The Curve Speaks

What Every Basis Point Is Screaming and Why Nobody Is Listening — March 14, 2026

The yield curve is the market's most honest speaker. It doesn't care about narratives. It doesn't read headlines. It prices the aggregate expectation of every bond buyer on Earth — central banks, pension funds, hedge funds, sovereign wealth funds — all compressed into a single line. When that line changes shape, something real has changed. Today, the shape is screaming.

I. The Curve Right Now

3.60%
3mo
3.87%
5yr
4.29%
10yr
4.91%
30yr
3.60%
3-Month
3.87%
5-Year
4.29%
10-Year
4.91%
30-Year

Key Spreads

SpreadValueWhat It Means
3mo-10y+68bpCurve is positively sloped (normal) — no longer inverted
5y-30y+103bpLong end significantly steeper than belly — term premium expanding
Fed Funds - 10y+54bp10y above Fed Funds — market expects rates to stay near current levels

II. The Four Voices of the Curve

Every point on the yield curve is a different speaker with a different message. Here's what each zone is saying:

Voice 1: The Short End (3mo-1yr) — "The Fed's Grip"

3-Month at 3.60% — This is the Fed's direct control zone. The 3-month yield tracks the federal funds rate (3.50-3.75%) almost exactly. The 3-month is saying: "The Fed is here and not going anywhere."

The short end barely moves day-to-day (-0.06% today). It's anchored by the most powerful force in markets: the Federal Reserve's balance sheet. When the short end is stable while the long end moves, all the information is in the spread.

Voice 2: The Belly (2yr-5yr) — "The Rate Path Expectation"

5-Year at 3.87% — The belly of the curve is where rate expectations live. A 5-year yield of 3.87% means the market expects the average Fed Funds rate over the next 5 years to be approximately 3.5-3.7% (yield minus term premium). That's barely any cuts from the current 3.75%.

The SHY ETF (1-3yr bonds) is down only -0.46% in the last month. The belly isn't moving much. Translation: the market has given up on meaningful rate cuts. This confirms the prediction market data: only 22% chance of a cut by December 2026.

Voice 3: The Long End (10yr-30yr) — "The Fear Premium"

10-Year at 4.29%, 30-Year at 4.91% — This is where the screaming is. The long end is where term premium lives — the extra compensation investors demand for uncertainty about inflation, fiscal deficits, and the future of the dollar.

TLT (20yr+ bonds) is down -3.01% in one month. IEF (7-10yr) is down -1.28%. SHY (1-3yr) is down -0.46%. The pain increases monotonically with duration. This is a bear steepener — the most feared yield curve regime.

Voice 4: The Spread (5y-30y) — "The Fiscal Alarm"

103 basis points — The gap between the 5-year and 30-year is over a full percentage point. This is the market's way of saying: "We'll lend you money for 5 years without much worry, but 30 years? That costs extra."

The "extra" is term premium — compensation for:

III. The Bear Steepener: Why It's the Worst Regime

There are four ways a yield curve can move. Only one is happening right now, and it's the worst one:

RegimeShort EndLong EndCauseImpact
Bull Steepener Falls fast Falls slow Fed cuts aggressively Best for stocks & risk
Bull Flattener Stable Falls Growth slowing, flight to safety OK for bonds, mixed for stocks
Bear Flattener Rises fast Rises slow Fed hiking Bad for risk but orderly
Bear Steepener Stable Rises fast Inflation fear + fiscal concern + term premium expansion Worst for everything

We are in a bear steepener. Here's the proof:

-0.46%
SHY (1-3yr) 1-Month
-1.28%
IEF (7-10yr) 1-Month
-3.01%
TLT (20yr+) 1-Month

The short end is barely moving. The long end is getting crushed. The curve is steepening because the long end is rising — not because the short end is falling. This is the bear variety of steepening, and it's the most dangerous because:

  1. No Fed rescue: In a bull steepener, the Fed is cutting — that's the cavalry arriving. In a bear steepener, the Fed is frozen (can't cut because of inflation, won't hike because of growth). There is no cavalry.
  2. Discount rate rises: The 10-year yield is THE discount rate for equity valuations. Every 10bp increase in the 10-year knocks ~1-1.5% off the S&P 500's fair value. From 4.0% to 4.29% is worth ~3-4% equity downside, which is... exactly what we've seen (SPY -4.3%).
  3. Mortgage rates follow the long end: 30-year mortgage rates track the 10yr + credit spread. Higher long yields = higher mortgage rates = housing pressure = consumer weakness.
  4. Fiscal spiral risk: Higher long yields = more expensive government borrowing = larger deficit = more issuance = higher yields. Self-reinforcing loop.

IV. The Auction Market vs. The Futures Market

One of the deepest contradictions in markets right now: speculators are short 5 million Treasury contracts while actual auctions keep clearing with solid demand.

DateSecurityYieldBid/CoverIndirect %Verdict
Mar 11 10-Year Note 4.217% 2.45x 74.3% STRONG
Mar 10 3-Year Note 3.579% 2.55x 59.6% SOLID
Mar 12 8-Week Bill 3.10x STRONG
Mar 11 17-Week Bill 3.19x STRONG

The Split Personality

Futures market (COT): -1.88M spec short on 10-year, -3.09M spec short on 5-year. Total: ~5 million contracts short. "Yields are going higher."

Auction market: 2.45-3.19x bid-to-cover ratios. 74.3% indirect (foreign/central bank) participation on the 10-year. "We want these bonds."

How can both be true?

  1. Different participants. The futures market is dominated by specs (hedge funds, CTAs, prop traders). The auction market is dominated by institutional buyers (central banks, pension funds, insurance companies). Specs trade on expectations. Institutions buy on mandates.
  2. Different time horizons. Specs are positioned for the next 3-6 months (tariff inflation, no Fed cuts). Institutional buyers are allocating for the next 10-30 years (need yield to match liabilities). Both can be "right" on their own timeframe.
  3. The basis trade. Many of the 5 million spec shorts are part of a leveraged basis trade: buy the cash bond at auction, short the futures contract, capture the basis spread. This makes the futures market look more bearish than it actually is. The short position is a hedge, not a directional bet.

Inversion Theory: The auction demand at 74.3% indirect tells you that foreign central banks are still showing up — not out of conviction, but out of role. They MUST hold Treasuries as reserves. This is the forced buyer that prevents the bear steepener from becoming a rout. But each auction, the yield they demand creeps higher. The erosion is 98% of yesterday's enthusiasm, not zero. Watch the tail (when high yield comes in above pre-auction trading level) — that's when even the forced buyers start flinching.

V. The Prediction Markets See Two Futures

MarketProbabilityWhat It Means
10y hits 4.4% by March 31 51.4% Coin flip — only 12bp away. Market expects the bear steepener to continue near-term.
10y hits 4.4% before 2027 58.0% Likely to happen at some point. Confirming upward pressure on long end.
10y hits 4.5% before 2027 48.0% Nearly even odds of another 21bp from here.
10y hits 4.8% before 2027 33.5% 1 in 3 chance of approaching the 5% threshold. Serious stress scenario.
10y dips below 3.7% before 2027 65.0% KEY: 2 in 3 chance yields eventually plummet. The rate-cut-eventually thesis.
10y dips below 3.5% before 2027 26.5% 1 in 4 chance of massive rally (recession signal).

The Dual-Peaked Distribution

Read those numbers carefully. The market is simultaneously pricing:

Both can be true if yields spike first, then crash. This is the classic yield curve pattern before a recession: long yields rise on inflation/term premium fears, then COLLAPSE as the economy breaks and the Fed is forced to cut aggressively. The prediction markets are pricing in the FULL ARC — up first, then violently down.

The historical parallel: In 2007, 10-year yields rose from 4.5% to 5.25% between February and June, then crashed to 3.5% by January 2008 as the recession hit. The top was the last gasp of the bear steepener before the bull steepener took over. The question is whether we're in the "February" or the "June" of that arc.

VI. TLT: The Long Bond's Confession

$86.54
TLT Price
$89
TLT Max Pain
13.9%
TLT ATM IV

TLT is trading $2.46 below max pain ($89 vs $86.54). The options market expected bonds to stay around $89, but the bear steepener pushed TLT lower. Market makers who sold puts near $87 are underwater.

But here's the most surprising number: TLT ATM IV is only 13.9%. Compare:

The bond market is pricing in LESS uncertainty than any other asset class. This is either:

  1. Complacency: Bond traders don't believe the bear steepener will accelerate. They think 4.3% on the 10-year is near the top.
  2. Anchoring: The Fed's rate guidance (data-dependent, patient) provides a floor for bond prices. Even if long yields rise, the short end is locked — limiting total portfolio loss for duration-heavy investors.
  3. Mispricing: If the FOMC on March 18 delivers a surprise (either direction), TLT's 13.9% vol will prove dramatically insufficient. With 5 million spec short contracts waiting to unwind or add, the move could be 2-3 standard deviations.

The inversion lens: Low TLT vol is the market's way of saying "we know the answer is binary (cut or hold) and we'll find out March 18, so why pay for volatility before the reveal?" But binary events are EXACTLY when volatility is most underpriced, because the option sellers model the expected move as an average of two scenarios. If one scenario is +3% and the other is -3%, the "expected" move is 0% — but the actual move is huge.

VII. What Each Yield Is REALLY Pricing

Let's decompose the curve. Every yield = Expected Future Short Rate + Term Premium. Here's the translation:

TenorYieldExpected Avg Fed RateTerm PremiumTranslation
3-Month 3.60% 3.60% ~0bp "The Fed is at 3.75% and I trust them for 90 days."
5-Year 3.87% ~3.50% ~37bp "The Fed averages 3.5% over 5 years. Small premium for uncertainty."
10-Year 4.29% ~3.30% ~99bp "I expect rates to average ~3.3% but I need a full point of premium to hold 10-year risk."
30-Year 4.91% ~3.20% ~171bp "I think the long-run rate is ~3.2% but I need 171bp of insurance against inflation, fiscal chaos, and the unknown."

The critical insight: Expected future rates are actually falling as you move out the curve (3.60% → 3.50% → 3.30% → 3.20%). The market DOES expect rates to come down eventually. But the term premium is EXPLODING (0 → 37 → 99 → 171bp). All of the bear steepener is in the term premium, not in rate expectations.

This means the bond market is saying two things simultaneously:

  1. "The Fed will eventually cut. We believe in reversion to ~3.2%."
  2. "But the RISK of getting from here to there is enormous. Inflation could stay high. The deficit could blow out. The political environment could become hostile to sound fiscal policy. For that risk, I need to be paid."

The term premium IS the uncertainty tax. And it's at its highest level since 2014.

VIII. The 5 Million Short: Who Gets Hurt

From the COT report (iteration 3), we know specs hold ~5 million contracts short across the Treasury curve. Here's how the yield curve shape determines who wins and who loses:

If Bear Steepener Continues (30% probability)

If Curve Flattens via Rally (55% probability — FOMC dovish)

If Curve Bear Flattens via Fed Hike (5% probability)

IX. The Curve's Recession Signal: Where Are We?

The Inversion-to-Steepening Sequence

The yield curve follows a predictable sequence around recessions:

1. Curve inverts
2022-2024 ✓
2. Curve un-inverts
Late 2025 ✓
3. Bear steepener
NOW ← We are here
4. Bull steepener
Recession + Fed cuts
5. Recovery
Normal curve returns

The 2s10s spread was inverted (negative) from mid-2022 through late 2024 — the longest inversion in history. It un-inverted in late 2025 as the Fed began cutting. Now it's positively sloped at ~+40bp but steepening via the long end rising (bear steepener).

The historical pattern: The recession typically arrives 6-18 months AFTER the curve un-inverts, during the bear steepener phase. We un-inverted ~4 months ago. If history rhymes, the recession window is Q3 2026 - Q1 2027.

The prediction markets agree: 33.5% chance of recession by end of 2026, 58% chance unemployment hits 5%+. Not a certainty, but not a tail risk either.

X. The Inversion Theory: Where the Steepener Destroys Itself

The Self-Destructive Loop

Long yields rise Mortgage rates spike Housing slows Consumer weakens Employment falls Fed forced to cut Yields crash

The bear steepener is self-limiting because higher long yields tighten financial conditions, which slows the economy, which eventually forces the Fed to cut, which collapses yields. The extreme of rising rates creates the economic damage that produces falling rates.

This is why the prediction markets can simultaneously price 58% chance of 4.4%+ AND 65% chance of sub-3.7%. The arc goes: up to 4.4-4.8% (bear steepener peak), then DOWN through 3.7% (recession/cut response). The question is timing and violence of the turn.

The Bond Vigilante Paradox

Bond vigilantes demand higher yields for holding duration risk (rightfully, given $38.6T debt). But higher yields INCREASE the deficit (more interest expense), which INCREASES supply, which requires EVEN higher yields to clear. The vigilantes are making the fiscal situation worse by demanding compensation for it. The concern about the problem IS the problem.

The Auction Paradox

Foreign central banks take 74.3% of the 10-year auction. Their buying keeps yields from truly exploding. But their buying is also what enables the $2T+ fiscal deficit to persist. Without their forced participation, Congress would be forced to cut spending or raise taxes. The forced buyers are enabling the fiscal behavior that increases the risk they need to be compensated for.

Every point on the curve is a conversation between fear and function. The short end is function (the Fed's machinery). The long end is fear (inflation, deficits, political instability). The belly is the market's attempt to average the two. Right now, fear is winning. But fear is self-correcting — it creates the conditions for its own resolution. The bear steepener always ends. The question is whether it ends with a whimper (gradual flattening) or a bang (yield collapse as recession arrives). March 18 won't answer this question, but it will tell us how much further the fear has to run before the function catches up.