Report Interpretation

Overview

Morgan Stanley examines why daily US Treasury futures moves can differ from the rate exposure investors expect from a single deliverable bond. It concludes that realized futures behavior combines current CTD sensitivity with probability-weighted delivery risk, and that USZ6 is especially exposed to a discrete duration extension if long-end yields rise further.

Core views

A Treasury futures contract references a basket of deliverable bonds with different coupons and maturities, so its rate sensitivity is not necessarily equivalent to that of one bond. Morgan Stanley contrasts two approaches: the current cheapest-to-deliver (CTD) approach, which assigns the contract the sensitivity of the bond that would be delivered today, and a probability-weighted approach, which averages potential CTD bonds according to their estimated likelihood of being delivered. The latter is described as the more complete theoretical framework because delivery occurs in the future, although the report notes that the cost and complexity of maintaining a CTD-probability model often exceeds its marginal benefit when the two approaches give similar results. For USZ6, the difference is currently material. Higher long-end yields are moving the contract closer to a CTD switch from the high-coupon, originally issued 20-year 5% May45s bond to the low-coupon, originally issued 30-year 1.375% Aug50s bond. Under the current-CTD framework, a roughly 20–30bp sell-off would raise futures PV01 by 46%. The probability-weighted framework also indicates an extension in rate sensitivity, but one that is smaller and occurs more gradually. Because USZ6 still trades partly according to the current CTD, the shorter-duration 5% May45s currently dampens the contract's realized rate sensitivity; a switch would create a discrete change in how the contract trades. The institution tests the two frameworks against realized daily US futures moves over the past two years. Both explain a high proportion of daily variation because each is based on changes in the deliverable basket, but neither perfectly captures realized moves, leaving potential tracking error. In an August 4, 2026 example for USU6, the futures price rise was greater than the current-CTD model implied but less than the probability-weighted model implied. Testing combinations of the two forecasts, Morgan Stanley finds that a 56% current-CTD and 44% probability-weighted blend produces the lowest RMSE, supporting its view that market participants collectively manage futures risk using both conventions. Morgan Stanley argues that a CTD switch should therefore drive USZ6 underperformance relative to the 5% May45s, while noting that valuations have moved closer to fair value. It maintains a conversion-factor-weighted long 5% May45s versus short USZ6 trade, citing continued long-end sell-off momentum and an attractive payoff around the switch. The stated entry is a 17.14-tick gross basis and 8.84-tick net basis using a 4.17% term repo rate.

Analysis framework

The report first explains the deliverable-basket mechanics of Treasury futures, then compares current-CTD and probability-weighted measures of futures sensitivity. It models sensitivity through parallel curve shifts, compares predicted and realized daily futures moves over a two-year sample, and selects the blend with the lowest RMSE before applying the results to USZ6 relative-value positioning.

Methodology notes

• Fixed Income and CreditDuration and Convexity Analysis

  PV01/DV01 analysis of Treasury futures and their deliverable bonds

  The report measures how much prices change for a one-basis-point yield move and shows how a CTD switch can materially extend USZ6's effective rate sensitivity.

• Other

  Current-CTD versus probability-weighted delivery-risk modelling, evaluated using RMSE

  Morgan Stanley predicts daily futures moves under each delivery-risk convention, combines them at different weights, and identifies the mix with the lowest root-mean-square prediction error.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• USZ6

  Short leg of Morgan Stanley's relative-value trade; expected to underperform if a CTD switch extends its rate sensitivity.

  Strengths

  Its current CTD exposure still dampens realized rate sensitivity before a switch.

  Weaknesses

  A likely transition to the 1.375% Aug50s could cause a discrete increase in PV01.

  Comparison

  Expected to underperform the 5% May45s as the latter ceases to be the CTD.

  Risks

  Neither CTD framework perfectly captures realized daily futures moves, creating tracking error.

• 5% May45s

  Long leg and current CTD bond in the recommended conversion-factor-weighted trade.

  Strengths

  Its shorter duration currently dampens USZ6's realized rate sensitivity and it is expected to outperform the futures contract following a CTD switch.

  Weaknesses

  It may lose CTD status if long-end yields continue to rise.

  Comparison

  Morgan Stanley favors it relative to short USZ6.

Key data

• Best-fitting realized-move blend56% current CTD / 44% probability-weightedProduced the lowest RMSE for realized US futures moves over the past two years.
• Potential USZ6 PV01 change46%Increase under the current-CTD approach over a roughly 20–30bp sell-off.
• Likely CTD transition5% May45s to 1.375% Aug50sExpected if long-end rates continue to sell off.
• Trade entry basis17.14 ticks gross; 8.84 ticks netNet basis uses a 4.17% term repo rate.

Impact & implications

The report says investors should not model US Treasury futures solely as either the current CTD or a fully probability-weighted basket. For USZ6, the hybrid trading behavior means a CTD switch could abruptly increase rate sensitivity and leave the futures contract lagging the current 5% May45s CTD bond.

Risks

• Neither the current-CTD nor probability-weighted framework perfectly explains realized futures moves, leaving tracking error for positions.

What to watch

• Whether continued long-end yield sell-off brings USZ6 to a CTD switch from the 5% May45s to the 1.375% Aug50s.
• The evolution of USZ6 PV01 and relative performance versus the 5% May45s as the contract approaches a CTD transition.
• Whether realized futures moves continue to fit a blended current-CTD and probability-weighted risk model.