Central Banks Are Failing to Communicate
Confusion over “forward guidance” has left markets guessing and policymakers in a bind. A return to data-dependence is overdue.
It’s complicated.
Photographer: Liu Jin/AFP
Over the past year, the Federal Reserve, the European Central Bank and the Bank of England have all raised interest rates briskly enough to make progress in curbing inflation. To finish the job, they need to get better at explaining themselves.
After the recession of 2008, the communication of monetary policy became especially important. Aggregate demand stayed weak even though central banks had lowered their policy rates to zero. Unable to cut further, policymakers needed other ways to loosen financial conditions. One was quantitative easing: The central banks bought bonds to push asset prices up and nudge long-term interest rates lower. The other, “forward guidance,” focused on communication: By promising to hold short-term rates at zero for an extended period, the banks delivered an extra shot of monetary stimulus.
Both methods had drawbacks. By suppressing long-term rates, QE encouraged investors to “reach for yield” and take bigger risks. Central banks are now wrestling with how to reduce their bond holdings without unsettling financial markets. The need to normalize this aspect of monetary policy is well understood.
Confusion over forward guidance is proving harder to dispel. Investors and analysts examine every scrap of information to predict the course of monetary policy — and they’ve grown accustomed to getting signals from the central banks. These days it’s hints and indications, real and imagined, not promises. The result is a muddle.
Following all three of the central banks’ latest policy announcements, the appetite for commitments was undiminished. Analysts greeted the Fed’s decision to leave its policy rate unchanged as a “hawkish pause” — meaning a pause plus the suggestion of another increase to come. The ECB’s decision to raise rates another 25 basis points was deemed a “dovish hike,” meaning an increase plus a hint that rates won’t be going any higher. Views differed on whether the Bank of England’s pause was hawkish or dovish, though analysts seemed to agree it must be one or the other.
The problem isn’t that financial markets are trying to forecast policy: That’s one of their most important jobs, after all. It’s that the central banks are thought to be providing information they aren’t in fact providing.
All three banks say they’re guided by the data. Fed Chair Jerome Powell, in particular, says conditions are hard to judge thanks to the pandemic, the energy-price shock, and other disruptions, and that things could look different by the Fed’s next policy meeting on Oct. 31 (let alone three months, six months or a year from now). From this it should follow that looking beyond current policy and being “hawkish” or “dovish” about the future makes little sense. None of the central banks knows where rates will need to go.
This crucial message keeps getting lost in the flurry of data, discussion, and rival interpretations among the banks’ policymakers. In the Fed’s case, the so-called dot plot — a “forecast” that isn’t a forecast — makes things worse.
Does it matter? Unfortunately, yes, because if central banks, despite their best efforts, are understood to have a plan for future rates, they might feel pressure to stick to it to avoid surprising investors — even if new information pushes another way. This hesitation, partly self-imposed, compromises monetary policy.
When the policy rate was stuck at zero, policymakers had a dilemma: Making promises on interest rates was risky yet necessary to supply added stimulus. This is no longer true. Interest rates are well above zero. They can be raised or lowered as conditions dictate. Central banks can again afford to be nimble, open-minded and wholly data dependent. They should be — and shouldn’t let investors think otherwise.
More From Bloomberg Opinion:
Want more Bloomberg Opinion? Terminal readers head to OPIN <GO> . Or you can subscribe to our daily newsletter .