1st Qtr. earnings out

[omagh]

Al,

 

This is classic Templeton.  Watsa has spent significant time learning his craft from Templeton who is no longer with us.  That education has clearly left its mark and serves FFH shareholders well.

 

-O

 

During the crash last year the only people with significant money to invest were BRK and FFH.  No one else had capital.  The insurance float allowed them to keep alot of money in cash for a long time and invest strategically when the time came.  Most hedge funds and value funds were frozen due to inability to raise cash at that point in time.  Not FFH though.

[vinod1]

When you have CR of 105 on 4b written, you lose 200m to underwriting.

 

But when you also are making 5% yield on 12b float, you wind up with 600m income.

 

This nets out to 400m a year in profit from insurance operations.

 

Some people think they are losing money from their insurance operations but I'm not in your club.

 

You need to look at the benefit of incremental float. I am not saying float has no benefit for fairfax. Say Fairfax is able to choke the 4b written at 105 CR by 25% so that you are writing only 3b at 100 CR. If they exhibit this level of discipline the float would be about 9b instead of 12b. (Calculating a crude average float tail of 3 years - 12b float/4b written). By giving up 3b float fairfax is able to save 200m. So the cost of incremental float is 6.7% after tax (200/3000) - so we need 10% pre-tax returns on the incremental float that need to be generated from the bond portfolio. Is this worth the cost and risk?0

 

Vinod

[vinod1]

The whole idea of reducing business in a soft market seems a little short sighted to me. All industries have their ups and downs, but one thing we all share in common is that customers are much easier to turn away than they are to recapture when markets turn hard. One soon realizes the value of retaining customers when you look at the cost in advertising dollars it takes to bring in each new customer. Before you start reducing your customer base, you better be darn sure you are not going to need them somewhere down the road.

 

 

    Three conditions that prevail in insurance, but not in most

businesses, allow us our flexibility.  First, market share is not

an important determinant of profitability: In this business, in

contrast to the newspaper or grocery businesses, the economic

rule is not survival of the fattest.  Second, in many sectors of

insurance, including most of those in which we operate,

distribution channels are not proprietary and can be easily

entered: Small volume this year does not preclude huge volume

next year.  Third, idle capacity - which in this industry largely

means people - does not result in intolerable costs.  In a way

that industries such as printing or steel cannot, we can operate

at quarter-speed much of the time and still enjoy long-term

prosperity.

- Buffett in one of his annual letters

[ERICOPOLY]

When you have CR of 105 on 4b written, you lose 200m to underwriting.

 

But when you also are making 5% yield on 12b float, you wind up with 600m income.

 

This nets out to 400m a year in profit from insurance operations.

 

Some people think they are losing money from their insurance operations but I'm not in your club.

 

You need to look at the benefit of incremental float. I am not saying float has no benefit for fairfax. Say Fairfax is able to choke the 4b written at 105 CR by 25% so that you are writing only 3b at 100 CR. If they exhibit this level of discipline the float would be about 9b instead of 12b. (Calculating a crude average float tail of 3 years - 12b float/4b written). By giving up 3b float fairfax is able to save 200m. So the cost of incremental float is 6.7% after tax (200/3000) - so we need 10% pre-tax returns on the incremental float that need to be generated from the bond portfolio. Is this worth the cost and risk?0

 

Vinod

 

105 CR on 4b -- that 200m that I'm talking about is pre-tax.  You seem to think it's after-tax...   Do I have it wrong?

 

I believe the hurdle on the incremental 3b of float is only 6.67% pre-tax.  Not 10% pre-tax.

 

When I worked at Microsoft, it amazed me that they launched product lines that never made money.  I could suggest to Mr. Ballmer that we should just stop doing this... I mean, why launch products that won't make money?

 

My point is that... nobody really belives that Fairfax intentionally writes an extra $3b of business on purpose where they need to make more than 6.67% investment performance to make it pay off.  I'm sure anyone admitting to doing this would already have been dealt with by Prem.  If they aren't doing it on purpose, then it's not going to be easy to identify exactly which policies to cut.  One year it's a quake in Chile, the next year a windstorm in Europe, another year it's 9/11 and another year it's Katrina.  Which lines do you cut... all of them?  They each contribute equally.  Next year it will be a hurricane in Hawaii or a quake in California.  Maybe it will come again this quarter.  I think it's whack a mole, where they can't just run in there and surgically cut out 3b of the most costly policies.

 

 

[ubuy2wron]

Only fairfax shareholders would post 40 complaints about FFH underwriting results the day that Mr. Mkt decides our favorite co. is worth almost 5%  more and the dow is off 150 points.

[Myth465]

I agree with Eric. Its a bit of a game of wack a mole and the interest and div cover the combined ratio to a degree.

 

Fairfax isnt like the other insurers and will never have Chubb or Lancashire CM ratios. You buy them for their investing skills. Its similar to asking why one manufacturing cant get the margins that another historically have. To me its the perfect stock for a core 10% position at book value.

 

Pair that with a trading position when it drops below 90% book value and sell that trading position at or slightly above BV. Thats my plan going forward for FFH.

 

I look to capture the 15% that FFH hopes to earn each year and consider excess gains, hard markets, growth in FFH Asia / Brazil, and revaluations to 1.5 BV free options that i am not paying for.

[ERICOPOLY]

If they still average 105 CR's, I maintain that they are far better off just being a hedge fund. Why bother with the insurance side if it does not provide free float, or pay for your float?

 

Well, let's go back to the beginning of time (when they started Fairfax) and let's ask whether shareholders would have been better off in a hedge fund run by Prem.

 

A couple of things go through my head.  

 

1)  book value per share for Fairfax has compounded far in excess of the pre-tax returns on their stock picking

2)  the hedge fund therefore needs to use leverage to beat the book value growth rate

 

Now, let's discuss these 105 CR's that you seem to think are worthless today...

 

On $4b of underwriting this costs them 200m a year.  But they have roughly 12b a year in float.  

 

For their hedge fund to invest 12b in bonds and have it only cost them 200m a year, they will need to borrow money at 1.67% per year.

 

Where are they going to find money at 1.67% per annum interest cost?  I won't wait for the answer, it's a rhetorical question.

 

And suppose somebody did offer them some crazy low rate like that... is it callable?  As in, a margin call?

 

Who borrows money at 1.67%?  Name a single hedge fund.

 

 

 

 

[Guest longinvestor]

     Three conditions that prevail in insurance, but not in most

businesses, allow us our flexibility.  First, market share is not

an important determinant of profitability: In this business, in

contrast to the newspaper or grocery businesses, the economic

rule is not survival of the fattest.  Second, in many sectors of

insurance, including most of those in which we operate,

distribution channels are not proprietary and can be easily

entered: Small volume this year does not preclude huge volume

next year.  Third, idle capacity - which in this industry largely

means people - does not result in intolerable costs.  In a way

that industries such as printing or steel cannot, we can operate

at quarter-speed much of the time and still enjoy long-term

prosperity.

- Buffett in one of his annual letters

 

Great post!

 

Besides the underwriting discipline, I walked away from the AGM remembering the gleam in Prem's eyes when he talked about the growth propspects in Brazil, India, etc. Hard markets will follow soft, in Canada, US, but there is only one way emerging economies are going...UP. By a lot. This is a one-time opportunity in which FFH is participating. Surely Prem is on the prowl to buy more such business. Discussion of hard vs soft markets etc without looking at what FFH is doing for the future is rather incomplete. In a sense, they appear to be letting market forces work in US/Canada but are firmly in the driver's seat elsewhere.

 

I particularly like the fact that BV is understated when it comes to Brazil, India etc. Cant wait for these entities to be the major part of FFH in the future. Aint selling a single share.

[niels12think]

I generally wouldn't buy FFh ahead of the hurricane season.  It could get cheaper FWIW.  Not advice, just what I would do.     

 

Loss reserves are set aside prior to the hurricane season (or indeed, prior to any claim). If you believe that the reserving is conservative, then on average it shouldn't matter or actually be slightly better to buy before hurricane season (claims)?

Trying to out-guess the market price of the issue could in my view be considered purely speculative and anecdotal on par with "sell in May ..."  - not to say it might not actually be best.

 

Couldn't it be likened with a wager, where you have the following options:

• to get $60 now (conservative loss reserve) and have a 50% risk of paying out $100 in some time from now and 50% probability of paying out $0.
  
• to pass on the event and get nothing.
  

 

And if one passes, then one also passes on the other everyday earnings in the period.

 

Cheers

 

[vinod1]

 

105 CR on 4b -- that 200m that I'm talking about is pre-tax.  You seem to think it's after-tax...   Do I have it wrong?

 

I believe the hurdle on the incremental 3b of float is only 6.67% pre-tax.  Not 10% pre-tax.

 

When I worked at Microsoft, it amazed me that they launched product lines that never made money.  I could suggest to Mr. Ballmer that we should just stop doing this... I mean, why launch products that won't make money?

 

My point is that... nobody really belives that Fairfax intentionally writes an extra $3b of business on purpose where they need to make more than 6.67% investment performance to make it pay off.  I'm sure anyone admitting to doing this would already have been dealt with by Prem.  If they aren't doing it on purpose, then it's not going to be easy to identify exactly which policies to cut.  One year it's a quake in Chile, the next year a windstorm in Europe, another year it's 9/11 and another year it's Katrina.  Which lines do you cut... all of them?  They each contribute equally.  Next year it will be a hurricane in Hawaii or a quake in California.  Maybe it will come again this quarter.  I think it's whack a mole, where they can't just run in there and surgically cut out 3b of the most costly policies.

 

 

 

You are correct. It should all be pre-tax. So the hurdle rate is only 6.7%.

 

I might be wrong in this, but drawing a line in the sand on pricing so that you are only writing business that have an expectation of profit and let the float fall to whatever level it may be. Doing something like this should be possible - basically underwriting discipline.

 

Overall the float does add value, I do not disagree with that. Paying 1.67% on float seems nice, but we know there are going to be surprises and they are only going to be negative surprises so there is always the possibility of much much higher costs than anticipated. This is basically what Buffett has emphasized so much in his annual letters:

 

We have no interest in writing insurance that

carries a mathematical expectation of loss; we experience enough

disappointments doing transactions we believe to carry an

expectation of profit.

 

I guess this boils down to one of risk tolerance thing. Prem seems to be saying, let float cost a little bit more, we know we are going to do much much better investing on the float, so let us not reduce float all that much. 

 

Vinod

[ERICOPOLY]

All right, so what you are saying is that you want lower CRs as a measure of margin of safety.

 

Currently, at about 5% yield the insurance operations break even around CR of 115%.

 

But they've been writing 105 CR despite some very nasty cat years.

 

I imagine to hit the break even point of 115%, you could rewrite history of the past decade to include a disaster nearly as large as the Chilean quake EVERY SINGLE QUARTER.

 

Fortunately, that 5% yield is just a conservative number and their actual returns were far in excess of that -- so they still made money on their insurance ops despite a Chile sized loss EVERY quarter.

 

It looks to me like their insurance business isn't really that bad, and that there is a fairly large safety margin before taking losses.

[ERICOPOLY]

Today I listened to the replay of the conference call.

 

Very interesting tidbit from Brad Martin:

On an accident year basis they wrote a CR of 95% if you exclude ALL cat losses.  This compares favorably to q1 2009 where they wrote 96.1% on an accident year basis, excluding all cat losses.

 

Translation:

They improved the CR by 1.1 points, exclusive of that which is beyond their control.  They are hitting the ball better, fielding it better.

 

Underwriting improved.

 

Lumpiness in cat losses hides this underlying trend.

 

They also said that they have $2.3b at holdco right now and will still have $1b after buying ZeniTh.

 

Prem said high CRs at Zenith due to extremely soft workers comp market in California. Zenith wrote $1b a few years ago and now writes $400 -- so CR crushed by expense ratio.

[twacowfca]

The whole idea of reducing business in a soft market seems a little short sighted to me. All industries have their ups and downs, but one thing we all share in common is that customers are much easier to turn away than they are to recapture when markets turn hard. One soon realizes the value of retaining customers when you look at the cost in advertising dollars it takes to bring in each new customer. Before you start reducing your customer base, you better be darn sure you are not going to need them somewhere down the road.

 

 

 

There is always a certain amount of attrition in the customer base.  When business is good it makes sense to replace these old customers by acquiring replacement business.  New business is

often unprofitable at first

if you don't view it from the usual practice of deferring the acquisition cost.  Also, new customers generally make more claims than seasoned policyholders because of adverse selection, not knowing the new customer as well as the old customer etc.

 

There is a balance here.  Passing on much new business has certain dis-economies too.  The expense ratio will increase.   WWWD?  WEB passes on much business when the market turns soft, and then makes out like a bandit by writing much more when the market turns hard.  :)

[twacowfca]

I generally wouldn't buy FFh ahead of the hurricane season.  It could get cheaper FWIW.  Not advice, just what I would do.    

 

Loss reserves are set aside prior to the hurricane season (or indeed, prior to any claim). If you believe that the reserving is conservative, then on average it shouldn't matter or actually be slightly better to buy before hurricane season (claims)?

Trying to out-guess the market price of the issue could in my view be considered purely speculative and anecdotal on par with "sell in May ..."  - not to say it might not actually be best.

 

Couldn't it be likened with a wager, where you have the following options:

• to get $60 now (conservative loss reserve) and have a 50% risk of paying out $100 in some time from now and 50% probability of paying out $0.
  
• to pass on the event and get nothing.
  

 

And if one passes, then one also passes on the other everyday earnings in the period.

 

Cheers

 

 

 

Please explain what you mean by saying that reserves are set aside before hurricane season.  The Bermuda Re companies I'm familiar with don't do this to the best of my knowledge, although the better ones are quick to set aside IBNR reserves.

[niels12think]

I generally wouldn't buy FFh ahead of the hurricane season.  It could get cheaper FWIW.  Not advice, just what I would do.    

 

Loss reserves are set aside prior to the hurricane season (or indeed, prior to any claim). If you believe that the reserving is conservative, then on average it shouldn't matter or actually be slightly better to buy before hurricane season (claims)?

Trying to out-guess the market price of the issue could in my view be considered purely speculative and anecdotal on par with "sell in May ..."  - not to say it might not actually be best.

 

Couldn't it be likened with a wager, where you have the following options:

• to get $60 now (conservative loss reserve) and have a 50% risk of paying out $100 in some time from now and 50% probability of paying out $0.
  
• to pass on the event and get nothing.
  

And if one passes, then one also passes on the other everyday earnings in the period.

 

Cheers

 

Please explain what you mean by saying that reserves are set aside before hurricane season.  The Bermuda Re companies I'm familiar with don't do this to the best of my knowledge, although the better ones are quick to set aside IBNR reserves.

 

In my experience it goes much like this

1. When the premium is earned, the insurer sets aside on a gross level, the net present value of estimated the ultimate claims that should be covered by the policies earning the premium.

Because the claims are not yet reported or may indeed not even be incurred yet, this reserving is on a gross and not on a case by-case level. Nevertheless this is part of the reserving process.

2. As the claims are afterwards reported, then reserves are set aside for the actual claims and the IBNR reserves are lowered according to the estimates. Reserves set aside for actual claims might be average reserves, or for larger claims it might be an actual estimate.

3. As the payments out on claims happen, the reserves are lowered by those payments.

4. When a claim is closed (all payments have been made), the remaining reserve is set to zero.

 

The reserves are evaluated regularly (for instance each year) and might be changed based on management or actuarial opinion/choices.

 

This means that for step one, the insurer underwrites at a false profit if it sets aside a too low estimate of ultimate claims in 1 and would run a greater risk of reporting a low, but false CR. On the other hand, the insurer which sets aside a conservative reserve for the ultimate claims will report a high CR initially, but will also have a higher likelihood of eventual actual payments on claims being lower than the reserve set aside originally as the premium was earned. 

Similarly in step 2, in setting aside reserves for actual claims and reducing the IBNR reserve accordingly, there can be a profit (loss) if the reserves for actual claims was lower (higher) than what was originally set aside to cover the claims. If the insurer sets aside a conservative reserve for actual claims then this will show up as an apparent loss (or lower profit) than the insurer which sets aside optimistically reserves for the actual claims.

2B, Also and very importantly, the actual claims are random, and for instance for hurricanes, it might appear that if there are more expensive hurricanes than what was estimated originally that there will be a loss exceeding the original estimate of the ultimate claims. On the other hand, if there are less costly hurricanes than what was originally estimated, then there will be a gain as the actual reserves to be set aside are lower than what was originally estimated. However, one might argue that the loss (or gain) in one year from such events should be seen with respect to the many other outcomes that might have been.

 

In step 3-4, the reality kicks in. As the claims are paid out, the actual payments might be lower than the reserve set aside - this will typically happen when the reserves set aside are conservative.

Or, the actual payments might be higher than the reserve set aside for those claims - this will typically happen for more optimistic reserves.

It is only in step 3-4 that the profit or loss is real. In steps 1 and 2, the profit or loss and the combined ratio is only based upon management and actuarial opinion and strategy/policies for estimating reserves.

But even in step 3-4, the 2B effect is very real and should also be considered.

 

This is why you will need to view claims on an underwriting year (not accident year) basis throughout many years. Consider for instance a single underwriting year, say 2004. If you have the reserves and payments split by underwriting year, then you can see the original reserve set aside on risk underwritten in 2004. This is step one from above. Then for each year following this, there would be some changes to the reserve set aside. This is caused by step two above (and possibly also changes to the way that the insurer chooses to estimate the IBNR reserves) and step 3 as the actual paymens are made. Eventually, after many years - possibly even 50 or 100 years down the road - as the last payment is made, it is possible to compare the actual accumulated net payments from the underwriting year with the original reserve set aside in 2004 and the actual premiums earned.

You might think that you would then - in year 2054 or something - be able to see if the original underwriting in 2004 was made at a loss or profit (CR for the underwriting year when all is said and done), but the 2B issue should in reality also be considered:

It might be that a lot of insured risk just never materialized (plain old luck, for instance that there were no severe hurricanes). In this case the underwriting might actually have been bad even though there was an eventual profit. The underwriting should really be measured on the many different outcomes that might have been, and not just the one eventual outcome which actually occured.

 

There are many estimates and models and opinions and choices involved in the reserving process at insurance companies, and it can be extremely difficult to see if it is the company reporting a CR of 85 or the company reporting a CR of 100 which is actually the profitable insurance enterprise :'(

 

And then we haven't even considered the investment side of the business nor the reinsurance ...  ::)

 

Cheers

 

PS: here are some links:

 

Claims reserving now means, that the insurance company puts sufficient provisions from the premium payments aside, so that it is able to settle all the claims that are caused by these insurance contracts

 

The insurer will conduct a reserving exercise with a view to assessing what this ultimate cost will be.

...

This balancing item between the incurred claims and the ultimate claims is commonly referred to as the IBNR or the IBNR reserve. In pure terms, it only allows for those claims that have occurred before the valuation date but have not yet been reported to the insurer either directly or through the broker, hence the name. This pure usage is not used in practice. The more common usage includes reserves for items such as reopened claims, future claims on exposures to be written within the projection period, salvage and subrogation.

 

 

 

[omagh]

http://ir.wrberkley.com/events.cfm

http://ir.wrberkley.com/common/download/download.cfm?companyid=BER&fileid=373085&filekey=4a097bec-37a7-4d7f-b459-07ef81ee585f&filename=UBS%205-11-2010.pdf

Bill Berkley presented recently at the UBS conference.  He estimates that the industry is writing business at 110 CR, but fudging with reserve releases.  Chart 8 in the 2nd link above shows the gap.  A couple of other charts show that we're headed for lows simliar to 1983 and 1999 prior to hard markets.  It will take a major cat to blow out a couple of insurers/reinsurers and the tide will turn.  If AIG is the one, and the leading writer of weak business, does the government provide backstop or let AIG go?  I suspect that they will continue to throw capital at AIG which will continue to distort the natural fallout of the marketplace.

 

-O

 

T-bone, you have hit the nail on the head. There is too much capacity today. That is causing the soft market. Until capacity falls materially then we will not see a hard market.