Automated Treasury Hedging Workbook

Can a duration-ratio short-futures hedge hold a $100M Treasury portfolio's market value stable through a rate-moving period, and where does the construction fall short of a clean hedge?

Objective

Build and evaluate a duration-neutral hedge of a simulated $100.0mm U.S. Treasury portfolio using CBOT 10-Year Treasury Note futures, tested over a defined management period against a mandate to minimize deviation from the portfolio's starting value.

Prepared as a Boston University AD713 (Derivative Securities and Markets) course assignment and for portfolio purposes.

Methodology

1. Construct a four-note Treasury portfolio (5Y, 7Y, 10Y, 20Y) and compute weighted-average modified duration and DV01
2. Size the initial short futures position by the duration-ratio method: portfolio dollar duration divided by futures dollar duration
3. Track daily mark-to-market on both the bond portfolio and the futures margin account over 31 trading sessions
4. Re-solve the hedge ratio and rebalance the futures position as rates move
5. Compute stability statistics (mean, standard deviation, range) on daily deviation from the starting portfolio value
6. Attribute the period result between bond price change, accrued interest, and futures margin P&L

Key Assumptions

Portfolio Composition

The duration profile is concentrated: the 20-year note is 43.0% of face but contributes 60.1% of portfolio duration. This is the origin of the curve-risk limitation noted below — the portfolio's risk sits materially further out the curve than the single instrument used to hedge it.

Execution

Eight transactions were executed over the period: one initial hedge and seven subsequent adjustments, all priced at end-of-day marks.

Total transaction cost across the period was $14,766, of which $14,330 (97.0%) was incurred on the initial hedge. As futures prices fell, the DV01 per contract declined and the required position rose from 1,261 to 1,290 contracts to remain duration-neutral — the hedge ratio is not static even when portfolio duration is unchanged, since the denominator moves with the market.

Results

Period attribution

The bond portfolio lost $2,548,316 in market value as rates rose over the period; the short futures position generated $2,946,296 in the margin account, more than offsetting that loss. Expressed as a capture ratio, the hedge recovered 115.6% of the underlying loss — directionally correct and modestly over-sized.

Conclusion: The stability mandate was met. A 0.46% standard deviation of portfolio value against a $100mm base, over a period that saw meaningful rate movement, reflects an effective duration-neutral hedge. The hedge is modestly over-sized (a 115.6% capture ratio) and carries unhedged curve risk from concentrating the hedge in a single 10-year instrument against a 5-to-20-year ladder — both addressed below.

Limitations

• The hedge uses a single 10-year futures contract against a ladder spanning 5 to 20 years. The 20-year note contributes 60.1% of portfolio duration but is hedged with an instrument of 6.50 modified duration; the sizing embeds a parallel-shift assumption, and any curve steepening or flattening passes through unhedged. A two-point hedge (for example, FV plus US futures) with front- and long-end DV01 matched separately would address this.
• Futures modified duration is held constant at 6.50 throughout. In practice the cheapest-to-deliver bond and its conversion factor shift as the curve moves, so the true contract DV01 drifts; holding it fixed means the position departs from genuine DV01-neutrality between re-solves.
• The rebalancing trigger is not documented in the workbook. Seven adjustments over 31 sessions is consistent with a threshold policy, but the threshold itself is not stated, so the rebalancing record cannot be independently reviewed or replicated.
• Margin financing is modeled one-directional. The margin balance ran positive for most of the period and earned interest at the assumed 5% rate; a sustained adverse move would invert this into a funding cost, which is not stress-tested.
• The hedge is first-order only. Duration-ratio sizing neutralizes linear rate exposure; convexity is unhedged, which is a real, if second-order, residual for a portfolio with a 20-year component under large rate moves.
• The portfolio, its marks and the hedging record are simulated. No independent term-structure model or yield-curve fitting underlies the marks; yields, prices and durations are taken as given at inception.

Deliverables

Technology Stack

• Microsoft Excel
• VBA
• Fixed income analytics

Skills Demonstrated

• Fixed income analysis and duration management
• Treasury futures hedge construction and sizing
• Interest rate risk management
• Portfolio hedging and rebalancing mechanics
• Performance attribution

Prepared for portfolio and educational purposes only. Figures are illustrative, derived from a simulated portfolio and hedging model, and do not constitute investment advice.