Fed says too little, RBI is too predictable. Both may have a problem

On July 29, US Fed left interest rates unchanged at 3.50-3.75%. Eight days later on August 5, RBI kept the rate unchanged at 5.25%. The second decision was more or less ignored; the first pushed 30-yr US treasury yields to their highest since 2007.

There was, as usual, no dissent in RBI’s Monetary Policy Committee (MPC). Traders had priced this in weeks in advance, not for lack of a case (June inflation hit 4.38%, an 18-mth high, with RBI’s forecast pointing to 5.1%), but because investors know MPC well enough to anticipate it. Such predictability can be a compliment for a central bank. It can also be a trap.

Also Read| Fed may talk less and let markets lead. That could spell more volatility for India


Nobody takes a 20-yr mortgage against this month’s rate. People act on where they expect rates to sit for years. A central bank that can shift those expectations barely needs to touch the rate at all.

Former Bank of England (BoE) governor Mervyn King put it well. Imagine a central bank so quick, it neutralises every shock on arrival.

Inflation would sit flat, rates would jump about with no visible relation to it, and an observer would conclude the bank was achieving nothing. He would be wrong. And if everyone believed him, that belief would destroy the stability producing it.

King called it the ‘Maradona Theory of Interest Rates’. In the 1986 World Cup, Diego Maradona famously ran 60 yards through the England defence in almost a straight line, beating 5 players without going round any of them, because each expected a swerve and moved first. The defenders were correctly ‘pricing’ what he would do if the space closed. It didn’t.

Markets in 2002 were likewise not wrong about BoE, which held rates all year and still hit its target. They were pricing hikes conditional on a recovery that never arrived. An expectation never triggered is not a false one. Which is why a central bank can bank it without ever spending it.

New Fed chair Kevin Warsh is restoring the mystique central banking spent 30 yrs dismantling. Policy statements have shrunk from 341 words to 130. Forward guidance is gone. He declined to submit his own rate projection, describing colleagues’ forecasts as written ‘with pencils [with] big erasers’. Economists called his last press conference internally contradictory.

But Warsh’s reasoning is better than reviews suggest. The Fed is not omniscient. Pretending otherwise is more dangerous. Dot plots dress up guesswork as arithmetic, spoonfeed markets, and merely echo your guidance back at you. Better to let markets discover the price. Fine, but discover what? Without a stated rule, they fall back on behaviour, and a debutant chair has none. So, they forecast the man.

A central bank can tell you two things, and Warsh has scrapped both: uPredictionRates will be 4% by December. That is hostage to every data release between now and then, and Warsh right to distrust it.

v Condition What would make us move. It promises no outcome, which is why it’s cheap to state, though not free to maintain. Only the first deserved to go.

Because Warsh is untested, investors price in a premium for guessing him wrong. Within the hour, 2-yr yields fell as traders cut odds of a nearterm hike, while the 30-yr hit a 19-yr high. Inflation fears alone cannot produce that shape. Only uncertainty about the reaction function can. Markets became less worried about policy than about policymaker.

The obvious defence: if cheap borrowing rested on false comfort, dearer borrowing is simply a more accurate price — correction, not cost. True, but two surcharges sit in that number, and only one is honest.

Charging extra because nobody knows what prices will do is information, and firms should pay it. Charging extra because nobody knows what Fed will do about prices is a fee for its reluctance to describe itself. Long yields price everything else, so that fee lands on firms insuring against what they used to plan around — and on workers not hired because the plan was shelved.

Former European Central Bank chief Mario Draghi’s ‘whatever it takes’ was cryptic about the means and unmistakable about resolve. It worked because credibility was already banked. Emphasising how little you know is a curious way to build the same account. The law requires only two explanations a year. Everything else is voluntary. Warsh is running down the balance.

None of which lets RBI off. An MPC can reasonably decline to raise rates against an oil shock it cannot control. What it cannot decline to defend is the condition: to raise if the shock spreads into second-round effects, which is precisely what Sanjay Malhotra flagged weeks earlier.

The objection, then, is not to the decision but to the inference. What investors have absorbed is not the condition but a blunter summary: this lot don’t raise rates. And the more firmly markets assume no rise is coming, the more damage an eventual rise does. Itself a reason to postpone it again.

Both banks have erred from opposite ends. When a central bank won’t say what would make it move, markets guess from its track record. And a track record yields either nothing, if it’s new, or a lazy summary, if it’s long. Warsh’s markets have nothing to read. Malhotra’s have stopped reading