A few weeks ago I actually sat down and read the fund list inside my pension savings account instead of just glancing at the total. Most of it made sense right away — an S&P 500 fund, a Nasdaq 100 fund, something tracking a basket of AI and power infrastructure names I added because it looked interesting at the time. Then there was a line I’ve scrolled past for two years without really asking about it: a fund holding U.S. 30-year government bonds.
I know what a bond is, in the textbook sense. I did not know why a 40-something engineer who’s supposedly investing for growth is holding one, or what actually makes its price move. So I did the thing I keep telling readers of this blog to do with anything they don’t understand: I looked it up properly instead of assuming it was fine because someone (in this case, me, two years ago) must have had a reason.
This week gave me a perfect, live example to learn from. On July 29, the Federal Reserve held its policy rate steady at 3.50–3.75% for a fifth straight meeting. The vote was 9–3, and — unusually — all three dissenters wanted to raise rates, not cut them. That’s the first time three members have dissented in the same direction since September 2016. You’d think a “steady as she goes” decision from the Fed would mean a quiet week for bonds. It didn’t. Longer-term Treasury yields have stayed elevated regardless, sitting in the high-4% range in early August, well above where they started the summer. The short-term rate stood still. The long end of the bond market did its own thing anyway.
That gap between “the Fed held” and “long-term bonds still moved” is exactly the mechanism I’d never bothered to learn.
What a bond price actually is
A bond is a loan with a fixed coupon. Say a government bond pays 4% a year, forever, until it matures. Once it’s issued, that 4% is locked in — it will never pay more, never pay less.
Now imagine new bonds start being issued paying 5%, because interest rates rose. Nobody is going to pay full price for your old 4% bond anymore, when a brand-new bond pays more for the same risk. So the price of your old bond has to fall — enough that its fixed coupon, divided by the lower price you’d pay for it today, works out to roughly the same 5% return a buyer could get elsewhere.
That’s the entire mechanism. Bond prices and interest rates move in opposite directions because a bond’s coupon is frozen at issuance, and the only thing that can adjust to match new market rates is the price. It’s not a mysterious market force — it’s just competition between old fixed payments and new ones.
The part I hadn’t appreciated is how much this effect is amplified for long-term bonds specifically.
Duration: the “how much,” not just the “which direction”
The tool that measures this sensitivity is called duration, expressed in years. As a rough rule of thumb: percentage price change ≈ duration × the change in yield, with the sign flipped (yields up, price down).
A 30-year bond has a lot more duration than a 2-year bond, because you’re locked into that old coupon for a lot longer before you get your money back. Small rate moves get magnified over a longer runway.
To make this concrete, I’ll borrow a number rather than guess one. I don’t have a verified duration figure for the specific fund sitting in my own account, so instead I’m using the published effective duration of TLT, a widely tracked U.S. long-term Treasury ETF that holds a comparable slice of the market (20+ year Treasuries) — about 15 years (15.31, per iShares’ own fact sheet). If you own a similar fund, look up its actual number rather than borrowing mine; funds differ.
Using duration ≈ 15 years as our stand-in:
- If yields rise by 1 percentage point: price falls by roughly 15% (15 × 1%).
- If yields fall by 0.5 percentage point: price rises by roughly 7.5% (15 × 0.5%).
That’s pure arithmetic, not a forecast — a rough approximation, not an exact price. Real bond math also has a smaller second-order effect called convexity that softens some of the sharpest moves, but duration alone gets you close enough to see the shape of the risk.
Here’s the real-world check on that shape: in 2022, when long-term U.S. yields rose sharply, TLT’s full-year total return came in at -31.41%. That’s a fund holding U.S. government debt — as safe an issuer as exists — falling by nearly a third in one year, purely because of how far and how fast rates moved. No credit risk, no company going bankrupt. Just duration doing exactly what the arithmetic above predicts it can do, at scale.
That number reframed the whole question for me. I’d been thinking of the bond fund in my account as the “safe” line in the portfolio. Duration says: safe from what, exactly? Safe from a company collapsing, yes. Safe from a bad year, no.
Why I hold one anyway
So why keep it? Not because I’m betting on where rates go next — I have no edge in predicting that, and neither does almost anyone else who tries. I hold it because of what it does next to the rest of the portfolio, not what it does by itself.
Stocks and long bonds don’t always move together. Sometimes they move in the same direction (when inflation is the shared worry), sometimes in opposite directions (when growth is the shared worry) — I wrote about that flip in an earlier piece on how stocks, rates, and currency connect. A bond fund with real duration is one of the few tools that can zig when equities zag, at least under some conditions. It’s a diversification ingredient, not a rate bet.
In my own numbers — and to be clear, this is structure and weighting, not account balances — government bonds sit at roughly 18% of my total invested mix across my retirement accounts, next to a larger core in S&P 500 and dividend-focused funds. My retirement setup is spread across four account types: two pension savings accounts, one ISA, and one workplace-style individual retirement pension (IRP). The bond fund lives in one of the pension savings accounts, sitting alongside U.S. equity index funds and a couple of thematic positions. It was never meant to be the return engine. It’s ballast.
Korea vs. the U.S.: how each system defaults you into bonds (or doesn’t)
This is where the comparison gets genuinely useful, because the two systems nudge people toward completely different starting points.
In the United States, the dominant default in a 401(k) or IRA is the target-date fund — a single fund that automatically blends stocks and bonds, and gradually shifts more of the mix into bonds as your target retirement year approaches. It’s called a glide path. Most American savers who never touch their allocation settings still end up owning some duration risk by their 40s and 50s, because the fund did it for them, quietly, on a schedule. I’ve written before about how that compares to the 401(k) side of the U.S. system.
Korea’s system defaults people somewhere very different. According to the Ministry of Employment and Labor’s 2025 retirement pension investment white paper, roughly 75% of all Korean retirement pension (퇴직연금) assets — about 378 trillion won of the 501 trillion won total — are still sitting in guaranteed-principal products, instruments that behave more like a savings deposit than a bond fund, with no duration exposure and no price swings, but also no real growth engine. The market-linked, performance-based share has grown to about a quarter of the total (24.6%), roughly doubling over three years, but guaranteed-principal products still dominate.
The personal pension savings account I actually use (연금저축) works differently from the broader 퇴직연금 statistic above — it’s self-directed, so I pick every fund myself, bond fund included. But the broader system-level default in Korea leans hard toward “guaranteed and flat,” while the broader system-level default in the U.S. leans toward “blended and gradually more conservative on a schedule.” Neither default is wrong. They’re just different answers to the same question: who decides how much interest rate risk you carry, and when?
| System | Default bond exposure |
|---|---|
| U.S. 401(k)/IRA | Auto glide path |
| Korea, workplace pension (퇴직연금) | Mostly guaranteed |
| Korea, personal pension (연금저축) | Investor’s choice |
The honest reading: most American savers get duration exposure by default, without deciding to. Most Korean retirement-pension savers get very little, also without deciding to. My bond fund isn’t the system doing its job automatically — it’s me doing manually what a U.S. target-date fund would have done for me on autopilot.
Where this goes wrong for me
I try to write one section like this in every post, because the reservoir of “this always works” content on the internet is already full, and it isn’t honest.
The biggest risk is exactly the number above: -31.41% in a single year is not a rounding error. If I needed that money on a fixed near-term date — say, a down payment due in eight months — a long-duration bond fund would be a genuinely bad place to park it. A fund has no maturity date; unlike holding one bond to term, you can’t just wait it out and get your principal back on a known day. The fund’s price is whatever the market says it is, on whatever day you need to sell.
There’s also a timing trap that looks like discipline but isn’t: buying a large lump sum of long bonds right now specifically because you expect the Fed to cut rates later this year would be a bet on the direction and timing of a very hard-to-predict variable — the same mistake as trying to time stocks around headlines, just wearing a more conservative-looking costume. Rate cut expectations are already partly priced into today’s bond prices; if they don’t arrive on schedule, or arrive smaller than expected, the “safe” trade can lose money too.
And duration cuts both ways with inflation. If inflation runs hotter than expected for a stretch, the fixed coupon on a long bond buys less in real terms every year it’s held, on top of any price move from rates. Long-duration bonds are not a hedge against every bad outcome — they’re a specific tool for a specific job, with a specific cost when the wind blows the wrong way. My own DCA routine, which I’ve written about in more detail here, is what keeps any single tool like this from becoming an all-or-nothing bet.
FAQ
Why do bond prices fall when interest rates rise?
Because a bond’s coupon is fixed at issuance. When new bonds start paying more, an old bond with a lower fixed coupon is only attractive at a lower price — one that brings its effective yield roughly in line with what’s newly available.
Is a bond fund the same thing as owning a bond?
No. A single bond held to maturity returns your principal on a known date, assuming the issuer doesn’t default. A bond fund never matures — it constantly buys and sells bonds to maintain its target duration, so its price can be permanently below what you paid, with no maturity date to wait out.
How much of a retirement account should be in bonds?
There’s no universal number. U.S. target-date funds typically scale bond exposure up with age using a glide path; a common rough guideline some advisors cite is roughly matching bond percentage to age, though that’s a starting heuristic, not a rule. My own bonds sit around 18% of the total mix — a level I chose for diversification, not because it matches any formula.
Do bond funds ever mature like a real bond?
No — that’s the core structural difference. A fund like the one in my account continuously rolls its holdings to stay near a target maturity range (in my fund’s case, long-dated Treasuries), so it never “matures” the way a single bond does. Its average duration stays roughly constant over time.
What’s the difference between duration and maturity?
Maturity is simply how far away the bond’s final payment date is. Duration is a more useful number for risk: it accounts for the timing of all the cash flows (coupons plus principal), and it’s what actually predicts how much the price moves for a given change in interest rates.
Back to the actual plan
None of this changes what I do on payday. I still buy the same funds, in the same proportions, on the same day of the month, regardless of what the Fed did or what long-term yields are doing this week. The bond fund is a small, deliberate piece of that routine — not a trade I’m making because I think I know where rates go next, because I don’t, and neither does the person selling you a confident answer.
This is not investment advice.