The Monetary System, Part 2: Bonds & Yields (2026)

Framework evergreen; figures dated. Data as of August 2026. How a bond works and how a yield is set are durable and read that way here. Every yield is a dated benchmark from the source named beneath its table — re-source it before reuse, because rates move daily. The China and Poland series are approximate and flagged; the euro-crisis spreads are the documented peaks. Inflation (Section 6) is annual World Bank CPI, and the bond total-return figures are indicative and rounded. Informational only, not investment advice; AI-assisted and human-reviewed (see Section 11).

The previous guide in this series ended with a government running a deficit. This one picks up the next question: how does it borrow the money? The answer is the government bond market — the largest, deepest financial market on earth, where sovereigns sell IOUs and the price buyers demand sets the yield, the single most important number in finance. This guide explains what a government bond is, how its yield is set, how to read the yield curve, and how yields have moved across nine big economies since 1995 — a forty-year fall, then a violent snap higher in 2022. It is the free primer; for the company- and project-level data that these macro rates ultimately discount, that is what Metal Pilot is for.

The headline: a bond’s price and its yield move in opposite directions, and almost everything about the bond market follows from that one seesaw.

This is Part 2 of “The Monetary System” — a three-part series on how government money works. The parts: Part 1 · Spending & Debt , Part 2 · Bonds & Yields (this guide), and Part 3 · Interest Rates & Inflation .

TL;DR & Key Takeaways

• Price and yield move opposite ways. When a bond’s price rises, its yield falls, and vice-versa. “Yields rose” and “bonds sold off” are the same sentence.
• A yield is built from four parts: the expected path of the central bank’s policy rate, plus a term premium for lending long, plus compensation for credit and liquidity risk. Most of a safe sovereign’s yield is the first two.
• The yield curve is a signal. A normal (upward) curve is the default; a flat or inverted curve — long rates below short — has historically preceded recessions, though it is a warning, not a trigger.
• Forty years down, then a snap. From the mid-1990s to 2020, 10-year yields fell relentlessly — into negative territory in Germany and France. Then 2022 inflation reversed it: yields spiked to their highest in over a decade.
• Who owns the debt matters. Roughly 30% of US Treasuries are held abroad and ~15% by the Federal Reserve — so foreign demand and central-bank policy both move the market.
• Inflation decides what you keep. After inflation, benchmark government bonds returned close to nothing in real terms in the US and UK across 2000–2024, versus roughly 1.4–2% a year across most of the other seven economies — and 2022 was the worst bond year in generations.

New to the topic? Read straight through from Section 1. Here for the payoff? Jump to the yield history in Section 5, the how-to-read checklist in Section 8, or the traps in Section 7. Every term is defined in the Vocabulary (Section 10).

1. What a bond is, and the price–yield seesaw

A government bond is a loan from an investor to a government, on fixed terms. The government sells the bond for a price, promises to pay a fixed coupon (interest) each year, and repays the face value (the principal) at maturity — 2, 5, 10 or 30 years later. A 10-year bond with a 4% coupon and $1,000 face value (a stylised bond, for illustration) pays $40 a year for ten years, then returns the $1,000.

The coupon is fixed for the life of the bond. The price, however, trades freely in the market every day — and that gap is where the single most important mechanic in fixed income lives: price and yield move in opposite directions.

Why? Because the coupon is fixed in cash. If a bond pays $40 a year and you buy it for $1,000, your yield is 4%. If fear or new issuance pushes the price down to $800, that same $40 coupon is now a 5% yield to whoever buys at $800 — the cash payment didn’t change, but the lower price you paid raises the return. So when prices fall, yields rise; when prices rise, yields fall. “The bond market sold off” and “yields rose” describe the identical event. Every headline about the bond market is really about this seesaw.

One more property follows from it: duration. The longer a bond’s maturity, the more its price swings when yields move, because more years of fixed coupons get repriced. A 30-year bond is far more price-sensitive to a change in yields than a 2-year one — which is why long bonds are where the biggest gains and losses happen, and why the shape of yields across maturities (the next sections) carries so much information.

Throughout this guide, “yield” means the annual yield to maturity on a country’s benchmark 10-year government bond unless stated otherwise, expressed in percent. The 10-year is the market’s reference point — long enough to reflect expectations about the economy, liquid enough to trade constantly.

2. How a yield is set, and how it’s issued

A safe sovereign’s 10-year yield is not one number handed down from on high; it is built up from four components, and separating them is how professionals read the market.

The expected policy rate. The biggest piece. A 10-year bond competes with rolling over short-term deposits for a decade, so its yield tracks where investors think the central bank’s policy rate will average over those ten years. If the market expects the central bank to keep rates near zero, long yields sit low; if it expects hikes, they rise. This is the channel that ties the bond market to central-bank policy — the subject of the next guide in this series.

The term premium. Extra yield investors demand for locking money up for a long time rather than staying short — compensation for the risk that rates, or inflation, surprise higher over the bond’s life. It can be small, and at times even negative, but it is the wedge between “expected average policy rate” and the actual long yield.

Credit risk. The chance the government doesn’t pay in full. For the US, Germany or the UK — issuers of their own currency — this is treated as near-zero, so it adds little. For weaker sovereigns it can dominate, which is exactly what the euro-area crisis exposed (Section 5).

Liquidity. How easily the bond trades. The most liquid benchmark bonds yield slightly less because that convenience is worth paying for; off-the-run and smaller markets pay a touch more.

For a safe sovereign, the first two — expected policy rate and term premium — explain most of the yield. That is why “the 10-year” is read as the market’s collective forecast of growth, inflation and central-bank policy over the decade ahead.

2.1 How governments actually issue

On the supply side sits debt management: the treasury or a dedicated Debt Management Office (the UK’s DMO, the US Treasury) decides how much to borrow, at which maturities, and sells it — overwhelmingly through auctions, where primary dealers (large banks obligated to bid) buy the bonds and distribute them. The manager chooses the maturity mix: issue short and borrowing is cheap but must be rolled over often (refinancing risk); issue long and it locks in a rate for decades but usually pays a higher one. Some managers also run buybacks, repurchasing old bonds to keep the market liquid. These are quiet, technical choices, but they decide how exposed a government is when rates rise.

Auctions run on a published calendar, not on demand. A government doesn’t sell bonds whenever it happens to need cash; it commits to a regular, pre-announced schedule and holds to it, because predictability is itself worth money — dealers who know exactly what is coming bid more confidently, and steadier demand shaves basis points off the cost of borrowing. The cadence follows the maturity. Short-term bills (a year or less) come to market most often: in the US the 4-, 8-, 13-, 17- and 26-week bills are auctioned every week, each on a fixed weekday, with the 52-week bill every four weeks. The medium notes — 2-, 3-, 5- and 7-year — are auctioned monthly. The long end is rarest: the benchmark 10-year note and the 20- and 30-year bonds are issued brand-new only about four times a year (the quarterly “refunding”) and then reopened — the identical bond sold again — in the months between, so a decade of borrowing is spread across dozens of separate sales rather than one.

Each auction sets one price, the same for everyone. US Treasuries — like most modern sovereigns — are sold at a single-price (or “Dutch”) auction. Buyers submit two kinds of bid: a competitive bid names the exact yield the bidder will accept and is mostly the primary dealers’ game, while a non-competitive bid (open to ordinary investors, capped at a small size) simply agrees to take whatever the auction decides. The treasury fills the competitive bids from the lowest yield upward until the whole offering is spoken for; the yield of the last bid it needs — the highest accepted yield, or “stop” — becomes the single clearing yield that every winner pays, aggressive and cautious bidders alike. That one number does the pricing: on a brand-new bond it fixes the coupon (set near the clearing yield, then frozen for the bond’s life); on a reopening, where the coupon is already fixed, it sets the price instead. The bid-to-cover ratio — total bids divided by the amount on offer — is the headline read on how strong the demand was.

So a bond’s terms are set once, at its auction, on that fixed calendar — but its price and yield are re-established continuously thereafter. The auction is the primary market, where new debt is created a few dozen times a year; nearly all the trading is in the far larger secondary market, where the trillions of dollars of bonds already outstanding change hands between dealers, funds and central banks every second the market is open. That secondary market is where “the 10-year yield” quoted in the news actually lives — not last week’s auction result but the live price of the current benchmark bond, moving tick by tick on every data release. In practice the market has usually priced a bond before its auction even happens: it trades on a “when-issued” basis in the days beforehand, so the auction tends to confirm the going rate rather than discover it from scratch. The seesaw of Section 1 runs the entire time — the coupon was fixed at one auction, but what that coupon is worth is repriced every trading second for the next ten years.

2.2 What “the 10-year” actually measures

So which bond does “the 10-year yield” actually measure? Just one: the most recently auctioned 10-year note, now trading on the open market. That newest note is the benchmark (or on-the-run) 10-year, and its yield — set by its price in daily trading — is the number the whole world watches. It holds that role only until the next 10-year auction a few months later, when the fresher note becomes the benchmark and the tracked yield shifts to it. Two things follow. First, the figure is the yield on a bond that is already issued and trading, not a rate fixed at auction: an auction doesn’t set the number, it just sells the new bond at whatever yield the market is demanding that day. Second, because “the 10-year” always means the current newest 10-year, it stays roughly ten years from maturity indefinitely — a constant-maturity rate that reads as one continuous series rather than a single bond counting down. In plain terms, it is the market’s live going rate to lend to the government for ten years. The U.S. Treasury publishes the official daily version as its par yield curve , and data providers quote it continuously through the trading day.

To follow the supply side directly — which maturity is sold when, and on what terms — the issuers publish it themselves. TreasuryDirect lists every upcoming US auction with its date and security type, sets out the standing auction schedule , and the Treasury posts a forward-looking tentative auction schedule each quarter; the UK’s Debt Management Office keeps the equivalent gilt auction calendar . Reading the calendar and the yield together is how the two halves of this section join up: the auctions decide how much new debt arrives and when, and the yield is the price the market puts on it.

3. Who lends, and how the borrower is graded

Every bond the treasury auctions, someone has to buy — and for US Treasuries the ownership split is itself a market signal. Roughly 55% sits with domestic private and public holders, about 30% with foreign investors, and around 15% with the Federal Reserve (Table 1).

Table 1. Estimated ownership of US Treasury securities, December 2024

Source: CRS “Foreign Holdings of Federal Debt” and PGPF , December 2024; shares of publicly held debt (~$28 trillion), approximate.

Figure 1. Three groups fund the US government.

Source: figures as in Table 1.

That ~30% foreign share is why a change in appetite from Japan, the UK or China moves the US market, and the ~15% Fed share is why central-bank bond-buying (the next guide’s subject) matters so directly here.

The foreign column is the one markets watch, because a shift in overseas appetite moves prices independently of the US economy. Ranked by country, the holdings are top-heavy and then splinter (Table 2).

Table 2. Top foreign holders of US Treasury securities (December 2025)

Source: US Treasury TIC “Major Foreign Holders of Treasury Securities,” end-December 2025, in US$ billion (top eight shown; total is all foreign holders). Holdings are attributed to the country of custody, not the nationality of the owner, so financial-centre jurisdictions () — Belgium (home of the Euroclear settlement system), Luxembourg, the Cayman Islands and Ireland — overstate their residents’ true ownership, holding large amounts for third-country investors. Of the ~$9.27tn total, about $3.9tn (~42%) is held by foreign official institutions (central banks and sovereign wealth funds), the rest by private investors.*

Three facts stand out. Japan is comfortably the largest foreign holder, near $1.2 trillion — the legacy of decades of trade surpluses and a yield-starved home market. The United Kingdom has overtaken China for second place (~$860bn against China’s ~$680bn): China has been steadily reducing its Treasury pile for years — to defend its currency and diversify its reserves — while the UK figure is swollen by London’s role as a global custody hub. That hub effect is the standing caveat — Belgium (Euroclear), Luxembourg, the Cayman Islands and Ireland together hold well over $1.6 trillion, much of it belonging to third-country investors routing through those jurisdictions. And only about 42% of foreign holdings now sit with official institutions; the majority is private money, which is more price-sensitive and quicker to leave — one reason foreign demand can swing a Treasury auction.

3.1 How the borrower is graded: credit ratings

The credit-risk component from Section 2 has a formal scoreboard. Three firms — Moody’s, S&P and Fitch, the “big three” agencies — publish a letter grade for almost every government’s debt: an opinion on how likely the issuer is to repay in full and on time. A rating is not a price, a yield or a buy/sell call; it is a relative ranking of default risk, and reading it well means knowing both the ladder and its limits. The three scales run in parallel, with only cosmetic differences between them (Table 3).

Table 3. The long-term credit-rating scale

Source: published long-term rating scales of Moody’s , S&P Global Ratings and Fitch Ratings . The dividing line that matters most runs between Baa3/BBB− and Ba1/BB+ — the boundary between investment grade and speculative (“high-yield” or “junk”). Each agency also appends an outlook (positive / stable / negative) that this table omits.

How to read a rating comes down to five moves:

• Find the investment-grade line. The one boundary that matters is between BBB−/Baa3 and BB+/Ba1: above it is investment grade, below it speculative — “high-yield,” or bluntly, “junk.” It matters because rules, not taste, sit on that line — many pension funds, insurers and index trackers may only hold investment-grade paper, so a downgrade across it can force selling regardless of price (a “fallen angel”). No major sovereign here is near it, but for weaker sovereigns it is the gap between cheap and ruinous borrowing.
• Read the outlook, not just the letter. Each rating carries an outlook (positive / stable / negative), or a formal watch, flagging the likely direction over the next year or two. A “negative outlook” on a top rating often moves the market more than the grade itself.
• Separate local- from foreign-currency ratings. Sovereigns are rated on both, and the local-currency rating is usually equal or higher — a government that borrows in its own currency can always print to repay, so default becomes a choice, not an inability. This is why big own-currency issuers keep high ratings despite heavy debt, and why the euro members — which gave that power up — were so exposed in the crisis of Section 5.
• Treat ratings as lagging, not leading. Agencies tend to move after the market: yields and credit-default-swap spreads had already re-priced US risk before S&P’s 2011 cut, and the euro periphery before its downgrades. Use a rating as a slow, structural gauge — not a timing signal.
• Expect the agencies to disagree. A one- or two-notch split rating is normal (France sits at Moody’s Aa3 but S&P/Fitch A+); the market usually takes the middle or the lower of the three.

The nine economies of this guide now span several rungs of the ladder, and the direction of travel says as much as the level (Table 4).

Table 4. Sovereign credit ratings of the nine economies (2026)

Source: long-term foreign-currency sovereign ratings from Moody’s , S&P Global Ratings and Fitch Ratings , as of 2026. Historical dates are the agencies’ published actions — US: S&P Aug 2011, Fitch Aug 2023, Moody’s May 2025 (its first-ever US downgrade, ending a top rating held since 1917); UK: Moody’s/Fitch 2013, S&P/Fitch post-Brexit 2016, further cuts in 2020; France: S&P from Jan 2012, with S&P and Fitch reaching A+ in autumn 2025 and Moody’s at Aa3; China: S&P and Moody’s to A+/A1 in 2017, Fitch to A in April 2025; Japan: below AAA since 2001–02, single-A since; Canada: Aaa/AAA at Moody’s and S&P, Fitch to AA+ in 2020; Australia: triple-A at all three; Poland: A2 (Moody’s), A− (S&P and Fitch), with negative outlooks in 2025. Ratings change — re-check before reuse.

The through-line mirrors the yield history. Germany and Australia are the only two rated triple-A by all three agencies — Germany the euro bloc’s anchor (and the reason the Bund is the risk-free reference), Australia a rare commodity-backed holdout; Canada sits just below, triple-A at Moody’s and S&P but AA+ at Fitch since 2020. The United States lost its final AAA in May 2025, when Moody’s followed S&P (2011) and Fitch (2023) — yet Treasury yields barely flinched, the clearest sign that for the world’s reserve issuer the market trusts its own judgment over the agencies’. France is the cautionary tale: triple-A as recently as 2011, it was cut to A+ by S&P and Fitch in autumn 2025 as political deadlock and a widening deficit eroded fiscal credibility — a slow-motion echo of the euro-periphery repricing of Section 5. Japan has sat in the single-A band since its early-2000s debt slump despite carrying the rich world’s highest debt ratio — the sharpest proof that an own-currency issuer rarely defaults — while the UK has settled two to three notches below its pre-2013 peak, China has drifted to single-A as growth slowed, and Poland, solidly investment-grade in the A range, drew negative outlooks in 2025 on a widening deficit. None of the nine is anywhere near the junk line — but for France, the US and (on outlook) Poland, the arrow points down.

4. Reading the yield curve

Plot a government’s yield at every maturity — 3 months, 2 years, 10 years, 30 years — and you get the yield curve. Its shape is one of the most-watched signals in macro, because it encodes what the market expects the central bank and the economy to do.

Normal (upward-sloping). Long yields above short. The default shape: investors demand a term premium for lending longer, and expect the economy to grow. A healthy, unremarkable curve.

Flat. Short and long yields converge. It signals that the market expects the central bank to stop raising rates — often a late-cycle sign, when growth is expected to slow.

Inverted (downward-sloping). Short yields above long. Counter-intuitive, and historically the most-cited recession warning: it means the market expects the central bank to be cutting rates in the future (because it expects a slowdown), pulling long yields below today’s short rate. The US curve inverted in 2022 and stayed inverted into 2024 — the deepest and longest inversion in decades.

Table 5. Yield-curve shapes and what they signal

Conceptual framework; not measured data. Curve shape is a signal, not a mechanism — see the traps in Section 7.

The caution that Section 7 develops: an inverted curve is a warning with a lag, not a switch. It has preceded US recessions reliably, but with variable and sometimes long delays, and it can invert for technical reasons that have nothing to do with a downturn.

5. Forty years in one direction, then a snap

The single biggest fact about the modern bond market is a four-decade fall in yields — from the double digits of the early 1980s to, astonishingly, below zero in much of Europe by 2020 — followed by a violent reversal in 2022. Across the nine economies the pattern rhymes, with local variations.

Table 6. Benchmark 10-year government bond yields (annual average)

*Source: OECD long-term (10-year) government bond yields via FRED , annual averages, for the US, UK, France, Germany, Japan, Canada and Australia. China 10-year (CGB) yields are approximate, from the People’s Bank of China /ChinaBond curve; China’s bond market became liquid only in the 2000s (2013 peak ~4.7%, record low ~1.6% in early 2025). Poland’s OECD series begins in 2001, so its 1995 and 2000 cells are blank.

Three episodes carry the story. The long decline (1995–2020) tracked falling inflation, aging demographics and, after 2008, central banks holding policy rates near zero and buying bonds — which pushed German and French 10-year yields below zero, a once-unthinkable state where investors paid to lend. The US anchors that arc.

Figure 2. The US 10-year fell for a generation, then snapped back.

Source: US row of Table 6.

The 2022 snap ended it. As inflation surged (the subject of the next guide), central banks raised policy rates at the fastest pace in forty years, and 10-year yields leapt — the US from ~0.9% in 2020 to above 4% by 2024, Germany from below zero to above 2%. Bondholders took heavy price losses (remember the seesaw), the worst in generations for long bonds. By 2024–25 yields sat at their highest in over a decade almost everywhere.

The standout divergences are China and Japan. China moved the opposite way — as the rest of the world’s yields spiked, its 10-year fell toward ~1.7% by 2025 as growth slowed and the central bank eased. Japan is the other exception: it pinned yields near zero for years under the Bank of Japan’s yield-curve control and only began normalising in 2024–25 (to ~1.6%) as it finally exited — the last major central bank to do so.

At the latest reading the spread is wide: Poland tops the field near ~5.7% and the US, UK and Australia sit near 4%, while Germany, France and Canada run 2.6–3.4% and China and Japan cluster around ~1.6–1.7% — the gap itself the market’s verdict on each economy’s growth and inflation path.

5.1 When credit risk comes back: the euro crisis

The table’s calm 2010–2012 columns for France and Germany hide a violent episode next door. In the euro-area sovereign crisis (2010–2012), the market suddenly re-priced the credit risk of the weaker members. Because they had given up their own currencies, Greece, Italy, Spain and others could not print to repay — so their bonds carried real default risk. Peripheral yields tore away from the German Bund: Italian and Spanish 10-year yields neared 7% in late 2011 while the Bund sat near 1.5%, and Greek yields spiralled into the double digits and beyond. The gap only closed after the ECB president pledged in July 2012 to do “whatever it takes” to hold the euro together. The episode is the clearest real-world proof of the credit-risk component in Section 2: same currency, wildly different yields, because the market judged some governments far likelier to default than others.

6. Inflation and the real return on bonds

Every yield in this guide so far has been nominal — the cash return printed on the bond. But a bondholder eats and pays rent in real goods, while the coupon is fixed in cash, so the number that actually matters is the real return: the nominal yield minus inflation. Inflation is the third character in the bond story — it feeds the term premium of Section 2, it triggered the 2022 snap of Section 5, and, over a lifetime of holding, it is the single biggest determinant of whether a “safe” government bond made its owner richer or quietly poorer.

6.1 How inflation is measured, and who does it

Inflation is not one number but a family of them, and which one you use changes the answer. At its core, every consumer-inflation measure works the same way: a national statistics office defines a basket of goods and services a typical household buys — food, rent, fuel, transport, haircuts, streaming subscriptions — weights each item by how much of the average budget it takes, collects tens of thousands of individual prices every month, and compares the basket’s total cost to a base period. The year-on-year change in that cost is the inflation rate. The weights are revised as spending habits shift, and statisticians adjust for quality changes (a faster laptop at the same price is a hidden price cut) and for shoppers substituting away from whatever has grown expensive — quiet technical choices that still move the headline by a few tenths.

The metrics differ mainly in what they cover and who they are built for. Headline CPI is the all-items consumer price index — the number in the news, and the one in Table 7 below. Core CPI strips out food and energy, whose volatile prices swing the headline from month to month, so central banks watch it for the underlying trend. In the United States the Federal Reserve actually targets a different gauge, the PCE (personal consumption expenditures) price index, which uses a broader, faster-updating basket and usually reads a few tenths below CPI. The euro area uses the HICP, a harmonised index built to identical rules across every member state so the figures are comparable — it is what the ECB’s 2% target refers to, and it runs a little above each country’s older national CPI. The UK long relied on the RPI but now headlines CPI (and CPIH, which folds in owner-occupied housing). Two cousins sit outside the consumer basket: the GDP deflator, the economy-wide price change that turns real growth into the nominal growth used in Part 1, and the PPI (producer prices), measured at the factory gate, which often moves before consumer prices do.

The measuring is done by independent statistical agencies, not by the central banks that target the results. In the US the Bureau of Labor Statistics publishes CPI and PPI while the Bureau of Economic Analysis produces the PCE index; the UK’s Office for National Statistics, Germany’s Destatis and France’s INSEE compile their national numbers, with Eurostat assembling the harmonised HICP; China’s National Bureau of Statistics reports its CPI. The World Bank series in Table 7 simply collates these national CPIs onto one comparable basis — which is why the figures there are standard all-items consumer inflation, and why the euro-area rows sit a touch below the HICP the ECB steers by.

6.2 The inflation record, 2000–2025

For two decades inflation was the dog that didn’t bark. From 2000 to 2020, consumer prices across the rich world rose at roughly 1–3% a year — low, stable and boringly predictable, which is exactly why nominal yields could fall so far without alarming anyone (Section 5). Then the post-pandemic surge broke the calm: inflation leapt to 8.0% in the US and 7.9% in the UK in 2022, 6.9% in Germany, and — most extreme of the nine — 14.4% in Poland, before receding through 2023–24. Two economies barely joined in: Japan, which spent most of the period in outright deflation (negative CPI in a dozen of these years before its 2022–25 pickup), and China, whose inflation dipped to 0.2% in 2023–24 as its economy slowed — the mirror image of both countries’ unusually low yields in Section 5.

Table 7. Consumer-price inflation, annual %, 2000–2025

Source: World Bank “Inflation, consumer prices (annual %)” (series FP.CPI.TOTL.ZG), national consumer-price indices, annual averages, rounded to one decimal. Euro-area figures are the national CPI; the harmonised HICP the ECB targets ran a little higher in 2022–23. The starred China row is shown for completeness but, like its yields, reads best as approximate. The US 2025 value was not yet posted in the World Bank series at the data-as-of date (—). A negative value is a fall in the price level.

Two features carry into the returns below. First, the 2021–23 spike is the whole game: a bondholder who locked in a ~1% yield in 2020 then watched inflation run to 8% in 2022 saw a small positive real return turn sharply negative in a single year. Second, the level differs enormously by economy — average inflation across 2000–2024 ran from about 0.4% in near-deflationary Japan to about 3.6% in Poland, with the US, UK and Australia in the mid-2s, Canada ~2.2%, and France, Germany and China nearer 2% — and that gap is most of what separates their real bond returns.

6.3 What government bonds returned since 2000

Put the two together — the nominal total return a government-bond investor earned (coupons reinvested, plus price gains as yields fell) set against the inflation that ate into it — and the forty-year bull market looks very different in real terms. The nominal figures below are indicative, drawn from benchmark government-bond total-return indices and rounded; the real figures are those returns less each economy’s average inflation from Table 7.

Table 8. What government bonds returned, 2000–2024 (annualized, indicative)

Source: nominal figures are indicative annualized total returns (coupons reinvested) on each economy’s benchmark government bonds, 2000–2024, from the standard total-return index families — the US anchored to NYU Stern (Damodaran) 10-year Treasury total returns, the euro-area and UK to Bloomberg and FTSE Russell government-bond indices, and China (approximate) to the ChinaBond /iBoxx China government series, which becomes meaningful only from the mid-2000s. Real return is the nominal return less the average annual World Bank CPI for 2000–2024 in Table 7. Rounded and indicative — index vintages and maturity mixes differ — so these show magnitude, not a tradeable track record.

Figure 3. After inflation, most of the bond bull market disappears.

Source: values from Table 8; China and Poland approximate.

The story splits into two eras. From 2000 to 2020, falling yields (Section 5) handed bondholders two decades of capital gains on top of their coupons — a genuine bull market, and the years when “bonds are safe” hardened into common sense. Then 2022 rewrote it: as inflation forced the fastest rate hikes in forty years, prices collapsed (the seesaw again), and the benchmark 10-year Treasury lost about 18% in a single year, German bunds around 12%, and UK gilts — hit by both global rates and the September 2022 domestic crisis — roughly a quarter of their value. It was the worst calendar year for government bonds in generations, and it is what drags the 25-year real return down toward nothing for the US and UK.

The punchline is the real column. Over the full 2000–2024 stretch, US and UK government bonds returned close to zero after inflation — a quarter-century of price risk for almost no gain in purchasing power — while the other seven all preserved a modest ~1.4–2% a year in real terms: Germany, France, China, Japan, Poland, Canada and Australia cluster there, whether helped by low inflation (Japan) or by higher starting yields (Poland, Australia). China is the outlier twice over: its bonds actually rose in 2022 (about +3%), because the People’s Bank of China was easing while everyone else hiked, and its low inflation left its real return among the highest of the nine. The lesson is the one the next section makes explicit — a headline yield tells you almost nothing until you subtract the inflation that will be running while you hold the bond.

7. Traps: what yields don’t tell you

Yields are informative, but they are routinely over-read. Four traps recur.

The inversion is a warning, not a trigger. An inverted curve has preceded US recessions, but with long and variable lags — sometimes well over a year — and the economy can keep growing throughout. Treating an inversion as a timing signal, rather than a probability shift, has cost many forecasters dearly.

Nominal is not real. A 4% yield with 3% inflation is a 1% real yield; a 4% yield with 6% inflation is negative in real terms. What a bond actually earns you is the real yield — nominal minus expected inflation. A rising nominal yield can still mean falling real returns if inflation is rising faster. Section 6 shows what this did to real bond returns since 2000: close to nothing in the US and UK.

Rising yields are not automatically “bad.” Yields rising because the economy is strong (healthy growth, normalizing policy) is very different from yields rising because investors fear a government’s solvency or inflation. The same move up can be benign or alarming depending on why — and the level alone doesn’t say.

“Bond vigilantes” are the exception, not the rule. Markets do occasionally punish a government’s fiscal recklessness with a sharp yield spike (the UK’s 2022 gilt episode is the textbook case). But for large, own-currency issuers, yields are driven far more by expected policy rates and inflation than by deficits alone — which is why high-debt countries can still borrow cheaply for years.

8. How to read a government bond yield — a checklist

Put the guide together and a single yield resolves into a short sequence of questions. None of these is a trade signal; together they turn “the 10-year moved” from a headline into something you can actually read. Run them in order.

• Which way did the price move? A yield and a price are the same fact told two ways (Section 1). “Yields rose” means “bonds fell” — so a rising yield is a loss for existing holders, and the loss is bigger the longer the bond’s duration. Start every reading by translating the direction.
• Decompose the yield. A safe sovereign’s yield is expected policy rate + term premium + credit + liquidity (Section 2). For the US, Germany or the UK the first two dominate, so “the 10-year” is mostly the market’s forecast of growth, inflation and central-bank policy for the decade — not a verdict on solvency.
• Read the curve’s shape, not its level. Normal, flat or inverted encodes what the market expects the central bank to do next (Section 4); an inversion is the classic recession warning, but with a long and variable lag (Section 7) — a probability shift, never a countdown.
• Ask whether credit risk is priced. An issuer that borrows in its own currency can always print to repay, so its credit component is near-zero; a euro member gave that up, which is why identical-currency bonds carried wildly different yields in the crisis (Sections 3 and 5). Check which kind of borrower you are looking at before reading a high level as danger.
• Subtract inflation. The real yield — nominal minus expected inflation — is what you actually keep (Sections 6 and 7). A 4% yield is +1% real at 3% inflation and negative at 6%; a rising nominal yield can still be a falling real one. Over 2000–2024 this gap turned two decades of US and UK bond “gains” into roughly zero.
• Ask why yields moved. Rising because growth is strong and policy is normalising is a benign story; rising because investors fear a government’s solvency or inflation is an alarming one (Section 7). The same move up can mean opposite things — the level alone won’t tell you which.

Table 9. Reading a yield: the six-test checklist

Conceptual framework, not investment advice — a way to read a yield, not a rule for trading one.

This is a lens, not a strategy: it tells you what a yield is saying, which is the necessary first step before any decision about what — if anything — to do about it.

9. Summary

A government bond is a fixed-coupon loan whose price trades freely — so its price and yield move in opposite directions, the seesaw that underlies everything else. A safe sovereign’s yield is built from the expected policy rate, a term premium, and small credit and liquidity components; the shape of yields across maturities — the yield curve — encodes what the market expects, with inversion the classic (if lagging) recession warning. Who lends and how the agencies grade the borrower is a signal in its own right: foreign demand and the Fed move the US market independently of the economy, and a downgrade across the investment-grade line can force selling regardless of price. Since the mid-1990s the dominant fact has been a four-decade fall in yields, into negative territory in Germany and France, ended by the violent 2022 snap higher as inflation forced central banks to hike — with China moving the other way as its economy slowed. The euro-area crisis showed what happens when credit risk returns to a market that had priced it at zero. And measured against inflation, that four-decade bull market flatters to deceive: after price rises, US and UK government bonds returned close to nothing in real terms across 2000–2024, versus a modest ~1.5–2% a year in Germany, France and China — proof that the number that matters is the real yield, not the headline one. The six-test checklist in Section 8 puts these pieces in the order a reader can actually apply them. The next guide picks up the force that now dominates all of this — central-bank policy and inflation, which set the expected-policy-rate component at the heart of every yield; for how these rate regimes ripple into real assets, see the macro-regime guide , and for the borrowing that this market finances, the fiscal guide . When you are ready to move from macro rates to the companies they discount, that is what Metal Pilot is built for.

Next in the series → Part 3 · Interest Rates & Inflation — the policy engine that sets the rates.

10. Vocabulary

11. Sources, methodology & disclaimer

11.1 Sources, methodology & data vintage

Benchmark 10-year yields are annual averages of OECD long-term government bond yields, retrieved via FRED (series IRLTLT01…A156N) for the OECD members shown. China’s 10-year (CGB) yields are approximate, drawn from the People’s Bank of China /ChinaBond curve, and are flagged as such because China is not in the OECD series and its bond market became liquid only in the 2000s; Poland’s OECD series begins in 2001, so its 1995 and 2000 cells are blank. US Treasury ownership shares (Table 1) are from CRS and PGPF (December 2024, publicly held debt). The top foreign-holder figures (Table 2) are from US Treasury TIC “Major Foreign Holders of Treasury Securities,” end-December 2025, which attributes holdings to the country of custody rather than the owner’s nationality (so custody-hub jurisdictions and the UK are overstated); about 42% of the ~$9.27tn total sits with foreign official institutions. Credit ratings (Tables 3–4) are the published long-term sovereign ratings of Moody’s , S&P Global Ratings and Fitch Ratings as of 2026, with the downgrade dates being those agencies’ announced rating actions; because ratings change with each action, they should be re-checked against the agencies before reuse. The euro-area crisis spread levels are the widely documented 2011–2012 peaks; see the ECB . Inflation (Table 7) is the World Bank “Inflation, consumer prices (annual %)” series (FP.CPI.TOTL.ZG), national CPI, annual averages; euro-area national CPI runs a little below the harmonised HICP the ECB targets, and the US 2025 value was not yet posted at the data-as-of date. Bond total returns (Table 8, Figure 3) are indicative annualized total returns on benchmark government bonds, 2000–2024, from the standard total-return index families — the US anchored to NYU Stern (Damodaran) 10-year Treasury total returns, the euro-area and UK to Bloomberg/FTSE Russell government indices, Japan, Canada and Australia to their benchmark government-bond indices, and China and Poland (both approximate) to the ChinaBond /iBoxx China and local-currency Polish government series; the real column is nominal less the average World Bank CPI in Table 7, and the single-year 2022 losses cited in Section 6 are from the same US series (≈−18%) and the iShares eb.rexx Government Germany fund (≈−12%). Yields, prices and returns are rounded to one decimal; curve-shape descriptions in Section 4 are a conceptual framework, not measured data. Data as of August 2026; yields move daily and the return figures are indicative, so re-source before reuse.

11.2 Disclaimer & disclosure

This guide is for informational purposes only and is not investment advice, a recommendation, or a solicitation to buy or sell any security, including government bonds. Yields are estimates as of the stated date and change continuously; readers should verify against the primary sources before relying on any number. Nothing here should be taken as a forecast of interest rates or any market outcome. This article was prepared with AI assistance: figures were gathered from the named public sources and human-reviewed, but readers should independently confirm before acting. The author holds no position that the content is designed to promote.