Rolling Down the Yield Curve

Yield Curve Riding Strategy — what even is it??

(also known as “Rolling Down the Yield Curve”)

But first, there’s a condition, lol.

(Condition) the interest rate term structure has to be upward-sloping.

(And this is talking about YTM — if spot is upward-sloping, then YTM and forward end up upward-sloping too.)

OK so, given an upward-sloping term structure, this strategy lets you squeeze out a tiny bit of extra return — earn a little more than the 1Y yield using a 1-year bond, a little more than the 2Y yield using a 2-year bond. Excess return in that sense.

The trick: you buy bonds with maturities longer than your actual investment horizon.

So if I’m only investing for one year, I go grab something like a 5-year bond, hold it for one year, and sell it. That kind of move is what we call a Yield Curve Riding strategy.

Let’s walk it. At $t=0$, I stare at the term structure up there and buy a 5Y coupon bond at par. One year goes by — I’ve collected the coupon, which is my return so far. My bond now has 4 years left on it. But — and here’s the magical assumption — the term structure I was looking at a year ago hasn’t budged. Same shape, same numbers. So my bond, now with 4 years remaining, is going to tell me its expected average annual return going forward is $y_4$. The coupon rate is the same as before, but the discount rate I’m using on it is now $y_4$ → meaning the bond’s price has popped above par → meaning if I sell it, I pocket some capital gain on top of the coupon. → Look at the example down at the very bottom and it’ll click. (Though, to be clear, this is leaning on the absurd premise that the term structure stays exactly the same..)

Going one layer deeper..

Rolling Effect: even if the bond’s interest rate level itself stays put, as the remaining time to maturity shrinks, the yield drops by that much —

this thing where the yield falls and the price climbs as the remaining maturity shortens — that’s the rolling effect.

So in the end, this whole strategy is basically a strategy that rides the Rolling Effect!!!

And it makes total sense that the effect is bigger when you’re rolling down across an interval where the YTM gap between $y_i$ and $y_{i+k}$ is large — bigger drop, bigger price pop. I was also assuming the term structure doesn’t move, but, uh, the degree to which it actually does move is exactly the risk I’m carrying, right? And since longer-maturity bonds have larger duration, the price swings from interest rate moves are more brutal on them — that’s the risk!!

One more thing — the Shoulder Effect keeps showing up as a related term too!!!

Shoulder Effect:

Compared to bonds with a long time still left to maturity,

bonds with a short time left → as they march toward maturity, their price rises by a bigger margin, which means an extreme drop in yield.

That bend on the yield curve is called the shoulder,

so the phenomenon where a short-term bond’s yield nosedives and its price jumps hard as maturity gets close — that’s the Shoulder Effect.

Now that I’ve also dragged the Shoulder Effect into the conversation, things probably got a little muddier, so let me close it cleanly:

does a rolling-down strategy that exploits the Shoulder Effect always spit out higher returns?

→ No no no no — once you factor in the price volatility coming from interest rate moves, which side of the curve to play depends on the situation!!

That’s how I’ll wrap it.

Let me copy down one worked example with actual numbers and then we’re done with this section —

A bond investor with a 5-year investment horizon could buy a bond maturing in 5 years and just collect the 3% coupon — no capital gains, that’s it. But, assuming the yield curve doesn’t shift over the horizon, that same investor could instead buy a 30-year bond for $63.67, hold it for five years, and sell it for $71.81 — earning an extra chunk of return on top of the 3% coupon over the exact same period.

Originally written in Korean on my Naver blog (2024-11). Translated to English for gdpark.blog.