MKL - Markel Corp

[GregS]

They called out $33.5m expense at Ventures for investigation and remediation for a manufacturing product.  This popped up earlier in 2018.  Did they ever say what unit or product this relates to?  I can't seem to find any further detail.

[Liberty]

https://concentratedcompounding.com/some-thoughts-on-markel-after-a-tough-year/

[Spekulatius]

I don’t own MKL, but have had a significant position in two insurance stocks - AXS and RE which I sold out. The reason was that  they haven’t been able to generwtenvianle returns on equity at least since 2015. It appears to me that this soft market lasts longer than prior ones and it could well be that the hybrid capital (catastrophe bonds etc) has permanently changed the market and reduced the returns in this segment. I also note that some specialty insurers like WRB still seem to be doing well.

 

FWIW, both AXS and RE had a history of producing underwriting earnings with 90-95% combined ratios,  but they don’t juice the returns  with an equity component like MKL or BRK does,  Both of them trade at about 1.1x tangible book, but they are been as low as 0.8x tangible book before. If the current return on equity persists, I don’t think they are even worth tangible book.

[scorpioncapital]

Do I understand correct that a negative float represents the equivalent of a positive interest rate on leverage? I.e. if combine ratio is 105 or 107, it represents an interest rate of 5% or 7% on the float leverage. And if it's 98% or 95%, you're getting paid to use leverage by 2 to 5% a year. But I wonder when most companies who are not in insurance borrow money at 5 to 7% (true high end is more junky) to buy real estate or finance other business assets, isn't it at least equivalent? On the other side of the equation, insurance companies earn money on the leveraged fixed income and equity portfolio. Should this not in itself still be a net positive, even if combined ratio is positive and there is a cost of borrow for a time? On average it may still end up being say 100% or zero cost.

[John Hjorth]

Combined ratio is defined as:

 

CR_n = [iL_n + IE_n] / EP_n,

 

where:

 

CR_n = Combined ratio in period n,

IL_n = Insurance losses in period n,

IE_n = Insurance operating expenses [costs] in period n, &

EP_n = Earned insurance premiums in period n.

 

Float is defined as:

 

F_n= LR_n + LAER_n + UPR_n - PAC_n - DFCAAR_n,

 

where:

 

F_n = Float end of period n,

LR_n = Loss reserves end of period n,

LAER_n= Loss adjustment expense reserves end of period n,

UPR_n = Unearned premium reserves end of period n,

PAC_n = Prepaid acquisition costs end of period n, &

DFCAAR_n = Deferred charges applicable to assumed reinsurance end of period n.

 

Insurance operations loss is defined as,

 

IOL_n = ILn + IEn

 

where :

 

IL_n and IE_n are defined above.

 

Cost of float is defined as :

 

COF_n = IOL_n / F_n

 

where:

 

IL_n and F_n are defined above.

 

- - - o 0 o - - -

 

Insert & combine the formulas with each other, reduce merged formula, and isolate either CR_n or COF_n expressed as function of the other variables, among them the opposite of the chosen one of CR_n and COF_n, and there you go.

 

- - - o 0 o - - -

 

Source: Berkshire Hathaway 1991 Letter & Motley Fool : "What is Combined Ratio?", here.

[Broeb22]

Do I understand correct that a negative float represents the equivalent of a positive interest rate on leverage? I.e. if combine ratio is 105 or 107, it represents an interest rate of 5% or 7% on the float leverage. And if it's 98% or 95%, you're getting paid to use leverage by 2 to 5% a year. But I wonder when most companies who are not in insurance borrow money at 5 to 7% (true high end is more junky) to buy real estate or finance other business assets, isn't it at least equivalent? On the other side of the equation, insurance companies earn money on the leveraged fixed income and equity portfolio. Should this not in itself still be a net positive, even if combined ratio is positive and there is a cost of borrow for a time? On average it may still end up being say 100% or zero cost.

 

If your question is...doesn't the fixed income interest income contribute to a negative cost of borrowing? I would say that the fixed income is there for the purposes of matching assets to liabilities. In other words, the fixed income should be set up to offset any inflation in legal costs/claims that the insurance portfolio would be subject to over time.

 

If you think about a workers comp policy that has a long-tail claim, i.e. someone gets hurt and is out of work for years, then the principal + income from your fixed income portfolio needs to be sufficient to meet those future claims.

 

So I would not consider income from a fixed income portfolio to be contributing to a "negative interest rate" of borrowing.

[John Hjorth]

In short, combined ratio is an efficiency measure, based on the denominator insurance premiums earned, while cost of float is an efficiency measure based on the the amount of float held. The difference between the two is the latter one takes into consideration the durability of the float held related to the actual insurance operation.

[scorpioncapital]

I think I see the problem.

 

Low rates and low equity held is insufficient to counter underwriting loss . But if rates rise Or equity exposure increases, results should get much better.

 

The one question is all things being equal, does underwriting combined ratios of an average run insurer go up as rates rise? The above premise is based on the idea that it is only financial repression holding earnings and roic back, even for poorly run companies with poor combined ratios. If cr can be held at no more than 105 , even a one or two percent rate rise should work wonders.

Are inflation expectations of long tail (or short tail for special insurers) backed into insurance loss reserves? I believe most companies say they don't expect inflation to be material or they use models in setting their reserves..don't quite trust this of course but is the idea that asset return in higher rate environment should move faster to the bottom line then Increase in liability revaluation?

[Cigarbutt]

Since Stephen Markel parted ways with Fairfax (in a friendly way) in 1990, Markel has done a good job at long term growth of book value by using a disciplined growth strategy in premiums keeping the CR around 95% and realizing superior investment returns.

 

With abundant capital flowing into reinsurance and indirectly permeating throughout the insurance channels, MKL has adapted by 1-maintaining a relatively high equity exposures in its float portfolio, 2-investing in Ventures and 3-actually participating in the alternative capital insurance market. The long term story while maintaining a strong culture caused the market to give a premium to its valuation, a rare accomplishment in the insurance world but the strategy now used by MKL is not without risks.

 

Note on the "permanent" change related to the easy-money capital: permanent it could be but it could also mean that the hard market that will eventually come will be stronger and longer lasting.

 

Note on the possibility of making up poor underwriting by growing and compensating: historically, this has been the norm but it has not resulted in rewarding results especially in insurance firms operating at high operating leverages.

 

From Mr. Buffett:

"Unfortunately, the wish of all insurers to achieve this happy result creates intense competition, so vigorous in most years that it causes the P/C industry as a whole to operate a significant underwriting loss. This loss, in effect, is what the industry pays to hold its float. For example, State Farm, by far the country’s largest insurer and a well-managed company besides, incurred an underwriting loss in nine of the twelve years ending in 2012… Competitive dynamics almost guarantee that the insurance industry—despite the float income all companies enjoy—will continue its dismal record of earning subnormal returns as compared to other businesses."

 

An interesting way to dissect the ROE can be found on page 8 of the following document and indicates how hard it can be to obtain a satisfactory ROE when underwriting profits are not forthcoming.

https://www.swissre.com/dam/jcr:2bc0a0e4-ead0-4c29-8a55-0ee9329c38b4/sigma4_2018_en.pdf

[Broeb22]

I think I see the problem.

 

Low rates and low equity held is insufficient to counter underwriting loss . But if rates rise Or equity exposure increases, results should get much better.

 

The one question is all things being equal, does underwriting combined ratios of an average run insurer go up as rates rise? The above premise is based on the idea that it is only financial repression holding earnings and roic back, even for poorly run companies with poor combined ratios. If cr can be held at no more than 105 , even a one or two percent rate rise should work wonders.

Are inflation expectations of long tail (or short tail for special insurers) backed into insurance loss reserves? I believe most companies say they don't expect inflation to be material or they use models in setting their reserves..don't quite trust this of course but is the idea that asset return in higher rate environment should move faster to the bottom line then Increase in liability revaluation?

 

https://www.google.com/search?q=historical+combined+ratio+1970s&rlz=1C1CAFC_enUS825US825&tbm=isch&source=iu&ictx=1&fir=vlx1HuDXXF12yM%253A%252C75w4qUCK231NwM%252C_&usg=AI4_-kSwu0iGed4X6n-WPf0UuyG05uv9vg&sa=X&ved=2ahUKEwiY8LmeprvgAhVmnuAKHVXJDNkQ9QEwBHoECAUQBA#imgrc=vlx1HuDXXF12yM:

 

I would point you to this visual of industry combined ratios over time 1970-2008, and based on a cursory look, there was not a huge relationship between interest rates and combined ratios but I would leave that up to you to decide.

 

In general, I would not expect there to be a very significant relationship between combined ratios and interest rates for a few reasons: 1) as interest rates rise, interest income on bond portfolios rises so float becomes more valuable which is a negative for insurance pricing; and 2) insurance risk is somewhat uncorrelated to interest rates because disasters happen at random and unexpected times, mostly independent of what is happening in any economoy. So if there is a soft market going into a disaster or heavy loss year, then the resulting hard market would be somewhat independent of what is happening in the economy.

[scorpioncapital]

I read a report that it is not expected or rising inflation that causes insurers to go belly up but unexpected large changes in inflation. This makes sense. If suddenly your policyholders are making claims at a much higher cost in a short period of time, you won't have made enough on your asset side to pay the claims. You could delay I suppose by agreement.

Would reinusrance combined ratio dynamics be the same as principle p/c? I would think reinsurance is potentially even more risky or volatile.

[walkie518]

I read a report that it is not expected or rising inflation that causes insurers to go belly up but unexpected large changes in inflation. This makes sense. If suddenly your policyholders are making claims at a much higher cost in a short period of time, you won't have made enough on your asset side to pay the claims. You could delay I suppose by agreement.

Would reinusrance combined ratio dynamics be the same as principle p/c? I would think reinsurance is potentially even more risky or volatile.

I would add here it will be policy dependent.

 

Chubb might pay you the day after an adjuster shows up your place to assess damage from a leaking toilet.

 

A policy that might be syndicated or any kind of commercial policy might take months to pay out

 

and depending on the policy, you might see an entire quarter of payments in one go that has been deferred by a quarter...

[Adam Value]

How should we view the divergence from the s&p since January? Can it be explained as simply multiple compression? If so where do we see fair value?

[Spekulatius]

How should we view the divergence from the s&p since January? Can it be explained as simply multiple compression? If so where do we see fair value?

I believe part of MKL’s halo has disappeared, due to recent missteps ( Reinsurance writeoff, annual loss - both are related) and hence the premium to other insurers in term of P/ B is shrinking. Fair value is in the eye of the beholder, but if this P/ B shrinks to the level of Y, AXS or RE then MKL has another~20% to go. MKL deserves to trade at a premium to above group imo, but I felt it was larger than deserved until recently. I feel that the price is getting interesting, but I haven’t put much work in it.

[obtuse_investor]

I mainly view the recent price decline as a much overdue multiple contraction. As of this writing, MKL is exactly it's 5yr average P/B ratio: https://ycharts.com/companies/MKL/price_to_book_value

 

[khturbo]

I did a couple of writeups on Markel if anyone is interested. The first one just goes through and adds up all the different buckets of value:

 

https://concentratedcompounding.com/some-thoughts-on-markel-after-a-tough-year/

 

Looks roughly fairly valued to me.

 

The second one looks at how they've done with their ventures investments and why the intrinsic value hasn't compounded much over the last few years:

 

https://concentratedcompounding.com/markel2/

 

IMO the intrinsic value growth should be better than it was the last few years over time but unless they really get going on the insurance side I don't think it compounds more than maybe 10% or so per year. At that point, I'd prefer BRK over MKL.

 

Would be interested to get people's thoughts.

[CorpRaider]

I will check out your posts perhaps this evening.  Thanks for sharing.  Does anyone have the current/MRQ investments per share figure handy?

 

I also want to heat check the buybacks over the last couple of years.  Kind of using that as a bs barometer with this one. 

[Spekulatius]

I will check out your posts perhaps this evening.  Thanks for sharing.  Does anyone have the current/MRQ investments per share figure handy?

 

I also want to heat check the buybacks over the last couple of years.  Kind of using that as a bs barometer with this one.

 

The analysis seems sound to me. Note that it comes essentially to the same conclusion, than the ychart P/B chart indicating that MKL trades at about its average valuation. I think the stock would be interesting at around $900 to me. That would be an about 10% discount to fair value. I think P/B still works well for MKL, because it is still basically an insurance company, unlike BRK, which generates the majority of its  earnings from business other than insurance.

[scorpioncapital]

How has Markel been investing its float? I remember they held a large amount of Berkshire stock. I think over time , the smart investment of the float is key to outperformance. Really insurers offer one key benefit - low, zero, or negative cost leverage. If this criteria is not met, there is no point for the structure to exist versus say a mutual, hedge, or investment conglomerate using shareholder equity or some small amount of paid debt to operate. If this criteria is met, then zero cost leverage + intelligent portfolio allocation = outperformance. I would also add another element which is the human desire to stray from the core and start doing extra, fancy things to accelerate things. This can be starting new startups that don't work and waste money and time.

[Liberty]

2018 letter:

 

http://www.markelcorp.com/-/media/investor-relations/letters-to-shareholders/2018.pdf

[chrispy]

Nice letter and the ILS example is very informative on how the industry works.

 

During the Google talk, gayner mentions he likes reading John Wooden's books. Well, I bought and read one last week. After reading this annual letter it is clear that Wooden has been influential on Tom and his writing style

[Guest notorious546]

how do you actually value this thing now

 

stock portfolio + bond portfolio + multiple on wholly owned businesses less debt less capital gains on stocks owned?

[Guest cherzeca]

do you net out unrealized gains on brk?

[Partner24]

Didn't like the letter. Too much "bla bla bla". Results have been so so over the last years. It was better before...more results, less words.

[Cigarbutt]

how do you actually value this thing now

 

stock portfolio + bond portfolio + multiple on wholly owned businesses less debt less capital gains on stocks owned?

I assume you mean "...less debt less {tax-related} capital gains on stocks owned…?

 

The valuation method you refer to then is fine, is some kind of a two-column approach which can be useful and rapidly computed from reported numbers but rests on significant underlying assumptions. You need to assume that MKL will continue to earn consistent underwriting profits, superior investment results (bonds, publicly-traded equities and ventures) and to continue astute capital allocation.

 

Another simple way to value MKL is to use a band of P/B (from 1,5 to 1,7 for instance) again assuming that the future will continue to rhyme with the past.

 

Chuck Akre and John Huber (Base Hit Investing) periodically come up with assessments that help to determine the drivers behind the return on capital for MKL. Here's an example:

http://basehitinvesting.com/markel-mkl-a-compounding-machine/