Correlation isn’t Causation: Commodity Prices and Interest Rates

Correlations interpreted as causation.

This represents one of the most dangerous premises in making predictions.

An article in the Bloomberg suggests that the low interest rate environment may continue to fuel a rally in commodity prices.

From Bloomberg,

Commodities will resume a rally after their worst month in a year as the Federal Reserve keeps its benchmark interest rate at a record low following the end of a $600 billion asset-purchase program, said HSBC Holdings Plc.

The Standard & Poor’s GSCI Total Return Index of 24 commodities tumbled 6.9 percent last month, the first loss since August and the biggest drop in a year on concern growth may slow. The gauge rose 16 percent this year through April after rising 9 percent in 2010 and 13 percent in 2009 as output trailed demand.

Fed Chairman Ben S. Bernanke has indicated the bank won’t remove stimulus immediately after the second round of so-called quantitative easing concludes this month. The central bank will hold its policy rate in a range of zero to 0.25 percent until the year-end, a Bloomberg survey of 72 economists showed.

The CHART OF THE DAY tracks the S&P GSCI Total Return Index and the Federal Reserve’s target rates in the past three decades. Commodities typically move counter to borrowing costs and the GSCI index has advanced 33 percent since the end of December 2008, the month the Fed cut its benchmark rate amid recession.

Though I would agree that commodity prices could rally, partly coincidental to an artificially suppressed interest rates climate over the interim, I’d say that such correlation does not always hold.

Again reference point matters.

The framing of the Bloomberg chart begins in the 1980s and spans through 2010. This is the epoch of globalization.

However, prior to 80s, as the above chart from economagic.com shows, the correlation was much about rising commodity prices (CRB-Reuters-blue) along with rising interest rates (Fed Funds rate-red, 10 year Treasury yield-green), i.e. 1960s until 1980.

That’s because this era marked the ‘guns and butter’ policies of the US (Vietnam War and entitlement programs as Medicare and easy money policies of the US Federal Reserve).

In other words, this period characterized the economic environment known as stagflation (high inflation, high unemployment-which wasn’t incorporated in the traditional Keynesian model).

Bottom line: Correlations of variables highly depends on the underlying forces which drives them. Occasionally, some correlations reflect on the causal nexus, but for most instances they don't.

In my view, the current level of interest rates reflects on the du jour actions of monetary policies. And this seemingly benign trend may radically change, which subsequently, could upset the balance of the low interest rate-high commodity price correlation premise.