Crop Losses

[JEast]

Not this quarter, but come next quarter expect some large losses on crop insurance.  The crop business looks good as one only has losses 2 out of every 10 years.  Of the two loss year, one is not too bad, but the second is usually a doosey.  Looks like we are starting out our 10 years with major losses in year 3.  Why we went into this business I do not understand and could not get an answer at the annual meeting this year as to why either.

 

http://news.yahoo.com/worst-drought-since-1956-shrivels-corn-soy-crops-001346529.html

 

 

Cheers

JEast

[Uccmal]

I have come to the conclusion after being an FFH shareholder for 15 years that they are mediocre underwriters and always will be. 

[oldye]

Is 3% really such a high cost of capital?  Berkshire has about 60 billion in debt that costs them more than 4%..

[HJ]

My impression is that crop insurance is largely an agency business, with USDA absorbing most of the risks, and the actual insurance carrier's loss is quite minimal.  Several years ago, when Iowa had a flood, impacting the region's corn crop, Partner Re had a conference call going through their potential losses, and how the losses were to be shared between themselves and the US government, the portion of losses that they actually needed to absorb turned out to be very minimal. 

 

You may have losses because of operating expenses, but unless Fairfax's business is meaningfully different, the loss may be more caused by operating expenses, rather than actual Cat.  There were some changes to the USDA program since then, but my impression is they largely took out the juice in a normal loss year, rather than shifting cat year losses to the insurance carrier.

[Uccmal]

Is 3% really such a high cost of capital?  Berkshire has about 60 billion in debt that costs them more than 4%..

 

No argument from me.  I have just come to accept it. 

[JEast]

Given that Fairfax's investment board (group) thinks there is a very good possibility of deflation, 3% is way too high a cost of capital for this company.  This is not a metric that I pay that much attention to, but the idea (metric) does bring up an important point in my view.

 

I am under the impression that I am associated with above average folks with above average talents.  As such and as a P/C company, we should have a cost in the 1% range (over the longer term).  For 10-years now C&F has been a mediocre underwriter, at best.  I suggest that they cap their policy count (not premiums) until they can prove underwriting discipline -- no more bolt on business.  Prem always says we will not chase market share, but some underwriter's pens appear to chase policy count with not much results to account for.  I am supportive of Doug Libby because of his fine work at Seneca, but C&F was a mess when we got it.  If the next five years can average 100%, then Doug will have a winner.

 

As follow-up to HJ's comment on that the 'USDA absorbing most of the risks' is a false premise.  It is true that you can put some of the riskiest policies in a different pool, but that does not account for the policies you keep during a drought.  This is the tricky part of crop insurance.  You think you are laying off risk when you are actually taking on more risk.  On top of that, we expanded our crop business this year as I was told at this year's meeting.  Ouch.

 

My intent is to be constructive, but understand the tone my be otherwise.  As you might now suspect, I do not like the crop business or consistently underwritten losses.

 

 

Cheers

JEast

[petec]

Is C&F that bad?  Annual meeting slides that I saved suggest it has a 2002-2010 average combined ratio of 99.8% and average annual reserve redundancies 8.9%.  Accident year triangles suggest that pre-2001 it was reserved poorly but since then has done well in hard markets but and only moderately poorly in soft ones.  Is that not better than average in a non-niche insurance business?

 

I'm happy to be corrected as I am fairly new at analysing insurers, but I am broadly in the camp that thinks FFH as a whole is an improving underwriter but that this has been hidden by unfavourable developments on old policies.  I think the next decade is likely better, especially because they're buying better stuff, and in a hard market they'll make hay.

[zippy1]

I found this story about how the crop insurance is being sold quite interesting. Also, one cannot compete on "price."

http://www.npr.org/blogs/money/2012/08/17/158933382/competing-against-the-nicest-guy-in-town

The federal government spends about $7 billion a year on crop insurance for U.S. farmers. Policies are sold by private companies, but the government sets the rates, so the companies can't compete on price.

 

That means the guys who sell crop insurance have to find other ways to compete. They try to out-nice each other. They are very charming. They wear polo shirts depicting hobbies. They have fun nicknames. And they know everyone in town.

 

Don "Dizz" Biefelt is the most interested, friendly neighbor you can imagine. He's 82 years old, and he sells crop insurance in Anchor Illinois.

 

I sat with him on Anchor's one public bench. His customers were everywhere. That guy over there, working on a truck — he's a customer. (And, by the way, the customer's wife just had a gallbladder out, Dizz says.) The guy in that house over there is another customer, as is the guy down at the end. Dizz knows their mothers, their nicknames, their wives' digestion problems.

 

But Don has competition. Brent "Hondo" Honneger works a few miles down the road. Brent also wears polo shirts and is charming and knows everyone.

 

The morning I meet Brent he has organized a Q&A session for farmers on how to file a crop insurance claim. Brent invited a bunch of Don's clients to this event. Just in case, he says, they're not getting all their questions answered.

 

Then he makes sure to stand by the door and personally greet each farmer:

 

There's one customer who Brent has been trying to poach from Don he's been trying to poach from Don for decades. "But he's like, 'I go to church with Don. I see him every Sunday.'"

 

Brent says he is always gracious. But he'll occasionally ask Don's customers a few questions about whether Don is keeping up with the times. "Does he mail everything still?" Brent will ask. "He's still operating like we operated 20 years ago."

 

Government subsidies for crop insurance have set the stage for thousands of tiny popularity contests in small farming communities all across the country.

 

Right now, in the worst drought since 1956, most farmers have generous insurance coverage, and I can confidently report they're getting very good service.

 

For more: Why does the government subsidize crop insurance in the first place? We try to answer that question in our latest podcast.

[Mikenhe]

I caught this on the way to work the other day - woth a quick listen but not very detailed.

 

http://www.npr.org/blogs/money/2012/08/23/159806895/why-do-taxpayers-subsidize-farmers-insurance

[HJ]

Endurance (ENH) has a pretty large crop insurance business.  Their recent presentation at KBW conference provides some insight into their crop insurance business as well as the USDA program. 

 

The presentation (the agricultural part starts at page 16):

http://www.sec.gov/Archives/edgar/data/1179755/000119312512353965/d397289dex991.htm

 

The webcast (the part about crop insurance starts at around 30 minutes):

http://wsw.com/webcast/kbw5/enh/

 

Doesn't mean Fairfax's business will necessarily look like this though.

[JEast]

I suspect that Odyssey's losses will indeed look similar and maybe worse depending on how they ceded.  I am counting on a loss in the $50-$70M range given that they wrote nearly $700-$800M this year.  El Nino is helping out the hurricane season, but most of this possible good fortune has gone to farmers instead.  Looks like a flat year :( on the whole for Fairfax's underwriting as we need 95 or below.

 

Cheers

JEast

[ourkid8]

I could not agree more! With the appointment last year of Andy Barnard to oversee all of Fairfax Insurance and Reinsurance operations, I can guarantee the focus will be on sub 100 CR! 

 

I'm happy to be corrected as I am fairly new at analysing insurers, but I am broadly in the camp that thinks FFH as a whole is an improving underwriter but that this has been hidden by unfavourable developments on old policies.  I think the next decade is likely better, especially because they're buying better stuff, and in a hard market they'll make hay.

[giofranchi]

I could not agree more! With the appointment last year of Andy Barnard to oversee all of Fairfax Insurance and Reinsurance operations, I can guarantee the focus will be on sub 100 CR! 

 

I'm happy to be corrected as I am fairly new at analysing insurers, but I am broadly in the camp that thinks FFH as a whole is an improving underwriter but that this has been hidden by unfavourable developments on old policies.  I think the next decade is likely better, especially because they're buying better stuff, and in a hard market they'll make hay.

 

 

+1

 

giofranchi

[HJ]

But a bad year doesn't make crop insurance a bad business.  One of the worst drought in history produces a combined ratio of 105-110 on this book of business, and most of the time, CR is in the 80's?  I'd think that's exactly the type of float that Fairfax wants more of? 

[JEast]

HJ -- where did you get your numbers??  (most of the time CR is in the 80s, 110 this year?)  My experience of watching the business for some time is much different.  Given that the governmental reimbursement is where the money is made, I would like to see your source.  I would very much like to be wrong in my assessment.

 

Cheers

JEast

[HJ]

I am actually very curious about your data point as well, but mine is based on that Endurance presentation.

 

On page 19 where they talk about their ARMtech business, there's a historical loss ratio results chart, and the commentary on the side, they state:

 

Historic average loss ratio post U.S. Federal cessions has been 78.7% [adjusted for the 2011 Federal reinsurance terms]

The best year was 2007 with a 69.8% net loss ratio and the worst 2011 with a 90.5% net loss ratio

ARMtech’s current expense run rate after the A&O subsidy is approximately 7%

 

So that gives a historical average of 85.7% (78.7%+7%) combined ratio during the 2004-2011 period?  In the webcast, (I think I remember hearing) them talking about this year that they are expecting loss ratio around 110.  They also stated in the webcast that in the absolute worst case in this line of business which is "never going to happen", their loss ratio will be capped at 160%.  In that same presentation, on page 23, they also lay out how the USDA program is structured in terms of loss sharing and gain sharing, in return for 41.2% of ceded premium.

 

I believe that pre 2011, the Fed reinsurance term actually left the carrier with a more volatile gain / loss profile?

 

 

[JEast]

The ARMtech presentation is cherry picking with starting data in 2004.  As mentioned earlier, crop insurance is a marginal business over a 10 year period with 8 good years, 1 bad year, and 1 year that wipes people out.  On the face of it, it looks similar to CAT Re but the ratios do not shake this out in my view.  One of the reasons is the regulation is to volatile with changes in the program every couple of years.  The chart below looks like a normal CAT chart, but due to changes in the reimbursement regulation and fixed premiums, move the numbers up another 5%+. 

http://s9.postimage.org/9go0uvcsf/MPIC.png

 

If the chart went back to 1988, you would see another 200%+ ratio.  It takes a lot of luck to get enough 80-90% years to recover a 200% year.  Since I am long FFH, just a terrible line of business from my seat.

 

 

Cheers

JEast

[HJ]

Thanks.  The chart was a bit small, I wasn't able to see it clearly.  Don't mean to beat this topic to death, but I'm genuinely interested in knowing what kind of business this line really is. 

 

I looked around, and was able to find on the National Crop Insurance Services website, several reports which studies the historical MPCI program profitability. 

 

http://www.ag-risk.org/SpecRpts.htm

 

The latest report provided by Grant Thornton prior to the drought news,

 

http://www.ag-risk.org/NCISPUBS/SpecRPTS/GrantThornton/Grant_Thornton_Report-2011.pdf

 

The conclusion here is that measured as a percent of net retained premium (net of expense ratios for the P&C companies), the MPCI program is slightly less profitable, with larger variability in profitability than what they call  the "P&C business" (using an aggregate data set published by AM Best, which includes a lot of auto and home insurance premiums). 

 

Although the numbers shown in the Grant Thornton study is different from what look like a similar study done in 1999 by PwC for the years that they overlapped.  The grant Thornton number seem to encompass a bigger data set:

 

http://www.ag-risk.org/ncispubs/specrpts/OIG5-99.PDF

 

 

One interesting factoid in reading these report:  The data for pre tax net income for P&C company includes investment income and capital gains on float, whereas the pre tax net income line for MPCI doesn't include any investment income, because according to Grant Thornton, the program participants remits premiums at harvest within 30 days to RMA, so the float is negligible for MPCI program.  Although I assume they collect from the farmers earlier than that deadline so do earn a little bit of float, and in addition, the earlier 1999 PwC report shows very meaningful estimated investment income through 1998 for the MPCI program, which maybe explained by program changes in terms of how these premiums are collected, but if added to the Grant Thornton net income number, would have pulled up return for the MPCI program meaningfully.

 

The other thing is an ROE comparison really should be as a percent of surplus capital that is required to be held.  The earlier PwC report in 1999 calculates an ROE, but the later Grant Thorton report doesn't. 

 

The third observation is that post 2004, it really has been a very good time for the business in aggregate.

 

Bottom line: using those aggregate numbers, I still come to the conclusion that the MPCI program, while not Geico, is still a solidly pretax low teen ROE business (assuming a generic 1:1 premium to surplus ratio) for the participants, as long as you stay within that program.  If you chose to retain risk though, it could get ugly in a hurry, like this year.

 

 

[JEast]

Updated/Enlarged the above chart.  In addition and to your comments, the business is indeed very short-tail so it is tough to use the float to any good investing purposes and another reason for the marginal business model in my view.

 

Cheers

JEast

[gaf63]

Corn states lead $25 bil crop loss payout,  and how much of this amount was covered by FFH's subs.? Will be an interesting qtr. report

 

http://www.bloomberg.com/news/2012-10-17/corn-states-lead-record-crop-loss-payout-bgov-barometer-1-.html