This year’s rise in bond yields has not been massive by historical standards.
Still, it’s more alarming because it came at a time when bond yields were already so elevated compared with the prepandemic years. The 30-year Treasury reached a 19-year high of 5.3% on Aug. 17, standing 2.4 percentage points above its 2017-19 average.
What’s behind the global bond selloff? Much ink has been spilled chalking it up to rising worries about the US government’s fiscal health. But we believe that shifting perceptions of the strength of the US economy and labor market have played a bigger role.
The answer matters because rising risk could be a harbinger of more bad news to come. If markets lose confidence in US debt, that can even become a self-fulfilling prophecy.
Additionally, for investors, there’s a big difference between bond yields being high because of the strength of the economy (which should lead to improved risk and inflation-adjusted returns on bonds in the future), and bond yields being high because of higher risk (which is neutral at best).
However, if you are worried about swelling US debt levels, there is some advice we can give based on our analysis.
The most likely way a debt crisis could resolve is an inflationary surge, which is, after all, why many investors have piled into gold the past few years. But rather than buying gold, you’ll probably get more bang for your buck betting on inflation directly via long-term bonds; that bet can be placed by rotating one’s nominal bond holdings into Treasury Inflation-Protected Securities.
Why Have Bond Yields Risen? It Depends on the Time Frame
For all the talk about government debt worries driving up long-dated bond yields, the yield increase in 2026 has been led by the short end, and the focus is squarely on the Federal Reserve.
The federal-funds rate expected by the first quarter of 2027 has shifted by around 1 percentage point. That’s driven the 5-year yield up by 0.8 points, while the 30-year yield is up by a smaller 0.4 points. The yield curve after five years has moved in a roughly parallel manner, with the five-year, five-year forward yield and 20-year, 10-year forward yield both increasing by 0.4 percentage points. Recall the 20-year, 10-year forward denotes the implied yield of a 20-year bond as of 10 years in the future. That points to an upward revision in the expected natural (or “terminal”) federal-funds rate.
Federal-funds rate expectations have increased owing largely to dissipating labor market fears as the unemployment rate has ceased rising. The inflationary shock from the Iran war plays a role, but more important for the Fed is sticky core inflation (which excludes energy prices). The latter indicates an economy still running slightly hot. AI’s impact on investment and stock prices is playing a key role here; insofar as the AI boost is expected to persist, it boosts the natural rate component of longer yields.
On the other hand, if we slightly lengthen the time frame, the chief development has been the steepening of the yield curve. Since the peak in October 2023, the 30-year/five-year spread has widened by over 0.5 percentage points. The US fiscal situation is a key contributor to this, but we’d argue it’s less about crisis risk and more about debt supply.
Gold’s ‘Debasement’ Trade That Wasn’t
But hasn’t there been a debasement trade over the past few years, where investors are running to gold from fiat-currency assets like bonds in anticipation of a collapse in purchasing power?
This story doesn’t make much sense given that long-run inflation expectations have been entirely flat over the past few years, including the period from the start of 2025 until February 2026 when gold prices more than doubled. Recall that we can decompose benchmark bond yields into the real component (for example, TIPS) and the breakeven inflation rate. 30-year breakeven inflation has hovered closely around a subdued 2.3% for the past five years. This is in line with the Fed’s target for 2% PCE inflation, considering that it usually runs below the CPI inflation upon which TIPS are based. The rise in yields has been entirely on the real side.
One might object that governments will impose “financial repression,” artificially restraining bond yields in combination with ramped-up inflation in order to reduce the real cost of debt. But if this is the case, one can hedge it by shorting nominal bonds and going long TIPS. Such a portfolio (if held to maturity) would yield a return equal to the actual inflation rate averaged over the holding period less the breakeven rate at the time of purchase. Investors worried about debasement should’ve piled into this trade, which would’ve manifested in rising breakevens. Dare we say that investors were snapping up gold more because of price momentum than sober calculations about future inflation?
After a long runup in prices, gold is probably no longer the best way to protect against a surge in inflation and debasement in the US dollar. Meanwhile, breakevens are dirt cheap if you think inflation will run high. Betting on breakevens means going long TIPS and short nominal bonds. One can capture that exposure by rotating a portfolio from nominal bonds into TIPS. Breakevens currently imply the Fed will essentially hit its 2% target over the next 30 years, so if inflation turns out higher, then you’ll come out ahead. If there’s an inflation crisis, then there’s a lot of upside.
Term Premium is Merely Converging Back to Historical Average
Another vantage point is to decompose yields into the expected short-rate over the life of the bond and the term premium. The latter compensates investors for risk. The term premium is not observed directly but is estimated by a model.
The term premium has risen in recent years—helping drive the steepening of the yield curve in 2025—but we don’t think the cause has been rising risks of a debt crisis.
In the event of a crisis, the main risk is a surge in inflation to wipe out the real value of the debt. But as we’ve seen via subdued breakevens, that risk isn’t a significant factor driving bond yields. Without inflation, the venue for wiping out debt in a crisis is outright default. But that’s extremely unlikely for an advanced economy with control of its currency like the US, and we doubt any default risk is being incorporated into the term premium.
Instead, we interpret the rise in the term premium over the past few years largely as a reversion to its historical average. This follows more than a decade in which the term premium was compressed by central bank purchases, along with macro risks being concentrated on the deflationary side for which long-term bonds are an attractive hedge.
Growth in the Term Premium Not Driven by Worsening Debt Outlook
The outlook for the federal debt is concerning, with the Congressional Budget Office projecting federal debt to rise from 98% of GDP at the end of 2025 to 175% by 2056. However, this outlook hasn’t actually changed much in recent years on net. The CBO’s current forecast is right in line with its February 2024 edition.
Astonishingly, the current projected debt/GDP in 2050 is around 20 percentage points below the forecast issued in January 2020 before the pandemic. There are some caveats, but this does weigh heavily against the notion that bad fiscal news has induced investors to begin heavily pricing in crisis risk.
It’s probably true that the growing US debt load has played a role in rising yields via the term premium. But rather than thinking about this as a crisis risk, it makes more sense to think of this as a supply/demand issue.
We can think of investors as existing along a spectrum, and some require a higher term premium to hold long-term debt than others. Government debt (along with AI and other factors) has boosted the supply of long-term bonds. Hence, the higher term premium is needed to bring demand in line with increased supply.
For existing bond investors, our explanation is a benign one, because it implies that the term premium’s rise isn’t mainly driven by rising risk.