MKL - Markel Corp

[twacowfca]

ap1234, I know you directed your question to JPerez, but I'll go ahead and give it a shot.  This is how I think of it:

 

For the reasons you listed below, it seems pretty clear to me that they will have difficulty returning double digits in BVPS in the near term.  However, we do know that MKL is using a good model, and it is one that has shown a good record over a long period of time, through many cycles.  In my opinion, as long as we trust the model and the people, we should let it operate as it does, and the long-term future expectation of growth should be similar, but lower due to size, than the past.  This is actually my main point, everything after is really just confirmations of this, including:

 

1) They are good underwriters, which should keep us moving positively;

2) They are good investors (though not Buffett, obviously);

3) Given the current environment, their actions seem very appropriate to me.  Bond yields are low, so they have kept a shortened duration (although the Alterra acquisition lengthened it again).  Value investors are becoming skittish in the equities market, so they have not deployed a lot of bonds into equities or maintained their historic equities allocation.  Perhaps they should be doing something different, but I'm not sure what it would be.

4) In the event of a significant correction, increase in bond yields, and/or a hard insurance market, they should be ready to act and get significant returns.

 

Future Returns

Given 1-4, above, at some point, they should be able to return to a normal operating environment, which will transition us from mid-high single digit growth to hopefully mid double digit growth.  Perhaps they will be able to have large returns on some large correction that will make up for the single digit growth. 

 

Thus, combining your shorter term analysis, and my long term analysis, we might expect something like 6-10% BVPS growth for 3-5 years, and then 12+% BVPS growth thereafter.  Averaging this over a very long period of time, I'm hopeful for 12+% BVPS growth over 10-20 years. 

 

Changes in Price to Book given those assumptions

So now, using the assumptions above, what would we expect from P/B values?  It seems clear to me that MKL is trading at the low end of P/B values now because of the lower growth prospects that you have identified.  However, we should consider what would happen to P/B when the lower growth prospects are raised to higher or even historic levels.  Looking at historic P/B levels for FFH/BRK/MKL when they were growing at double digit clips, they seem to be north of 1.4 (sometimes over 2x book).  Thus, my expectation is that when growth prospects return, the P/B will have expanded a fair amount from the current ~1.2 of book to a higher number, perhaps 1.5 (or even Gayner's 1.7 "fair" value, I think he has said).  So, over the period of time of investment, you get 25% in gains in P/B expansion as well as the underlying growth in BVPS.  Given the current environment, this seems like a pretty fair deal.  However, if you have a >15% discount rate, you probably can't invest in it with a margin of safety. 

 

 

I would appreciate any faults in the above thinking.  Incidentally, I have similar views for FFH, except that it is even more positioned for a downturn, so the near term expectations are even lower.  However, with its leverage and exposure to very profitable asian insurance, I could see them having 15-20% growth if it ever uncoiled/unhedged.  They are trickier, however, and there is a ton of discussion on that topic in other threads.

 

Here's a listing of P/B ratios from the Brooklyn Investor, as well as a good discussion:

http://brooklyninvestor.blogspot.com/2011/10/markel-at-book.html

 

Joel,

great post! :)

 

I am not really comfortable with insurance and reinsurance regulations… So, I ask you a question: if, like ap1234 suggests, MKL cannot invest wherever it finds value (because is obliged to always invest in bonds), how could it be so much different for a reinsurance company like GLRE? For what I know, in fact, GLRE has no capital invested in bonds.

 

Thank you,

 

Gio

 

By conventional wisdom, GLRE intends to write business that isn't expected to have volatile results.  Then, rating agencies should allow them to invest in securities (stocks ) that are volatile .

[no_free_lunch]

Although, I wonder, does the bond-bull really matter for these insurers?

 

As the bond prices rise, even if it is unrealized, does that not allow them to shift a greater proportion over to equities?

[ap1234]

Thank you racemize and skanjete!! Very interesting insights.

 

Overall, I agree that Markel’s management is very astute and their current investment positioning is very sensible. At the end of the day, Markel has several levers at their disposal to grow BV/share (certainly more levers than a traditional insurer). And I trust that management over time will not run the business statically but rather opportunistically take advantage of opportunities both on the underwriting and investment side of the business.

 

That said, the one part that I still struggle with is your assumption that the “long-term future expectation of growth should be similar, but lower due to size, than the past.” I personally disagree. I don’t think the company’s biggest headwind is size. Markel is still small enough that they could compound BV/share at 10-15%/annum. I think the biggest headwind is that Markel derives most of their BV growth from investing, the investment portfolio has and will likely have a significant allocation to bonds, and lower bond returns are a near certainty when compared to historic results. This is a HUGE headwind that hasn’t been there over the past 20 years. If int. rates stay low for the next 5+ years, this will be a material drag on the company's growth prospects.

 

Anyways, the fact that Markel’s growth might be lower than the past isn’t such a bad thing when you consider that the entry price today is much lower than other times in history.

 

 

[JPerez]

Thank you JPerez. I really appreciate your insights!! You seem to be very knowledgeable about the business and quite detail oriented. Given that you have a significant % of your net worth tied up in Markel, I would love to hear your thoughts on the following question.

 

I apologize in advance for a long-winded post but I want to provide some color to help in the discussion. I should also mention that I am a MKL shareholder so my only goal is a constructive dialogue about the business and its prospects. If you think I am overlooking anything, I would be very interested to know what I am missing in my analysis.

 

Question: What is your expectation for Markel’s BV/share growth over the next 5+ years? I don’t care about the next 1 or 2 years. I want to know what you think they can grow BV over an investment time horizon? More importantly, I’m curious to hear your thought process. For example, if you expect 12% BV/share growth, how do you get to 12%? If you expect 15% BV/share growth, how do you arrive at 15.

 

The primary drivers of BV/share growth are below:

 

Insurance: $3.5-4B in net premiums – what is your expectation for premium growth and a normalized CR?

 

Investments: $16-17B in investments – what is a reasonable asset mix for cash, bonds and equities? Markel is underweight equities today which will go up over time but they are also restricted in how much they will need to have in bonds (regulated insurance entity). 

 

Given you normalized asset mix above (cash, bonds, stocks), what is your return expectation for each asset class? For example, if int. rates stay low, do you think the bond portfolio can do more than 3%/year over the next 5 years?

 

Ventures: $64 million YTD. What multiple do you assign to this business?

 

Putting this all together, how do you get to your BV/share growth assumptions over the next 5 years?

 

Why am I asking this? The difficulty I have is that most discussions about Markel seem to be high level and backward looking. Investors will say things like: “Markel has an enviable track record compounding BV/share at 17% for 2 decades.” Or “Markel is trading at the lowest P/B in 2 decades.” All the company has to do is compound at 15% or even 12% and you will get outsized returns. It sounds like a bargain!

 

I can’t argue with Markel’s track record. The company’s past is remarkable. The business is a rare combination of a high quality insurer and capital allocator. And I have tremendous respect for Tom Gayner (I have been to the Markel breakfast many times). If Markel can continue to compound BV/share in the double digits per annum, shareholders will be delighted.

 

My issue is that none of this has to do with the future. A discussion of the past should include a discussion of WHY they got the returns they did. The insurance operations are great. But the real source of returns is not from insurance (the premiums to equity ratio is only 0.6x). The primary driver of BV/share growth is the investment returns (investments to equity ratio of 2-3x). My primary concern with Markel is as follows:

 

1. The VAST majority of Markel’s investment portfolio has been in bonds

2. As an insurance company, Markel will always need to have a material % of their investments in bonds (after all they are using policy holder money to invest). Even if you think they will transition to Berkshire’s model of private businesses, it could take years before this becomes the majority of the investment portfolio. 

3. Since going public in 1986, interest rates have been in a secular decline. They have operated in only one environment, a secular bond BULL market.

4. Since 1986, corporate bonds have delievered an annual return of 8%!!

5. Given where int. rates are today, they can’t come close to getting the same returns in bonds over the next 5+ years. As such, BV/share growth will likely be much lower in the future. 

 

Nobody today would look at a bond manager and say “Wow. You did 8% per annum over the past 30 years. And you took on very little risk. Sounds fantastic. Where do I sign up?” Investors will instead say, “Congratulations. 8% per annum on bonds is great. But what do you think we can earn over the next 5-10 years?”

 

Markel doesn’t have unlimited flexibility. They can’t invest 100% of their investment portfolio in equities as they are using policyholder’s money. So at their core, they are still a bond manager that will allocate more to equities (public and private over time).

Hello ap1234,

brk and mkl are the core of my portfolio and my thinking here is not to look for huge returns but to compound at a rate of around 9-10% per year with little risk on those 2 companies and then try to increase the returns with other investments. I am in the process of looking for more of those other investments as I try to build a more balance portfolio.

About expectations in 5 years or so, it depends on so many factors like soft/hard insurance market, stock market, bond market, significant acquisitions, buy-backs, etc. that it would be too hard for me to say. I could have a go but with so many factors at play my guess would be pretty meaningless.

I think as things stand today they could have a combined ratio of 93% as the reinsurance business should have a better average combined ratio than the old markel. That would give us around 280 million yearly profit pretax. (I might be a little aggressive here but I based the number on management expectations)

Investments are having a run rate of 110 million per quarter  before amortization so 440 million yearly profit pretax.

And finally ventures my guess is a run rate of 65 million annual profit pretax and before amortization of intangibles.

All of the parts should add 785 million pretax or 550 million after tax at a 30 % tax rate.

That is a 8.5% increase in BV per share if they keep the same rates next year and that does not include any other comprehensive income.

At the moment they have investments in equities of close to 3 billion and giving them a reasonable return of 6% give us 180 million pretax and 117 million after tax (35% rate) so that would add another 1.8% so that gives me a 10.3 % increase in BV per year.

Obviously I think they can do better than that, they should increase their equity investments to get closer to 75% of book value and equity returns will hopefully be better than 6%. Also my hope is that they add to markel ventures becoming a bigger part of the company.

Something around this 9-10% is my base case and anything above is just gravy.

Obviously as you say they can not increase their equity investments recklessly.

There was a good few insurance companies that were on the brink of bankruptcy at the turn of the century because of having too many equity investments, an insurance company has to keep their float in solid investments like cash and bonds to be able to pay their claims. If they would have a big percentage of their investments in equities and you get a 50% crash in the market you would need to raise capital and that could destroy the company or put you back 10 years in bv per share.

Do you all think my expectations are reasonable? Am I missing anything in the numbers?

 

 

[ap1234]

Thank you Jperez! It's very interesting to hear your perspective. I have a couple questions/observations on your #'s:

 

1. How do you get comfortable with a 93% CR for the combined entity? Markel has historically generated a 96-97% CR. To arrive at a consolidated 93% CR implies Alterra will operate at a a 90% CR.

 

I  know that these are the #'s that were used in the merger document that you referenced earlier. However, I'm just curious to know if you have analyzed Alterra's CR track record as a public company? Did they generate a CR of 90% over their history? Is this a reasonable assumption going forward?

 

2. You did not include interest expense on Markel's corporate debt. If you're going to add up all the income sources (underwriting and investing), I assume you have to substract the interest expense as well as any corporate overhead not included in the CR.

 

3. I know it makes sense to add back bond amortization to arrive at the normalized pre-tax investment income.' Amortization from Alterra's bond portfolio is not a real 'cash charge.' That said, will the amortization expense not have an impact on BV/share growth over the next several years (i.e. if we are trying to get a sense for what the actual BV/share will be 3 or 5 years out)? In other words, if you're going to add the 6% of capital gains from equities that flow through OCI, should we not hit the company's BV over the next few years from the amortization charges? After all, I assume BV/share will be lower because NI is reduced each year by the amortization charges?

[JPerez]

Thank you Jperez! It's very interesting to hear your perspective. I have a couple questions/observations on your #'s:

 

1. How do you get comfortable with a 93% CR for the combined entity? Markel has historically generated a 96-97% CR. To arrive at a consolidated 93% CR implies Alterra will operate at a a 90% CR.

 

I  know that these are the #'s that were used in the merger document that you referenced earlier. However, I'm just curious to know if you have analyzed Alterra's CR track record as a public company? Did they generate a CR of 90% over their history? Is this a reasonable assumption going forward?

 

2. You did not include interest expense on Markel's corporate debt. If you're going to add up all the income sources (underwriting and investing), I assume you have to substract the interest expense as well as any corporate overhead not included in the CR.

 

3. I know it makes sense to add back bond amortization to arrive at the normalized pre-tax investment income.' Amortization from Alterra's bond portfolio is not a real 'cash charge.' That said, will the amortization expense not have an impact on BV/share growth over the next several years (i.e. if we are trying to get a sense for what the actual BV/share will be 3 or 5 years out)? In other words, if you're going to add the 6% of capital gains from equities that flow through OCI, should we not hit the company's BV over the next few years from the amortization charges? After all, I assume BV/share will be lower because NI is reduced each year by the amortization charges?

Thank you for your comments.

Alterra combined ratio from 2007-2011 was 91.9%, in any case to make the case more conservative we can say the combined ratio for the new markel is 95% so we can subtract 80 million from pre tax profits.

You are absolutely right about interest expense which is quite an oversight, it will subtract 120 million pretax so if we take 200 million pretax that would mean 160 million less after tax so that is 2.5% of BV so that would give us an increase of close to 8% per year.

About the amortization of intangibles and fixed investments they should be subtracted to get to BV according to GAAP but they should be included for the calculation of the increase of intrinsic value in a given year because they are non cash items.

We can look at it like if the BV has been artificially increased in a one-off as per the day of the acquisition and from there on it is amortized. For me a more realistic look at the company would be achieved if we deduct that amount that the bonds are marked in the books above par from BV and we add the amortization of the bonds back into income every quarter.

 

[giofranchi]

That said, the one part that I still struggle with is your assumption that the “long-term future expectation of growth should be similar, but lower due to size, than the past.” I personally disagree. I don’t think the company’s biggest headwind is size. Markel is still small enough that they could compound BV/share at 10-15%/annum. I think the biggest headwind is that Markel derives most of their BV growth from investing, the investment portfolio has and will likely have a significant allocation to bonds, and lower bond returns are a near certainty when compared to historic results. This is a HUGE headwind that hasn’t been there over the past 20 years. If int. rates stay low for the next 5+ years, this will be a material drag on the company's growth prospects.

 

Hi ap1234,

we have already talked about your worry of a secular bear in bonds, and you know that generally I agree with you. Anyway, this is exactly what great entrepreneurs do: to shift resources from low yielding assets with poor future prospects into higher yielding assets with much better future prospects. How? Well, it is very difficult to say… Chances are you will never know their businesses as well as they do, and therefore it is very difficult to predict what their next move might be… let alone their next two, three, four, etc. moves! Yet, the value of those companies is the combination of all their future moves discounted to the present! Think of FFH and its acquisition of Thomas Cook: from there they might go on acquiring some of the best businesses in a nation of 1.2 billion people, whose median age is just 27! Can you imagine the potential for growth that is built in there? How do you take that potential into consideration, while trying to value FFH? Once again I repeat what I am doing: I look for great entrepreneurs, whose process I understand and agree with, who already have an outstanding track-record, and who are still young enough to go on building wealth for a very long time. I suppose returns that are in line with those achieved in the past, or somewhat less. If the price is low enough to make possible a 15% compounded annual return for years to come, I invest and stay with them. In a diversified portfolio (10 to 15 names) some will surprise on the downside, some on the upside, but overall I guess I will get where I want to go. :)

 

And, by the way, at Q3 2013 end on BRK’s balance sheet there were assets of “Insurance and Others” worth $302 billion, of which only $29 billion were invested in bonds: 9.3%. While Total Assets were $458 billion, therefore bonds represented only 6.55% of Total Assets.

Why cannot MKL and FFH follow suit?

 

Gio

 

[skanjete]

In my view 8% compounding seems a little low.

 

It's true that their bonds will perform terrible if rates stay as low as they are. But, if rates stay as low as they are, it's fair to assume that stocks are still relatively undervalued, and so their stock portfolio will outperfom and thus at least partly compensate for the bond portfolio.

 

But rate can't stay low forever, thus at some point in the future, I guess they will go up again.

 

If intrest rates go up, it will be a drag on the short-term performance of their bonds, but at the same time, the value of their float goes up. And at that moment, the value of the Alterra acquisition will out itself, because with the acquisition they grew the float per share considerably.

 

One can also look at it this way : suppose they have an equity portfolio of 75% of book value. If Gayner can get 12% a year on his stock portfolio, that alone translates itself into 9% compounding (pre-tax). 12% a year doesn't seem unrealistic for Gayner, especially if intrest stay as low as they are.

One has to take into account some taxes of course, but on the other hand the bonds, the underwriting performance and ventures will contribute something as well.

 

 

[skanjete]

Another thing I forgot to mention : it is important to consider the situation not in a static, but dynamic way.

 

If one makes an extrapolation of the current circumstances, 15% growth of BVPS doesn't seem not easy to get.

 

But if one is trying to estimate the BV over 5 years, it is unrealistic to think that things (intrest rates, premium rates,  disaster/casualty rate, stock market dynamics) will stay static or even move in a straight way. Situations and circumstances wil be volatile. This volatility is a big risk for less able management, but creates big opportunities for capable managements.

 

I think that the quality of Markel's management alone in combination with the fact that opportunities present themselves, will guarantee some extra percentage points of growth. So from this point of view 15% doesn't seem to be that unrealistic. If I remember correctly, I think that Gayner also suggested a growth figure of something like 15%.

 

A simple illustration of the above fact is the effect of dollar cost averaging. Even if an index is exactly the same at the end of a period as it was at the beginning, one can have a positive investment result by taking advantage of the volatility through dollar cost averaging.

[ap1234]

Gio and skanjete, thank you for sharing your thoughts! It is very interesting to hear how you think about the business. I agree with the core of your thesis – Markel has MANY levers at its disposal AND the management team runs the business opportunistically as opposed to statically. This is why I own MKL stock.

 

I just happen to think that getting to a 15% BV/share growth is A LOT harder than you do. After all, they generated 17% BV/share growth over the past 2 decades when the business was much smaller AND more importantly bonds as an asset class were in a secular bull market. I am hoping that one of you will convince me I am wrong.

 

Gio, I agree with your overall thesis but I have a few points of clarification to your post:

 

1. Racemize already showed a chart from Brooklyn Investor which highlighted that 100% of Berkshire’s float is invested in cash & bonds. Why do you think Markel will be allowed to invest more than 100% of their BV in equities & private businesses? After all, the majority of Markel’s investment portfolio is ‘borrowed money’ as oppose to their own capital. As an aside, Markel has said several times in the past that their LT goal is holdings equities at 80% of BV.

 

2. I agree that Markel management (especially Tom Gayner) fits the category of a great entrepreneur. But MOST businesses have complete flexibility to allocate capital WHEREVER they want. My entire point is not that Gayner is not a very astute capital allocator. He is. It is that his business model restricts him from investing the money in whatever asset class he chooses.You can’t get the benefit of borrowed money AND then invest in whatever  way you want. If you want float, you need to set aside some money in cash & bonds.

 

3. You reference bonds as a % of assets and highlight that they are low. I think the more appropriate metric is to include cash + bonds. After all, in a low rate. environment, cash does not earn anything.

 

4. More importantly, I don’t think the correct metric is bonds (and cash) as a % of assets. The appropriate metric is cash + bonds as a % of book value. If we are talking about ROE or BV/share growth, then we should look at the investment portfolio as a % of BV.

 

Skanjete, I agree with your investment this. But I am less convinced about the direction of int. rates or equity markets.

 

1. If int. rates gradually go up over the next 5 year, Markel will benefit tremendously. However, when I invest in a business, I assume that the ROE has to be adequate even if rates never go up. I don’t want to rely on something that is outside my control. And I have no view on the direction of rates (other than it’s more likely to go up than down).

 

2. Getting a 12% equity return is a very difficult feat. I happen to think US equities as an asset class will deliver 7-8% over the next 10-20 years (I could be way too conservative). For Markel to get to 12% will be a remarkable feat in that environment. As an aside, I believe the reason that Markel is underweight equities today is not simply that they acquired Alterra’s bond portfolio. Tell me if you have a different opinion. But it seems to me if Tom Gayner thought he could do 12%/year in equities from current levels, I suspect he would have moved up the equities weights much faster especially given the paltry returns on bonds.

 

[benhacker]

Ap, excellent post.  I feel similarly on most of your points above.

 

Ben

[skanjete]

 

Skanjete, I agree with your investment this. But I am less convinced about the direction of int. rates or equity markets.

 

1. If int. rates gradually go up over the next 5 year, Markel will benefit tremendously. However, when I invest in a business, I assume that the ROE has to be adequate even if rates never go up. I don’t want to rely on something that is outside my control. And I have no view on the direction of rates (other than it’s more likely to go up than down).

 

2. Getting a 12% equity return is a very difficult feat. I happen to think US equities as an asset class will deliver 7-8% over the next 10-20 years (I could be way too conservative). For Markel to get to 12% will be a remarkable feat in that environment. As an aside, I believe the reason that Markel is underweight equities today is not simply that they acquired Alterra’s bond portfolio. Tell me if you have a different opinion. But it seems to me if Tom Gayner thought he could do 12%/year in equities from current levels, I suspect he would have moved up the equities weights much faster especially given the paltry returns on bonds.

 

 

You're correct : the current figures don't warrant a BVPS growth of 15%. That's why their book value is priced relatively low in the market.

If a BVPS growth of 15% or higher would be obvious, the market would price the share a lot higher of course. In the second half of the 90 the market price was about 2x - 2,5x book value.

 

Take 1998 for example. The market priced book value at 2,35 times. Apparently, things looked very bright back then. The track record at that point was also quite fenomenal : in the preceding decade, book value had compounded at 23,6%!

 

However, lots of thing happened since then. A difficult acquisition, disasters (2001, 2005,2011), stock market crashes with sub par stock returns...

Result : in the ensuing 15 years, BVPS compounded "only" at 12,7%/year.

Today, the market is obviously less enamored with Markel (or Fairfax, or BRK, for that matter), resulting in a premium of only 20% on book value.

So although book value grew at 12,7% a year for 15 years, the share price only grew at 7,8% a year.

 

So what counts is not the prospects (were very good in 1998), but the valuation of the prospects. The prospects at this point aren't brilliant, but the valuation is a lot more reasonable than say 15 years ago.

 

PS. I didn't express any opinion on the future of the stock market or intrest rates. I only said that if you assume that these low intrest rates are here to stay for a long time, the stock market valuation isn't excessive at all.

 

 

[giofranchi]

Gio, I agree with your overall thesis but I have a few points of clarification to your post:

 

1. Racemize already showed a chart from Brooklyn Investor which highlighted that 100% of Berkshire’s float is invested in cash & bonds. Why do you think Markel will be allowed to invest more than 100% of their BV in equities & private businesses? After all, the majority of Markel’s investment portfolio is ‘borrowed money’ as oppose to their own capital. As an aside, Markel has said several times in the past that their LT goal is holdings equities at 80% of BV.

 

2. I agree that Markel management (especially Tom Gayner) fits the category of a great entrepreneur. But MOST businesses have complete flexibility to allocate capital WHEREVER they want. My entire point is not that Gayner is not a very astute capital allocator. He is. It is that his business model restricts him from investing the money in whatever asset class he chooses.You can’t get the benefit of borrowed money AND then invest in whatever  way you want. If you want float, you need to set aside some money in cash & bonds.

 

3. You reference bonds as a % of assets and highlight that they are low. I think the more appropriate metric is to include cash + bonds. After all, in a low rate. environment, cash does not earn anything.

 

4. More importantly, I don’t think the correct metric is bonds (and cash) as a % of assets. The appropriate metric is cash + bonds as a % of book value. If we are talking about ROE or BV/share growth, then we should look at the investment portfolio as a % of BV.

 

Hi ap1234,

why bonds as a % of BV? Total assets is what matter imo. I mean: BRK is leveraged 2.5x, while FFH is leveraged 4.5x, and total assets give me the idea of how them both used not only their own capital but also “borrowed money”. They both have put to use their own capital plus “borrowed money”, and what they earn on their own capital plus “borrowed money” is for them to keep, and goes to increase BV.

I agree that what matters is bonds + cash, but I don’t see why the sum must be always the same as a percentage of float… once again, for the man with only a hammer, or better, with only two hammers: stocks and bonds, that percentage should be constant indeed… But what about BRK, which buys entire businesses, and probably today relies more on this type of capital allocation that the other two? The earnings of entire businesses are much less volatile than stock prices, and BRK can count on a steady stream of cash, that certainly helps it to face any unexpected insurance or reinsurance claim which might arise. I mean, we know that BRK’s risk profile today is different than FFH’s or MKL’s, therefore why should they keep the same % of float in bonds + cash?

For BKR, in fact, even if you add cash to bonds you get to $64 billion, while equity securities + other investments are worth $129 billion.

I am not saying that FFH or MKL won’t be subject to regulation… They most surely will be! I am only saying that they have many “hammers”, or “levers” as you call them, and they are led by people who will constantly recognize which lever to use at any time, given the surrounding circumstances.

 

Gio

 

[ap1234]

Hi Gio, I will try to answer your questions below:

 

1. How should we think about Berkshire’s leverage relative to Fairfax and Markel?

 

Berkshire operates with significantly LESS leverage relative to Fairfax and Markel. In simplest terms, the size of the investment portfolio is a function of book value + float + debt. Markel has $2.5 of investments for $1 of book value. Fairfax has $3.5 of investments for $1 of book value.

 

Since 1994, Berkshire has had an annual investment leverage of 1.2x and float is just 40% of BV.  Compare this to Fairfax. At the end of 2012, Fairfax had investment leverage of 3.4x and float was 200% of BV!! 

 

There is a table from Greg Speicher’s post which highlights Berkshire’s historical levearge:

 

http://gregspeicher.com/?p=2247

 

2. Does Berkshire need to set aside cash + bonds or can they operate with 100% of their portfolio invested in high quality private businesses that generate stable cash flows?

 

Brooklyn Investor’s blog did this analysis. From 1997-2010, cash & bonds was approx. 100% of Berkshire’s float. In other words, Berkshire was only willing to invest their BV in equities NOT their entire investment portfolio. All of their float (borrowed money) was invested in cash & bonds.

 

http://brooklyninvestor.blogspot.ca/2011/12/so-what-is-berkshire-hathaway-really.html

 

3. Why look at bonds as a % of BV as opposed to assets?

 

Because the entire discussion has been focused on gauging the company’s ‘normalized’ ROE or BV/share growth over the next 5-10 years. If we are talking about ROE, we need to put every number in relation to BOOK VALUE (not assets).

 

When an insurer is levered 2-3x (investments divided by BV), looking at investments as a % of asset understates the impact to BV. For example, Markel has $17 billion in investments, $6.5 billion in BV and $24 billion in assets. If you look at investments as a % of assets, it looks like the leverage is low. But if you look at investments as a % of BV, you can see how small changes in the return assumptions for bonds or equities has a huge impact on BV (which is what the discussion is trying to capture).

 

4. What is the impact of leverage on BV/share growth? And how do low int. rates impact BV/share growth?

 

P&C insurers have 2 types of leverage: underwriting leverage (premiums/equity ratio) and investment leverage (investments/equity ratio). Given that the majority of the leverage comes from investment leverage, this has been the focus of the discussion.

 

It’s easiest to illustrate with an example. Let’s take Markel. If we did this analysis for Fairfax, you would see that Fairfax has even more investment leverage.

 

BV = $6.5 billion

Net premiums = $4 billion

Investment portfolio = $17.2 billion

Cash + bonds  = $14 billion

Equities = $3 billion

 

Markel has a premiums/equity ratio of 0.6x. For every 1 point improvement in their CR, they only increase the ROE by 0.6% pre-tax. In contrast, Markel has an investment/equity ratio of 2.6x. For every, 1% increase in the total return, Markel increases their ROE by 2.6% pre-tax. Changes to investment returns have a HUGE impact on the BV/share growth of the business because of this leverage.

 

Why do we care about bond returns?

 

Because the cash + bonds/equity ratio is 2.2x today. So if we think the cash + bonds will earn 3%, we will get a pre-tax ROE of 6.6%. If historically cash + bonds generated 5%, they generated a pre-tax ROE of 11%. This is a 400 bps difference to the ROE. Simply by having lower int. rates (assuming nothing else in the business changed), we just lost 400 bps of ROE per annum.

 

But what happens if Markel went back to their long-term asset mix?

 

You might argue that I’m overstating the impact of low int. rates as Markel currently is overweight bonds and underweight equities. Let’s assume that Markel re-weights their equity portfolio to 75% of BV. Equities would increase from $3B to $5B and cash + bonds would decline to $12B. At $12B, cash + bonds/equity ratio would be 1.85x. So every 1%  decline in int. rates still has a 1.8% impact on the ROE of the business.

 

I hope this helps.

 

[vinod1]

ap1234,

 

Good post! This is pretty much in line with my thinking as well.

 

A minor quibble on a few items.

 

Since 1994, Berkshire has had an annual investment leverage of 1.2x and float is just 40% of BV.

 

IMHO, Berkshire's insurance subsidiaries should really be thought of as independent standalone units. This is because the statutory surplus that is used to support the insurance business for regulatory purposes is segregated from say the Railroad or Utilities business units. Insurance subsidiaries have about $80 billion in statutory surplus (roughly shareholders equity) supporting them and we need to measure leverage, etc based on this number.

 

Does Berkshire need to set aside cash + bonds or can they operate with 100% of their portfolio invested in high quality private businesses that generate stable cash flows?

 

To continue on from my comments above, the private businesses are held outside of the insurance subsidiaries. So there is really no requirement from a regulatory perspective. The benefit of this is that should insurance segments be wiped out in some unfortunate scenario, Berkshire would still have its other subs that retain value.

 

Please let me know if my understanding is incorrect.

 

Vinod

 

 

[giofranchi]

Hi ap1234,

 

first of all we should never forget that float is just a kind of “cheap and safe” (if you know what you are doing!) leverage. And a good investor should be able to compound at high rates of return (mid teens) unlevered! Therefore, no matter how you look at it, I don’t think float will ever become a drag on performance. Instead, it will always be an aid... maybe, not as much of an aid as it was before!

 

Second, to calculate ROE, we use the following formula:

Return ROE = (P/E) * (100%-CR) + (IA/E) * (IR)

where:

P = Earned Premium

E = Equity

CR = Combined Ratio

IA = Invested Assets

IR = Investment Return

Where Investment Return is the return on Invested Assets, or the return on (capital + float).

That’s why imo what matters are Invested Assets.

 

Third, I think the answer lies in BRK’s balance sheet: I don’t think it is very useful to look at FFH’s and MKL’s balance sheet today, instead BRK’s balance sheet today is how FFH’s and MKL’s balance sheet might look like 10 years from now. And I think a glance at BRK’s balance sheet today is enough to understand how little bonds count. As they go on investing in private businesses, which generate earnings to increase BV, float will become less and less relevant.

FFH and MKL have enjoyed huge success being very different from BRK, and they could have never followed suit. But, as you have pointed out, a secular bear in bonds will force them down that path, if they want to keep compounding capital at high rates of return. Or else you are right and their performance will be hampered.

 

Now I have a question: BRK has Total Assets of $458 billion, Equity of $211 billion, and bonds + cash of $64 billion: $458 - $211 - $64 = $183 billion of assets, that must correspond to an equal amount of liabilities… If float truly is 100% invested in bonds + cash, to which kind of liabilities do those $183 billion correspond to?

 

Finally, and most importantly, I would never bet against those people… We say “never bet against America”, right? Well, those people are America on steroids! ;)

 

Gio

 

[ap1234]

Thank you JPerez. A couple quick follow-up questions:

 

1. What would Markel’s BV be today if they didn’t get the one-time bump up from acquiring Alterra’s investment portfolio? If we are going to add back the amortization, I assume we should adjust the BV/share downward. As you suggest, we should deduct the amount that bonds are marked in the books above par from BV.

 

2. Do you know where I can find a breakdown of Markel’s AOCI? I’d be curious to know how much of the unrealized gains on the balance sheet are from equities vs. bonds?

 

I am not worried about the unrealized gains on equities. The unrealized gains on bonds will presumably go to zero over time if they hold the bonds to maturity.

 

[ap1234]

Thanks vinod1! Interesting observation.

 

[Kiltacular]

ap1234,

 

Good post! This is pretty much in line with my thinking as well.

 

A minor quibble on a few items.

 

Since 1994, Berkshire has had an annual investment leverage of 1.2x and float is just 40% of BV.

 

IMHO, Berkshire's insurance subsidiaries should really be thought of as independent standalone units. This is because the statutory surplus that is used to support the insurance business for regulatory purposes is segregated from say the Railroad or Utilities business units. Insurance subsidiaries have about $80 billion in statutory surplus (roughly shareholders equity) supporting them and we need to measure leverage, etc based on this number.

 

Does Berkshire need to set aside cash + bonds or can they operate with 100% of their portfolio invested in high quality private businesses that generate stable cash flows?

 

To continue on from my comments above, the private businesses are held outside of the insurance subsidiaries. So there is really no requirement from a regulatory perspective. The benefit of this is that should insurance segments be wiped out in some unfortunate scenario, Berkshire would still have its other subs that retain value.

 

Please let me know if my understanding is incorrect.

 

Vinod

 

Vinod.  Good points and I don't disagree with them but I think that Burlington Northern is actually held by National Indemnity. 

 

If correct, I believe Buffett did this because he lacked flexibility during the crash as his financials -- AXP and WFC -- were getting destroyed. 

 

Anyone know for sure -- globalfinanceparnters, you out there?

[giofranchi]

Another observation I think is important: to have invested float in bonds during the last 30 years had a very clear meaning, that was to invest “money we hold, but do not own” SAFELY. Am I right?

Now, of course, we are worried about a secular bear in bonds… And a secular bear in bonds means the probability to lose money investing in bonds for the next 20 years will be much higher! Therefore, can someone explain easily what’s the point of going on investing float in bonds?! If it gets to be as risky an investment policy as investing in equities? And probably riskier than investing in private businesses?

I don’t understand.

Imho, it is not enough to look at what happened in the past… especially in a period of transition between a secular bull and a secular bear in bonds, like the one we are living through.

 

Gio

 

[skanjete]

Gio,

You state correctly that float is money you hold, but don't own. This means that it can be called for. The problem is you don't know when you'll have to pay it. This means there's a big liquidity/volatility risk.

 

Indeed, stocks and private businesses offer a lot less risk on the long term. But there is a big caveat : you have to be sure that nobody comes asking you for that money when you need it the most. You don't want to be hurt by the inherent volatility of the stock market or an illiquid private equity market.

 

So an insurer has to be sure that the float money is available when claims have to be paid. That makes it difficult to expose oneself to the stock market volatility or the illiquidity of private businesses.

 

That's also why Markel only invests a portion of it's book value into stocks.

[giofranchi]

Gio,

You state correctly that float is money you hold, but don't own. This means that it can be called for. The problem is you don't know when you'll have to pay it. This means there's a big liquidity/volatility risk.

 

Indeed, stocks and private businesses offer a lot less risk on the long term. But there is a big caveat : you have to be sure that nobody comes asking you for that money when you need it the most. You don't want to be hurt by the inherent volatility of the stock market or an illiquid private equity market.

 

So an insurer has to be sure that the float money is available when claims have to be paid. That makes it difficult to expose oneself to the stock market volatility or the illiquidity of private businesses.

 

That's also why Markel only invests a portion of it's book value into stocks.

 

Hi skanjete,

Your reasoning is correct, but once again from a “secular bull market in bonds” point of view! In a secular bear for bonds things look much different! In an environment of rising interest rates, who wants to be the one to predict how and when those rates will rise?! Not me! Think about the ‘70s: you would have lost a lot of money investing in bonds back then! And I think in a rather unpredictable way!

I think also Mr. Buffett wrote in a letter of his (I don’t remember exactly if it is a partenrship letter or a BRK letter, I will check) about the danger of investing in bonds in the wrong climate.

Just look at what happened recently: in a matter of weeks bond yields moved up more than 100 basis points: which means that bonds with a duration of 10 years declined in value more than 10%… In a long secular bear for bonds do you really want to have your float invested that way?!

 

Gio

 

[skanjete]

Gio,

You state correctly that float is money you hold, but don't own. This means that it can be called for. The problem is you don't know when you'll have to pay it. This means there's a big liquidity/volatility risk.

 

Indeed, stocks and private businesses offer a lot less risk on the long term. But there is a big caveat : you have to be sure that nobody comes asking you for that money when you need it the most. You don't want to be hurt by the inherent volatility of the stock market or an illiquid private equity market.

 

So an insurer has to be sure that the float money is available when claims have to be paid. That makes it difficult to expose oneself to the stock market volatility or the illiquidity of private businesses.

 

That's also why Markel only invests a portion of it's book value into stocks.

 

Hi skanjete,

Your reasoning is correct, but once again from a “secular bull market in bonds” point of view! In a secular bear for bonds things look much different! In an environment of rising interest rates, who wants to be the one to predict how and when those rates will rise?! Not me! Think about the ‘70s: you would have lost a lot of money investing in bonds back then! And I think in a rather unpredictable way!

I think also Mr. Buffett wrote in a letter of his (I don’t remember exactly if it is a partenrship letter or a BRK letter, I will check) about the danger of investing in bonds in the wrong climate.

Just look at what happened recently: in a matter of weeks bond yields moved up more than 100 basis points: which means that bonds with a duration of 10 years declined in value more than 10%… In a long secular bear for bonds do you really want to have your float invested that way?!

 

Gio

 

The investment in bonds and cash is not primarily for the return. The first concern (up to a certain point of course) for the insurer is to have access to his money when he needs it to fund claims. The important thing is not return on your money, but return of your money when you need it. You remember the quote from Buffett about cash/liquidity being as oxygen : when there's plenty around, you don't think about it, until you're short of it, and then it's the only thing you can think about.

 

Of course the extent of the liquidity needs varies from business to business, depending on the type of business the insurer writes (long tail- short tail f.e.), but in my view, liquidity (again, up to a certain point) is more important than return on the invested money. That's why a prudent insurer will always have a sufficient source of liquidity.

 

You mention the bond bear market of the '70. Indeed a terrible climate for bonds. Of course, not all bonds are equal. Long term bonds fared a lot worse than short term bonds.

I remember a comical anecdote of Buffett in this context. It went along these lines : Buffett was concerned about the 30y bonds in the portfolio of Geico. He gently tried to make Byrne aware of the risk of these bonds during a golf game, but Byrne apparently wasn't focused and didn't get the clue. At one of the last holes Buffett got so frustrated that he, in a very un-Buffetteque way, blurted out that you should be an idiot to hold 30y bonds in the Geico investment portfolio...

 

So bonds as an investment category aren't good at the moment, but for most insurers a necessity. But there's a world of difference between short duration bonds and long term bonds. That's why it is very wise of Markel to have shortened it's bond maturity profile so much. They give up some return, but they limit risks and create a lot of optionality.

 

 

[giofranchi]

The investment in bonds and cash is not primarily for the return. The first concern (up to a certain point of course) for the insurer is to have access to his money when he needs it to fund claims. The important thing is not return on your money, but return of your money when you need it. You remember the quote from Buffett about cash/liquidity being as oxygen : when there's plenty around, you don't think about it, until you're short of it, and then it's the only thing you can think about.

 

Of course the extent of the liquidity needs varies from business to business, depending on the type of business the insurer writes (long tail- short tail f.e.), but in my view, liquidity (again, up to a certain point) is more important than return on the invested money. That's why a prudent insurer will always have a sufficient source of liquidity.

 

You mention the bond bear market of the '70. Indeed a terrible climate for bonds. Of course, not all bonds are equal. Long term bonds fared a lot worse than short term bonds.

I remember a comical anecdote of Buffett in this context. It went along these lines : Buffett was concerned about the 30y bonds in the portfolio of Geico. He gently tried to make Byrne aware of the risk of these bonds during a golf game, but Byrne apparently wasn't focused and didn't get the clue. At one of the last holes Buffett got so frustrated that he, in a very un-Buffetteque way, blurted out that you should be an idiot to hold 30y bonds in the Geico investment portfolio...

 

So bonds as an investment category aren't good at the moment, but for most insurers a necessity. But there's a world of difference between short duration bonds and long term bonds. That's why it is very wise of Markel to have shortened it's bond maturity profile so much. They give up some return, but they limit risks and create a lot of optionality.

 

Well, I agree that the problem is return OF capital… But, if you hold 10yr bonds, you now have less than 90% of what your capital was barely 8 months ago… That is not exactly my idea of “return of capital”… And we are not yet in a secular bear for bonds! Instead, a low rates environment will probably still persist for some more years!

Of course, you could decrease the duration of your bonds portfolio, but the shorter the duration the more similar bonds become to cash… Therefore, why to keep gathering float, if I cannot use it anyway?! It makes no sense… I assume the risk of making some mistakes in writing insurance contracts, without the benefit of using float as a lever… No, thank you!

Instead, float invested in a combination of ready cash for emergencies and free cash flow generating, wholly owned businesses, which each month spawn new cash to be added to the cash reserve, makes a lot of sense to me. It might not be conventional wisdom… but it is 30 years now that a secular bear market for bonds has disappeared from the face of the earth… And conventional wisdom might not be of great help, when a secular bear in bonds finally starts!

If you want, as a comparison think about what happened to bond investments in Europe in 2011, when interest rates for Spain and Italy spiked up… Float invested that way is not safe at all! And “return of capital” is not guaranteed at all! I would much prefer to hold a substantial cash pile for any unexpected event, and to buy something like Ferrero, which generates tons of cash each month, than to invest in Italian bonds… wouldn’t you?!

 

Gio

 

[jay21]

Gio, think of it this way:

 

I know I need to pay out a claim for $100 ten years from now. I buy a $100 10 yr, which will return my $100 after 10 years (plus I get interest along the way).  It doesn't matter how much rates move, at the end of the period I am guaranteed to have my $100 back.

 

If I put the $100 in stocks for 10 years, who knows what it will be worth when I need the $100 10 years from now.