Can central bankers talk too much?

IMAGINE YOU are a journalist trying to reassure your bosses that you will hit a tight deadline. What would be more effective: a forceful but brief commitment that you will do whatever is needed to get the job done, which leaves them in the dark on all the things that might go wrong along the way? Or a plan detailing every step you will take—but in which they can spot unnerving risks?

That resembles the choice central banks face as they try to convince financial markets and the public that they will meet their goals. Over the past decade their preference has been clear: the more transparency and detail, the better. In 2011 America’s Federal Reserve began holding press conferences after its monetary-policy meetings. It started publishing the range of rate-setters’ economic forecasts the following year. Across the rich world, forward guidance on the path of interest rates has become part of the toolkit. Central bankers make ever more speeches, bringing once-hidden debates out into the open. Some tweet their views.

The theoretical justification for all the talk is strong. The more markets understand how the central bank will react to events, the better they anticipate future policy. Conditions in financial markets should immediately tighten or loosen in response to economic news, making central bankers’ jobs easier. It is as if setting out your plan to your boss makes it easier to implement.

Today, however, the theory is being tested. The European Central Bank (ECB) meets on October 24th, after The Economist goes to press, amid a very public row about monetary policy. In September the ECB said that it would restart quantitative easing (QE), the purchase of bonds with newly created money, and that it would keep buying assets until inflation picks up from its current level of 1% towards the bank’s aim of close to 2%. Hawks such as Klaas Knot, the head of the Dutch central bank, have been unusually vocal in their dissent. Bond yields, which move inversely to prices, first fell as markets digested the ECB’s guidance. But the bickering has since sent them in the other direction. Market pricing now also reflects expectations of how the political struggle over open-ended QE will play out. Investors have spotted a flaw in the plan.

In America the Federal Reserve may cut interest rates for a third time this year on October 30th. It has been accused by economists at Goldman Sachs, a bank, of constructing a “hall of mirrors” in its communications with markets. The Fed, the argument goes, has this year simultaneously signalled its intentions to bond markets while taking its cues from them. But bond yields are a prediction of what the Fed will do, not an instruction. As a result, the Fed and the markets have entered a pessimistic spiral, while the real economy has been ignored. In its eagerness to be in touch with markets, the Fed has forgotten that it is in the lead.

Central banks everywhere must also work out how to offer forward guidance when facing sharply divergent forks in the road. A trade truce between America and China could transform the economic outlook. A no-deal Brexit could cause chaos in Britain that spills over to the rest of Europe. Telling markets what to expect of policy is much harder when prediction involves choosing between black and white.

Might it help, therefore, for central banks to talk a little less? Microeconomists have long known that ambiguity can have strategic uses. Employment contracts, for example, do not specify every action an employee must take, nor all the obligations of an employer, possibly because it may be better to leave room for either side to punish the other’s bad behaviour. In recent years Bengt Holmström of MIT, who in 2016 won the Nobel prize for economics, has argued that central-bank opacity has its uses in credit markets. Most of the time, he argues, these markets, unlike stockmarkets, are “information-insensitive”—they do not respond much to news. In contrast to stocks, there is no upside for the lender when things go especially well, and default is a remote risk, especially when loans are adequately collateralised. “A state of ‘no questions asked’ is the hallmark of money-market liquidity,” he argues.

In a panic, however, money-markets dry up as the risks loom larger. Lenders find themselves having to scrutinise every transaction. Restoring stability might require a promise that is light on detail, and thus hard to pick apart. At the worst of the euro zone’s sovereign-debt crisis, for instance, Mario Draghi, the head of the ECB, pledged to do “whatever it takes” to keep the single currency safe. Mr Holmström also notes that when the Fed provided emergency lending to banks during the financial crisis, it did not disclose which institutions received support, for fear that any associated stigma could provoke bank runs.

Too much information

Might a similar logic carry across to central bankers’ everyday goals, such as targeting inflation? Inflation expectations, like financial panics, can prove self-fulfilling. Some economists reckon that central banks’ promises to keep inflation low may have become so credible that the public rarely revises its expectations in light of economic news—another case of “no questions asked”.

But the analogy breaks down when it comes to interest rates. Rates vary and markets have to expect something. Central banks might as well steer such expectations. The limits of communication are best seen as the latest round in the decades-old battle between those who want monetary policy to be set by rules, and those favouring discretion. The clearest forward guidance would be a fully transparent algorithm that relates interest rates to economic data. But such a mechanical “reaction function” exists only in economic models. In reality, policymakers have to use their judgment, meaning their decisions are inherently uncertain.

As long as that is true, there is a limit to how much more transparency can make interest rates predictable. And, as the recent experience of central banks shows, talking can have its downsides. It is worth pondering when silence might be golden. ■