The Macro Factors Driving Steeper Yield Curve Risks

Macro Drivers of Steeper Yield Curves

Long-term rates not only reflect the level at which the policy rate is expected to settle, but also any additional compensation investors may demand for holding a long-term bond vs. a series of short-term notes. That additional compensation can be affected by structural macro drivers such as inflation risks, growth shocks, and bonds’ supply and demand drivers.

This macro channel conveys numerous factors that can influence long-term yields, including: price level shocks, perceptions of central bank independence, central bank balance sheet operations (such as Quantitative Tightening), government bond issuance, financial regulation, investors’ preferred habitats, and flight-to-safety behavior.  The takeaway is that there are good reasons to believe that macro drivers can contribute to further increases in long-term yields that result in steeper yield curves. The remainder of this post sequentially assess these factors. 

Price Shocks: Higher, more volatile inflation is reflected in a rising inflation-risk premium. Exhibit 1 illustrates how broad trends in inflation risks are mirrored in the U.S. 2s/10s curve.  Factors driving a higher inflation-risk premium include larger and more frequent one-off price level shocks that drive large and sustained deviations from the inflation target. For example, tariff-related supply chain distortions, energy weaponisation, and climate events leading to higher food prices could easily blow central banks off course from achieving their 2% targets. The recent oil and LNG price spike offers a sharp reminder that these risks are live.

Exhibit 1

U.S. 2/10s Yield Curve and the Inflation Risk Premium

Central Bank Independence: Another factor that could weigh on the medium-term inflation outlook is an erosion in perceived central bank independence. Academic Research finds that central bank independence has deteriorated in most jurisdictions.¹ More recently, Fed Chair Powell and other FOMC members have come under mounting political pressure. For example, the rise in long-term U.S. inflation expectations is consistent with a less-confident view that the Fed will achieve its 2% inflation target beyond its current forecast horizon (Exhibit 2).

Exhibit 2

U.S. 10-year Average Inflation Expectations

In addition, large fiscal deficits risk driving the inflation-risk premium higher.  Low growth and higher interest rates raise debt sustainability concerns (Exhibit 4). Combined with diminished central bank independence, fiscal laxity could lead to elevated perceived risks of fiscal dominance, driving investors to demand greater compensation for holding longer-term debt.  This live risk was underscored by the passage of the most recent U.S. budget in 2025, which is set to keep the U.S. deficit close to 6% of GDP, according to estimates from the Congressional Budget Office.

Supply-Demand Dynamics: Higher real-risk premia, including those indicated by the bond market’s supply-demand dynamics , can be affected by factors including an increase in the size and frequency of supply shocks, such as the global financial crisis, Brexit, protectionism/de-globalization, and COVID. 

Following the post-COVID inflation surge and higher energy prices resulting from Russia’s invasion of Ukraine, central banks raised interest rates and began to shrink their balance sheets, a process referred to as quantitative tightening (QT; see Exhibit 3). QT is the reverse of QE, whereby central banks commit to buying safe assets, effectively removing duration risk from the market.  With central banks stepping back from QE, duration risk is returning to the market.  In addition, some large central banks that accumulated U.S. Treasuries (USTs) as part of their reserves—which for years helped compress bond yields even as the Fed was hiking rates (i.e., a savings glut as referred to by former Fed Chair Ben Bernanke)—are reducing their allocations. This dynamic is especially worth watching going forward as events across the Middle East unfold. 

Alongside central banks stepping back from QE and carrying out QT, we are seeing adjustments to bond issuance schedules. For example, the UK Debt Management Office and Japanese Ministry of Finance have made recent announcements regarding issuance reductions at the ultra-long end in favor of shorter-maturity issuance.²

Other factors that could affect supply-demand dynamics in the bond market include the potential for increased barriers to foreign investment, such as the proposed Section 899 in a draft of the 2025 U.S. budget, which was subsequently removed from the final bill.

Exhibit 3

Central bank balance sheets recently crested in size amidst QT effects

Financial Regulation: Increased political appetite for financial deregulation, which would generally enable greater risk taking on the part of the financial sector—potentially reducing demand for safe assets (e.g. on behalf of insurers and pension funds), could be another source of uncertainty. As a counterpoint, market participants have noted that the loosening of the Supplementary Leverage Ratio (SLR) and recent passage of the GENIUS Act, which would support the growth in dollar backed stable coins, are factors supporting UST demand.

Investor Behavior: Finally, preferred-habitat investors could dampen the decline in long end rates.³ According to this theory, preferred-habitat investors continue to hold securities on the front of yield curves even as short-end interest rates fall, incentivizing arbitragers to borrow short and buy long.  However, the theory also stipulates that long yields may fall by less than short rates as risk averse arbitragers require compensation for the risk that short-end yields could rise, thus reducing the appeal of their long-end positions.

In totality, these factors suggest that the balance of risks from structural macro drivers points to the potential for further curve steepening.

An Uncertainty-driven Model of the U.S. 2/10-year Yield Curve

The section above sets out the structural drivers of the U.S. yield curve, and recent trends suggest the likelihood for continued volatility and further upside risks. In this section, we aim to quantify that risk from a subset of the various identified drivers. 

We first present a regression model of the 2/10s UST curve based on three explanatory variables: 

• Inflation Risk Premium (IRP): A measure of inflation uncertainty. It is measured by subtracting a Federal Reserve derived long-term inflation forecast from the 5y5y U.S. Inflation swap. Investors seek compensation for increased inflation uncertainty over long time horizons.
• Economic data surprises relative to consensus expectations (Citi Economic Surprise Index, ESI): A measure of economic growth uncertainty. The larger the deviations of incoming data from consensus, the more perceived uncertainty in economic activity. In a more uncertain growth environment, investors will demand higher compensation for rates exposure at longer tenors. 
• 10-year Swap Spreads (swap rates minus UST yields, denoted SS): A measure of  changes in supply or demand for USTs, linked, for example, to Treasury issuance or regulatory changes that affect banks purchases of USTs. In particular, increased UST issuance or central bank balance sheet reduction would be associated with a narrowing swap spread.

The regression is estimated using weekly data from 2010 onwards, and explanatory variables are z-scored to allow interpretation in standard deviations (sd).⁴ Given our focus on the impact of macro drivers over a longer-term horizon, our model includes the Goldman Sachs Cyclicals vs Defensives Index (CvD) to control for short-term cyclical impacts on the 2/10s curve. For example, a rise in cyclical versus defensive stocks captures improving nominal growth prospects (reflation expectations). Those reflation expectations are usually captured by a combination of central bank easing expectations and improved growth and inflation prospects further out. These expectations lead to a steeper yield curve due to cyclical reasons that are unrelated to the structural factors captured by the other variables. The results are summarised in Exhibit 4 and are as expected.

Exhibit 4

Regression Results

For example, a one standard deviation increase in the inflation-risk premium is associated with a 3.7 bps steepening in the 2s/10s yield curve. Higher growth uncertainty, proxied by economic surprises versus consensus expectations, also steepens the yield curve as investors become more uncertain about economic prospects and demand a higher compensation for risk at the long-end of the curve. Finally, a one standard deviation decrease in swap spreads (i.e., expectation of higher UST supply) steepens the curve by nearly 7 bps.

Exhibit 5 provides an illustrative example of how our regressors of interest rates may respond to plausible shocks in a range of scenarios. They are based on potential U.S. structural changes that we have written about previously—inflation shocks (including tariffs), central bank independence, fiscal largesse, and an AI-driven productivity boost.  For example, a 5-standard deviation shock to inflation compares to the 4-standard deviation increase observed in response to the energy shock triggered by Russia’s invasion of Ukraine.  Similarly, we saw a 5-standard deviation fiscal expansion during the pandemic period, as governments stepped in to support economies under lockdown.  

Exhibit 5

Structural Scenarios: How changes in drivers may affect the U.S. yield curve

Our range spans from a ~10 bp decrease in the 2/10s yield curve on the back of a positive supply-side impact of AI, to a ~90 bp increase if all three of our upside scenarios materialise. This stylised example illustrates that risks to the U.S. curve based on these structural changes are plausibly skewed to the upside, and if they were to crystalise, would take the slope of the yield curve closer to its pre-GFC average (Exhibit 6). It should be emphasised that this analysis is focused specifically on structural changes, and were these to bring about a recession or lower cyclical growth expectations, the yield curve would be steeper than otherwise, given the positive coefficient of our CvD control variable.

Exhibit 6

U.S. Curve Scenarios Relative to Average Curve Steepness

Conclusion: Structural macro drivers could plausibly push the yield curve steeper

After experiencing a downward trend in the aftermath of the GFC, the U.S. yield curve has risen substantially, a pattern seen in other global bond markets.  In the face of a relatively resilient U.S. economy, a key question for investors is whether structural macro drivers could nevertheless drive the long end higher, leading to a further steepening in the yield curve.

This note highlights a number of structural reasons why this issue warrants analysis: inflation is higher and supply shocks make it less predictable (e.g. via tariffs); governments continue to issue debt despite fiscal largesse; and monetary policy independence is being challenged. These factors have been drivers of the slope of the yield curve in recent years. Moreover, we illustrate that plausible scenarios for further such structural shocks suggest that the balance of risks for yield curves is to the upside.  If these were to crystallise, they could push the slope of the yield curve closer to its pre-GFC average.