the stock is trading at 5.2 x 2023 earnings. That is not ‘fairly cheap’… that is crazy cheap. Do you not think $4.3 billion is a reasonable estimate for operating earning for 2023? 
 

Or is it more a weighting issue… where Fairfax is getting too big and you want to lighten up to rebalance your overall portfolio? Regardless of fundamentals or what the stock might actually be actually worth?

 

 

I'm not Pete, but I'll take a run at this.

 

If you want to value a security using PE as a metric, you need to do so on the assumption that earnings are neither unusually high nor unusually low and that they are sustainable for a prolonged period.  A PE is essentially a mental short-cut for assessing the value of a perpetuity.  To make the argument that a 5.2 PE is cheap and that the company should have a PE of, say, 12, you need to assume that the current excellent operating conditions for an insurance company will persist for many years on end.

 

To do this, you need to argue that FFH has some sort of special sauce that enables it to write a 94 CR while buying US treasuries yielding 5%, but nobody else can/will do so.  So, in essence, the argument needs to be that $1 of capital in Crum or Odyssey can be used to write $2 of premium, the underwriting earnings will be 12 cents (94 CR) and riskless investment income will be 10 cents (a US Treasury yielding 5%), providing a slick return of 22% on that equity, BUT no other company can replicate that.  No other company will see this, obtain new capital, expand their book of business, and competition will not push the CR rate up and squeeze FFH's books of business.  If you can hammer out this argument in your own mind why FFH can do this and nobody else can/will, then your earnings are sustainable and you can simply slap some sort of market average PE onto current earnings to arrive at a valuation estimate.

 

Setting aside the argument about the sustainability of earnings, the comment saying, "And I think the stock looks fairly cheap on that basis" is in my view a reasonable and valid comment.  You have quite rightly pointed out that FFH has locked in some fairly attractive investment returns for the next few years.  You've done the arithmetic through to develop pro forma earnings estimates going forward 2.5 years and shown that there will be big earnings coming down the pipe, even if a guy gives a moderate haircut to underwriting profitability for 2024 and a  massive haircut to underwriting profitability for 2025 (but, hey if they actually continue to write a 94 CR, so much the better!).  If you do this, it is difficult to envisage a scenario where adjusted BV (after accounting for the excess of market over book for certain associates) doesn't hit $1,100 by Dec 31, 2025.  If operating conditions in the insurance market continue to be as wonderful as they currently are, with a CR of 94 and a treasury of 5% being SIMULTANEOUSLY available, that Dec 2025 BV could be higher, but it seems to be a no-brainer that they'll make the $1,100 BV given that the returns on the fixed income portfolio are largely locked in.  So, someone who doesn't buy the argument that FFH ought to currently trade at PE12x$180EPS=US$2,000+ can quite reasonably believe that it could trade at somewhere between 1x and 1.2x BV on Dec 31, 2025.  With the shares currently trading at ~US$830, a price on Dec 31, 2025 of $1,100 to $1,300 is quite plausible and is fully consistent with the observation, "And I think the stock looks fairly cheap on that basis."

 

It really amounts to a bit of a differing view of just how far into the future you are comfortable to predict outstanding insurance results.  I am assuming that we are at the peak of the insurance cycle and that conditions will deteriorate as capital enters the industry and companies competing to expand their books of business push the CRs higher (probably to slightly above 100 before it's all said and done).  If it actually does work out that FFH can routinely obtain a 22% return on an incremental dollar of capital, so much the better.  But, personally, I am unwilling to assume that today's wonderful insurance conditions will persist for a prolonged period.  I would be happy in 10 years if I am wrong today!

 

 

SJ

[SafetyinNumbers]

13 minutes ago, StubbleJumper said:

 

 

I'm not Pete, but I'll take a run at this.

 

If you want to value a security using PE as a metric, you need to do so on the assumption that earnings are neither unusually high nor unusually low and that they are sustainable for a prolonged period.  A PE is essentially a mental short-cut for assessing the value of a perpetuity.  To make the argument that a 5.2 PE is cheap and that the company should have a PE of, say, 12, you need to assume that the current excellent operating conditions for an insurance company will persist for many years on end.

 

To do this, you need to argue that FFH has some sort of special sauce that enables it to write a 94 CR while buying US treasuries yielding 5%, but nobody else can/will do so.  So, in essence, the argument needs to be that $1 of capital in Crum or Odyssey can be used to write $2 of premium, the underwriting earnings will be 12 cents (94 CR) and riskless investment income will be 10 cents (a US Treasury yielding 5%), providing a slick return of 22% on that equity, BUT no other company can replicate that.  No other company will see this, obtain new capital, expand their book of business, and competition will not push the CR rate up and squeeze FFH's books of business.  If you can hammer out this argument in your own mind why FFH can do this and nobody else can/will, then your earnings are sustainable and you can simply slap some sort of market average PE onto current earnings to arrive at a valuation estimate.

 

Setting aside the argument about the sustainability of earnings, the comment saying, "And I think the stock looks fairly cheap on that basis" is in my view a reasonable and valid comment.  You have quite rightly pointed out that FFH has locked in some fairly attractive investment returns for the next few years.  You've done the arithmetic through to develop pro forma earnings estimates going forward 2.5 years and shown that there will be big earnings coming down the pipe, even if a guy gives a moderate haircut to underwriting profitability for 2024 and a  massive haircut to underwriting profitability for 2025 (but, hey if they actually continue to write a 94 CR, so much the better!).  If you do this, it is difficult to envisage a scenario where adjusted BV (after accounting for the excess of market over book for certain associates) doesn't hit $1,100 by Dec 31, 2025.  If operating conditions in the insurance market continue to be as wonderful as they currently are, with a CR of 94 and a treasury of 5% being SIMULTANEOUSLY available, that Dec 2025 BV could be higher, but it seems to be a no-brainer that they'll make the $1,100 BV given that the returns on the fixed income portfolio are largely locked in.  So, someone who doesn't buy the argument that FFH ought to currently trade at PE12x$180EPS=US$2,000+ can quite reasonably believe that it could trade at somewhere between 1x and 1.2x BV on Dec 31, 2025.  With the shares currently trading at ~US$830, a price on Dec 31, 2025 of $1,100 to $1,300 is quite plausible and is fully consistent with the observation, "And I think the stock looks fairly cheap on that basis."

 

It really amounts to a bit of a differing view of just how far into the future you are comfortable to predict outstanding insurance results.  I am assuming that we are at the peak of the insurance cycle and that conditions will deteriorate as capital enters the industry and companies competing to expand their books of business push the CRs higher (probably to slightly above 100 before it's all said and done).  If it actually does work out that FFH can routinely obtain a 22% return on an incremental dollar of capital, so much the better.  But, personally, I am unwilling to assume that today's wonderful insurance conditions will persist for a prolonged period.  I would be happy in 10 years if I am wrong today!

 

 

SJ

What if the equity portfolio returns 5-10% per year on average for an indefinite period of time? Wouldn’t an ROE of 15-20% be achievable then even if combined ratios inevitably go higher? 

[MMM20]

On 8/20/2023 at 8:10 AM, SafetyinNumbers said:

What if the equity portfolio returns 5-10% per year on average for an indefinite period of time? Wouldn’t an ROE of 15-20% be achievable then even if combined ratios inevitably go higher? 

It seems like the stock just hasn’t come close to adjusting to a more normalized environment for investment returns b/c of an anchoring / recency bias. If if fact ~5% risk free rate and ~6% underwriting profit = ~11% spread turn out to be persistent for the decade or more, the stock might literally be worth $4,000 per share right now. I don’t think anyone is saying that. Even with half or a quarter of that, the stock is worth a whole lot more than $820 b/c of the value of a large and well managed investment operation generating even 60/40 beta-like returns, and an insurance operation generating roughly costless float to leverage that up without risk of a margin call. IMHO. 
 

You all have done a great job showing that it’s likely somewhere in the middle - really meaningful structural changes leading to durably higher earnings power through cycles, and that we are in a particularly favorable operating environment right now.

 

To be clear I am not saying the stock is worth $4000. Even I am not that bullish. But I do think there is now a totally plausible upside case in which you could pay $4000 today and make ~10% returns long term from there. Who would’ve said that 5 years ago? Definitely not THIS random guy on the internet…

 

Thanks again all for the great FFH forum. Best thing on the internet
 

Edited August 21, 2023 by MMM20

[John Hjorth]

@SafetyinNumbers,

 

From the Markel Topic in the Investment Ideas forum a few years ago :

 

 

and

 

Just to elaborate on the basis return calculations in the considerations above, where cost of float in period n isen't 1-CR_n.

[StubbleJumper]

19 minutes ago, SafetyinNumbers said:

What if the equity portfolio returns 5-10% per year on average for an indefinite period of time? Wouldn’t an ROE of 15-20% be achievable then even if combined ratios inevitably go higher? 

 

 

Sure, but the traditional thesis for investing in FFH is that you'd hope that their equity returns would be a little higher than 5-10%.  The constraint on what this can actually do for a shareholder is that, by necessity, an insurance company keeps two-thirds or three-quarters of its portfolio in fixed income.  So, in the case of FFH, they currently have about US$12.5 billion in common stocks and associates.  A 10% return on that would be $1.25b, or about $50 or $60/sh.  It definitely helps, but the 15% ROE is a big hurdle over the long term (just look at the table that Prem publishes in his letter every year).

 

 

SJ

[MMM20]

1 hour ago, StubbleJumper said:

 

 

Sure, but the traditional thesis for investing in FFH is that you'd hope that their equity returns would be a little higher than 5-10%.  The constraint on what this can actually do for a shareholder is that, by necessity, an insurance company keeps two-thirds or three-quarters of its portfolio in fixed income.  So, in the case of FFH, they currently have about US$12.5 billion in common stocks and associates.  A 10% return on that would be $1.25b, or about $50 or $60/sh.  It definitely helps, but the 15% ROE is a big hurdle over the long term (just look at the table that Prem publishes in his letter every year).

 

 

SJ

The traditional math is less applicable now that they have upwards of $30b in float, isn’t it? ~5-10% equity returns, ~3-6% fixed income returns, ~0% cost of float, and the dividend and ongoing buybacks add up to a ~15-20% return for shareholders now, right?

[StubbleJumper]

41 minutes ago, MMM20 said:

The traditional math is less applicable now that they have upwards of $30b in float, isn’t it? ~5-10% equity returns, ~3-6% fixed income returns, ~0% cost of float, and the dividend and ongoing buybacks add up to a ~15-20% return for shareholders now, right?

 

No, the math is what it is.  The long-term financing differential (ie long bond rate minus the cost of float) is a shade over 4 percent.  So if you hypothesize a 0 percent cost of float you should be thinking about a 4 percent treasury bond rate.

 

If you have 25 percent equities that return 10 pct and 75 pct fixed income, your float yields 0.75x4 + 0.25x10 = 5.5 pct.  Lever it up by 2 assuming that your premiums to surplus ratio is that high (usually ffh runs it lower) and your insurance subs might get an roe as high as 11 pct if they jettison excess capital.  Ffh's subs usually carry more capital, so drop your expectation accordingly.

 

The investment thesis is really that they will generate some alpha on the investment side.  

 

 

SJ

[gfp]

If you are taking underwriting profit out of the equation in your example above, the simple math in your example is 5.5% on $56.6B of invested assets against an $18.6B common shareholders' equity.  The leverage is closer to 3 than 2.  Which is what gets you to 15% on equity.

[StubbleJumper]

22 minutes ago, gfp said:

If you are taking underwriting profit out of the equation in your example above, the simple math in your example is 5.5% on $56.6B of invested assets against an $18.6B common shareholders' equity.  The leverage is closer to 3 than 2.  Which is what gets you to 15% on equity.

 

You are right.  The basic math above was about the roe you might expect from one of the insurance subs. 

 

For the corp as a whole, you need to add in the return from other non insurance assets held by the hold co, and then strip off interest, taxes and Holdco admin costs.

 

It is worthwhile looking at the table depicting growth in BV that Prem publishes every year the annual letter.  Hitting 15%+ is a notable event.  We will probably get three or four of those types of years in rapid succession, so we should be pleased about what we see.  But people should project that forward over many years at their own risk and péril!

 

 

SJ

Edited August 20, 2023 by StubbleJumper

[SafetyinNumbers]

12 minutes ago, StubbleJumper said:

 

You are right.  The basic math above was about the roe you might expect from one of the insurance subs. 

 

For the corp as a whole, you need to add in the return from other non insurance assets held by the hold co, and then strip off interest, taxes and Holdco admin costs.

 

It is worthwhile looking at the table depicting growth in BV that Prem publishes every year the annual letter.  Hitting 15%+ is a notable event.  We will probably get three or four of those types of years in rapid succession, so we should be pleased about what we see.  But people should project that forward over many years at their own risk and péril!

 

 

SJ

How many years in that table depicting growth in BV was the float 2x the BV?

[TwoCitiesCapital]

3 hours ago, StubbleJumper said:

 

To do this, you need to argue that FFH has some sort of special sauce that enables it to write a 94 CR while buying US treasuries yielding 5%, but nobody else can/will do so.  So, in essence, the argument needs to be that $1 of capital in Crum or Odyssey can be used to write $2 of premium, the underwriting earnings will be 12 cents (94 CR) and riskless investment income will be 10 cents (a US Treasury yielding 5%), providing a slick return of 22% on that equity, BUT no other company can replicate that.  No other company will see this, obtain new capital, expand their book of business, and competition will not push the CR rate up and squeeze FFH's books of business.  If you can hammer out this argument in your own mind why FFH can do this and nobody else can/will, then your earnings are sustainable and you can simply slap some sort of market average PE onto current earnings to arrive at a valuation estimate.

 

 

I think it's possible the earnings may be more sticky than otherwise assumed. 

 

Other companies MAY see it, but may NOT be able to capitalize on it due to capital constraints after they're fixed income portfolios went down ~10+% last year. 

 

If interest rates do NOT go down, other companies would have to either issue new equity to raise capital OR organically wait for current earnings and amortization of bond discounts to refill capital holes over the course of the next 2-3 years just to get back to where they were at YE 2021. 

 

And that assumes no major catastrophes during that 2-3 year period so CRs can remain positive and not contribute to loss of capital. A large catastrophe year could further extend the timeline. 

 

I don't think insurance CRs and earnings can continue at this level of growth and profitability into perpetuity, but Fairfax did itself a large favor on the fixed income side by protecting capital. When Prem mentioned in the past the ability to 2-3x premiums written in a hard market, who here thought it'd happen in an environment where nobody else could? I definitely underestimated that potential and the lucrative profitability that results from it. 

[Viking]

13 hours ago, StubbleJumper said:

 

 

I'm not Pete, but I'll take a run at this.

 

If you want to value a security using PE as a metric, you need to do so on the assumption that earnings are neither unusually high nor unusually low and that they are sustainable for a prolonged period.  A PE is essentially a mental short-cut for assessing the value of a perpetuity.  To make the argument that a 5.2 PE is cheap and that the company should have a PE of, say, 12, you need to assume that the current excellent operating conditions for an insurance company will persist for many years on end.

 

To do this, you need to argue that FFH has some sort of special sauce that enables it to write a 94 CR while buying US treasuries yielding 5%, but nobody else can/will do so.  So, in essence, the argument needs to be that $1 of capital in Crum or Odyssey can be used to write $2 of premium, the underwriting earnings will be 12 cents (94 CR) and riskless investment income will be 10 cents (a US Treasury yielding 5%), providing a slick return of 22% on that equity, BUT no other company can replicate that.  No other company will see this, obtain new capital, expand their book of business, and competition will not push the CR rate up and squeeze FFH's books of business.  If you can hammer out this argument in your own mind why FFH can do this and nobody else can/will, then your earnings are sustainable and you can simply slap some sort of market average PE onto current earnings to arrive at a valuation estimate.

 

Setting aside the argument about the sustainability of earnings, the comment saying, "And I think the stock looks fairly cheap on that basis" is in my view a reasonable and valid comment.  You have quite rightly pointed out that FFH has locked in some fairly attractive investment returns for the next few years.  You've done the arithmetic through to develop pro forma earnings estimates going forward 2.5 years and shown that there will be big earnings coming down the pipe, even if a guy gives a moderate haircut to underwriting profitability for 2024 and a  massive haircut to underwriting profitability for 2025 (but, hey if they actually continue to write a 94 CR, so much the better!).  If you do this, it is difficult to envisage a scenario where adjusted BV (after accounting for the excess of market over book for certain associates) doesn't hit $1,100 by Dec 31, 2025.  If operating conditions in the insurance market continue to be as wonderful as they currently are, with a CR of 94 and a treasury of 5% being SIMULTANEOUSLY available, that Dec 2025 BV could be higher, but it seems to be a no-brainer that they'll make the $1,100 BV given that the returns on the fixed income portfolio are largely locked in.  So, someone who doesn't buy the argument that FFH ought to currently trade at PE12x$180EPS=US$2,000+ can quite reasonably believe that it could trade at somewhere between 1x and 1.2x BV on Dec 31, 2025.  With the shares currently trading at ~US$830, a price on Dec 31, 2025 of $1,100 to $1,300 is quite plausible and is fully consistent with the observation, "And I think the stock looks fairly cheap on that basis."

 

It really amounts to a bit of a differing view of just how far into the future you are comfortable to predict outstanding insurance results.  I am assuming that we are at the peak of the insurance cycle and that conditions will deteriorate as capital enters the industry and companies competing to expand their books of business push the CRs higher (probably to slightly above 100 before it's all said and done).  If it actually does work out that FFH can routinely obtain a 22% return on an incremental dollar of capital, so much the better.  But, personally, I am unwilling to assume that today's wonderful insurance conditions will persist for a prolonged period.  I would be happy in 10 years if I am wrong today!

 

SJ

 

@StubbleJumper My point with the PE in my post was to highlight that it is absurdly low for Fairfax right now. Fairfax's stock price today of $828 makes sense if Fairfax was earning about $80 per year (and assuming earnings grow modestly in the future). It is a well run P&C insurer so trading at a PE of 10 is hardly an aggressive multiple to attach.

 

My current estimate is Fairfax will earn $160 this year. And with slightly conservative assumptions, earnings will grow in 2024 and 2025. That is not in the same universe as $80 in earnings.

 

So a buyer of Fairfax's stock today at $828 is getting $80 in estimated 2023 earnings for free ($160-$80). That is one hell of a discount for something that might or might not happen in 2026 or later. It doesn't make any rational sense. It is too large.

 

Yes, my earnings estimate for 2023 might be a little high. And it also might be a little low. We are almost 8 months through the year.

 

My thesis is investors are way underestimating what a 'normalized' amount of earnings is for Fairfax today. Yes, the future is uncertain. There are risks. But there are also opportunities. Some income streams will face headwinds. At the same time other income streams will experience tailwinds.  

Edited August 21, 2023 by Viking

[UK]

6 hours ago, Viking said:

 

@StubbleJumper My point with the PE in my post was to highlight that it is absurdly low for Fairfax right now. Fairfax's stock price today of $828 makes sense if Fairfax was earning about $80 per year (and assuming earnings grow modestly in the future). It is a well run P&C insurer so trading at a PE of 10 is hardly an aggressive multiple to attach.

 

My current estimate is Fairfax will earn $160 this year. And with slightly conservative assumptions, earnings will grow in 2024 and 2025. That is not in the same universe as $80 in earnings.

 

So a buyer of Fairfax's stock today at $828 is getting $80 in estimated 2023 earnings for free ($160-$80). That is one hell of a discount for something that might or might not happen in 2026 or later. It doesn't make any rational sense. It is too large.

 

Yes, my earnings estimate for 2023 might be a little high. And it also might be a little low. We are almost 8 months through the year.

 

My thesis is investors are way underestimating what a 'normalized' amount of earnings is for Fairfax today. Yes, the future is uncertain. There are risks. But there are also opportunities. Some income streams will face headwinds. At the same time other income streams will experience tailwinds.  

 

Viking, again, thank you very much for sharing your work and thoughts!

 

It is really almost impossible anything to add to it, but one thing, I am not sure if really you assumptions could be called "slightly conservative". I do not think at all that you have to be conservative or that these numbers are impossible, or that there could not be further positive upsides or surprises etc, but incorporating CR of 96 or less for longer term, I think is quite optimistic (possible but not sure if probable). Regardless I agree very much with your general thinking and however you look at FFH today (or even at CR 98 or 100), either on absolute or on relative, it is still to cheap and you still do not need scales to see if patient is way too fat here:). I also like very much that FFH (as also BRK but with lesser degree) is very well positioned if rates will stay higher/normal for longer or ever. But most importantly, it finally seems that at this stage, in order for them to do really well, say to earn their 15 per cent target for longer term, all they have to do is just not to do anything really stupid, to paraphrase Munger:). And if they will do something clever, as they did many times in different areas in recent 5 years, then even better! Given that, I do not understand how it is not selling at least 1.2-1.3 BV already and would not be shocked to see them trade at some 1.5 BV in mid term and my plan is just to hold it. Unless something really stupid is done, but I do not expect this at all:)

 

Edited August 21, 2023 by UK

[Dudley]

Whether @Viking's earnings estimates are exactly right or approximately right is not important. He is directionally right. There is no doubt that earnings have skyrocketed due to a Munger-esque lollapalooza confluence of factors that Viking has outlined. The stock has not kept up and after a few more impressive earnings results and maybe some buybacks, the stock price will eventually reflect the true earnings power of Fairfax.

 

The beauty of this setup is that even if Wall Street continues to be slow to appreciate the story here, Fairfax has the ability and desire to buy back stock, which would make the stock even more attractive. It's a win-win scenario that could only possibly be improved by more cash at the Holdco.

[jfan]

On 8/20/2023 at 7:52 AM, StubbleJumper said:

I am assuming that we are at the peak of the insurance cycle and that conditions will deteriorate as capital enters the industry and companies competing to expand their books of business push the CRs higher (probably to slightly above 100 before it's all said and done). 

The Insurance Market Cycle: Hard Versus Soft Markets - Cottingham & Butler (cottinghambutler.com)

 

I understand your thought process here about the cyclical nature of the insurance business as well as interest rates affecting it. It would be lovely to have a hard market that last as long as the recent soft market. I certainly can come up with a narrative that interest rates will likely remain higher for longer vs returning back to their lows over the past decade. With many insurance companies mal-positioned over the past decade (in terms of their investment portfolio), how many of them will have the capacity to write more policies, or raise capital at attractive valuations to do so? Furthermore, with higher interest rates, will alternative sources of capital (eg private equity) have increasing difficulties raising funds effectively to buy these poorly performing insurance companies?

 

Am I way off base if I assume that the average hard market typically last 4 years? And if this one started in 2018, the cycle should have moderated at this stage and more premium growth unlikely to persist? If the above narrative plays out, how much longer could a semi-hard market last? one? two? more years?

 

 

 

 

Edited August 21, 2023 by jfan

[UK]

1 hour ago, jfan said:

The Insurance Market Cycle: Hard Versus Soft Markets - Cottingham & Butler (cottinghambutler.com)

 

I understand your thought process here about the cyclical nature of the insurance business as well as interest rates affecting it. It would be lovely to have a hard market that last as long as the recent soft market. I certainly can come up with a narrative that interest rates will likely remain higher for longer vs returning back to their lows over the past decade. With many insurance companies mal-positioned over the past decade (in terms of their investment portfolio), how many of them will have the capacity to write more policies, or raise capital at attractive valuations to do so? Furthermore, with higher interest rates, will alternative sources of capital (eg private equity) have increasing difficulties raising funds effectively to buy these poorly performing insurance companies?

 

Am I way off base if I assume that the average hard market typically last 4 years? And if this one started in 2018, the cycle should have moderated at this stage and more premium growth unlikely to persist? If the above narrative plays out, how much longer could a semi-hard market last? one? two? more years?

 

 

 

 

 

This is very interesting info, thanks! As I think about all this, I am coming to maybe such conclusions: a. insurance cycle impossible to predict over longer term, but it is unlikely to remain always beneficial, however b. even with underwriting results at zero, FFH is still cheap, and c. as with other important things (investments etc), when insurance cycle turns negative, the extent of negative consequences will also depend on what decisions FFH will take and how they will execute, and I think there is some basis for optimism here, looking at how they managed everything in the last 5 years? 

 

Edited August 21, 2023 by UK