The American baseball player Yogi Berra supposedly said that "It gets late early around here."  We're not quite sure what that means but we thought about it when we saw the two deals Berkshire Hathaway did with The Hartford and AIG respectively.  

The biggest deal by far is the legacy reinsurance transaction  between Berkshire's National Indemnity Company and AIG.  In exchange for about $9.8 billion in premium, which will be paid in full by June 30th, National Indemnity is assuming 80% of both the net losses and net allocated LAE that will attach at a layer of $25 billion. AIG has carried subject reserves of $34billion in place against claims from the policies which will be maintained for its own 100% share of the initial $25 billion of claims and its 20% share of future claims.  The Nico responsibility is 80% of the net losses and LAE in excess of $25 billion with a limit on its own payouts capped at $20 billion.  

By our math this means that if AIG is spot on in its reserving, and claims don't exceed the     $34 billion against which it's reserved, then Nico could be on the hook for 80% of $9 billion or $7.2 billion.  The premium, as noted is about $9.8 billion (it could be higher as there is a 4% interest rate on premiums not paid by the mandatory June 30th deadline) so it would seem that Nico would see a net gain of at least $2.6 billion. 

However, in the worst case scenario, if Nico begins to pay 80% at the $25 billion claims level claims could increase to $50 billion for Nico to exhaust its $20 billion limit having paid 80% of the $25 billion in claims from its $25 billion attachment point to its limit which would be reached at $50 billion. 

AIG believes that $34 billion should cover the claims but the desire for certainty again is here. If AIG is off on its reserving by what would be a 47% underestimate, and the eventual total comes to $50 billion instead of $34 billion, then you can see why AIG's Peter Hancock did the deal.  If claims did reach $50 billion AIG would be out the whole amount.

 

However, with the Nico arrangement AIG is out the first $25 billion in claims it continues to be responsible for; the $9.8 billion premium to Nico and its own 20% share of the layer from $25 billion to $50 billion or $5 billion. That potential liability, if claims reach the $50 billion level, under the reinsurance arrangement will "only" cost AIG $39.8 billion.  Without the reinsurance deal in place AIG would be on the hook for $50 billion assuming that Nico limit would have been reached.

The book Nico is taking on is substantially all of AIG's US commercial long-tail exposures for accident years 2015 and prior and supposedly includes the largest part of AIG's US casualty exposure during the period 2015 and earlier. According to the most complete article about the deal we've read --co-authored by  Leslie Scism in the Wall Street Journal, the deal includes product lines such as workers comp, D&O liability, PI, med mal, commercial auto liability and other liability policies.  According to the Journal "AIG is protecting itself in case its US liability claims will exceed the $34 billion, the amount it has set aside for payouts covered by the transaction."

Estimates are that it will be years before that $25 billion limit is reached --some estimates are as long as ten years--which will trigger Nico's reimbursement to AIG of 80% of the claims and LAE up to its own $20 billion payout cap --or $50 billion in total claims. That's a long time for National Indemnity to have that $9.8 billion premium.  Since 2005 National Indemnity has averaged a 6% investment return. If Nico maintained that 6% return in ten years the AIG premium would be worth $17.5 billion.  The math continues to work in Berkshire's favor as the length of time it would take to settle and pay the claims included in the $25 billion to $50 billion layer (another ten years?) would only produce more investment return on the admittedly shrinking premium principal amount.  

Even in a worst case $50 billion claim scenario which blows through the reserves by 47% it would seem that Ajit Jain and Nico are well positioned.  As for AIG, certainly prudent reserving left them comfortable with their $34 billion set aside but paying the $9.8 billion in premium to reduce their exposure by 80% on claims exceeding that amount has them breathing easier.  Risks and uncertainties as we have seen are inbound from areas outside the normal course of business.  This is one less thing for them to worry about.

Berkshire of course is well familiar with run off deals having agreed to a $14 billion limit when it accepted $7 billion in premium for the transfer of Lloyd's pre-1993 liabilities when the assets of Equitas were transferred to it in 2006.  Presumably that kind of experience allowed Berkshire to quickly ascertain the validity of that $34 billion AIG had set aside and conclude that it was an adequate amount. Still, per the Journal article, the deal came together fairly quickly, consummated with a lunch between Warren Buffett and Peter Hancock.  As has been reported in the press there supposedly had been friction between AIG and Berkshire since Berkshire hired AIG executives to start-up its own Berkshire Hathaway Specialty Insurance.

The deal with The Hartford, just earlier this month, saw Nico provide $1.5 billion of reinsurance for certain legacy asbestos and environmental liability exposures above estimated loss reserves of $1.7 billion. The premium The Hartford paid for that additional $1.5 billion in protection was $650 million.

We saw this piece from The Motley Fool summarizing the 5 Operating Principles behind Berkshire Hathaway's Reinsurance business. It's worth reading.