# Yield Curve Analysis > A Claude skill from brainbytes-dev/everything-claude-trading. - **Type:** Skill - **Install:** `agentstack add skill-brainbytes-dev-everything-claude-trading-yield-curve-analysis` - **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior - **Seller:** [brainbytes-dev](https://agentstack.voostack.com/s/brainbytes-dev) - **Installs:** 0 - **Category:** [Finance & Payments](https://agentstack.voostack.com/c/finance-and-payments) - **Latest version:** 0.1.0 - **License:** MIT - **Upstream author:** [brainbytes-dev](https://github.com/brainbytes-dev) - **Source:** https://github.com/brainbytes-dev/everything-claude-trading/tree/main/skills/macro/yield-curve-analysis ## Install ```sh agentstack add skill-brainbytes-dev-everything-claude-trading-yield-curve-analysis ``` Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent. ## About # Yield Curve Analysis ## When to Activate - Analyzing term structure of interest rates for trading signals or macro assessment - Designing curve trades (steepeners, flatteners, butterflies) - Interpreting yield curve inversions and their economic implications - Calculating key rate durations and understanding curve risk - Evaluating central bank policy impact on different parts of the curve ## Core Concepts ### Yield Curve Shapes **Normal (Upward Sloping):** - Short rates long rates - 2s10s inversion has preceded every US recession since 1955 (with variable lead time of 6-24 months) - Reflects market expectation that rates will be cut due to economic slowdown - Can be driven by aggressive Fed tightening (front end) or flight to safety (long end) - False positives are rare but exist (1966, arguably 1998) **Flat:** - Minimal spread between short and long rates - Transition state between normal and inverted, or vice versa - Reduces bank profitability (banks borrow short, lend long) **Humped (Bear Flattener):** - Belly of the curve (5Y-7Y) yields higher than both ends - Can occur during transitional monetary policy periods - Less common, but creates butterfly trade opportunities ### Term Structure Theories **Pure Expectations Theory:** - Forward rates equal expected future spot rates - Long rates are geometric average of expected short rates - Implication: inverted curve = market expects rate cuts - Reality: ignores risk premium, but useful as baseline **Liquidity Premium Theory:** - Investors demand extra compensation for holding longer maturities - Term premium should always be positive (not always true empirically) - Term premium estimation: Adrian-Crump-Moench (ACM) model, Kim-Wright model - Current term premium is observable via Fed models **Preferred Habitat Theory:** - Different investors have preferred maturities (pension funds want long, banks want short) - Supply/demand imbalances at specific tenors drive relative pricing - Explains humps and kinks in the curve - QE disrupts natural supply/demand by removing duration from the market ### Key Curve Metrics **Spreads:** ``` 2s10s = 10Y yield - 2Y yield (most watched recession indicator) 2s30s = 30Y yield - 2Y yield (steeper, more sensitive to term premium) 3m10Y = 10Y yield - 3M T-bill yield (Fed's preferred recession indicator) 5s30s = 30Y yield - 5Y yield (long-end steepness) Fed funds - 2Y = front-end tightness ``` **Term Premium:** ``` 10Y yield = Expected average of short rates over 10 years + Term premium Term premium drivers: - Inflation uncertainty (higher uncertainty = higher premium) - Supply of duration (more Treasury issuance = higher premium) - Central bank holdings (QE reduces premium, QT increases it) - Foreign demand (strong foreign demand compresses premium) - Historically: -0.5% to +2.5% range; negative term premium from 2015-2021 due to QE ``` **Key Rate Durations:** - Sensitivity of a bond's price to a 1 bp change at a specific maturity point - Used to decompose curve risk: how much do I lose if the 5Y rate rises 10 bps but 10Y stays flat? - Essential for hedging: match key rate duration profile, not just overall duration ### Curve Fitting Methods **Bootstrapping:** - Derive zero-coupon (spot) rates from observed coupon bond prices - Sequential: use shortest maturity first, solve for each spot rate - Produces exact fit to observed prices **Nelson-Siegel Model:** - Parameterizes the curve with 4 parameters: level, slope, curvature, decay - y(tau) = beta0 + beta1 * ((1-exp(-tau/lambda))/(tau/lambda)) + beta2 * ((1-exp(-tau/lambda))/(tau/lambda) - exp(-tau/lambda)) - Smooth, parsimonious, widely used by central banks - Svensson extension adds a second hump parameter **Spline Methods:** - Cubic splines fit piecewise polynomials through observed points - More flexible but can overfit - Used when precision at specific tenors matters ## Methodology ### Curve Trade Construction **Bull Steepener (expect rate cuts, long end stable):** ``` View: Fed will cut rates; front end rallies more than long end Trade: Long 2Y futures, Short 10Y futures (duration-weighted) DV01 ratio: 2Y DV01 / 10Y DV01 to equalize interest rate sensitivity Profit if 2s10s spread widens (steepens) Risk: if long end sells off more than front end (bear steepener) ``` **Bear Flattener (expect rate hikes, long end anchored):** ``` View: Fed tightening; front end sells off, long end anchored by slowing growth expectations Trade: Short 2Y futures, Long 10Y futures (duration-weighted) Profit if 2s10s spread narrows (flattens) Risk: if curve steepens instead (growth surprises to upside) ``` **Butterfly:** ``` View: belly of curve is mispriced relative to wings Trade: 2s5s10s butterfly - Long wings (2Y and 10Y), Short belly (5Y) - Or vice versa Weighting: match DV01 so total portfolio duration is zero Example: Long 1x 2Y, Short 2x 5Y, Long 1x 10Y (approximate weights) Profit: if 5Y yield moves relative to interpolated 2Y-10Y line Barbell vs bullet: butterfly trades express view on curvature ``` ### Curve Signal Analysis 1. **Plot current curve** against historical percentiles for each tenor 2. **Decompose moves** into level, slope, and curvature using PCA (first 3 components explain ~99%) 3. **Compare implied forwards** with market expectations (Fed funds futures, OIS) 4. **Assess term premium** using ACM or Kim-Wright models 5. **Monitor curve momentum** — curve steepening/flattening trends tend to persist 6. **Cross-reference with macro** — leading indicators, PMI, inflation expectations ### Recession Probability from Curve ``` Models: - NY Fed model: uses 3m10Y spread to estimate recession probability Probit regression: P(recession in 12 months) = Phi(beta0 + beta1 * spread) When 3m10Y inverts: ~30-40% recession probability When 3m10Y at -100bps: ~60-80% recession probability - 2s10s alternative: similar signal but historically less reliable than 3m10Y Caveats: - Lead time varies: 6-24 months from inversion to recession - QE may distort term premium, making inversions less meaningful - "This time is different" arguments emerge every cycle (and are usually wrong) ``` ## Examples ### Example 1: Post-Inversion Steepener ``` Scenario: 2s10s has been inverted at -80bps for 6 months 3-month T-bill: 5.25%, 2Y: 4.80%, 10Y: 4.00% Fed signaling end of tightening cycle Trade: Bull steepener - Long 2Y Treasury futures (2 contracts, DV01 = $40/bp each) - Short 10Y Treasury futures (1 contract, DV01 = $80/bp) - Net DV01: zero (duration neutral) Thesis: 2Y will rally more than 10Y as Fed pivots to cuts Target: 2s10s moves from -80bps to 0bps P&L: 80bps * $80 DV01 = $6,400 per unit Risk: if Fed delays cuts and front end stays elevated Stop: if 2s10s widens to -120bps (loss = $3,200 per unit) Risk/reward: 2:1 ``` ### Example 2: Butterfly on Humped Curve ``` Current yields: 2Y = 4.50%, 5Y = 4.70%, 10Y = 4.40% 5Y looks rich relative to wings Implied 5Y rate from 2Y-10Y interpolation: 4.45% 5Y is 25bps above theoretical (butterfly spread = +25bps) Historical average butterfly: +5bps Trade: Sell the butterfly (short belly, long wings) - Long 1x 2Y futures - Short 2.2x 5Y futures - Long 1x 10Y futures - DV01 neutral overall Target: butterfly normalizes to +5bps (20bps profit) Stop: butterfly widens to +40bps (15bps loss) Catalyst: front-end rally on dovish Fed, or long-end selloff on supply concerns ``` ### Example 3: Curve and Equity Signal ``` Signal analysis: - 2s10s inverted 14 months ago (currently at -20bps, steepening) - Historical pattern: equities peak 3-6 months after inversion - Steepening from inversion = recession approaching (Fed about to cut) - S&P 500 has been flat for 3 months (topping pattern?) Framework: - Phase 1: Curve inverts -> equities continue rising (6-18 months) - Phase 2: Curve begins steepening from inversion -> equities peak - Phase 3: Recession begins -> curve steepens aggressively -> equities decline 20-40% - Phase 4: Recovery -> curve normalizes -> equities bottom Current assessment: Late Phase 2. Reduce equity exposure. Position for further steepening. Build cash for Phase 3 opportunities. ``` ## Quality Gate Before implementing curve trades, verify: - [ ] Duration-weighted properly — curve trades should be DV01-neutral unless intentionally taking directional risk - [ ] Carry and roll-down calculated — what does the trade earn if curve stays unchanged? - [ ] Financing costs included — repo rates for funded positions, margin costs for futures - [ ] Macro thesis is explicit — what economic scenario drives the trade? What would invalidate it? - [ ] Historical context checked — where is the spread relative to historical range and percentiles? - [ ] Term premium decomposition performed — is the spread driven by rate expectations or term premium? - [ ] Central bank positioning assessed — QE/QT schedule, forward guidance, and how they affect specific tenors - [ ] Liquidity considered — on-the-run vs off-the-run, auction cycle effects on pricing - [ ] Roll dates accounted for — futures rolls create basis risk and transaction costs - [ ] Cross-market confirmation — do credit spreads, equity vol, and FX carry confirm the macro view? ## Source & license This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code. - **Author:** [brainbytes-dev](https://github.com/brainbytes-dev) - **Source:** [brainbytes-dev/everything-claude-trading](https://github.com/brainbytes-dev/everything-claude-trading) - **License:** MIT Install and usage instructions live in the source repository linked above. ## Pricing - **Free** — Free ## Security capabilities Automated source analysis of v0.1.0 — what this tool can access: - **Network access:** no - **Filesystem access:** no - **Shell / process execution:** no - **Environment & secrets:** no - **Dynamic code execution:** no *"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.* ## Versions - **0.1.0** — security scan: passed — Imported from the upstream source. ## Links - Listing page: https://agentstack.voostack.com/l/skill-brainbytes-dev-everything-claude-trading-yield-curve-analysis - Seller: https://agentstack.voostack.com/s/brainbytes-dev - Browse the marketplace: https://agentstack.voostack.com/browse --- Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.