Estimating Treasury Futures Hedge Ratios with Empirical DV01

Summary

The document considers estimating hedge ratios for spreads between Treasury futures when historical DV01 data are unavailable. It describes empirical DV01 as an estimate from a regression of futures price changes against changes in the yield of the cheapest-to-deliver bond. Since DV01 measures price sensitivity to yield, the estimated sensitivity can be used to approximate a hedge ratio between contract legs.

The answer notes that shorter estimation horizons, such as three or six months, are commonly used in this context and cites a rolling three-month estimate as an illustration. It cautions that computing actual historical DV01 is preferable when feasible: a contract specification change can cause a large shift in duration and hedge ratio that an empirical model may struggle to capture. The material offers practitioner guidance rather than a general validation of window selection, and it does not provide enough detail to establish a universally appropriate rolling window.

Key ideas

• DV01 measures futures price sensitivity to yield changes.
• An empirical DV01 can be estimated by regressing futures price changes on cheapest-to-deliver yield changes.
• Shorter rolling windows, including three- or six-month horizons, are used as empirical estimation choices.
• Actual historical DV01 is preferable when available because contract specification changes can shift hedge ratios sharply.
• The document does not establish one optimal rolling window for every spread or market regime.

Tags

# How can I approximate the hedge ratio for Inter Commodity Treasury Spreads?


# How can I approximate the hedge ratio for Inter Commodity Treasury Spreads?












Looking at the excellent CME Treasury Analytics tool, I can see that the hedge ratio for spreads betweend diff treasury futures is derived from the DV01 of each leg.

I can get treasury futures data and test strategies, however I don't think is that easy to get historical DV01 in order to get the historical hedge ratio. Is it possible to approximate the Hedge Ratio using using the minimum variance approach?....what would be an appropriate rolling window for the calculation?

Thanks

## Answer by Helin (score 3, accepted)

https://quant.stackexchange.com/a/43676

DV01 is defined as $$ \text{DV01} = -\frac{dP}{dy}, $$ so technically you could run a regression of futures price changes vs (CTD) yield changes. The resulting DV01 is known as empirical DV01. In the context of trading bond futures, shorter-term horizons such as 3m and 6m are typically used. The chart below shows the actual TY/WN hedge ratio and an empirical version (using a rolling 3m window):

But to be honest, it's probably better to compute the actual DV01 over time... The chart below shows the historical TY/US hedge ratio. The jump results from the US contract spec change in 2015 that drastically changed the duration of the US contract. Empirical models would struggle a great deal.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.