Berkshire Hathaway & Variable Annuities

Sometimes I miss stuff in the news… last week Berkshire Hathaway reinsured the remainder of the variable annuity business that Cigna reinsured in the 1990s.  Four articles:

It was the last that attracted my attention, and led me to the rest of the articles cited.  To that article I commented:

This is very similar to the reinsurance deals that Buffett has done regarding asbestos.  Long-tail, upfront premium, with capped downside to Buffett.  He can invest the $2.2B of proceeds as he likes, and make additional money.  My guess is this leads to a profit of $500 million or so for Berkshire, plus any incremental from reinvestment of the $2.2B.

CIGNA was the patsy of the poker game of variable annuity reinsurance in the ’90s — this ends that ugly performance.  

Full disclosure: long BRK.B

In the past, CIGNA gave disclosures on its variable annuity liabilities.  From the most recent 10-K (pages 58-59):

Future policy benefits – Guaranteed minimum death benefits (“GMDB” also known as “VADBe”)

These liabilities are estimates of the present value of net amounts expected to be paid, less the present value of net future premiums expected to be received. The amounts to be paid represent the excess of the guaranteed death benefit over the values of contractholders’ accounts. The death benefit coverage in force at December 31, 2011 (representing the amount payable if all of approximately 480,000 contractholders had submitted death claims as of that date) was approximately $5.4 billion.

Liabilities for future policy benefits for these contracts as of December 31 were as follows (in millions):

2011 – $1,170

2010 – $1,138

Current assumptions and methods used to estimate these liabilities are detailed in Note 6 to the Consolidated Financial Statements.


Accounts payable, accrued expenses and other liabilities, and Other assets, including other intangibles – Guaranteed minimum income benefits

These net liabilities are calculated with an internal model using many scenarios to determine the fair value of amounts estimated to be paid, less the fair value of net future premiums estimated to be received, adjusted for risk and profit charges that the Company anticipates a hypothetical market participant would require to assume this business. The amounts estimated to be paid represent the excess of the anticipated value of the income benefit over the value of the annuitants’ accounts at the time of annuitization.

The assets associated with these contracts represent receivables in connection with reinsurance that the Company has purchased from two external reinsurers, which covers 55% of the exposures on these contracts.

Liabilities related to these contracts as of December 31, were as follows (in millions):

2011 – $1,333

2010 – $ 903

As of December 31, estimated amounts receivable related to these contracts from two external reinsurers, were as follows (in millions):

2011 – $712

2010 – $480

Current assumptions and methods used to estimate these liabilities are detailed in Note 10 to the Consolidated Financial Statements.

At 2010, the net liability was less than $1.6B.  2011, near $1.8B.  Buffett gets a $2.2B premium, and from what was said on the conference call, the net liability declined in 2012.  Thus I estimate Berkshire Hathaway as being $500M to the good on this transaction, subject to future market conditions.

Okay, so assume the duration on these liabilities is 10 years on average.  At the limit of $4 billion in claims, Berkshire Hathaway would have to earn somewhat less than 7%/yr in order to fund payments.  That’s more than achievable for Buffett.

From Roddy Boyd’s work on AIG, by the late 80s, Hank Greenberg was running out of places to expand in P&C insurance.  As such, he began to expand with life insurance and financial services.  Greenberg liked the fact that it diversified his risk exposures, and while it was small enough, did not threaten the solvency of the whole.

Buffett has always had a wider field to play on than AIG.  In some ways, life & annuity reinsurance does diversify BRK.  It is a little more complex for Buffett, because he has a decent number of bets against the equity markets & credit falling dramatically.  This deal is similar, in that the liabilities decrease as the equity market rises, and increase as the market falls.

Buffett does not have to maintain the hedges that Cigna currently does, but if it doesn’t Buffett might have to make some promises to the regulators like a BRK parental guarantee of the subsidiary reinsuring the variable annuities.

Buffett tends to think more in book value terms — he will try to over accumulate the implied returns in reinsuring the variable annuities.  From my angle, odds are he will succeed.  The only real concern is if he does not continue the hedge, and the markets fall for a considerable period, like 1973- 1982, or 2000-2008.  That would drive up the cost of the liability considerably, if left unhedged.  That said, Buffett, of all investors tends to do well after market declines, often finding compelling investments while others are afraid to commit.

Of itself, this deal is not large enough to materially materially harm BRK if the markets do badly.  It does add on to the “long bias” of BRK if left unhedged.

A note to those who get to question Buffett at his annual meeting: Ask him about this deal.  Cigna wrote a lot about this, held a teleconference, etc.  It was a big deal for them.  BRK said nothing.  Ask Buffett if he decided to continue hedging, or whether his investing in portfolio companies or wholly owned companies constitutes an internal pseudo-hedge that he thinks will outcompete hedging.

Personally, I expect that he does not hedge.  After all, he didn’t on his equity and credit index wagers.  Buffett is a bull on the US an the world; it would not be like his past behavior to hedge.  Besides, that’s why he keeps cash around, and hey, this gave him $2.2B more cash.

Full disclosure: long BRK.B

PS — one more article about variable annuity guarantees — they give me the willies.  Much as I like insurers, I avoid insurer that offer a lot of variable annuities.  I believe they are mis-reserved.  When I listened to Cigna’s conference call and their reserving off of “thousands of scenarios” I could not help thinking of how such methods could be tweaked or abused.  There is no good underlying theory for long dated options with uncertain payoff patterns.