SHOPPING FOR DEFLATION / THE FINANCIAL TIMES

Shopping for deflation

By: Brendan Greeley


Alphaville used to spend a lot of time listening to people in markets talk about the Dallas Fed's trimmed mean PCE and the Cleveland Fed's median CPI. These are alternate ways to measure inflation. Both are completely legitimate and interesting in their own way, but in early 2015 they also happened to show a higher inflation rate than the Federal Open Market Committee's benchmark measure, core PCE. A generous interpretation of these conversations would conclude that people were really, really interested in finding the best way to measure inflation.

It's also possible, however, that they were just shopping for inflation.

People want the Fed to do certain things, but in theory the Fed doesn't pay attention to people. It pays attention to inflation. So people find the inflation measure that suggests the Fed policy they'd prefer, and then talk about that instead. That's what was happening five years ago, when the Fed's policy rate was still on the floor, and Alphaville had to consider the median CPI and the trimmed mean PCE:  





We thought about this as we read an argument from Stephen Moore, Trump's new nominee for the Fed, that the FOMC should be watching changes in the price of oil and wheat and aluminum. "The Fed’s goal should be to avoid excessively loose or tight money by seeking stable commodity prices," he wrote in an oped for the Wall Street Journal. This would be a radical shift, one no one at the Fed has talked about in decades.

But: Commodity prices have dropped sharply since October, and by some measures are now technically deflating. Central banks respond to deflation first with panic, then with dramatically accomodative monetary policy. So again here, a generous interpretation is that Mr Moore is really, really interested in finding the best way to measure inflation. It's also possible, however, that he's just shopping for a way to say the word "deflation."

"No serious monetary economist thinks it's a good idea to target commodity prices," says Mark Gertler. Commodity prices can provide some information, he says, and the Fed certainly takes them into account, but they're not a great signal of future inflation. Mr Gertler, a macroeconomist at NYU who's advised the New York Fed, says there was a time when economists considered using commodity prices as a benchmark for monetary policy, but that time was several decades ago, and they abandoned the idea.

"The problem," says Gertler, "is commodity prices are determined in these highly competitive markets that are subject to all sorts of shocks." Some prices are "sticky" -- it's hard to get them to change. Ice cream prices, for example: very sticky. Commodity prices are the least sticky of all. Wheat and crude and aluminum and natural gas are sold by sophisticated traders in large volumes on global markets. And because those markets are global, shocks anywhere in the world can shift prices: droughts, coups, embargos, cartels, tariffs.

Here's what that looks like. We took price changes to a commodity index and ran it against the Fed's preferred inflation measure, core PCE:




It's... volatile. And becoming more so.

There are lots of different commodity indices. We chose the Producer Price Index for Intermediate Demand -- what companies in the US pay for stuff that they turn into other stuff. Processed and unprocessed fuels are weighted together at 40 per cent of the index, because it just takes a lot of gas and lubricants to make things. Even the prices of copper and aluminum reflect the price of energy. "You need a lot of energy to go and look for copper or to create aluminum" says Francisco Blanch -- he runs global commodities research for Bank of America Merrill Lynch:

Aluminum is solid power, by the way... it takes 15 megawatt hours to create one ton of aluminum, plus a couple tons of alumina. It's hugely energy intensive.

And so what you're actually looking at up there is a history of shocks to global petroleum prices. Demand rose from China in the early 2000s, with an immediate recovery after the recession. Prices collapsed after the innovations of fracking, followed by a price war with Saudi Arabia, and even more efficiency from US frackers. Commodity price indexes swing on a bunch of stuff the Fed can't control, and they swing too quickly for the Fed to react.

Mr Blanch points out that commodities are becoming a smaller part of the consumer "basket" -- what people buy. With fracking, there's more certainty of a return every time a company drills a hole, so the price of oil no longer shows the risk of a dry hole. Over time, he said, this has been dragging crude prices down, and will drag it lower in the future.

That's all on the supply side of energy. There will be changes to the demand side as well. "We are transitioning to a very different economy in the next ten years," says Mr Blanch. "And we're going to see a meaningful increase in fuel efficiency rates in the next five years."

So: energy and in particular oil are a huge part of any commodities index. They are incredibly volatile now, and will be a smaller part of any consumer basket in the future. "Most economists agree, what you want to target is not just a fraction of the basket of commodities," says Mr Gertler. "You want to target a suitably weighted average of all of the goods and services, not just the component that's most volatile, that's nuts."

Again, Mr Moore could be really interested in how best to measure inflation. Plenty of people are. The Cleveland Fed is holding a whole conference on the subject in May. Perhaps Mr Moore is worried in particular about supporting commodity prices for oil-producing states like Texas, or soyabean-producing states like Iowa.

Or perhaps he's just shopping for deflation.


Alphaville called Mr Moore last night and this morning. We'll update the post if we hear back.