L3VLUP
ProSign in
Rates Lab

The curve as it is. Then pull the levers.

Central banks set the front end. Deficits and QE fight over the long end. The shape that falls out of that argument decides what every asset on earth is worth. Start with the real US Treasury curve, then break a model of it until the mechanism sticks.

Below the simulator: five dated moments from Treasury’s own data, how a move in the curve reaches valuations, banks, credit, buyouts, mortgages, currencies and gold, and the questions interviewers ask about all of it.

The US Treasury curve

Par yields at the close of 9 October 2026

The real curve first, so the simulator below has something to argue with. The comparison dates are the published closes nearest one month and one year before the latest.

3m10y

+99bp

10-year minus 3-month

2s10s

+44bp

10-year minus 2-year

5s30s

+58bp

30-year minus 5-year

3.50%4.00%4.50%5.00%5.50%6.00%3M2Y5Y10Y30Y
Latest, 9 October 2026A month earlier, 9 September 2026A year earlier, 9 October 2025Tenors evenly spaced, not to scale.
Yields by tenor
TenorLatest9 Oct 2026A month earlier9 Sept 2026A year earlier9 Oct 2025
3M4.25%3.95%4.03%
2Y4.80%4.43%3.60%
5Y5.02%4.61%3.74%
10Y5.24%4.83%4.14%
30Y5.60%5.28%4.72%
Spreads, basis points
Spread9 Oct 20269 Sept 20269 Oct 2025
3m10y+99bp+88bp+11bp
2s10s+44bp+40bp+54bp
5s30s+58bp+67bp+98bp

Source: U.S. Department of the Treasury, Daily Treasury Par Yield Curve Rates, data to 9 October 2026. Treasury’s table. Spreads are the longer tenor minus the shorter, from Treasury’s two-decimal yields.

The curve you just built

Yield by maturity. The faint line is where you started, so you can see the shape change.

-0.4%1.0%2.4%3.9%5.3%3M2Y5Y10Y30YYoursStart

10-year

4.24%

2s10s

+41bp

Real 10-year

1.74%

Starting curve

A normal upward-sloping curve: you get paid more to lend for longer. Move a lever and watch what happens to the shape.

The levers

3.5 %

What the central bank sets. Owns the front end. Zero is ZIRP; 20 is Volcker in 1981.

0 tn

Central-bank net purchases. Positive is QE, negative is QT. The Fed bought roughly $4.6tn in 2020-21.

6 % GDP

Government borrowing as a share of GDP. More paper, more term premium. 2020 ran near 15%.

2.5 %

What the market thinks inflation averages from here. Negative is deflation; 15 is the 1970s.

2 %

Drives the neutral real rate the front end converges to. Deep recession to China-in-2005.

Share your result 📈

𝕏in💬🤖

Instagram and TikTok have no desktop share link, so copy the caption and paste it into the app.

Who wants which asset now

Rates are a relative-price machine. Move them and you have moved the demand for everything else.

Long-duration bonds (30Y)

Little changed

Long yields down, so prices up. The longer the bond, the bigger the move.

Equities (long-duration growth)

Little changed

Growth stocks are long-duration assets. Their value sits in distant cash flows, so the 10-year discount rate hits them hardest.

Credit (IG spreads)

Little changed

Central-bank buying crowds private money out of governments and into credit, which tightens spreads.

Gold and real assets

Little changed

Gold trades off real yields, not nominal. Your real 10-year here is 1.74%.

Name that curve move

0/0 correct

Four shapes, and interviewers ask for them by name. Read the scenario, pick the move.

The Treasury announces a large increase in long-dated issuance to fund a widening deficit. The central bank holds the policy rate flat.

This is a teaching model, not a pricing model. Yields here are the expected average short rate over each tenor plus a term premium that responds to issuance, central-bank purchases and inflation uncertainty. Real curves also carry convexity, swap spreads, foreign demand and the small matter of what everyone else is positioned for. For where rates actually are today, see the Macro Chartbook.

Expected path or term premium?

Every long yield can be split into two parts. The first is the expected path: the average short rate investors expect over the life of the bond. The second is the term premium: the extra yield they demand for locking money up for that long instead of rolling over bills, as compensation for not knowing what inflation and policy will do.

When the path moves

Data that changes the expected run of policy, a jobs report or an inflation print, moves the front end most. The two-year is almost all path. A rally led by the two-year is usually the market pricing cuts.

When the premium moves

Supply, QE or QT, and uncertainty about inflation move the long end without changing expected policy much. A ten- or thirty-year sell-off with a quiet two-year usually points here.

Neither part is published. Both are estimated with models, and the estimates differ, so treat any single term-premium number as one model’s view. In an interview the useful move is to name which part you think drove the change, and why: “the two-year barely moved, so this was premium, not policy” is an answer; “yields went up” is not. The simulator above uses the same split: the policy, inflation and growth levers move the path, while issuance, QE and inflation uncertainty move the premium.

What moves the curve

Each driver below is a lever in the simulator. Read the mechanism, then try it.

Policy rate

Policy owns the front end

Three-month bills trade almost on top of the policy rate, and the two-year is close to the average policy rate the market expects over the next two years. A hiking cycle lifts the front end first and furthest, which is how curves flatten and then invert.

In the lab: Raise the policy rate on its own and watch the 3M and 2Y climb while the 30Y barely moves: a bear flattener.

Net issuance

Issuance is a supply shock to duration

Deficits are funded by selling Treasuries, and the mix matters as much as the size. Bills add little duration; ten- and thirty-year bonds add a lot, and investors want a higher term premium to absorb it. The Treasury announces the mix each quarter at its refunding, and long-end supply surprises tend to steepen the curve.

In the lab: Push issuance up with policy flat: the long end rises and the front end stays put, a bear steepener.

QE / QT

QE and QT move the term premium

When the central bank buys long bonds it removes duration from private portfolios, so investors accept a lower term premium for what is left: the portfolio-balance channel. Quantitative tightening runs the same channel in reverse, more slowly, by letting holdings mature without replacement so the private sector has to absorb more supply.

In the lab: Add QE and the long end falls with the front end pinned: a bull flattener. Switch to QT and it reverses.

Expected inflation

Inflation lifts the whole curve, and the long end twice

Higher expected inflation raises the path of policy rates the market expects, which lifts every tenor. It also raises the compensation investors want for the risk that inflation surprises, and that risk compounds with maturity, so the long end carries an extra premium.

In the lab: Raise expected inflation: the curve shifts up and the real 10-year readout shows how much of the move is inflation.

Trend growth

Growth sets where rates settle

Stronger trend growth raises the neutral real rate, the level policy drifts back to once the cycle passes, and that is what long yields price. A growth scare does the opposite: long yields fall as the market prices a lower destination and earlier cuts.

In the lab: Cut trend growth and the long end falls toward a lower neutral rate while the front end waits for the central bank.

Five moments the curve teaches from

Each is a single close in Treasury’s daily par yield data, with the dated figures that frame it. Spreads are in basis points; negative means inverted.

The 2006-07 inversion

Close of 15 November 2006

3M
5.09%
2Y
4.80%
5Y
4.62%
10Y
4.61%
30Y
4.69%
3m10y -48bp2s10s -19bp5s30s +7bp

The Federal Reserve stopped raising rates in mid-2006 and then held them, so the front end sat high and still while ten-year yields drifted lower. Bills paid more than ten-year notes for months on end. Every September 2007 close in Treasury's file shows a positive 2s10s again, the month the Fed began cutting, and the NBER dates the recession that followed from December 2007.

  • First negative 2s10s close of 2006, 31 January 2006-1bp
  • Deepest 2s10s close of 2006-07, 15 November 2006-19bp
  • Deepest 3m10y close of 2006-07, 27 February 2007-64bp

Treasury's 2006 file has no 30-year yield before 9 February 2006, so the January 2006 closes have no 30-year yield or 5s30s to show.

Source: U.S. Department of the Treasury, daily par yield curve rates for 2006 and 2007.

August 2019: 2s10s goes negative

Close of 27 August 2019

3M
1.98%
2Y
1.53%
5Y
1.40%
10Y
1.49%
30Y
1.97%
3m10y -49bp2s10s -4bp5s30s +57bp

The 3m10y had already inverted in the spring, and the Fed made its first rate cut since 2008 at the end of July. Through August, ten-year yields fell faster than two-year yields, a bull flattener, until the 2s10s closed below zero in Treasury's end-of-day series. A recession did follow, from February 2020, but it was set off by the pandemic, which is why this inversion is argued over as evidence.

  • First negative 3m10y close of 2019, 22 March 2019-2bp
  • 2s10s at the close on 14 August 2019, 14 August 2019+1bp
  • First negative 2s10s close of 2019, 27 August 2019-4bp

The 2s10s inversion widely reported in mid-August happened during the trading day. Treasury publishes end-of-day par yields, and in that series the first negative close comes later in the month.

Source: U.S. Department of the Treasury, daily par yield curve rates for 2019.

March 2020: the dash for cash

Close of 9 March 2020

3M
0.33%
2Y
0.38%
5Y
0.46%
10Y
0.54%
30Y
0.99%
3m10y +21bp2s10s +16bp5s30s +53bp

As the pandemic spread, the Fed cut rates in an emergency move on 3 March and took them to near zero on 15 March, alongside large purchases of Treasuries and mortgage bonds. Bill yields collapsed toward zero. The ten-year did not move in a straight line: after its low on this date it sold off sharply for over a week as investors sold even safe assets to raise cash, before the Fed's buying steadied the market. The curve ended the month steeper, with the front end anchored at zero.

  • Lowest 10-year close of March 2020, 9 March 20200.54%
  • Highest 10-year close in the rest of March, 18 March 20201.18%
  • Steepest 3m10y close of March 2020, 18 March 2020+116bp
  • First 3-month close at zero, 25 March 20200.00%
  • Lowest 10-year close of 2020, 4 August 20200.52%

Source: U.S. Department of the Treasury, daily par yield curve rates for 2020.

July 2023: the deepest 2s10s inversion

Close of 3 July 2023

3M
5.44%
2Y
4.94%
5Y
4.19%
10Y
3.86%
30Y
3.87%
3m10y -158bp2s10s -108bp5s30s -32bp

The Fed raised rates by 4.25 percentage points in 2022 alone and kept going into 2023, making its last increase of the cycle in late July 2023. Two-year yields priced policy staying high; ten-year yields priced it coming down eventually, so the gap went deeply negative. The same shape squeezed lenders that fund short and lend long, and Silicon Valley Bank failed that March after losses on long-dated securities.

  • Start of the unbroken run of negative 2s10s closes, 6 July 2022-4bp
  • Deepest 3m10y close of 2023, 4 May 2023-189bp
  • Deepest 2s10s close of 2023, 3 July 2023-108bp
  • Run ends: first 2s10s close at or above zero, 27 August 20240bp
  • Business days of negative 2s10s closes in the run537

The day count is the number of business days in Treasury's files from the first close of the run up to the first close at or above zero.

Source: U.S. Department of the Treasury, daily par yield curve rates for 2022, 2023 and 2024.

September 2024: the dis-inversion

Close of 6 September 2024

3M
5.13%
2Y
3.66%
5Y
3.50%
10Y
3.72%
30Y
4.03%
3m10y -141bp2s10s +6bp5s30s +53bp

Through the summer of 2024 the two-year yield fell faster than the ten-year as markets priced rate cuts, a bull steepener. The 2s10s turned positive in late August and stayed positive from this date to the end of the year; the Fed began cutting on 18 September. The 3m10y, tied to the policy rate itself, stayed inverted until December. In 2007 the recession began after the curve had already dis-inverted, which is why this moment drew attention rather than relief.

  • First positive 2s10s close since the 2022 run, 28 August 2024+1bp
  • First close of the positive run that held through 2024, 6 September 2024+6bp
  • 3-month yield the day the Fed began cutting, 18 September 20244.84%, from 4.95%
  • First 3m10y close at or above zero, 13 December 2024+6bp

Source: U.S. Department of the Treasury, daily par yield curve rates for 2024.

How a curve move reaches everything else

Treasuries are the base rate the rest of finance is priced over. Move them and the effect travels.

DCF valuations

The risk-free rate in CAPM is usually the ten-year Treasury yield, and the cost of debt is priced off Treasuries plus a spread, so a higher ten-year raises WACC on both legs. The terminal value takes the biggest hit because its cash flows are furthest out, which is why long-duration growth companies reprice most. Ask whether the rate move also changes the cash flows before you mark only the discount rate.

Banks and net interest margin

Banks fund short, with deposits and wholesale borrowing, and lend or hold securities longer. A steeper curve usually widens net interest margin; an inverted one squeezes it, because funding costs follow the front end up while longer assets were locked in at older, lower yields. Rising long yields also cut the market value of fixed-rate securities, which is how duration risk became a bank-run story in 2023.

Credit spreads

A corporate bond yields the Treasury of matching maturity plus a credit spread, so the all-in cost of borrowing moves with both. Investment-grade yields move mostly with Treasuries; high-yield yields move mostly with the spread. A curve pricing recession tends to come with wider spreads, because default risk rises when growth falls.

LBO debt costs

Leveraged loans pay a floating rate, SOFR plus a margin, and SOFR tracks the policy rate, so the front end sets the cash interest bill of most buyouts. Higher base rates shrink how much debt a business can carry at a given interest cover, which lowers leverage, raises the equity cheque and compresses returns. Fixed-rate high-yield bonds price off the Treasury of the same maturity instead.

Housing and mortgages

A US thirty-year fixed mortgage is usually refinanced or repaid long before thirty years, so lenders price it against the ten-year Treasury rather than the thirty, plus a mortgage spread. When the ten-year rises, monthly payments on new loans rise with it, affordability falls, and owners holding older low-rate mortgages are reluctant to move.

FX

Currencies respond to rate differentials, and the front end matters most because it reflects expected policy. Higher US yields relative to other countries tend to support the dollar, all else equal. The exception is a rise driven by fiscal or risk-premium worries, which can weaken a currency even as its yields climb.

Gold

Gold pays no coupon, so its opportunity cost is the real yield: the yield on inflation-protected Treasuries. Higher real yields have usually weighed on gold and lower ones supported it. The link is loose, and it weakens when other buyers dominate, central banks among them.

Interview questions, with model answers

Say your answer out loud before you open the model one. Then compare structure, not wording.

The curve is inverted. What is the market telling you, and does inversion cause recessions?

It says the market expects short rates to be lower in future than they are today, which usually means policy is seen as restrictive and cuts are expected. Inversion has preceded every US recession since the 1970s, usually by many months and sometimes by around two years. It is a signal about expectations, not a mechanism by itself, although it does feed back: an inverted curve squeezes bank margins, which can tighten lending.

Define a bull steepener and a bear steepener, and give an example of each.

Bull and bear refer to prices, so a bull move means yields are falling and a bear move means they are rising. A bull steepener has short yields falling more than long ones, typical of the start of an easing cycle: the 2024 dis-inversion is one. A bear steepener has long yields rising more than short ones, usually because of supply, inflation risk or a higher term premium while policy stays put.

The Fed cuts by 25 basis points. What happens to the ten-year yield?

It depends on what was priced. If the market fully expected the cut, the ten-year may barely move, because it already reflected the expected path. If the cut is a surprise, the ten-year usually falls, but by less than the front end. It can even rise if investors read the cut as raising future inflation risk, which pushes up the term premium.

What is the term premium, and why can you not just look it up?

It is the extra yield investors demand for holding a long bond instead of rolling short bills over the same period. A long yield is the expected average short rate plus the term premium, but neither component is observed: both have to be estimated with a model. Estimates such as the New York Fed’s ACM model and the Fed Board’s Kim-Wright model can differ materially.

Why does a 100 basis point rise in rates hurt a high-growth company more than a mature one?

Duration. Most of a growth company’s value sits in cash flows many years out, often in the terminal value, and the present value of a distant cash flow falls much more than a near one for the same rise in the discount rate. A mature company returns more of its value sooner, so its valuation is less sensitive.

Why do some economists prefer the 3m10y spread to 2s10s as a recession signal?

The three-month bill tracks the policy rate today, so 3m10y compares where policy is now with where the market expects rates to settle. The two-year already embeds two years of expected moves, so 2s10s can turn before policy has been tight for long. The New York Fed’s recession-probability model uses the 10-year minus 3-month spread for this reason.

Walk me through what a higher ten-year yield does to a DCF.

The risk-free rate rises, so the cost of equity rises through CAPM, and the pre-tax cost of debt rises too, so WACC goes up. Every discounted cash flow is worth less, and the terminal value falls most because it sits furthest out. Then check whether the cause of the move, stronger growth or higher inflation, should also change revenue and cost assumptions, so the valuation stays consistent.

What does quantitative tightening do to the curve?

It lets the central bank’s bond holdings run off, so more duration has to be held by private investors, who want a higher term premium to hold it. The effect falls mostly on the long end and builds gradually. It also drains bank reserves, which is why central banks slow or stop it before reserves become scarce.

Take it further

Valuation

Debt and credit

Markets and interviews

Rates sit underneath most of the labs and the models library.

Research behind the curve

You can explain the curve. Can you defend the call?

A discovery session finds the gap between knowing the mechanism and holding a view under pressure. The bootcamps close it.

Book time with Suro Explore the bootcamps

The curve, weekly

One email a week on what the curve did, what the crowd is pricing next, and how to talk about it under questioning.

One email a week, free, and you can leave at any point.