The bond market is closed today (for Armistice Day), but other markets are open. So it’s a working day for me, as is every Monday, and I was poking around at E*Trade when I noticed a link to a paper published by Bond Desk on “estimating the number of individual bonds required to minimize the impact of default risk”. Here’s a snippet:“Another implication is that credit quality is more important than individual security selection. A diversified portfolio of 10 or more investment grade bonds (rating grades between AAA-BBB) should be protected against extreme default losses. “ http://www.bonddeskgroup.com/sites/all/themes/bonddesktheme/...As you work your way through the paper, you cannot but be impressed with the thoughtfulness of their methodology. But when you reach their conclusion (that just ten bonds might be enough to keep you out of trouble), you should conclude that the paper is either wishful thinking, or mere cleverness, but it certainly is not a responsible effort to manage risk. In short, it is nothing that you’d want to use as the cornerstone of your risk-management policy. To see why/how this is so, merely think back a couple of years to 2007-2008. What was the rating on Lehman’s debt before they went under? On WAMU’s debt? On CIT’s debt? And what were the recovery-rates to bond-holders? All were rated invest-grade before they failed. But if any of them had been part of a ten-bond, fixed-income portfolio, sizable losses would have been suffered. Roughly speaking, Lehman, WAMU, and CIT could be categorized as ‘financials’, and it wouldn’t have been unreasonable to have had some exposure to ‘financials’ in a ten-bond portfolio (along with exposure to retailers, industrials, energy, utilities, etc.) In other words, the fact that it was Lehman, CIT, and WAMU that went under, rather than JPM, BAC, etc. is incidental. Any of them could have suffered Lehman’s fate, because all of them were screwing around with counter-party risks. That’s what ‘financials’ do. They know the Fed will backstop them. They know they have been given a license to “privatize profits and socialize losses”. So they do the rational thing and exercise that license to the max, meaning, they will blow themselves up sooner or later, and bond-holders will take a haircut (though the principals will make out quite nicely). So, create a properly-diversified, invest-grade, ten-bond portfolio, and do your buying at par. Assume that one of your positions blows up and that the recovery-rate is a generous, $0.25 on the dollar. What will be the impact on P/L across the whole portfolio? Across any realistic time-frame, that single default puts the whole portfolio underwater, as in, you’re not making any money and would have been better off burying your capital in the backyard. In other words, as I’ve said repeatedly about the bond-game as it is conventionally explained to would-be, fixed-income investors: What might seem safe really isn’t. What, then, is the magic number of bonds such that suffering defaults could be tolerated? The honest answer is that the number of bonds is only one of the factors that has to be considered, and that ’ten’ is merely a good start. In other words, responsibly-done bond-investing is no different than any other securities game. No matter the asset-class, you’ve gotta find an edge that creates, on average and over the long haul, a positive expectancy. A ten-bond portfolio will do that. The paper is right in that respect. But such a portfolio carries more risk than most investors should accept, because it fails to minimize to a tolerable level the consequences of a failure when it does happen. Ask yourself this: “Would you rather lose one position in ten, or one in forty?” “Would you rather lose a position for which you paid par (or a premium to par) as opposed to one that you bought a steep discount to par?” Sometime, when you’re truly bored, work out for yourself the consequences of suffering such-and-such failure and recovery-rates, over a portfolio of X number of bonds, bought at an average price of Y, offering an average YTM of Z. What you’ll find is that ten bonds isn't your final target, but merely a good beginning. Charlie
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |
BATS data provided in real-time. NYSE, NASDAQ and NYSEMKT data delayed 15 minutes.
Real-Time prices provided by BATS.