When Monetary Policy Surprises Stop Translating

The workhorse framework for measuring monetary policy surprises was developed by Gürkaynak, Sack, and Swanson in a 2007 paper,² using high-frequency movements in federal funds futures around Fed announcements to isolate the unexpected component of policy decisions.

Prior to this work, economists often measured monetary policy surprises using only the unexpected component of changes in the federal funds target rate. The problem with this was that by the early 2000s the Fed had become increasingly transparent, allowing markets to anticipate the next meeting’s decision in advance with a high degree of certainty. So, a richer dataset was needed to measure how markets update views on future monetary policy around key events such as monetary policy statement and minutes releases. To do this, economists began examining high-frequency changes around monetary event windows in fed funds and eurodollar (now the Secured Overnight Financing Rate or SOFR) futures contracts, along with U.S. Treasury yields and other assets.

Subsequent analysis of these surprises found large moves in long-dated forward interest rates in response to monetary policy surprises – a puzzling result since longer-dated interest rates should in theory be relatively insensitive to cyclical economic variation and short-run monetary policy shifts.

Using the San Francisco Fed’s database of high-frequency changes in asset prices around Fed statements and minutes releases, we also find that historically, hawkish surprises – defined as increases in near-dated forward fed funds futures contracts around a monetary event – have been associated with similar yield changes further out the curve.

For example, the 5-year, 5-year-forward real U.S. interest rate over the last decade has tended to rise (fall) by roughly 10 basis points (bps) for every 25-bp hawkish (dovish) monetary policy surprise (see Figure 2). Narrowing the sample to 2016–2022, the sensitivity rises further to a roughly 18-bp move in the 5-year, 5-year-forward for every 25-bp surprise.

In explaining the elevated sensitivity of intermediate- and long-dated forward rates to short-run meeting-by-meeting surprises, economists have argued³ that these changes result from markets viewing the surprise not as random noise, but as informative about more persistent shifts in how the Fed is likely to respond to macro conditions going forward. It’s the evolution of what’s become known as the “Fed reaction function.”

Fed Chair Alan Greenspan began using forward guidance in the 1990s, after a surprise hike in early 1994 kicked off a large and volatile move higher in yields across the curve. But the more prominent use of forward guidance after the global financial crisis (GFC) in 2008–2009 likely solidified this behavior, as Fed communications more clearly conveyed not only the reasoning behind current policy actions, but how policymakers would likely react at future meetings to evolving economic fundamentals.

The Fed’s use of large-scale asset purchases (known as quantitative easing or QE) likely also contributed. Surprises around QE announcements – which coincided with monetary policy rate surprises – likely directly reinforced the relationships between near-dated fed funds futures and yields further out the curve. QE had direct implications for both short- and long-term rates given its use to reinforce forward guidance and compress the premium that investors required to buy longer-maturity bonds. A significant body of research finds that QE programs did indeed reduce yields through these channels.

More recent research from the San Francisco Fed further extends this market reaction-function learning framework in an important direction: It shows that surprises are not only affected by how the Fed communicates, but what markets learn from a surprise depends on the state of the economy at the time.⁴ Specifically, the researchers find that a hawkish move during a demand-driven inflation episode teaches investors a great deal about the Fed’s likely future behavior, while the same move during a supply-driven episode teaches them very little 

The post-2022 relationship between monetary policy surprises and long-term rate moves has declined notably. Unlike the 1990s through the post-GFC period, which was characterized mainly by demand shocks, a series of supply shocks have characterized the post-pandemic macro environment. Kristin Forbes, former external member of the Bank of England’s Monetary Policy Committee, and others find that the role of global shocks has increased sharply since the pandemic, and that these shocks tend to be more supply-focused and have greater volatility.⁵

Combined with broader research literature that has long argued for a more cautious monetary policy response to supply shocks, this suggests that markets have become increasingly sophisticated in distinguishing between different types of inflation – and they price monetary policy surprises accordingly. (Read more about supply-driven versus demand-driven inflation in last week’s Macro Signposts, “Fed Policymaker Comments Raise the Stakes for Inflation Data.”)

What are the implications for today? While one never wants to overinterpret very short-run market behavior, the recent stability of longer-dated forward interest rates to changing near-term policy expectations is consistent with economic and policy shifts.

The sticky inflation over the past year (and more broadly since 2021) has coincided with a series of supply shocks and changing global trends. Energy prices have reacted to geopolitical developments and supply disruptions. Tariffs have raised the cost of imported goods across a range of categories at the same time that rapid investment in AI infrastructure has created bottlenecks in semiconductors, data centers, and related capital goods.

These factors are likely boosting demand as well. Governments, companies, and households race to invest and build resilience against a more uncertain world, while greater wealth – at least at the top of the income distribution – has supported consumption.

The bond market appears more sanguine, however. The fact that the 5-year, 5-year-forward real rate barely moved around the June meeting despite one of the largest recorded surprises in recent history is notable, but not inexplicable. Less forward guidance coinciding with elevated energy market volatility meant that markets discounted the meeting’s signals about durable medium-term shifts in policy rates.

As these macro fundamentals continue to prevail – a higher frequency of global supply shocks with less of a forward guidance policy anchor – then short-dated interest rate volatility is likely to rise. However, based on recent relationships, that may not translate into more volatile longer-dated rates. And this is yet another factor, along with generally elevated yield levels and a more volatile macroeconomic environment, making high quality bonds look attractive on a risk-adjusted basis.