Category Archives: Asbestos

Prior Line Development: It’s all A&E

Pawing through 2012 financials, I noticed that the development on claims older than 10 years grew by $2.3B. Interesting given the total development on prior was $12.3B favorable. I guessed asbestos and environmental reserves drove the old years, so I put together this chart:

prior vs A&E 2012The chart is a little busy. The blue columns are total development on claims 10 years and older, by calendar year. The red dots are asbestos losses. The green line signifies all non-A&E losses on claims 10 years and older.

(From here on, I’m calling development on claims 10 years and older prior line development, because that’s the line it is taken from in Schedule P, Part 2. Please don’t confuse prior line development with prior year development, the latter being development on all prior year claims, not just the oldest.)

All amounts are billions of dollars. I forgot to label the y-axis – sorry.

First, notice to the far right: For the past two years, prior line development has been close to zero, once you exclude A&E claims.  (I’m assuming all A&E claims are more than 10 years old. ISO’s GL policy excluded A&E around 1986, so this is a pretty safe assumption for every year in the table, except maybe the earliest one or two.)

Next, notice that in every year except 2001, the red dots are almost always above the green line. That means almost every year, A&E losses are more than half all prior-line losses. Toting up all years, A&E is 78% of the prior line total.

A&E emergence has slowed in recent years, averaging $2.4B across the past seven years back to ’96, vs. $3.9B for the seven years before that. Even with that, A&E drags on the industry combined ratio by about six-tenths of a percentage point.


Berkshire gobbles up AIG’s asbestos reserves

Shouldn’t have been surprised by this one. Bloomberg with the call, via Business Week:

April 20 (Bloomberg) — Warren Buffett’s Berkshire Hathaway Inc. will get $1.65 billion from American International Group Inc. for assuming the risk of asbestos policies from the bailed- out insurer.

The deal with Omaha, Nebraska-based Berkshire’s National Indemnity will result in a deferred pretax gain of $200 million this quarter, AIG said today in a statement.

Warren takes on $4.45B of gross asbestos losses to go with the billions he’s already holding from CNA and Lloyd’s and XL and, hell, probably somebody’s grandma. (If he corners the market, will Justice file an antitrust action?)

In addition to the cash, he also gets the benefit of $2.8B in reinsurance protections that AIG had in place. Berkshire is responsible for losses stemming from uncollectible reinsurance. In all, the limit on this retroactive policy is $3.5B.

Meanwhile, AIG will book $200M profit. (Bloomberg makes it sound like the profit is booked this quarter, but accounting rules force it to be spread across a couple of years.) So it sounds like those asbestos reserves were sitting on AIG’s balance sheet for $1.85B net and $4.65B gross, though there might be some discounting action I’m missing.

At year-end, AIG carried gross asbestos losses of $5.53B and net of $2.2B. So AIG will still carry some asbestos reserves – maybe $880M gross and $375M net. Business Insurance notes the remaining asbestos reserves are booked at policy limits, so they can’t hurt AIG any more than they already have.

And unless everyone is being obtuse, AIG will still have about $127B in net environmental claims.

Wall Street Journal and others note this is part of a big last-minute cleanup before the federal government begins to sell its stake in AIG.

And that’s what makes the asbestos dump not-so-much of a surprise. When investors sink their money into a firm, they want it to go toward making tomorrow profitable. They don’t want it to pay for yesterday’s problems.

And with AIG’s recent bout with under-reserving ($4B+ last year), the popular press was reporting a bit of nervousness among investors.

Wednesday’s announcement is here. AIG’s 10K is here, with all the dirt around page 108. I bored you with this stuff before here.

Tagged , ,

Berkshire’s $1.8B reserve release

Business Insurance shows us an analyst with a sell recommendation on Berkshire Hathaway, pointing to a $1.8 billion reserve release so far this year.

I’m not an analyst, I’m an actuary. So I can’t tell you whether this means you should buy into or sell off Warren’s empire. I will point out that the $1.8B release is pretty close to the $1.6B in CNA’s asbestos reserves that Berkshire took on earlier this year.

So the reserve releases sort of cleared the way for the CNA deal. Berkshire was able to take on the reserves without overleveraging its surplus.


Fairfax bets on low inflation

Fairfax Holdings spent about $175 million this year, betting that prices won’t rise much over the next decade, and has already made $100 million on the deal. WSJ has details (sub. req.):

Fairfax paid $174 million in upfront fees to protect $22 billion of its investment portfolio against the possibility of deflation over the next decade. In exchange, Fairfax will receive a payment amounting to the drop in CPI below 2%—the level of inflation when Fairfax bought its contracts—multiplied by the $22 billion.

If deflation averages 2% annually over the next 10 years, Fairfax’s contracts would rise in value the equivalent of 4% of $22 billion, or $880 million, each year over the next decade, according to traders familiar with Fairfax’s trades.

In that scenario, if Fairfax holds on to its investments during the 10-year period, it would reap nearly $9 billion from its $174 million investment.

The company wouldn’t get anything for its bet if inflation turns out to be higher than 2% over the next 10 years.

The whole story sounds like the company is betting on deflation – prices actually falling. Even the excerpt above contradicts itself. And betting that prices will actually fall across an entire decade is pretty risky. Even Japan only had six years in the last 19 where inflation fell below 0%.

But a bet on low inflation at the start of this year makes a lot more sense, as 10-year Treasury yields were rising and the consensus held that government overspending would force prices to take off. Fairfax was playing the contrarian.

Now the 10-year is around 2.5%, and more investors are concerned about deflation. So Fairfax’s position has already netted it $100 million in capital gains.

The move is portrayed as protection on Fairfax’s investment portfolio, but I have trouble seeing that. Much of Fairfax’s investments is tied up in property-casualty reserves, which are a natural hedge against deflation. The method of P&C reserving <wonk> loss triangles </wonk> have long-term inflation built into them. So deflation between now and the date claims are paid would reduce your ultimate loss from what you thought it would be. Anybody smart enough to short the CPI would know that.

What this really sounds like is a clever play on Fairfax’s $1 billion asbestos and environmental exposure. (See pages 147 and 49 of the 2009 annual report.) That money won’t be paid out for years and years and years. Asbestos cases take forever to settle. It’s like Bleak House in slo-mo. Just this week, Travelers won a $262 million judgment on a case that’s 17 years old. And the loser is appealing. Who knows when that cash will change hands?

Putting 20% of your A&E reserves into a contrarian derivative would be odd and daring, but, so far, profitable.

Update: Hard to tell what kind of derivative Fairfax bought, but it behaves somewhat like an inflation swap.

Tagged , ,

CNA throws the asbestos yoke

One of my favorite insurance stories begins in the 1950s when CNA got $80,000 for a general liability policy written without an aggregate limit.

(To the lay reader, that means that no matter how many claims were made, CNA would continue to pay them. Normally, an insurer places an aggregate limit on the policy, so that if a lot of claims emerge, its liability is limited.)

The policyholder was a drywall manufacture called Fibreboard, and the way I  heard it, the underwriter went to his grave believing it was a profitable policy.

Unfortunately, Fibreboard laced its drywall with asbestos. In 1993 – more than 30 years after the original policy, CNA settled up, $1.5 billion poorer. Hey, actuaries, $1.5 billion on $80,000 –  that’s a that’s a loss ratio of 1.88 million percent!

I dust off that chestnut because CNA has finally dumped its asbestos and environmental exposures; the happy recipient is Warren Buffett, whose Berkshire Hathaway has been buying asbestos reserves for years.

And with the structure of this deal, it’s not hard to see why. CNA has $1.6 billion in A&E reserves. It is handing Berkshire $2 billion to be rid of them, plus ceding them $200 million of reinsurance collectibles. Assuming CNA has done a good job setting its reserves, Berkshire is ahead $400 million.

That’s a dicey assumption. Historically, CNA has not done well at the difficult job of estimating A&E reserves. According to its 2009 10-K (see p. 6 of the pdf), the company has taken $2.3 billion in A&E hits since December 1999, including $155 million last year. At 12/09, the three-year asbestos paid survival ratio – year-end reserves divided by average paid loss over the past three years – was 8.0, according to the 10-K (p. 105). That’s a tad low in my opinion, though not ridiculously so. And the rate of new claims hasn’t slowed (p. 104).

In short, the track record ain’t great, which is why Warren was able to command a fat margin.

Usually this sort of a deal is a play on the time value of money. Asbestos is notoriously long-tailed. When I worked with asbestos reserves, we figured we’d be paying claims for another 40 years. And the claims guys always told me Berkshire was notoriously slow to pay asbestos claims.

Anyhow if you assume the average date of payment on CNA’s $1.6 billion is 20 years from now and you could earn 5% after taxes, you’d have to put aside about $600 million today to cover all the claims. Normally, the difference between the booked reserve and the money you actually need to set aside – $1 billion here – would become profits for the company that assumed the reserves.

This deal doesn’t seem to be that way. $2.2 billion cash (representing $2 billion cash plus $200 million reinsurance collectible) will be held in a collateral trust, with CNA as beneficiary. I read that as saying CNA will receive the investment income from the trust.

It would seem they will get 10% more investment income this way than they would have gotten from the $2 billion cash they held. That’s because the $200 million receivable has been turned into a cash asset and CNA will get the income from that, too.

Berkshire makes $400 million (before taxes) when the deal settles. CNA books a $375 million after-tax loss.

Can’t blame CNA too much, as they’ve had an A&E problem since Nixon was president. Must be nice to be out from under it.