The Confession

The yield curve — the simplest chart in finance, just a line connecting yields across maturities — is currently shaped like a confession written under duress. It says two contradictory things simultaneously, and both are true. This report is about learning to read the contradiction rather than resolving it.

Read this curve from left to right and it tells two stories. Story One (1-month to 2-year): the front end is inverted. The 3-month T-bill at 4.33% yields more than the 2-year note at 3.56%. The market expects the Fed to cut — it just doesn't know when. Story Two (2-year to 30-year): the back end is bear steepening. The 30-year at 4.87% towers 131 basis points above the 2-year. Long-term investors demand compensation for risks the front end pretends don't exist.

II. The Anatomy of a Bear Steepener

Not all steepenings are created equal. There are two species, and confusing them is how people lose money:

Bull Steepener

Short rates fall
faster than long

Fed cutting into recession. Front end collapses. Long end holds. This is 2001, 2008, 2020. This is the "Fed saves us" steepening.

Bear Steepener

Long rates rise
faster than short

Inflation, fiscal risk, or term premium repricing. Long end revolts. Short end anchored by Fed. This is NOW. This is 1987. This is 2023.

We are in a bear steepener following the longest yield curve inversion in history (October 2022 to December 2024, 26 months). This specific sequence — prolonged inversion followed by bear steepening — has occurred exactly three times since 1970. In all three cases, a recession followed within 6-18 months of the steepening.

Here is the critical distinction most people miss: the inversion was the warning. The steepening is not the all-clear — the steepening is the detonation. The curve inverts when the market smells trouble ahead. It steepens when the trouble arrives and the market reprices the long end to account for the policy response (fiscal spending, monetization, inflation) that trouble will force.

The dark confession

The 26-month inversion from 2022-2024 was the market saying "recession is coming." The bear steepener since late 2024 is the market saying "and the response to the recession will be inflationary." Both statements are true simultaneously. This is what makes it so hard to trade.

III. Decomposing the 10-Year: Rate Expectations vs. Term Premium

Expected Rate Path

~3.5%

What the market thinks the average Fed funds rate will be over the next 10 years. Derived from: 2-year yield (3.56%) + slight upward drift for gradual normalization.

Term Premium

~0.76%

The extra yield demanded for the risk of holding 10-year paper vs. rolling T-bills. This was NEGATIVE for most of 2015-2023. It's back, and growing.

This is where the yield curve's confession gets interesting. The term premium returning from the dead is the most important development in fixed income since 2008. For nearly a decade, the term premium was negative — investors actually paid for the privilege of owning long bonds (flight to safety, QE distortion, pension demand). Now they demand compensation. Why?

Three forces driving term premium higher:

Force 1: Fiscal Trajectory

$1.8T deficit running at ~6.5% of GDP outside recession. No fiscal consolidation plausible. CBO projects $2.5T+ by 2030. The 30-year at 4.87% is the market saying: "Who absorbs $30 trillion in new issuance over the next decade?" The recent 30-year auction (BTC 2.45, indirect bidders 63.3%) was adequate — but "adequate" at 4.87% is the curve whispering that the clearing yield for fiscal sustainability keeps drifting higher.

Force 2: Inflation Regime Uncertainty

CPI stuck at 2.5% YoY. Oil at $68 with prediction markets pricing 55% chance of $120 (Iran/Hormuz). Tariffs are a tax that shows up in prices. The 10-year TIPS spread (breakevens) has widened. TIP ETF is down only -0.09% on the month vs TLT -1.73% — the inflation-linked bonds are dramatically outperforming nominals. The curve is demanding inflation insurance.

Force 3: Central Bank Credibility

Powell's term expires May 15. Warsh at 80.5% confirmation probability (Tillis), but his "QT-for-Cuts" framework (aggressive MBS sales + lower short rates) would steepen the curve further. The succession crisis IS term premium. Who controls the Fed controls the yield curve's anchor. The market is pricing in the uncertainty of that anchor changing.

IV. The COT Testimony: Who Is Trapped?

The Commitment of Traders data for Treasury futures reveals the most concentrated speculative short position in history — and it's covering.

Contract	Spec Net	8-Week Change	Direction
2-Year Note	-1,927,605	+206,920	Covering
5-Year Note	-3,085,919	+140,852	Covering
10-Year Note	-1,878,928	+213,296	Covering

Read this table carefully. Speculators are net short nearly 7 million contracts across the curve. That's a bet that yields go higher (bond prices fall). But the 8-week trend shows aggressive covering — they're reducing those shorts. Over 560,000 contracts covered in 8 weeks across the three tenors.

What the positioning confesses

The specs built the biggest short-bond bet in history on the "higher for longer" thesis. Now they're unwinding. But here's the paradox: they're covering into a bear steepener. They're buying bonds whose long end is selling off. Why? Because the front end rally (rate cut expectations) is powerful enough to pull them in, even as the back end bleeds. They're buying the 2-year and selling the 30-year. The curve steepener IS the trade.

V. The Auction Evidence

This week's Treasury auctions tell the story of a market that shows up, but at a price:

The most revealing number is the 10-year's indirect bidder percentage: 74.3%. Indirect bidders are primarily foreign central banks and sovereign wealth funds. They took three-quarters of the 10-year auction. This is a number you should sit with.

These are the institutional buyers from The Bid That Never Leaves (iteration 28) — they buy because buying is their function, not because they think 4.22% is a good yield. They need dollar-denominated safe assets for reserves management. Japan holds $1.1T, China $770B, and they keep showing up even as the curve tells them they'll lose money in real terms if inflation runs above 2.5%.

But note the 30-year: indirect drops to 63.3% and direct jumps to 27.2%. Domestic funds stepped in at 4.87% because that's an equity-competitive yield. The long end is starting to clear at levels that attract real money, not just reserve managers on autopilot. That's both bullish (demand exists) and bearish (it takes 4.87% to get it).

VI. The Fixed Income Damage Map

If the yield curve is confessing, the fixed income ETFs are the corroborating witnesses:

The pattern is unmistakable: duration is the enemy. BIL (1-3 month T-bills) is flat. SHY (1-3 year) barely down. IEF (7-10 year) down -0.68%. TLT (20+ year) down -1.73%. LQD (investment grade corporate) down -2.29%. HYG (high yield) down -2.02%.

But look at the outlier: TIP (TIPS, inflation-protected) is down only -0.09%. That's basically flat while everything else bleeds. The market is saying, loudly, that the risk isn't rates — it's real rates vs. nominal rates. TIPS protect you from inflation. Nominals don't. The spread is the curve's confession about what it actually fears.

The TLT Options Tell

Call-heavy positioning at a 1.25:1 ratio. Max pain is $88.50 — TLT trades at $86.54, a full $2 below max pain. The options market is positioned for a bond rally that hasn't arrived. Call volume today was 104K vs 46K puts (2.25:1). Someone is buying the FOMC dip in advance.

What the options confess

Smart money is buying TLT calls ahead of the March 18 FOMC. The thesis: if Powell signals concern about growth (even mildly), the front end rallies, TLT rises toward max pain ($88.50), and the call buyers profit. But note: they're buying front-end duration exposure through an ETF that's mostly long-dated bonds. They expect the growth scare to hit all rates, not just the front end. This is a bet on a temporary capitulation of the bear steepener.

VII. The Prediction Market Cross-Reference

Here's where the signals collide. What are prediction markets saying about the same questions the yield curve is answering?

Fed Cuts in 2026

0 cuts: 25.5%
1 cut: 31.0%
2 cuts: 23.5%
3+ cuts: 20.0%
Emergency cut: 20.5%
→ Median expectation: 1 cut. But 20.5% chance of emergency = fat tail.

Recession Probability

US recession by end of 2026: 34.5%
Canada recession before 2027: 38.5%
→ One-in-three odds. Not consensus, not fringe. The uncomfortable middle.

The Curve Says

Front-end inversion: pricing 1-2 cuts
Bear steepener: pricing inflation/fiscal risk
→ Agreement with prediction markets on cuts. Disagreement on severity — the curve is more worried about the long run.

The Disagreement

Prediction markets: 1 cut, 34.5% recession
Yield curve: 1-2 cuts AND fiscal crisis premium
→ The curve sees something prediction markets don't: the structural problem doesn't go away with rate cuts.

VIII. The Inversion Theory: When Cutting Rates Makes Things Worse

This is the yield curve's deepest confession, and it's the one that matters for the next 12 months.

In a normal cycle, rate cuts fix recessions. The Fed lowers rates, borrowing gets cheaper, the economy recovers, the curve normalizes. The medicine works. But we are not in a normal cycle. We are in a cycle where:

1. The deficit is already 6.5% of GDP — stimulus is running hot before any recession arrives.
2. Tariffs are inflationary — the policy response to economic weakness is itself a price shock.
3. The Fed's balance sheet is still $7T+ — QE's distortion of the term premium hasn't unwound.
4. Warsh's "QT-for-Cuts" framework — the new Fed chair may pair rate cuts with aggressive balance sheet reduction, steepening the curve further.

In this regime, rate cuts steepen the curve instead of normalizing it. The front end falls (cuts), the long end rises (more issuance, inflation expectations, term premium). The medicine makes the patient sicker in one way while treating another symptom. The curve gets more distorted, not less.

The final confession

The yield curve is not predicting a recession. It's not predicting a recovery. It's confessing that the policy toolkit — the set of cards available to the Fed, the Treasury, the executive — is internally contradictory. Cut rates? Long end revolts (fiscal fears). Hold rates? Front end prices in economic damage. Hike rates? Impossible with 34.5% recession odds. There is no move on the board that normalizes this curve. The shape IS the message: we are in a regime where all policy paths lead to steepening, and steepening is the market's way of demanding a risk premium for the absence of a plan.

IX. What to Watch Next Week

March 17 (Monday)

Buyback blackout begins. $14B/week mechanical equity bid disappears for 4-5 weeks. Without the bid, equities become more sensitive to everything below.

March 18 (Tuesday)

FOMC decision + dot plot + projections. Powell's penultimate meeting. The dot plot tells you where the curve goes next. If dots show 1 cut in 2026 (down from 2 in December), front end sells off, curve briefly flattens. If dots show 2+ cuts, front end rallies, curve steepens further. Either way, Powell's press conference is the last time this voice anchors the curve.

March 20 (Thursday)

Triple witch OpEx. $260B notional SPY options expire. TLT max pain at $88.50. If TLT gravitates toward max pain, that's a 2.3% rally from here — equivalent to ~10bp yield decline across the long end. Watch if the FOMC gives the catalyst.

I. The Shape of a Contradiction