Q3 results

[Luke]

https://www.fairfax.ca/press-releases/fairfax-announces-conference-call-2023-10-27/

[SafetyinNumbers]

I expect we’ll get some more previews this week before they report but here’s the lay of the land with respect to analyst expectations 

[Luke]

Pretty excited!

[SafetyinNumbers]

I think they can beat consensus on the back of IFRS17 which I think most analysts are ignoring. More important though will be the earnings power demonstrated in investment income, associates/dividend income and the combined ratio. Analysts will likely increase their 2024 estimates and introduce 2025 estimates which will make Morningstar’s pessimistic prognostication less relevant. 
 

It would nice to have growing earnings forecast like every other P&C insurer but FFH is too cheap and unloved for that. 
 

 

[netcash1]

On 10/28/2023 at 5:59 AM, SafetyinNumbers said:

I expect we’ll get some more previews this week before they report but here’s the lay of the land with respect to analyst expectations 

 

[netcash1]

what website are you using to get the earnings estimates?

Thanks

[SafetyinNumbers]

15 minutes ago, netcash1 said:

what website are you using to get the earnings estimates?

Thanks

I pay for Koyfin.

[StubbleJumper]

53 minutes ago, netcash1 said:

what website are you using to get the earnings estimates?

Thanks

 

I prefer to get my earnings estimates from @Viking on COBF....

 

 

SJ

[TwoCitiesCapital]

19 minutes ago, StubbleJumper said:

 

I prefer to get my earnings estimates from @Viking on COBF....

 

 

SJ

 

[SafetyinNumbers]

1 hour ago, StubbleJumper said:

 

I prefer to get my earnings estimates from @Viking on COBF....

 

 

SJ

 

I guess if you like them realistic but that's not why I look at analyst estimates. I just want to see where expectations are. Quants really like estimates that are going up and earnings growth. Maybe analysts will get there in the next few years. I think it's part of how multiple expansion can get out of hand. 

 

Edited October 29, 2023 by SafetyinNumbers

[UK]

6 hours ago, StubbleJumper said:

 

I prefer to get my earnings estimates from @Viking on COBF....

 

 

SJ

 

True:), but it is also good to know that the general market thinking is:)

 

Edited October 30, 2023 by UK

[Hoodlum]

Below is from Scotia Bank analyst.

 

https://www.theglobeandmail.com/investing/markets/inside-the-market/article-mondays-analyst-upgrades-and-downgrades-for-oct-30/

 

Quote

Mr. Hardie reaffirmed Fairfax Financial Holdings Ltd. (FFH-T +1.48%increase) as his “top pick,” seeing it “well-positioned to navigate the current environment.” He kept a “sector outperform” rating and $1,500 target for its shares. The average target on the Street is $1,485.09.

 

“Fairfax has demonstrated resilience through the business cycle and turbulent financial markets, but we view it as a less-defensive play than more traditional publicly listed insurers,” he said. “At this stage of the market cycle, this likely provides an attractive balance: downside protection thanks to the relative resilience of insurance operations through a potential recession, and upside potential when markets recover. There have been significant changes at Fairfax that we believe investors have yet to fully recognize.”

 

Edited October 30, 2023 by Hoodlum

[Viking]

With Fairfax set to report Q3 results after markets close on Thursday here are a few of the things i will be looking for:

 

1.) Topline growth? 

• Over or under recent trend of 8%?
• What is outlook for remainder of 2023 and outlook for hard market for 2024?

2.) underwriting profit/CR?

• CR was 100.3% in Q3 2022
• CR was 93.9% in 1H 2023

3.) Interest and dividend income for Q3?

• Q1 = $382.3 million
• Q2 = $464.6
• Q3E = $521?
• Q3 2022 = $256.5

4.) Average duration of bond portfolio?

• At Q2, it was 2.4 years

5.) Share of profit of associates?

• What is build: Eurobank, Poseidon, Exco, other?

6.) Investment gains (losses):

• Equities: tailwind
• Fixed income: headwind
• IFRS 17 offset: small tailwind

7.) Share buybacks during quarter?

• Is there any commentary about future buybacks on conference call?

8.) Book value?

• Q2 = $834/share

9.) GIG update - still expected to close in 2023?

Edited October 31, 2023 by Viking

[StubbleJumper]

That nicely summarizes the principal levers.  A couple of comments:

 

25 minutes ago, Viking said:

3.) Interest and dividend income for Q3?

• Q1 = $382.3 million
• Q2 = $464.6
• Q3E = $521?
• Q3 2022 = $256.5

 

I am looking forward to seeing the interest and dividend number for Q3.  I think you've significantly improved the method for estimating interest income, so it will be interesting to see how closely reality tracks your estimate.  I suspect it will track reasonably well.

 

27 minutes ago, Viking said:

2.) underwriting profit/CR?

• CR was 100.3% in Q3 2022
• CR was 93.9% in 1H 2023

 

Thinking back to Q3, wasn't it a reasonably light Q3 for cats?  Usually Q3 is the worst for hurricanes and other nasty weather, but I'm having trouble recalling much in the way of really large storms.  I guess Hilary was a remarkable storm, but my recollection was that the insurable damage wasn't too bad.  I think we will be pleasantly surprised for Q3's CR.

 

 

SJ

[Parsad]

5 minutes ago, StubbleJumper said:

That nicely summarizes the principal levers.  A couple of comments:

 

 

I am looking forward to seeing the interest and dividend number for Q3.  I think you've significantly improved the method for estimating interest income, so it will be interesting to see how closely reality tracks your estimate.  I suspect it will track reasonably well.

 

 

Thinking back to Q3, wasn't it a reasonably light Q3 for cats?  Usually Q3 is the worst for hurricanes and other nasty weather, but I'm having trouble recalling much in the way of really large storms.  I guess Hilary was a remarkable storm, but my recollection was that the insurable damage wasn't too bad.  I think we will be pleasantly surprised for Q3's CR.

 

 

SJ

 

+1!  I think Viking's got it pretty fine-tuned now.  I think CR will come in modestly higher around 94.9 or so.  BV will be around $855-860 USD.  

 

Cheers!

[Tommm50]

Impressive volume today for Fairfax on both the OTC and Toronto Exchange. Maybe they've been reading Vikings posts...

[Viking]

5 hours ago, StubbleJumper said:

That nicely summarizes the principal levers.  A couple of comments:

 

 

I am looking forward to seeing the interest and dividend number for Q3.  I think you've significantly improved the method for estimating interest income, so it will be interesting to see how closely reality tracks your estimate.  I suspect it will track reasonably well.

 

 

Thinking back to Q3, wasn't it a reasonably light Q3 for cats?  Usually Q3 is the worst for hurricanes and other nasty weather, but I'm having trouble recalling much in the way of really large storms.  I guess Hilary was a remarkable storm, but my recollection was that the insurable damage wasn't too bad.  I think we will be pleasantly surprised for Q3's CR.

 

 

SJ

@StubbleJumper , i am with you. Interest and dividends is the number i am looking forward to seeing the most. Duration? Composition (any shift into corporates)? The problem with a quarterly forecast is Fairfax is probably pretty active each quarter and we don’t know what amount is actually maturing each quarter. So we will see…
 

Joseph Wang (the Fed Guy) has nailed the move higher in interest rates over the past year. He thinks interest rates might be in the peaking at current levels (for now). Might be a good time to lock in some duration. Pretty much everyone is in the higher for longer camp today… and a year ago everyone thought a recession was imminent (they were completely wrong). Locking in duration looks like such fat pitch right now. 

Edited November 1, 2023 by Viking

[SafetyinNumbers]

40 minutes ago, Viking said:

@StubbleJumper , i am with you. Interest and dividends is the number i am looking forward to seeing the most. Duration? Composition (any shift into corporates)? The problem with a quarterly forecast is Fairfax is probably pretty active each quarter and we don’t know what amount is actually maturing each quarter. So we will see…
 

Joseph Wang (the Fed Guy) has nailed the move higher in interest rates over the past year. He thinks interest rates might be in the peaking at current levels (for now). Might be a good time to lock in some duration. Pretty much everyone is in the higher for longer camp today… and a year ago everyone thought a recession was imminent (they were completely wrong). Locking in duration looks like such fat pitch right now. 

How long would you extend duration? 

[TwoCitiesCapital]

44 minutes ago, SafetyinNumbers said:

How long would you extend duration? 

 

Not the guy you asked, but....

 

At least to the expected duration of their liabilities. That would be neutral positioning. I think it was said else where that was around ~4 years. 

 

My preference would be that they overweight it at this point, so call it 5-6 years on average. Whether via ladder or barbell I'd leave up to them and their discretion. 

[Crip1]

1 hour ago, SafetyinNumbers said:

How long would you extend duration? 

No need, IMHO, to go very long or very short. I like the idea of not trying to make too much in terms of big macro moves here so a barbell strategy makes sense. Locking in a big chunk longer term at current rates with another big chunk still earning solid returns at the shorter duration which can be reinvested in short-term instruments or moved into Corporates/Munis/Etc as opportunity presents itself.

 

-Crip

[Viking]

4 hours ago, SafetyinNumbers said:

How long would you extend duration? 

Do we know what the average duration of Fairfax’s insurance liabilities are?
 

As an investor, a risk to Fairfax is if short term interest rates crater. How? Bad recession in 2024. Or something breaking in financial system causing central banks to aggressively ease. Not a base case. This would likely cause interest income to decline.
 

It makes sense to me to get average duration on fixed income closer to average duration of liabilities. Because, with the move up in yields further out on the curve, you are (finally) getting paid for duration today. (And you were not for much of the past decade.)

As an investor i like certainty - so i am hopeful they extend duration. 2.6 years? 2.8 years? WR Berkely, who is also at 2.4 years average duration telegraphed on their call that they may begin to extend duration of their fixed income portfolio.

 

Having said that, Fairfax has many factors to consider when deciding on average duration of their fixed income portfolio. What i might want as an investor is not on their list of factors (and it shouldn’t be). They have a very good fixed income team. They will have good reasons for whatever average duration they report with Q3 results.

Edited November 1, 2023 by Viking

[SafetyinNumbers]

8 hours ago, Viking said:

Do we know what the average duration of Fairfax’s insurance liabilities are?
 

 

This is from the Q123 call. So all we know is at least 4 years. We should also note that the claims liabilities are in various currencies (but likely disproportionately USD) while the portfolio is in USD. 

 

 

[Spekulatius]

8 hours ago, Viking said:

Do we know what the average duration of Fairfax’s insurance liabilities are?

I use insurance liabilites / Premiums as a quick measure. I get $57B (2022 insurance liabilites) /22.6B (2022 net premiums) = 2.6 years roughly. It's probably not totally correct but in the correct ballpark.

[Cigarbutt]

5 hours ago, Spekulatius said:

I use insurance liabilites / Premiums as a quick measure. I get $57B (2022 insurance liabilites) /22.6B (2022 net premiums) = 2.6 years roughly. It's probably not totally correct but in the correct ballpark.

Putting discounting of insurance contract liabilities aside for a second (2023 results will be IFRS17ed).

In your quick measure, why do you include the reinsurance part in the numerator but not in the denominator (otherwise the measure becomes about 2.1-2.2)?

Retention can vary from year to year and over time.

Any way to improve upon that measure?

Just take a look at reserve triangles:

About 50% milestone of the ultimately paid payments versus adjusted reserves happens in the third year (sometimes in fourth year).

A picture about statistics and graphic reserve triangles and interpreted from an undiscounted perspective:

Reserve triangles typically look like the above (amounts paid over time) and tend to be somewhat skewed to the right (skewness even more pronounced with longer tail lines). Conceptually (undiscounted), the way duration is meant to mean in the insurance reserves world (opinion), the 50% milestone then underestimates to some degree the duration as recognized in insurers' parlance.

-----

Short version, P+C insurers' duration of insurance contract liabilities is reported to be between 2 and 4 years (more towards 4 years with long tail business). So with FFH, you'd expect the duration to be around 4 years because of their very significant exposure to long tail lines.

-----

So, i would suggest, as a quick measure, to use your measure, to call it the Spekulatius quick measure, to correct for the harmony between the numerator and denominator and to multiply by 2 and then i think you'd be in the right ballpark.

[SafetyinNumbers]

4 hours ago, Cigarbutt said:

Putting discounting of insurance contract liabilities aside for a second (2023 results will be IFRS17ed).

In your quick measure, why do you include the reinsurance part in the numerator but not in the denominator (otherwise the measure becomes about 2.1-2.2)?

Retention can vary from year to year and over time.

Any way to improve upon that measure?

Just take a look at reserve triangles:

About 50% milestone of the ultimately paid payments versus adjusted reserves happens in the third year (sometimes in fourth year).

A picture about statistics and graphic reserve triangles and interpreted from an undiscounted perspective:

Reserve triangles typically look like the above (amounts paid over time) and tend to be somewhat skewed to the right (skewness even more pronounced with longer tail lines). Conceptually (undiscounted), the way duration is meant to mean in the insurance reserves world (opinion), the 50% milestone then underestimates to some degree the duration as recognized in insurers' parlance.

-----

Short version, P+C insurers' duration of insurance contract liabilities is reported to be between 2 and 4 years (more towards 4 years with long tail business). So with FFH, you'd expect the duration to be around 4 years because of their very significant exposure to long tail lines.

-----

So, i would suggest, as a quick measure, to use your measure, to call it the Spekulatius quick measure, to correct for the harmony between the numerator and denominator and to multiply by 2 and then i think you'd be in the right ballpark.

 

I accepted what Prem said and it was consistent with the triangles but I'm not an insurance expert. 

 

I assume reserves also follow a similar chart pattern as above so would it be reasonable to expect that Fairfax always has a lot of reserves they can and do release in Q3 of every year given the normally high CAT activity? 

 

Last year Fairfax had 15% of cat losses in the quarter but the combined ratio was only 100.3%. It's part of why I'm optimistic about the underwriting profit given how well most peers have performed. It's impossible to know what business they actually wrote or what came back to haunt them. Markel reported a 99% combined ratio after the close today and looks like they had both problems.