10-Year Treasury Expected Range: Does It Work on Rates?

A 10-year Treasury futures contract moves by a fraction of a percent on an ordinary day, far less than an equity index. An expected-range band built in return space adapts to that: on the published ZN series, the next close landed inside the inner band 70.8% of the time and inside the outer band 94.6%, historically, in line with the other asset classes.

What has been published for rates?

The Proof page lists ZN (10-year Treasury futures) over 2005–2024 at 70.8% inner and 94.6% outer. Past behavior is not a guarantee of future results. Rates is one of the five asset classes in the 40-instrument set, covered as a group in the cross-asset article.

Why does a small-moving market not break the band?

A band drawn a fixed number of points from price would be meaningless on a contract that rarely moves a point. The Behavioral Transform Model measures the standard deviation of recent returns, so a market whose typical day is a few hundredths of a percent gets a correspondingly narrow band. The same rolling calculation, with the same window of roughly 60 bars, was run on ZN as on Bitcoin; return space vs price space explains why the unit is what makes that possible.

What does the figure not say?

It says nothing about yields, duration or where rates are going. The band describes how far the futures price has recently moved and projects that onto the next close. It is an expected range, not a claim that price reverses at its edges.

What are the limits?

The ZN sample covers two decades, much shorter than the 97-year equity series; the paper reports a confidence interval per instrument. The figure is for the futures contract, so how the long series was joined at roll dates is a property of the data; the live pages do not state how the published series was built. The band runs too narrow in the first days of a fast volatility jump, the outer band is slightly optimistic in the deep tails, and the paper validates calibration, not the profitability of any use.

Frequently asked questions

Does it work on a yield chart such as TNX?

The published figure is for the ZN price series. A yield series is a different quantity, and no figure is published for it.

Does the band work on bond ETFs such as TLT or LQD?

The construction runs on any series of closes, but no bond ETF row appears among the 16 shown on the Proof page, so no claim is made here.

Why is the ZN outer band figure slightly higher than the others?

Each instrument has its own figure and sample period, and the paper reports confidence intervals; a difference of a few tenths of a point is not treated as meaningful on its own.

Details are in the working paper. The 30-day free trial lets you view the range on a Treasury chart yourself: start free, cancel anytime.

‍