Personally, I did not have an outline when I began this series.  If I had decided to create a “story arc” it would ended with part six, which led to my becoming the corporate bond manager, but I will end this series on a different note, and with different lessons.

1)  After the merger, post 9/11, we decided on heresy.  We were going to sell a large amount of the CMBS I had acquired  over the prior three years for a large capital gain, and redeploy it into the bonds of hotels, airline EETCs, and every other area negatively affected by 9/11.  We did a huge down in credit trade.  Some of the tale is told here.

As far as mortgage bonds went, the sale was a stunning success.  The execution levels for the sales were great, and what we reinvested in were areas of the market that were dramatically oversold.  What could be better?  (A client who knew how use the results would be better.)

2) One day in 2000, the client came to me and said, “We’d like to do a bond indexed annuity.”  After reviewing product design, which allowed holders a one time option to increase their rate over the term of the annuity, and doing a little bit of game theory work, I said, “Here’s the good news: given what we know about policyholder behavior and what we know about bonds, this is a cinch to hedge.”  As I explained the dynamics to them they realized the risks were minimal, and they decided to proceed ahead.  Sadly, the product was not attractive enough, and it was killed.  Equity products got attention in that era, not income products.

3) In 2002, when I was a corporate bond manager, I had to do what a mortgage bond manager does on occasion: read thick prospectuses.  Bear markets in credit often offer the most interesting deals — as an example, the Prudential “C” bonds.  In this case I had to read through the prospectus of a Dominion subsidiary that had an Enron-like financing structure post-Enron.  If Dominion’s credit was downgraded, and its stock traded below a certain level for a certain number of days, the bonds would have to be redeemed at par-plus through an issuance of preferred stock.

We became one of the larger holders of those bonds. Enron-like structures are good for bondholders if they are a small part of the capital structure.  They are bad if they are big, because you can’t protect everything.

We bought the bonds at a significant discount to par for a 3-year bond.  Our research showed that Dominion the parent company was on the hook.  The larger holders negotiated (we were in the top 10), and eventually the bonds were tendered for a 10%+ gain, plus 7% of carry over the less than one year period.  I ended up sharing the the experience in real time with Cramer, who wrote a post about it in the midst of the furor.  Yes, bond markets can affect stock markets, and vice-versa.

4) There was a large debate in that era over what to do about Qwest / US West.  Amid corporate troubles, there was one easy play — buy long-dated US West bonds.  We all agreed on that.  But should we own Qwest bonds?  That was harder.

But then one day, because of our high yield contacts, we came into contact with the offering documents of a Qwest subsidiary, which had done badly.  It was a private placement, so when we inquired about it they asked for our documents, and we faxed a copy to them, though we had not been on the original deal (good thing).

As it was the debt was trading at under 50 cents on the dollar.  As I read through the prospectus, I realized that the debt was guaranteed by Qwest.  Given the short term of the debt, it made sense to trade our lower yielding Qwest positions for it.

And what did we do?  Nothing.  What happened to the debt two years later?  It was paid off at par.  To misphrase Mr T., “I hate it when a plan doesn’t come together!” (I am certain that comment dates me.)

5) (the end of the end) While I was the less-restricted manager of bond assets, both corporate and mortgage, for my client, I would sometimes meet with my former boss for lunch.  I would tell him what I was doing, and he would say, “Don’t you realize the risks you are taking?”  I would tell him, “Yes, but I have to evaluate risk and return relative to the other risks and returns available elsewhere in the market.  You have to pick the best of them.”

One thing I do know, I was far more willing to accept bond market risk than my boss, who had a strong teaching in the 90s, prior to hiring me, a neophyte.

There are two odd things here: why should an actuary turned bond manager take risk to make money amid panic?  Because he learned to be contrarian — this is something that must be experienced in order to teach it.  Second, why should I do far better than the guy who taught me?  I was more willing to take risk when it seemed to be rewarded.  Don’t get me wrong, when spreads and yields are narrow, I am not there.  I don’t take uncompensated risks.

There is wisdom in trying to understand the credit cycle.  It is one of the few constants in terms of economics.  If you follow credit, you will understand the economy in the short run.  If you follow the credit cycle, you will invest better than most.

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.