How will a rising 10‑year Treasury yield impact other assets?

Our Asset Allocation Committee (AAC) is positioned for the U.S. 10-year Treasury yield to rise. In bond parlance, we are short duration. Inflationary uncertainty, a flat yield curve, pro-growth policies, and significant deficits all point in that direction.

“...there is a nuanced relationship between asset classes and interest rates.”

Predicting the impact of a rising Treasury yield on other asset classes, however, is not simple: there is a nuanced relationship between asset classes and interest rates. Below, I explain how to manage interest rate risk across asset classes by evaluating the sensitivity of asset class pairs to changes in the 10-year yield. Given their negative duration and median relative valuation, floating rate loans currently present a compelling opportunity to capitalize on rising rates, while long‑duration bonds offer an effective tool for duration management.

Estimating durations

We estimate the empirical duration of asset class pairs used for tactical asset allocation, such as value versus growth stocks, as follows:

Value_t – Growth_t = α + -Duration x Δ10y_t¹

Bets duration vs. history

(Fig. 1) Duration estimates and historical ranges

As of February 7, 2025.
Analysis by T. Rowe Price. See “Additional Disclosures” section for additional information on the sources used.
Represents daily data from 1/8/2010 to 2/7/2025. First date for estimated empirical duration was 1/6/2012. The look-back window for percentiles is from 1/6/2012 to 2/7/2025. Past Performance is no guarantee of future results.

Figure 1 shows the current duration (green bars) and historical range (blue lines) for 13 asset class pairs. A positive number means that the asset class on the left tends to underperform the one on the right when the 10-year yield goes up, and vice versa for a negative number.  Starting from the left, the duration of Treasury Long vs. U.S. Agg. (core investment grade) is 8. This means that when the 10-year yield increases by, say, +0.1%, Treasury Long underperforms U.S. Agg. Bonds on average by -0.8%. Also, the taller the blue line, the wider the historical range and—generally—the noisier the estimate.

The duration of the stocks versus bonds pair—the last one on the right—is notoriously unstable. It is currently -1.5 but has been as negative as -15 over our 1/8/2010–2/7/2025 period. A negative duration means that historically stocks have outperformed bonds when the 10-year yield increases. This tends to happen because rates often rise together with growth expectations (on the flipside, rates tend to come down and bonds rally when stocks sell off due to the flight-to-safety effect).

“Higher rates can hurt stocks even more than bonds....”

However, the negative duration is a historic average. It does not always occur. Higher rates can hurt stocks even more than bonds, especially when stock valuations are high, and there’s an inflation (rather than growth) shock. Many research articles have been written on the instability of the stock-bond correlation.²

What current valuations and duration levels are telling us

Currently, the duration of U.S. growth vs. value is negative, which means growth stocks have been outperforming with rising rates. But this is an anomaly by historical standards, most likely due to the artificial intelligence theme. In theory, the cash flows of growth companies are further out in the future, which means a longer duration. Plus, value often outperforms with higher rates as financials boost their net interest margins and materials and commodities companies tend to benefit from rising inflation.

Current asset class valuations vs durations

(Fig. 2) Two pairs stand out as attractive

As of January 31, 2025.
Analysis by T. Rowe Price. See “Additional Disclosures” section for additional information on the sources used. Forward earnings yield is used for equity valuation, and yield to worst is used for fixed income valuation. Current valuation is based on valuation data on 1/31/2025. The look-back window for valuation percentile is based on monthly data, from 1/2012 to 1/2025. Percentiles are calculated by sorting the valuation differentials for each bet. Past Performance is no guarantee of future results.

Bond market duration estimates are more reliable, as evidenced by the shorter blue bars for bond asset class pairs. Right now, the most interesting duration number is for floating rate versus U.S. aggregate. It has the most negative duration of all the pairs and shows a narrow historical range. This suggests that overweighting floating rate loans could be an efficient and precise way to benefit from rising rates.

Moreover, while spreads are generally tight across credit markets, floating rate trades at a median valuation relative to the U.S. aggregate. It has been out of favor due to expectations of Fed rate cuts. However, the current trend is toward fewer rather than more cuts as inflation rears its ugly head again.

Figure 2 maps relative valuation percentiles against empirical durations for asset class pairs. Two trades appear reasonably priced and offer efficient duration management: Floating Rate vs. U.S. Agg. and Treasury Long vs. U.S. Agg, (here, you would reduce the allocation to Treasury Long). Of course, you could also consider a move from Treasury Long to Floating Rates to help reduce duration even more efficiently.

And if you think rates are coming down, reverse these trades.