For years, I’ve argued that central banks fundamentally misunderstand their own monetary policy transmission mechanism. They believe the federal funds rate drives nominal variables, when increasingly it’s the other way around.
In this blog post, I present empirical evidence through a structural VAR analysis spanning 1973 to 2025 that demonstrates this complete reversal of causality.
The results in my view are nothing short of remarkable. The federal funds rate – the Fed’s supposed primary policy instrument – has transformed from a leading indicator that once shaped economic outcomes to a lagging variable that merely follows market-determined nominal developments.
This analysis stems from my repeated attempts over the past few weeks to re-estimate interest rate policy rules for both the Fed and the ECB. Perhaps intellectually dulled, I’ve been “playing along” with the New Keynesian game of estimating Taylor rules.
Paradoxically, this “betrayal” of market monetarist thinking has only strengthened my conviction about the need to examine causality in the monetary transmission mechanism.
The conclusion: we market monetarists have been right, and we may have convinced both the Fed and the markets – even though no one has officially acknowledged it. This analysis is presented below.
The Methodology: Letting the Data Speak
I’ve conducted a structural Vector Autoregression (VAR) analysis using monthly data from January 1973 to June 2025, examining five key macroeconomic variables:
- Trade-weighted US Dollar Index (nominal, combined series)
- M2 Money Supply
- Personal Consumption Expenditures (PCE) — my proxy for NGDP
- Federal Funds Rate (FFR)
- 10-Year Treasury Yield
The analysis employs a Hodrick-Prescott filter (λ=129,600) to extract cyclical components after log-transforming the level variables.
I then estimate a VAR(3) model and conduct Granger causality tests to identify which variables lead and which follow.
The Smoking Gun: Federal Funds Rate as Follower, Not Leader
Look at the FFR row in the table below – it’s a sea of statistical insignificance!
The federal funds rate fails to Granger-cause any variable except the 10-year yield at conventional significance levels. Meanwhile, look at the FFR column – it is caused by nominel spending (PCE) and 10 year yields!
Figure 1 below adds up these results.
The federal funds rate exhibits:
- Low out-degree: Causes only 2 variables significantly (M2 and 10-year yield)
- High in-degree: Is caused by 3 variables (PCE, 10-year yield, and indirectly through other channels)
- Net causality score of -1: Negative, indicating follower status
But the real story is PCE’s dominance:
- Highest out-degree: PCE causes 4 out of 4 variables
- Moderate in-degree: Only caused by 2 variables (M2 and Dollar)
- Net causality score of +2: The clear leader of the system
This reveals the complete inversion of monetary transmission. PCE – our proxy for nominal spending -drives everything including the Fed’s own policy rate.
The federal funds rate doesn’t lead nominal developments; it follows them. Meanwhile, PCE acts as the system’s true anchor, causing movements in money supply, exchange rates, interest rates, and even monetary policy itself.
A true policy instrument should lead the system as PCE does. Instead, the FFR is relegated to follower status – validating the nominal path that PCE has already determined. The supposed conductor of monetary policy is actually just another instrument in PCE’s orchestra.
The Historical Transformation: The Rise and Fall of FFR Leadership
Perhaps most striking is how the federal funds rate’s role has completely reversed over our sample period.
Table 2 and figure 2 provide an overview of these changes.
In the 1970s-1990s, the FFR led the dance. It Granger-caused movements in money supply, nominal spending, exchange rates, and bond yields. This was the era when the Fed moved first and markets reacted.
But something fundamental shifted around 2000 (I suspect it actually started with the introduction of Treasury Inflation Protected Securities (TIPS) in 1997 – thank you Bob Hetzel).
The federal funds rate increasingly became a follower – reacting to developments in nominal variables rather than causing them.
By the 2010s, the transformation was complete: the FFR had become almost entirely endogenous.
Variance Decomposition: Who Explains Whom?
The Forecast Error Variance Decomposition (FEVD) at 24 months provides another perspective on this transformation:
Table 3 reveals a striking truth: the Federal Funds Rate (FFR) explains almost none of the variation in key nominal variables. It accounts for just 1.7% of dollar movements, 4.8% of money supply, and 2.4% of PCE — our proxy for NGDP.
By contrast, PCE alone explains over 20% of the FFR’s own variance, with the 10-year yield contributing another 6%. In other words, the Fed’s policy rate follows the nominal economy; it doesn’t lead it.
PCE and M2 exhibit far more explanatory power than the rate central banks claim to control. The conclusion is unavoidable: interest rate policy is no longer a tool – it’s a follower. The FFR reads the temperature of the economy but no longer changes it.
Why Did the FFR Lose Its Leadership?
The transformation from leader to follower wasn’t accidental – it emerged from the very evolution of modern central banking itself.
The Taylor Rule trapped the Fed in endogeneity. Once markets understood that the Fed systematically responds to inflation and output gaps, they began anticipating these responses.
The federal funds rate became endogenous by design, transforming the Fed from active leader into predictable follower of economic conditions.
Forward guidance paradoxically accelerated this transformation. Greater transparency and communication, intended to enhance policy effectiveness, instead completed the endogeneity circle.
When central banks telegraph exactly how they’ll react to future data, they’ve effectively announced that their policy rate is a dependent variable, not an independent force.
The 2008 crisis delivered the coup de grâce. When rates hit the zero lower bound, the Fed’s resort to quantitative measures exposed the federal funds rate’s fundamental limitations. Monetary aggregates briefly assumed leadership during this period, shattering any remaining pretense that interest rates represented an all-powerful policy tool. The emperor wasn’t just naked—he was powerless.
Today’s high-frequency financial markets represent the final stage of this evolution. Information flows instantly, expectations adjust continuously, and nominal variables shift before the FOMC can even convene.
Markets no longer wait for Fed decisions; they anticipate, price in, and effectively determine where rates must go. The Fed arrives at each meeting to find markets have already done the heavy lifting, leaving only the ceremonial announcement of what everyone already knows must happen. By the time the FOMC meets, markets have already moved nominal variables to where they “should” be, and the Fed merely validates these moves.
This isn’t a bug in the system – it’s the inevitable feature of a transparent, market-based monetary regime operating without explicit nominal anchors. The Fed engineered its own irrelevance through the very reforms meant to enhance its effectiveness.
What does it mean for monetary policy when the federal funds rate has transformed from leader to follower?
Our VAR evidence reveals that nominal variables and market rates drive the federal funds rate, not vice versa. The FOMC meetings have become theatrical performances – committee members debate quarter-point adjustments whilst markets have already determined where rates must go based on nominal conditions.
The federal funds rate now functions as a thermometer rather than a thermostat. It reads the temperature of nominal conditions but cannot change them.
Since the FFR follows rather than leads, actual monetary influence must operate through expectations. Markets coordinate around implicit nominal targets, and the Fed’s role has been reduced to validating these market-determined paths.
This reality is deeply inconvenient for the monetary policy establishment. Central bankers imagine themselves as maestros conducting the economic orchestra, yet our analysis demonstrates they’re dancing to the market’s tune. The Federal Reserve believes it conducts policy through rate adjustments; the data shows it merely reacts to developments beyond its control.
Conclusion: The New Monetary Reality
The transformation from leader to follower is complete and irreversible. Rather than attempting to restore the federal funds rate’s lost leadership – a futile endeavour in modern financial markets – we must design institutions that acknowledge how monetary systems actually operate.
This follower status isn’t inherently problematic. Markets excel at processing information and coordinating expectations. The danger lies in maintaining the illusion of active policy whilst passively rubber-stamping market outcomes. Every quarter-point debate at the FOMC perpetuates this charade, wasting credibility on decisions that markets have already made.
When nominal expectations drive outcomes and policy rates follow, the path forward requires embracing market-based nominal targeting with the credibility that shapes those very expectations. As I’ve long argued, effective monetary policy works through nominal expectations, not mechanical rate adjustments.
This VAR analysis confirms that even the Fed’s primary tool has become subordinate to the market forces it once commanded.
The emperor has no clothes. It’s time we designed monetary institutions that acknowledge this reality.
Postscript: When Fiscal Dominance Destroys Market-Based Nominal Targeting
The VAR analysis reveals that markets now effectively coordinate nominal expectations, with the Fed merely following.
This could be benign – markets are rather good at implicit NGDP targeting. But as I’ve warned in my latest post “The Fiscal Dominance Trap,” we’re approaching the point where this market leadership transforms from feature to catastrophic bug.
The unpleasant monetarist arithmetic is brutal: with US federal debt at 119% of GDP, each percentage point rise in rates adds $350 billion to annual interest costs. The fiscal dominance threshold – where raising rates to fight inflation becomes self-defeating – sits around 7-8%. Above this level, higher rates create larger deficits, more debt issuance, higher term premiums, and an explosive feedback loop that no amount of Fed credibility can break.
We’re witnessing the Carter-Burns-Miller nightmare in real time, but with constraints that make the 1970s look like a monetary picnic. When Carter pushed out Burns for Miller in 1977, federal debt was 35% of GDP. Volcker could take rates to 20% because the fiscal arithmetic still worked. Today, with debt at 119% of GDP, the Volcker solution is mathematically impossible.
The terrifying irony is that just as markets have learned to coordinate nominal expectations efficiently, fiscal dominance threatens to turn this virtue into catastrophe.
When velocity accelerates – as it did 4% within months of Miller’s appointment in 1978 – markets will still lead, but they’ll be leading us into an inflationary spiral. The Fed will still follow, but it will be following markets as they price in fiscal insolvency.
The window for avoiding this fate is maybe measured in months, not years. Either we achieve fiscal adjustment of 5-6% of GDP – politically impossible in any democracy – or we learn that even the most sophisticated market-driven monetary arrangements cannot overcome basic arithmetic. When governments spend beyond their means indefinitely, something must give. What gives is always the currency, and this time, there’s no Volcker waiting in the wings to save us.
Bibliography
Christensen, L. (2011). “Market Monetarism: The Second Monetarist Counter-Revolution.” Working Paper.
Granger, C.W.J. (1969). “Investigating Causal Relations by Econometric Models and Cross-spectral Methods.” Econometrica 37(3): 424-438.
Hodrick, R.J. and Prescott, E.C. (1997). “Postwar U.S. Business Cycles: An Empirical Investigation.” Journal of Money, Credit and Banking 29(1): 1-16.
Sims, C.A. (1980). “Macroeconomics and Reality.” Econometrica 48(1): 1-48.
Sumner, S. (2012). “The Case for Nominal GDP Targeting.” Mercatus Center Research Paper.
