[Macroeconomic Survival Ch. 3] I Extended Duration While Waiting for the Peak in Rates
Once I Could Wait, I No Longer Wanted To
After moving money into SGOV, Korean CD-rate ETFs, and cash management accounts, my portfolio became quiet.
Small moves in interest rates no longer produced the losses I had experienced in long-duration bonds. As the Federal Reserve raised rates, the income from ultra-short instruments rose instead. Time seemed to be working on my side while I decided what to do next.
The calm did not last.
US consumer-price inflation, which had reached 9.1% year over year in June 2022, fell to 3.0% in June 2023. The Federal Reserve raised its policy-rate target to 5.25%–5.50% in July 2023, then held it steady in September.
The Fed continued to say inflation was too high and further tightening remained possible. I focused on a different part of the picture.
Inflation had fallen from its peak, and rate increases had paused.
Was this finally the top?
Earning interest in ultra-short bonds while rates remained high was satisfactory. But if the next stage was rate cuts, SGOV's income would decline. Long-duration bonds, on the other hand, could gain substantially as yields fell.
Cash had given me permission to postpone a decision.
Once conviction returned, waiting began to look less like safety and more like an opportunity cost.
Duration Was the Leverage Inside a Bond
Bond prices generally move in the opposite direction from market yields.
When new bonds offer higher coupons, older lower-coupon bonds must become cheaper to compete. When market yields fall, the relative value of existing bonds rises.
Not all bonds move by the same amount.
Duration is a central measure of how sensitive a bond's price is to changes in yields. It is not simply the number of years to maturity. It also reflects when the investor receives principal and interest.
As a rough approximation, a bond fund with a duration of ten years might lose about 10% if yields rise by one percentage point, or gain about 10% if they fall by one point. Actual results depend on the shape of the yield curve, convexity, and changes in the portfolio, but the basic relationship remains.
Longer duration magnifies both the benefit of falling rates and the damage of rising rates.
In 2022, that sensitivity worked against me. In 2023, I thought I could use the same sensitivity in the other direction.
The duration that had magnified my losses when rates rose could magnify gains when they fell.
My First Choice Was TLT
The first long-duration fund I chose after leaving ultra-short assets was TLT.
TLT invests in US Treasury bonds with more than twenty years remaining to maturity. Because it holds government bonds, it carries little corporate-default risk. Its long average maturity, however, makes it highly sensitive to changes in rates. It pays monthly distributions while maintaining concentrated exposure to long duration.
I saw two reasons to own it.
The first was capital appreciation. If market yields had passed their peak and began to fall, TLT's long duration could amplify the price gain.
The second was equity protection. In a recession, corporate earnings and stock prices might fall while central-bank easing and demand for government bonds pushed Treasury prices higher. Long bonds could then offset part of the equity loss.
The reasoning appeared sound.
But it combined two distinct purposes.
- Insurance intended to soften a stock-market decline.
- A directional position intended to profit from lower rates.
Insurance should be held only in the amount required, including through the uneventful periods when it does not pay off. A directional position invites the investor to buy more as conviction grows.
I told myself I was buying TLT for protection. As the expected date of rate cuts seemed to move closer, I became more interested in its potential upside than in its defensive role.
From TLT to EDV
Once I understood TLT, another question followed.
If rates were going to fall, could I obtain more duration with the same amount of money?
That question led me to Vanguard's EDV.
EDV invests in STRIPS backed by US Treasury securities with remaining maturities of roughly twenty to thirty years. STRIPS separate a Treasury bond's principal and individual coupon payments into securities that can trade on their own. Each piece pays a single cash flow at maturity and therefore behaves like a zero-coupon bond.
One can imagine buying for $30 today the right to receive $100 far in the future. Instead of paying regular coupons, the security earns its return through the gap between the discounted purchase price and the final payment.
Because all cash arrives far in the future, STRIPS generally carry more duration than comparable coupon bonds. Vanguard materials have shown the duration of EDV and its benchmark around twenty-four years at certain dates. TLT's effective duration has generally been closer to fifteen to seventeen years.
In a rough duration calculation, a one-percentage-point move in yields could therefore move EDV's price by more than 20% in the opposite direction. The actual change will not match the approximation exactly, but the point was clear: small changes in yields could produce large gains or losses.
I thought of this as capital efficiency.
If a smaller allocation to EDV could provide the same duration exposure as a larger allocation to TLT, the remaining money could stay in equities. I could retain more exposure to stock-market growth while allowing EDV to react more strongly in a crisis.
It looked like a precise way to obtain more protection with less capital.
It was also a way to create a larger loss with the same capital if my rate forecast was wrong.
How a Zero-Coupon Bond Earns a Return
The term zero coupon initially sounded contradictory.
If the bond paid no interest, where did its return come from? The answer was the difference between its purchase price and the amount paid at maturity.
A STRIP trades below the cash amount it will eventually pay. As time passes and maturity approaches, its price tends to move toward that final value, all else equal. Interest effectively accumulates inside the price rather than arriving as a periodic coupon.
EDV, however, is not the same as buying one STRIP with a fixed maturity and holding it until payment.
To track its index, the ETF sells securities as their remaining maturities shorten beyond the eligible range and buys new STRIPS in the twenty- to thirty-year range. Holding EDV for a long time does not cause the whole fund to mature on a specific date and return a fixed principal. The fund continuously maintains long duration.
That distinction mattered.
An individual Treasury bond eventually reaches a date when the promised principal is paid. An EDV investor remains exposed to the market price at the time of sale. The US government may honor every payment, yet an investor who must sell while yields are high can still realize a large loss.
Credit safety did not guarantee price stability.
It was the same lesson I had encountered with the Korean 30-year fund. This time I did not take the risk unknowingly. I accepted it because I was confident rates would fall.
Convexity Did Not Remove Duration
One feature of EDV that drew my attention was convexity.
Duration approximates the relationship between yields and prices with a straight line. The actual relationship is curved. All else equal, a bond with greater positive convexity can rise slightly more than a simple duration estimate when yields fall and decline slightly less when yields rise.
Long-maturity zero-coupon securities generally have substantial convexity. I viewed that asymmetry as another reason EDV could be an efficient defensive asset.
Here too, I focused first on the part that supported my expectation.
Convexity does not eliminate duration. EDV combines favorable convexity with extremely long duration. A statement that convexity may soften a loss relative to a linear estimate does not mean the absolute loss will be small.
Nor do yields at every maturity move together. Even if the Federal Reserve cuts its policy rate, long-term yields can respond differently because of inflation expectations, fiscal deficits, Treasury supply, and the term premium.
I had not given enough weight to all the steps between “the Fed cuts rates” and “EDV rises substantially.”
The Boundary Between Defense and Return-Seeking Faded
I had returned to long-duration bonds to protect an equity portfolio.
If a recession caused stocks to fall, I hoped long Treasuries would move in the opposite direction and reduce the shock. In that role, bonds should have helped me withstand outcomes I had not predicted, rather than prove that I had predicted them correctly.
My behavior moved elsewhere.
I estimated the timing of cuts, searched for more duration than TLT, and calculated how to capture more price appreciation with less capital. I still spoke about defense, but I was increasingly thinking about the return available from lower rates.
The stated purpose was insurance. The way I held the position was becoming speculative.
The problem was not simply that long bonds were risky. I had also assigned a defensive asset the job of proving my market forecast.
Moving from TLT to EDV is not necessarily wrong. If the required hedge is calculated and the investor can withstand prolonged volatility, EDV may be a useful way to obtain substantial duration with limited capital.
My mistake was to think first about how likely my forecast was to be right, rather than how much risk I could live with if it was wrong.
Rate Hikes Paused, but Long Bonds Did Not Recover
By the second half of 2023, the Federal Reserve had effectively stopped raising its policy rate.
The bond rally I expected did not immediately follow. Inflation had retreated but had not disappeared, and the economy and labor market remained stronger than expected. Long-term market yields rose even after the policy rate stopped moving.
Once again, the problem was timing.
Rate cuts might still arrive, but I did not know how long I would have to withstand EDV's duration before then. I had given up the high income available from ultra-short bills while long-bond prices were still falling. The combination made me impatient.
That was the moment to stop and reconsider the size of the position.
Instead, I added a long-duration inflation-linked bond fund, LTPZ, because inflation still concerned me. I also bought the triple-leveraged Treasury fund TMF, reasoning that a delayed easing cycle might eventually produce a larger rebound.
I had returned to long bonds for portfolio protection.
By adding larger and more complex exposures, the position had moved closer to an attempt to profit from falling rates.
The next chapter examines LTPZ, which I bought to prepare for persistent inflation, and TMF, which magnified my confidence in a rate forecast. It also looks at how the daily compounding of leverage changed what waiting meant.
Sources Consulted
- US Bureau of Labor Statistics: June 2023 Consumer Price Index
- Federal Reserve: September 2023 FOMC Decision
- FINRA: Duration and the Price Impact of Interest-Rate Changes
- iShares: TLT Structure and Effective Duration
- US Treasury: How STRIPS Work
- Vanguard: EDV Structure and Average Duration
- SEC Investor Education: Zero-Coupon Bonds
🌊 [Macroeconomic Survival Series]
- Prologue: The Asset I Trusted Most Was the One I Understood Least
- Chapter 1: I Expected Slower Growth in Korea, but the Market Moved on a Different Clock
- Chapter 2: Only After Retreating to Cash Did I Learn the Value of Waiting
- Chapter 3: I Extended Duration While Waiting for the Peak in RatesCurrent
- Chapter 4: Preparing for Inflation Added More Risk
- Chapter 5: I Tried to Forecast Rates and Currencies Together
- Epilogue: Asking Again What My Investing Is For