The DV01-weighted package will provide the market with a way to trade a duration-neutral spread in a single transaction. To achieve this, the cash and futures leg quantities will be adjusted to account for DV01 differences. The leg quantities will be set at listing and fixed until a new spread package is listed following the next auction. The cash leg quantity will be fixed at $5M for all packages (except 2-year packages which will be fixed at $10M) and the futures leg quantity will be set such that the leg quantity ratio closely approximates the prevailing hedge ratio. This package size provides an effective hedge without being large enough to hinder liquidity. In addition, standardization of package sizes across the various tenor spreads is consistent with sizing in the underlying outright markets.

We invite the public to comment on this discussion paper either through their existing CME Group representatives or by emailing TreasuryLink@cmegroup.com.

New spread instruments will be available for trading on the business day following a Treasury auction, consistent with how new on-the-run Treasuries are listed for trading on BrokerTec. Spreads involving 2-, 5- and 7-year cash will be listed monthly, following auctions at the end of the month. Spreads involving 10-, 20- and 30-year cash will be listed quarterly following auctions in February, May, August and November^[1].

On any given day, spreads will be available between the relevant OTR Treasury and front-month futures contract. In addition, spreads against deferred month futures contracts will be listed following the auction that precedes the futures roll and will vary by tenor:

• Deferred month spreads involving 2-, 5- and 7-year cash will be listed following auctions at the end of the month prior to the roll month (i.e., January, April, July, October)
• Deferred month spreads involving 10-, 20- and 30-year cash will be listed following auctions during the roll month (i.e., February, May, August, November).

For example, for the 2yr-ZT package, only the spread against the Sep-2026 (U6) futures would have been listed through July 23, the day of the 2-year auction. Beginning July 24, the day after the 2-year auction, spreads against both the Sep-2026 and the Dec-2026 would have been listed (Figure 2). The Sep-2026 spread would remain listed until the last trade date of the Sep-2026 contracts, September 30.

The pricing of Treasury Link spread packages will align with the Treasury basis quoting convention, where buying the spread (the basis) is equivalent to buying cash and selling futures:

The price ratio is equal to the leg quantity ratio:

For 1:1 packages, the price ratio is equal to one and the spread price simply reflects the difference between cash and futures prices:

For DV01-weighted spreads, the spread price is adjusted by the price ratio, which effectively sterilizes the spread price from parallel shifts in yields. For a discussion on price ratio configuration, see Appendix B.

This spread pricing is similar to that used in Treasury basis trading, where prices are quoted in terms of the gross basis between futures and the cheapest-to-deliver (CTD). Under the assumption that futures DV01 is equal to the DV01 of the CTD divided by its conversion factor (CF), the duration-neutral hedge ratio for a basis trade is equal to the conversion factor of the CTD.

For a spread between on-the-run cash, which is not usually CTD for most tenors, and futures, a different hedge ratio is needed that more accurately takes into account the differences in DV01 between cash and futures.

To calibrate the DV01-weighted leg quantities for a given cash/futures pair, we will:

1. Calculate the hedge ratio (HR) on a given day (t) using the prevailing cash and futures DV01s, where we assume futures DV01 is equal to the DV01 of the CTD divided by its conversion factor (CF). See Appendix A for the derivation:

2. Given a fixed cash leg quantity of $5 million ($10 million for 2-year), calculate the futures leg quantity such that the leg quantity ratio approximates the hedge ratio on the day of the auction. If the resulting ratio is equivalent to 1:1, then the futures leg quantity will be adjusted either up or down by 1 contract, depending on whether the hedge ratio is above or below one:

While the assumption we use to calculate futures DV01 is commonly used, it is not the only way that it can be calculated. In practice, traders may instead use forward-DV01s or option-adjusted DV01s, which account for embedded delivery options, to calculate hedge ratios.

For example, using the results from the July 27, 2026 5-Yr auction (Figure 3), the hedge ratio on the day of the auction for the 5yr-ZF DV01-weighted package is approximately 1.0529 (Figure 4).

^{Source: TreasuryDirect, Bloomberg, author’s calculations}

There exist many pairs of leg quantities that approximate a given hedge ratio. What differs across these pairs is the size of the resulting package and how closely the leg quantity ratio matches the hedge ratio. Some examples of leg quantities that approximate the hedge ratio calculated above are shown in Figure 5.

In our analysis, we use the term tail to describe the difference between the prevailing hedge ratio on a given day and the leg quantity ratio that was fixed at listing^[2]. We represent tails in terms of the number of futures contracts that would be needed to hedge the tail for a $5 million ($10 million for 2-year) cash position for apples-to-apples comparison across package sizes.

If a trader wants to trade a package of $5 million in cash using five $1 million packages, they will have a tail of over two futures contracts to manage. Increasing the package size to $5 million reduces the tail to 0.4 futures and increasing it to $10 million ($20 million for 2-year) reduces it further to 0.1 futures per $5 million cash. These examples demonstrate that greater package size leads to greater hedge precision.

However, there is a trade-off. While larger packages have smaller tails, they may reduce matching opportunities and hinder liquidity. This trade-off is well balanced by a package size of $5 million in face value ($10 million for the 2-year package).

For additional context, Figure 6 shows the hypothetical leg quantity ratio for the 5yr-ZF DV01-weighted package involving the 5yr OTR auctioned in June 2026. Following the auction, the hedge ratio would have been calculated as 1.0275 and the futures leg quantity would have been set to 51 contracts, implying a leg quantity ratio of 1.02. Over the life of the spread (until the 5yr OTR rolled to the July 2026 issue), the hedge ratio trended up towards 1.0337 with a maximum tail of 0.69 futures contracts. The spread price moved between -9.66 and -9.46, a range of 0.2 points, which is in the 93rd percentile of absolute daily spread changes since 2022 (Appendix C).

Figure 7 shows hedge ratios and (hypothetical) corresponding leg quantities for all cash/futures pairs, which were calculated on the auction date of the respective on-the-run Treasury. Note that for the 2yr-ZT spread, the futures leg quantity was adjusted up to 51 contracts since the hedge ratio implied a futures leg quantity of 50, which would not differentiate the DV01-weighted package from the 1:1 package.

Generally speaking, a smaller package size means larger tails. In addition to the tails that can arise when leg quantities are set at listing, they can also grow over time as hedge ratios change, while leg quantities remain fixed.

To analyze this, we constructed a historical time series of tails for various package sizes across tenors. In this series, hedge ratios are calculated on a daily basis, while leg quantity ratios are reset when new packages would have been listed. In addition, futures are rolled to the next contract month either at the same time as a new package is listed – for 2-, 5- and 7-year packages – or on first notice date – for 10-, 20- and 30-year packages – due to the timing of auctions relative to the Treasury futures roll.

The charts in Figure 8 provide a historical view of the trade-off between package size and hedge precision. Each chart shows tails expressed in terms of the number of futures contracts per $5 million cash position over time for $1 million, $5 million and $10 million package sizes^[3].

Hedge ratio movements over time often generate tails in excess of one futures contract for a $1 million package. Increasing the package size to $5 million significantly improves hedge effectiveness, where tails rarely exceed one futures contract. Sizing up to $10 million, the benefit is less clear. The frequency of tails that exceed one futures contract remains low while the large tails that we see with a $5 million package remain with a $10 million package.

While maintaining a standard package size of $5 million in cash face value along the curve (with the exception of $10 million for 2-year packages) provides the benefit of simplicity, it overlooks how spread volatility increases further out the curve (Appendix B). Traders already navigate this dynamic when trading outright futures, where contracts from the 5-year point and beyond feature a $100K notional despite variation in duration and price return volatility. Traders manage this through appropriate position sizing. Given that spreads usually trade in larger size than outrights and the impact of package size on tails, we consider this straightforward approach to package sizing to be appropriate.

We find that large tails are primarily driven by switches in the CTD. Figure 9 shows the average and maximum of daily changes in tails for a $5 million package ($10 million for 2-year), broken down by days when there was a CTD switch and all other days. On average we find that hedge ratio changes are much larger on days when CTD switches occur and that the largest changes also occur on those days. These changes are more significant for longer tenors, where CTD switches that lead to material changes in DV01s are more prevalent.

Based on historical analysis of hedge ratios and tails, we feel that a package size of $5 million in face value ($10 million for 2-year) is appropriate. Due to CTD switches or divergence in views over hedge ratios, the DV01-weighted packages may not always serve every trader’s needs. However, the ability to simultaneously execute cash and futures as a spread, either through 1:1 or the DV01-weighted packages, will greatly improve execution efficiency and serve as a strong foundation from which tails can be managed in the deeply liquid outright markets.

Let HR denote the (constant) hedge ratio such that the spread is duration-neutral, i.e., the spread price will not be affected by a small parallel shift of in the Treasury yield curve:

For a given small change in yields, Δy:

Under the conventional assumption DV01_Futures = DV01_CTD/CF_CTD, where CF_CTD is the Conversion Factor of the CTD:

The price ratio can be set independently from the leg quantity ratio, which is designed to approximate the DV01-neutral hedge ratio subject to leg quantities being multiples of $100K ($200K for 2-year packages). Because of this rounding approximation, parallel yield shifts can cause slight movements in the spread price. This creates operational friction for passive order entry, as traders may need to update working orders even when the DV01-neutral spread remains unchanged. 

Setting the price ratio equal to the hedge ratio resolves this operational issue. However, doing so introduces a mismatch between the spread price and the underlying position's actual P&L, creating scenarios where a trader incurs P&L without any corresponding movement in spread price.

Assume that the hedge ratio is 1.024 at listing (t = 0), implying a leg quantity ratio of 51 futures contracts to $5M cash (1.02). Given cash and futures prices at t = 0, we calculate the implied spread price under two scenarios:

1. Price ratio = leg quantity ratio
2. Price ratio = hedge ratio

For simplicity, tick size constraints are ignored. Suppose that between t = 0 and t = 1, a parallel shift in the yield curve causes the futures price to fall by 1/32nd and the cash price to fall by 1.024 1/32nd. As shown in Figure 10, the implied spread price calculated using the hedge ratio remains unchanged over this period. 

In contrast, the spread price calculated using the leg quantity ratio decreases due to the divergence between the leg quantity ratio embedded in the spread price and the actual hedge ratio. Because the spread position was not perfectly hedged, a buyer of the spread at t = 0 would have experienced a small loss of $6.25 by t = 1.

This example demonstrates that while matching the price ratio to the hedge ratio keeps spread prices stable during parallel yield shifts, it creates a disconnect between the spread price movement and the true economic P&L of the position.

^{[1] Auction dates can be found on the Treasury tentative auction schedule
[2] Similar to a tail in the context of Treasury futures roll trading, representing the additional quantity of futures contracts needed to be bought or sold in the back month to maintain duration through the roll. Not to be confused with a tail in the context of Treasury auctions, representing the difference between the high yield awarded at auction and the when-issued yield.
[3] Note that for the 2-year package, tails are expressed in terms of the number of futures contracts per $10 million cash position over time for $2 million, $10 million and $20 million package sizes.}