By Dan Hallett on December 17, 2012
Type the phrase “bond bear market” into Google and you’ll see a long list of articles proclaiming the certainty of the oncoming bear market in bonds. I agree that bond yields are artificially low and that bonds are expensive relative to stocks. But I don’t agree with some of the fundamental arguments in favour of abandoning bonds altogether.
Rates must rise
Without failure, for my entire 18+ year career the overriding expectation has been for interest rates and bond yields to rise from their “historic lows”.
My start was in 1994 – notable since that year stands alongside 1979-80 as the worst bond market in North American history. That year, however, was immediately followed by one the best calendar years on record for North American bonds – comfortably recouping 1994 losses and then some. After all this time, the same “rates must rise” argument continues to proliferate.
It’s true that the lower that yields fall, the less they can continue falling. That’s a mathematical fact. But that alone doesn’t mean that yields are bound to rise. And save for one grim scenario, a rise in bond yields is no death sentence for bonds.
Hyper-inflation, bonds’ greatest threat
The most frightening scenario for bonds is hyper-inflation – as happened in Germany in the early 1920s. When a a country experiences hyper-inflation, its currency and bonds becomes worthless. While accelerated inflation is a very real long-term risk, few expect it to materialize.
How might bonds fare, then, if we get a repeat of brisk rate hikes that we saw in 1994? To answer this question, I examined short term, mid-term, long-term and our firm’s laddered bond solution in the context of quickly rising rates.
Long bonds carry the most risk
The table below summarizes the yield increases experienced through 1994.
This leap in yields caused low double-digit losses in mid- and long- term bonds and mid-single-digit losses for short bonds. Notice that in late 1993 and early 1994, the yield curve was steep. After the rate hikes, it had flattened considerably. With yields at a much lower starting point today, what kind of damage might we see if we saw a repeat of these brisk yield spikes?
To answer this question, I used three individual government bonds (i.e. short, medium and long term) for price and coupon information. I then assumed that, in year one, bond yields jumped 200 basis points (i.e. 2 percentage points) across the board – without affecting the steepness of the curve. Subsequent to this big jump I further assumed that yields would fluctuate but remain close to the ‘newly elevated’ yield level until maturity.
The first table below shows hypothetical annualized rates of return over various periods assuming that bonds are held to maturity. While strongly rising rates push bond returns into the red in the first year, note that holding each of the bonds for the term delivers an annualized return that is very close to the yield-to-maturity (YTM) at purchase.
The above table applies to bonds that are bought and held to maturity. If we’re instead dealing with a bond portfolio that maintains a constant term to maturity (or duration), the results change only slightly. The next table shows the same calculations, this time adding figures for the laddered bond solution we use for our clients and assuming constant terms to maturity.
Bond bears focus on the initial impact of rate hikes or on the very long end of the yield curve. But even when terms are held constant, holding a bond portfolio for a time frame equal to the initial average term still produces positive returns despite sharply rising rates at the outset on bonds that never mature.
The greatest risk is in holding very long-dated bonds. In the scenario that I cooked up, it took several years for long bonds to recover from their initial steep losses. But I’ve yet to see any investor hold most or all of the bond exposure in long bonds. And short- to mid- term bonds recovered their losses within 3-4 years.
Rather, simply matching the term or duration of an investor’s bond portfolio to the time frame of a stated goal is an effective way of avoiding losses in bonds. And this works even assuming yields rise extremely quickly. If yields rise more gradually, those initial losses are skinny or non-existent but long-term returns are not materially impacted by the speed of rate changes.
Drivers of rising yields
Rising bond yields can occur in the face of events including a growing economy and a currency or debt crisis. If inflation is constrained to single-digit price increases, then it is likely a sign of healthy inflation – perhaps resulting from sustainable economic growth.
But for this to happen, the economy must become healthier and start posting higher rates of growth. In that scenario, bonds will suffer initially but corporate bonds and stocks will likely prosper. (We have allocated 60% of our bond ladder to corporate bonds.)
If you are loaded up on long bonds, expect accelerated or hyper inflation then fleeing bonds may well be the right move. Everybody else can relax in the comfort of knowing that even if bonds post losses, they will be short-lived, capital will be preserved and at least a modest return is likely simply by holding for a few years.
Bond bears’ growl is sounding more like a whimper.
Post Script – To see our year-by-year calculations for each bond exposure see:
Note that an earlier version of this post showed more dramatic rate hikes, mirroring activity seen in 1994. In retrospect, we moderated the potential rate hikes give the much lower yield starting point today compared to 1994. The end points and conclusions are not vastly different.