Reuters logo
COLUMN-Fed must pay more attention to oil prices: John Kemp
September 13, 2012 / 11:27 AM / 5 years ago

COLUMN-Fed must pay more attention to oil prices: John Kemp

By John Kemp

LONDON, Sept 13 (Reuters) - Concern about triggering another rise in oil prices is unlikely to deter the Federal Reserve from embarking on another round of quantitative easing.

Officials believe the stimulus from lower interest rates will outweigh the drag on growth from higher fuel costs, as my colleagues Jonathan Spicer and Matthew Robinson have explained (“Benefits of QE3 seen outweighing oil sting” Sep 12).

In any event, the Fed sees rising oil prices as a microeconomic phenomenon, a shift in relative prices which has only a temporary impact on the headline inflation rate, rather than a macroeconomic problem that signals capacity constraints and deeper inflationary pressures in the economy.

“Accommodative monetary policies ... both traditional and nontraditional, have provided important support to the economic recovery while helping to maintain price stability,” Fed Chairman Ben Bernanke told his audience at Jackson Hole.

“Despite periodic concerns about deflation risks, on the one hand, and repeated warnings that excessive policy accommodation would ignite inflation, on the other hand, inflation (except for temporary deviations caused primarily by swings in commodity prices) has remained near the Committee’s 2 percent objective and inflation expectations have remained stable” he explained.

But the chairman’s analysis understates the impact of oil prices on economic performance as well as their significance for monetary policy.

Crude oil prices have both microeconomic and macroeconomic aspects. The price of crude and refined products is set in competitive markets by microeconomic forces of supply and demand. But price-setting in futures markets is also influenced by expectations, which have a macroeconomic component.

More importantly, spending on gasoline and other oil products comprises a large enough part of household and business spending that it has inevitable spillovers to the rest of the economy. Changes in oil prices have an impact on real incomes and not just relative prices.

It is therefore not safe to ignore oil prices when setting monetary policy. The Federal Reserve, and other central banks, needs to do much more to integrate oil into its model of the economy to improve its forecasting performance and provide more accurate estimates of how various policy options will affect the economy.


In the last twelve months, the cost of filling up at the pump has been running at around $400 billion per year, according to the Bureau of Economic Analysis (BEA), amounting to about 3.7 percent of all personal consumption expenditure.

The amount spent on motor fuels is small compared with housing (17 percent of total personal consumption expenditures) and healthcare (16 percent) but still more than on any other item other than financial services (7.5 percent), food (6.5 percent) and spending on eating out and drink (5.4 percent).

Consumer spending on motor fuels is much more variable than on any other category of expenditure, reflecting the huge variability in oil prices.

The BEA compiles estimates for spending each quarter at seasonally adjusted annualised rates. The overall seasonally adjusted annualised rate of personal consumption expenditure has varied by about $56 billion from one quarter to the next, taking the standard deviation of quarterly spending rates.

But the rate of spending on gasoline and other motor fuels has varied by almost $23 billion. No other category comes close. Annualised spending on new motor vehicles has varied by around $14 billion. Electricity and gas by about $6.5 billion. Food and drink for home consumption, hospital fees, and spending on financial services have all tended to vary by less than $5 billion.

The relatively high share of household spending on motor fuels, combined with its extreme variability, mean that gasoline prices have a unique macro impact. Oil prices have a comparable impact to tax changes and interest rates.


There is considerable evidence that previous rounds of quantitative easing (QE) have been associated with, and perhaps caused, broad-based increases in the price of commodities. Policymakers have acknowledged the possibility of a link, but insist the impact is one-off and temporary.

However, that is a rather incomplete picture. In each case, commodity prices rose until the effects of QE wore off and the economy slowed. In fact, many observers have pinned part of the responsibility for the slowdowns on the rising cost of fuel and food, which squeezed household spending, maximised uncertainty, and offset some or all of the early boost from QE policy actions.

In each case, commodity prices stopped rising because QE failed to have a sustained impact on the economy. The policy only had a temporary effect on commodities because it largely failed to live up to expectations. If QE had actually succeeded in putting the economy onto a stronger growth trajectory, the price of fuel and food would almost certainly have continued to rise, intensifying the income squeeze and with it the hit to household and business confidence.

Any prediction that QE will have only a temporary impact on commodity prices, and macro performance, is therefore conditioned on the expectation that the policy will fail. If it eventually succeeds, the impact on oil and other commodity prices would be much larger.


Thinking about forecasts conditionally helps explain the huge range of views about whether or not QE will result in rampant inflation.

For much of the last four years, inflationists have worried low interest rates and money creation would result in a surge in prices, as more money chases the same volume of goods and services.

In practice, however, inflation has remained mostly tame, especially in the United States as reserves have piled up idly in the banking system.

So were the inflationists just plain wrong, as Nobel Economics Laureate Paul Krugman has argued openly and Bernanke appeared to imply?

Not necessarily.

Inflationists predicated their forecasts on the assumption QE would actually work, with extra reserves moving out of the banking system and shifting the economy onto a faster growth path.

In practice, QE has failed to have the impact its advocates hoped. The result is no inflation, but not much job creation and growth either.

The anti-inflationists were implicitly forecasting QE would not have much effect on the real economy and the liquidity trap would continue to impound excess reserves in the banking system. QE would not prove inflationary but it would not be much help either.

The conditional approach to thinking about forecasts helps explain why the models used by the Fed and other central banks have proved so hopelessly wrong over the last five years.

The Fed and other central banks treat oil prices as largely exogenous. Prices are mysteriously determined by non-monetary forces. They then enter the Fed model via the forward curve.

Understood properly, however, the Fed should treat oil prices as a partly macroeconomic variable. The Fed should be able to estimate how much its conventional and unconventional monetary policies might move oil prices, and how that change might feed back to the economy through changes in real household and business incomes.

It needs to do something similar with liquidity preference and the demand for cash balances and bank reserves.

The current generation of models do not cope well with the sort of complex conditionalities and indeterminacies raised by oil prices and liquidity preference. There is no easy way to incorporate them. But as long as the Fed and other central banks continue to treat them as exogenous, the models will remain next to useless as forecasting tools.

0 : 0
  • narrow-browser-and-phone
  • medium-browser-and-portrait-tablet
  • landscape-tablet
  • medium-wide-browser
  • wide-browser-and-larger
  • medium-browser-and-landscape-tablet
  • medium-wide-browser-and-larger
  • above-phone
  • portrait-tablet-and-above
  • above-portrait-tablet
  • landscape-tablet-and-above
  • landscape-tablet-and-medium-wide-browser
  • portrait-tablet-and-below
  • landscape-tablet-and-below