Macroeconomics
Real Interest Rate
Nominal minus inflation — the rate that actually decides whether your savings grow
The real interest rate is the nominal rate adjusted for inflation: r ≈ i − π, exactly (1+i) = (1+r)(1+π). It is the decision-relevant rate behind every savings account, mortgage and bond — and the one that went negative across most of the developed world from 2009 to 2022.
- Exact form(1+i) = (1+r)(1+π)
- Linear approxr ≈ i − π
- Decision-relevantEx-ante: r = i − πe
- US 10y TIPS yield≈ 1.5% (2024)
- Negative real era2009-2022, US 1y to −7%
- Natural rate r*≈ 0.5% late-2010s (LW)
Interactive visualization
Press play, or step through manually. The visualization is yours to drive — try it before reading on.
Watch the 60-second explainer
A condensed visual walkthrough — narrated, captioned, under a minute.
From dollars to goods
If a bank pays you 5 percent on a savings account and prices over the year rise by 3 percent, you do not gain 5 percent in any economically meaningful sense. You gain roughly 2 percent. The real interest rate is the formalisation of that fact: the inflation-adjusted return on a financial contract, measured in goods rather than in dollars. Written exactly,
(1 + i) = (1 + r) × (1 + π)where i is the nominal interest rate quoted on the contract, r is the real rate, and π is the inflation rate over the same horizon. Solving:
r = (1 + i) / (1 + π) − 1Expanding the right-hand side gives the working approximation almost everyone uses informally:
r ≈ i − πThe approximation drops a cross-term equal to r·π. At r = 2 percent and π = 2 percent the error is 0.04 percentage points — utterly negligible. At Argentine-style 80 percent inflation and a 5 percent real rate the error is 4 percentage points — bigger than the real rate itself. The rule of thumb: use linear below 10 percent inflation, multiplicative above.
Ex ante and ex post — the role of expectations
A loan is signed today; inflation realises tomorrow. The real rate splits into two distinct quantities depending on whether π means expected or realised inflation:
| Form | Equation | What it does | What is observable |
|---|---|---|---|
| Ex-ante (decision-relevant) | re = i − πe | Drives saving, borrowing, investment | i; πe only via surveys or TIPS |
| Ex-post (accounting) | r = i − π | Determines actual purchasing-power return | Both i and π directly, after the fact |
| Long-run identity | Δi = Δπe, r constant | "Fisher effect" prediction | Tested by co-integration on multi-decade data |
The ex-ante rate is the one that drives behaviour. A retiree deciding how much to consume looks at the real return she expects on her bond portfolio. A firm deciding to build a factory weighs the expected real cost of capital against the expected real cash flow. The ex-post rate, by contrast, is what gets settled after the fact. The two differ whenever inflation surprises. Surprise inflation hurts lenders (ex-post real return is lower than agreed) and helps borrowers (ex-post real cost is lower than agreed). Surprise disinflation does the opposite — the great 1980s windfall for long-bond holders.
Worked example: TIPS, break-evens and a 30-year mortgage
Suppose on a Tuesday morning the 10-year nominal Treasury yields 4.20 percent and the 10-year TIPS yields 1.80 percent. By inversion of the Fisher identity, the market's pricing of average annual inflation over the next decade is
π^e ≈ i − r = 4.20% − 1.80% = 2.40%This is the break-even inflation rate: the inflation rate at which a marginal investor is indifferent between holding nominals and TIPS. Just above the Fed's 2 percent target.
Now consider a household borrowing $400,000 on a 30-year fixed-rate mortgage at 6.5 percent. The market's πe = 2.4 percent implies a real mortgage cost of
r_mortgage ≈ 6.5% − 2.4% = 4.1%If inflation surprises to the upside — say it averages 4 percent over the life of the mortgage — the household's ex-post real rate falls to 6.5 percent − 4 percent = 2.5 percent. The lender suffers a 1.6-percentage-point per year loss in real return over thirty years. The total real-dollar windfall to the borrower exceeds $100,000 in present-value terms. This is the silent mechanism by which inflation transferred wealth from US bondholders to homeowners over 1965-1980, and again partially in 2021-2022.
The negative-real-rate era, 2009-2022
Across most of the advanced world for over a decade after the 2008 financial crisis, ex-post real interest rates on short-dated government debt were persistently negative. The mechanism is mechanical: nominal policy rates pinned near zero (or below — the ECB, Swiss National Bank, Sveriges Riksbank, Bank of Japan and Danmarks Nationalbank all went below zero at various times) combined with inflation running around 1-2 percent produced ex-post real rates of roughly −1 to −2 percent on short-dated debt for most of the decade.
The 2021-2022 inflation surge widened the gap dramatically. With the Fed holding the federal funds rate near zero through March 2022 while CPI inflation hit 9.1 percent in June 2022, US one-year ex-post real rates fell to roughly −7 percent — the deepest negative real rate since the Carter-era oil shocks. Real yields on TIPS at the 10-year maturity reached approximately −1.2 percent in late 2021 before policy normalisation pushed them back to about 1.5 percent in 2024. The era ended once central banks normalised, but it had already done its work: real debt burdens across the advanced world were eroded by 10-20 percent of GDP, much as in the post-WWII repression episode.
| Period | Avg US 10y nominal yield | Avg US CPI inflation | Avg ex-post 10y real yield | Comment |
|---|---|---|---|---|
| 1970-1979 | 7.5% | 7.4% | 0.1% | The Great Inflation; real rates near zero |
| 1980-1989 | 10.6% | 5.5% | 5.1% | Volcker disinflation; high real rates |
| 1990-1999 | 6.7% | 3.0% | 3.7% | Great Moderation begins |
| 2000-2008 | 4.6% | 2.8% | 1.8% | Pre-crisis decline |
| 2009-2020 | 2.4% | 1.8% | 0.6% | Zero lower bound, deeply low real rates |
| 2021-2022 | 2.6% | 5.4% | −2.8% | Inflation surge, real rates collapse |
| 2023-2024 | 4.1% | 3.3% | 0.8% | Normalisation; 10y TIPS ~1.5% |
The natural rate of interest — r-star
Behind the observed real rate sits a theoretical concept: the natural rate or neutral rate, denoted r*. It is the real rate consistent with output at potential and stable inflation — the rate at which monetary policy is neither stimulating nor restraining the economy. It cannot be directly observed; it has to be estimated from data using state-space filters, DSGE models, or term-structure decompositions.
The most-cited US estimates come from Thomas Laubach and John Williams (now at the New York Fed), who decompose the observed real rate into a trend (r*) and a cyclical component using a Kalman filter. Their results show r* falling steadily from roughly 3 percent in the late 1990s to about 0.5 percent by the late 2010s, with a partial rebound to 1-1.5 percent post-pandemic. Three explanations dominate the literature:
- Demographics. Ageing populations save more (precautionary, life-cycle motives) and invest less (slower workforce growth). Lawrence Summers's secular-stagnation hypothesis (2013) ties the decline to demographic shift in the global North.
- Productivity slowdown. Slower trend productivity growth lowers the marginal product of capital and therefore the real return investors will accept.
- Global saving glut. Ben Bernanke's 2005 lecture: emerging-market reserve accumulation (China, oil exporters) flooded the global market with savings, depressing real yields worldwide.
A lower r* means the zero lower bound binds more often: if the equilibrium real rate is 0.5 percent and inflation expectations sit at 2 percent, the equilibrium nominal rate is 2.5 percent — leaving only 250 basis points of conventional cutting room before hitting zero. Half the post-1960 US recessions saw rate cuts of more than 250 basis points; nearly all required additional unconventional tools (QE, forward guidance) when r* was low.
Where real rates show up — and where nominal ones mislead
| Application | What to use | Why |
|---|---|---|
| Comparing a 1980 savings account to a 2024 one | Real rates | 1980 was 13% nominal but 0% real; 2024 is 5% nominal and 2% real |
| Capital budgeting / NPV | Real if cash flows are in real terms; nominal if nominal | Consistency — don't mix |
| Taylor rule policy prescriptions | Real rate gap from r* | Monetary policy bites through real rate |
| r-vs-g debt sustainability | Real rate r minus real growth g | Determines automatic debt dynamics |
| Pension and insurance liabilities | Real (for CPI-linked) and nominal separately | Match liability character to discount rate |
| Cross-country comparison of monetary stance | Real policy rates | Otherwise high-inflation countries look hawkish |
| Saving decisions in retirement | Real after-tax return | Purchasing power, net of taxes, is what funds consumption |
Three historical episodes that show real rates in action
1. The post-WWII repression, 1945-1980
The US emerged from WWII with debt-to-GDP above 100 percent. The Federal Reserve, under explicit Treasury pressure, capped long-bond yields near 2.5 percent through 1951 while inflation averaged about 5 percent. Even after the 1951 Treasury-Fed Accord nominally restored Fed independence, real rates stayed deeply negative through the inflationary late 1960s and 1970s. Carmen Reinhart and Belen Sbrancia (2011) estimate that financial repression — the combination of below-market interest rate caps, capital controls, and forced bank holdings of sovereign debt — eroded US debt-to-GDP by about 4 percent of GDP per year on average between 1945 and 1980. Over 35 years that compounds to over 80 percent of GDP — far more than any primary surpluses could deliver.
2. The Volcker disinflation, 1979-1982
Paul Volcker took office as Fed Chair in August 1979 with US CPI inflation running 11 percent and the federal funds rate at 11 percent — real rates near zero. By June 1981 the federal funds rate had been pushed to 19 percent while realised CPI inflation was 9 percent, producing ex-post real rates of 10 percent — the highest in modern US history. The aim was to break inflation expectations, which had become entrenched. It succeeded: inflation collapsed from 14 percent in 1980 to 3 percent by 1983. But the cost was a deep recession, unemployment peaking at 10.8 percent in 1982, and a savings-and-loan crisis amplified by the rate move. Long-dated bondholders who had bought 12-15 percent nominal coupons in the late 1970s reaped enormous ex-post real returns over the 1980s as inflation surprises ran below expectations.
3. The 2021-2022 inflation shock
The Fed held the federal funds rate at 0-0.25 percent from March 2020 through March 2022. CPI inflation, which had averaged 1.8 percent in the 2010s, climbed from 1.4 percent in January 2021 to 9.1 percent in June 2022. Ex-post real rates collapsed: the 1-year Treasury real yield touched −7 percent in mid-2022. The Fed then tightened aggressively, lifting the federal funds rate from 0.25 percent to 5.5 percent over 16 months — the steepest pace since Volcker. By late 2024 real rates had returned to mildly positive territory, with the 10-year TIPS at roughly 1.5 percent. Households who locked in 3 percent fixed-rate mortgages in 2020-2021 and then faced 8 percent inflation in 2022 saw their real mortgage rates fall to roughly −5 percent for that year — a massive wealth transfer from lenders to homeowners.
Common pitfalls
- Confusing ex-ante and ex-post. The decision-relevant rate is the one priced in at origination, not the one realised after the fact. Many policy errors come from imputing inflation surprises to deliberate policy.
- Using the linear approximation at high inflation. Above roughly 10 percent inflation the cross-term r·π becomes non-trivial. Argentine and Turkish analyses require the multiplicative form.
- Treating real rates as observable. Only ex-post real rates are; ex-ante requires a model of expectations. TIPS yields are the closest market-observable proxy, but they include a small liquidity premium (lowering them) and an inflation risk premium (raising them).
- Assuming r* is constant. It isn't; it has trended down across decades. Models that assume a fixed neutral rate misread monetary-policy stance.
- Ignoring the term premium in long real yields. A 10-year TIPS yield embeds expected future short real rates plus a term premium for bearing duration risk; only the first component is the underlying real rate.
- Mixing real and nominal in NPV calculations. Discount real cash flows by real rates and nominal by nominal — never mix. A common source of capital-budgeting error.
Frequently asked questions
What is the difference between nominal and real interest rates?
The nominal rate i is the rate printed on the contract: a savings account paying 5 percent, a mortgage at 6.5 percent, a 10-year Treasury yielding 4 percent. The real rate r adjusts for inflation π over the same horizon: it is what those dollars are worth in goods. Exactly (1+i) = (1+r)(1+π); approximately r ≈ i − π. If you earn 5 percent nominal while prices rise 3 percent, your purchasing power grew by roughly 2 percent. The real rate is the economically relevant quantity for every intertemporal decision.
What is the difference between ex-ante and ex-post real rates?
Ex ante means at origination — when the contract is signed, inflation is still in the future and only expected. re = i − πe is the real return the parties think they are agreeing to. Ex post means after the fact — once inflation actually realises, r = i − π is what the lender actually earned. The two differ whenever inflation surprises. The ex-ante rate drives behaviour; the ex-post rate determines who wins and loses (surprise inflation redistributes wealth from lenders to borrowers).
How did real interest rates go negative in the 2010s and early 2020s?
Through a combination of near-zero nominal rates and positive inflation. After the 2008 financial crisis the Federal Reserve held the federal funds rate near zero for seven years; the ECB, Bank of Japan, Swiss National Bank and several Scandinavian central banks took nominal rates explicitly below zero. With inflation running around 2 percent, ex-post real rates on short-dated government debt were negative for most of 2009-2022. The 2021-2022 inflation surge widened the gap dramatically: US 1-year real rates fell to about minus 7 percent in mid-2022 before aggressive Fed tightening pushed them back above zero in 2023.
Why does the real rate matter more than the nominal one for decisions?
Because consumption and investment decisions are about real goods, not dollar balances. A retiree withdrawing 4 percent a year from a savings account paying 5 percent nominal but losing 4 percent to inflation is losing purchasing power. A firm deciding whether to build a factory cares whether the cash flow it earns in goods exceeds the real cost of borrowing. Consumption-Euler equations, capital-budgeting NPVs, the Taylor rule, the natural rate r* — all are written in real terms. The nominal rate is just the contractual surface.
What are TIPS and how do they measure real rates directly?
Treasury Inflation-Protected Securities, issued by the US Treasury since 1997, pay a fixed real coupon on a principal that grows with the CPI. Their yield is therefore a direct read on the market-required real rate at each maturity. As of late 2024 the 10-year TIPS yields roughly 1.5 percent — that is the market's required real return for lending the US government money for a decade. Subtracting from the 10-year nominal Treasury yield gives the break-even inflation rate. Other countries have similar instruments: UK index-linked gilts, French OATi, German Bunds inflation-linked, Japanese JGBi.
What is the natural rate of interest, or r-star?
r* is the real interest rate consistent with output at potential and stable inflation — the rate at which monetary policy is neither stimulative nor contractionary. It is a theoretical quantity that has to be estimated from observed data using filters and DSGE models. The most-cited estimates, by Laubach and Williams at the New York Fed, show r* falling from roughly 3 percent in the 1990s to about 0.5 percent by the late 2010s, with a partial rebound in the post-pandemic period. A lower r* makes the zero lower bound bind more often — which is why monetary policy has become so much harder.
How does the real rate connect to the Fisher equation?
The Fisher equation is the identity (1+i) = (1+r)(1+πe). Solving for r gives the exact real rate: r = (1+i)/(1+πe) − 1. The linear approximation r ≈ i − πe drops the cross-term r·πe; the error is r·πe itself, which is tiny at low inflation and substantial at high inflation. Fisher's broader 'effect' is the empirical claim that a one-point rise in expected inflation raises nominal rates one-for-one, leaving the real rate unchanged in the long run. It holds approximately at horizons of a decade or more.
What is financial repression and how do negative real rates produce it?
Financial repression is the deliberate suppression of real interest rates below the level that would clear free markets — typically by holding nominal policy rates below the inflation rate while constraining banks to hold sovereign debt. The mechanism transfers wealth from savers and creditors to debtors, especially governments. The classic post-WWII example: the US held nominal rates capped near 2 percent during the late 1940s while inflation ran 5-15 percent, eroding the WWII debt-to-GDP ratio from over 100 percent to below 50 percent without significant primary surpluses. The 2010s saw a milder repetition.