Wednesday, April 06, 2011

Everything Old is New Again in Bonds

Unconstrained strategies for bonds are hot now with yields so low.  But wait. Let’s take a step back.  What do we mean by a constrained strategy?
A constrained strategy is one that limits the investments one can engage in either through:
  • Specifying an index that the manager is charged with beating
  • Specifying percentage limits for investments, split by categories such as credit quality, interest rate sensitivity, asset subclasses (ABS, RMBS, CMBS, Corporates, Agencies, etc.), and other variables
  • Barring investment in more funky fixed income instruments such as preferred stock, trust preferreds, junior debts, CDOs, ABS, RMBS, CMBS, etc.
  • Or some combination of the above.
There have been unconstrained strategies in fixed income before — they just weren’t called that.  Many value investors in the old days didn’t care what the legal form of the investment was — they only looked for an adequate margin of safety.  Their portfolios were a hodgepodge of debt and equity instruments.  Specialization in only doing debt instruments wasn’t common.
Most debt-only investments were constrained, particularly those from bank trust departments.  Of course, this was an era where investing in junk debt was not respectable for all but the most intrepid of investors.
With the advent of the 1980s we had two innovations: junk bonds and bond index funds.  The first took the world by storm with the demand for yield; I experienced that at the first insurance company that I worked for — they overloaded on junk bonds.  This was before the regulators began regulating bond credit quality more strictly.
The second took a longer time to germinate.  The first bond index fund came into existence in 1986 at Vanguard.  They couldn’t call it a bond index fund, because they could not exactly replicate the index.  There were too many bonds that were illiquid, and they could not buy them at any reasonable price.  Instead, they took an approach that we would call “enhanced indexing” today.  Match the interest rate sensitivity of the index, and the credit quality, but choose bonds that had more potential than the bonds in the index.
In that sense, though the SEC allows bond funds to be called index funds today, all bond index funds are enhanced index funds because there is no way to source all of the bonds.  And from my own days as a corporate bond manager, I learned that bonds in major indexes always trade rich.  From my piece, The Education of a Corporate Bond Manager, Part IX:
There was another example where I crossed bonds where it was legitimate — if it was done to help a broker in distress.  One day, someone offered me a rare type of Capital One bonds at a normal level, and I asked whether the bonds in question were the ones that were in a major bond index, without saying that per se.  After figuring that out, I bought them at the level, and called a broker that was likely to be short the bonds to see if he wanted them.  He certainly did, and offered them at a three basis point concession to where I bought them, as opposed to ripping the eyeballs out (as the technical term went).
The whole set of two transactions took 15 minutes, and made $15,000 for my client.  What was funnier, was that my whole family came to visit me that day, my wife and at that time, seven kids.  They heard the two transactions, though I had to explain it to them later. To the second broker, I had each of the kids say “Hi,” ending with the then three-year old girl who squeaked “Hi.”  He said something to the effect of, “I knew you had a large family, but it only really struck me now.”
That three-year old is now a beauty at twelve, and bright as anything, but I digress.  (They grow so fast… the nine-year old girl is cute as a button too.)
Bond management was once unconstrained by those who looked for total returns in the old days, and constrained in the old days by those who looked for yield.  (Many managers would not buy bonds that traded at a premium.)  Then the bond indexes became popular as a management tool.  In one sense, it freed bond management, because rather than hard constraints, they matched credit and interest rate sensitivities of the index.
But what that constrains is credit policy and interest rate policy.  One managing to beat a benchmark index has limited options.  What if you want to position for:
  • Widening credit spreads
  • Narrowing credit spreads
  • Rising interest rates
  • Falling interest rates
  • Yield curve steepening
  • Yield curve flattening
  • Outperformance/underperfomance of a given sector
Any sort of directional bet could go wrong, and more often than bonds that fit the idea of replicating the index parameters, but are special in ways that the index does not appreciate.  So rather than going “whole hog” with the bet, you merely lean toward it, such that if you are wrong, you won’t destroy the outperformance versus the index.
But in this modern world where derivatives are widely accepted as fixed income instruments, a la Pimco, fixed income managers can do a lot more.  There is more freedom to make or lose a lot of money.
The unconstrained strategy can be thought of  in two ways: always trying to earn a positive return with high probability (T-bills are the benchmark, if any), or being willing to accept equity-like volatility while the bond manager sources obscure bonds, or takes large interest rate or credit risks.
I prefer the first idea, because it is more conservative, and fixed income management should aim for safety on average.  As I have said before, I only believe in taking risks that are well-compensated.
But here’s a hard one.  With the yield curve so wide, shouldn’t a bond manager with an unconstrained mandate put a little into long bonds or long zeroes?  I would think so, but I wouldn’t put a lot there unless the momentum started to favor it.
I like the concept of the unconstrained strategy; indeed, it is what I am doing for clients, but it is of the first variety, try to make money for clients in all markets, and not just be a wild man in search of yield or total return.
I find the move to unconstrained mandates to be a return to what value managers did long ago, but in a more complex fixed income environment.  I wonder though, as to whether the future failures will invalidate the idea for most.  It is tough to manage any asset class while adjusting the risk level to reflect what should not be done in a given era, whether in equities or debt.  The danger comes from trying to maintain yield levels that are higher than what is sustainable.

No comments:

Lunch is for wimps

Lunch is for wimps
It's not a question of enough, pal. It's a zero sum game, somebody wins, somebody loses. Money itself isn't lost or made, it's simply transferred from one perception to another.