The Negative Convexity Project
Me: We can’t buy the majority of Residential Mortgage Backed Securities [RMBS] anymore.
Boss: What! That is a staple asset class of ours. There’s nothing illegal about life companies owning RMBS.
Me: nothing illegal, yes, but because of new cash flow testing rules which our client is subject to, the negative convexity of RMBS will force our client to put up more risk-based capital than they would otherwise have to. Most RMBS will require so much additional capital that the additional yield is uneconomic, and that assumes we get the yield when we want it, ignoring prepayment and extension risks.
Boss: I can’t believe that we can’t buy any RMBS… are there any exceptions?
Me: There are a few. You know about the odd RMBS classes that have positive convexity, but little yield?
Boss: Yes, Yes… but why would we want to buy that? Our client needs yield!
Me: I know that. Would that they could do something other than need yield to sell yield. There is one type of RMBS that still fits, and it is the NAS bond [Non-accelerating security], last cash flow structure. Also, some of the credit-sensitive RMBS bonds rated less than AAA don’t affect the convexity issue, but we might not want to buy them, because the additional yield per unit risk is not compelling.
Boss: So what do we do?
Me: Buy NAS bonds when they are attractive, and buy CMBS that is attractive, after that look to corporates that our analysts like.
Boss: You are right, but I hate to lose a staple asset class.
What I wrote there took longer than a single conversation, and involved contact with the client as well. The client was very conservative with capital, because they levered up more than most life insurers. The results of detailed cash flow testing would affect large annuity writers like my client, and negative convexity would make them put up more capital, constraining the amount of business they could write.
Wait: negative convexity simply means your bond portfolio hates interest rate volatility — it does better when things are calm. That is certainly true with residential mortgages, where people refinance easily when rates fall, and in that era, no one faced falling property prices.
It took some effort, but I made my case to the client and my boss, and we stopped buying most RMBS. As an aside, it made asset-liability management tighter.
I was not totally hidebound with respect to derivatives. I bought our first asset-swapped convertibles, and synthetic corporates. If the risks associated with getting additional yield were small, I would take those risks. In both cases, they converted other asset into straight corporate debt (plus counterparty risk).
But I wouldn’t do anything. I grew to hate CDOs, as I saw how perverse the structure was. I remember one weird CMBS deal structure that added a note that combined the AAA, BBB & BB CMBS of the deal. What a nice yield, but the riskiness was greater than my models would allow for the incremental yield.
Finally, for this piece, the “piece of work” broker that I have previously described pitched me a private placement debt deal for a power producer affiliated with his firm. After hearing the initial spiel, I said, “Okay, soft-circle me for $25 million, subject to due diligence; send me all of the hard data via email and paper.”
My request should not have obligated me to buy the deal. Indeed, when I got the hard data, and began estimating the counterparty risks, I thought the deal was a loser, so I contacted the “piece of work,” and said, “Sorry, but we are dropping out of the deal — it just doesn’t offer enough value for the yield.” After some arguing, he eventually said, “Look, stay in the deal, and I promise you that I will get you out at par at minimum on deal day. Okay?”
Sigh, even though he was number eight with us, he served an important firm that could potentially do a lot for us, so I agreed. The day of the deal came, and indeed, he got us out at a teensy premium (I would have accepted par, maybe even a slight scrape). The deal did horribly, at least initially, though I have no idea of what the eventual credit result was.
As my boss who taught me bonds would say, “On Wall Street, if you want a friend, get a dog.” There are some honorable people on Wall Street, but the economics of Wall Street often leads to suboptimal results for clients, and indeed, the salesmen may be sweet enough, but they live to distribute paper; they don’t live to be your friend in any true sense. Professional duty to company trumps friendship.