Time for a change at the Fed

Plan A didn’t work. Plan B didn’t work. I suggest the Fed get going on Plan C.

The Bureau of Labor Statistics announced last week that the seasonally adjusted consumer price index fell by 1% during the month of October, implying an annual deflation rate around -12%. That’s the biggest monthly drop in the CPI since publication of seasonally adjusted changes began in February 1947. The core CPI (excluding food and energy) saw its first decline in a quarter century.

Does this mean that deflation is now upon us? Mike Bryan argues that despite the indication from the headline and core CPI, actual decreases in prices were not that widespread in October. One of the ways that Bryan proposes to measure this is to order the different components of the CPI by how much the price changed, with those whose price fell the most at the bottom and those whose price increased the most at the top. Suppose you threw out items in the bottom category until you’d eliminated 10% of the total spending by the typical consumer, threw out the 10% in the top category as well, and calculated the sample mean of the remaining 80%. That basket of “typical items” did not change in price during October. If you threw out more than 10% of the bottom and top categories, you’d end up with a slightly positive inflation rate for the month. The graph below shows the resulting average October inflation (reported at an annual rate) when you throw out the bottom x% and top x%, plotted as a function of x. The value x = 0 corresponds to the usual sample mean (the headline CPI itself), while x = 50 corresponds to the sample median.

trimmed_cpi_nov_08.jpg

Alternatively, we might look at the cumulative inflation readings over the last year, rather than the single month of October in isolation. These suggest a 3.7% inflation rate from the CPI and 2.2% for the core CPI. Trimming x = 8% from the bottom and top yearly changes gives an annual trimmed mean of 3.0%.

cpi_nov_08.gif

So maybe everything’s OK? I think not. Two forward-looking indicators are profoundly troubling. First, the yields on inflation-indexed Treasuries for medium-term maturities are actually higher than those for regular Treasuries. If taken at face value, that means investors anticipate an average deflation over the next 5 years at a -1.29% annual rate. Perhaps one might dismiss this as another indication that the usual arbitrage activity is completely absent in current markets, so that the nominal-TIPS spread is no longer a meaningful indicator.

exp_inf_nov_08.gif

But a second and equally troubling suggestion of expected deflation is the extremely low yields on short-term Treasury bills. Again there may be those who interpret this not as a harbinger of deflation but instead as a reflection of the astonishing (and equally frightening) flight to quality that we have been witnessing.

Even if you don’t interpret the October CPI, TIPS yields, and nominal T-bill yields as warning flags of deflation, they nonetheless raise what is to me the core question: If the Fed wanted to use monetary policy to stimulate the economy at the moment, as I believe it should, what would it do?

The traditional answer would be to lower the fed funds target. But surely further cuts in the target rate can accomplish nothing in the current environment. The effective fed funds rate and 3-month T-bill rate are already more than 50 basis points below the current target, so cutting the target by 50 basis points is supposed to accomplish what, exactly? If you’re counting on another couple of rate cuts as the last arrows in your quiver, I’m afraid to report that continuing this battle rather desperately calls for a Plan B.

dff_dfedtar_nov_08.png

Nor should there be any enthusiasm for yet another lending facility. What started as a few billion dollars is now in the trillions, and credit spreads continue to widen. If we keep doing the same thing we’ve been doing for the last year and a half, why should we expect any different result?

So here’s my suggested Plan C. The goal of monetary policy should be to achieve a core inflation rate of 3.0% (at an annual rate) over the next 6 months. That’s something that can be accomplished without rate cuts or lending facilities, and here’s how.

Step 1 is for the FOMC to form a clear determination that a 3% core inflation rate is indeed their immediate goal. If you hope to get somewhere, it’s a good idea to start with a plan of where you’re trying to go.

Step 2 is to communicate the goal to the public. Bernanke and Kohn should state clearly that they’re worried by the October fall in the CPI, that they see a danger of too much slack in the economy developing, and that they will now be adopting quantitative easing with the goal of preventing further declines in the overall price level.

Step 3 is to start creating money and use it to buy up assets until the goal set out in Step 1 is achieved. What sort of assets? My answer here would be the exact opposite in philosophy of the kind of purchases and loans that the Fed has been implementing over the last year. The Fed has been trying to sop up the illiquid assets that nobody else wants. But I think what the Fed should be doing is instead acquiring assets of a type that would allow it to quickly reverse its position if a sudden shift in perceptions causes inflation to come in above the intended 3% target. The Fed can’t afford to dump the illiquid securities it’s been taking on recently, and that leaves it with substantially less flexibility to ease out of an expansionary policy once it starts to be successful. My goal would therefore be to buy assets for the Fed that won’t lose their value with a reversal of expectations and whose sell-off by the Fed wouldn’t be itself an additional destabilizing force.

What specifically would such assets be? I’d start with those clearly undervalued TIPS. Next I’d buy short-term securities in the currencies relative to which the dollar has been appreciating. Here again if the Fed has to sell these off in a sudden change in perceptions, the Fed will have both made a profit and, by selling, be a stabilizing force. If we’re still seeing no improvement, the Fed can start to buy longer-term Treasuries.

What if the policy is unsuccessful, and we still get severe deflation despite the 3% inflation target? In some ways, that’s the best outcome of all, since, as I explained previously, in that scenario the Fed has solved the nasty problem of all that debt owed by the Treasury.

Targeting inflation is not just another arrow in the quiver; it’s a bazooka, at least for purposes of preventing deflation. Time to take aim and fire.


Originally published at Econbrowser and reproduced here with the author’s permission.

One Response to "Time for a change at the Fed"

  1. Guest   November 27, 2008 at 10:45 am

    After 75 years of inflation, which has apparently worked out ever so well, God forbid we should have a year or two of falling prices. Let’s reward the deadbeats and speculators, and punish the frugal and sensible among us, who are admittedly a minority.