Skip to main content

This post appeared earlier today on RealMoney. Click here for a free trial, and enjoy incisive commentary all day, every day.

Image placeholder title

Experience is a great teacher if you are good at forgetting things. I can recall the crude oil price crash of December 1985 to February 1986 and how bonds traded tick-for-tick with crude oil, and then how stocks traded tick-for-tick with bonds. A little over a year later, we changed the chain of causation from the yen to bonds to stocks.

Now the yen is good for a few laughs every year when the 13-year-old yen carry trade, last visited here in

September 2007

, gets unwound globally in a two-minute period, and bonds pretty much ignore crude oil, and stocks and bonds more often than not trade inversely. If old dogs refuse to learn new tricks, their owners stop taking them to the vet's.

So imagine my amusement when the Ritalin-for-lunch bunch starts hyperventilating about how lower crude oil prices -- and that's "lower" in a very relative term; the present price level would have been shockingly high even eight months ago -- will pull this market higher single-handedly.

Shocks for the Long Term

First, let's stipulate the obvious: Stocks and crude oil have been correlated negatively over the past year. The operative part of that sentence is, "over the past year." Let's take the long-term history of crude oil and the

S&P 500

and divide it along two very obvious fracture dates, May 6, 2003, when the

Federal Reserve

bravely began its war on deflation, and Aug. 17, 2007, when that war entered its surge phase.

S&P 500 and Crude Oil:
Three Regimes Since 1983

Click here for larger image.

Source: Bloomberg

Three sets of ordered pairs emerge from that division and are mapped above. The January 1983-May 2003 period, marked in blue, has an essentially random relationship. The May 2003-August 2007 period, marked in red, has a strong positive relationship; both stocks and crude oil rose simultaneously under a combination of strong global growth and generally stimulative policies. The August 2007-August 2008 period, marked in green, has a negative relationship; stocks sank under the weight of the credit crunch while crude oil prices doubled.

Restated, we have one year out of almost 26 years of data in which we can demonstrate a negative relationship between crude oil and the S&P 500, and yet the fast-trigger crowd thinks this is all you need.

Rolling Three-Month Correlation of Returns
S&P 500 and Crude Oil Since January 1983

Click here for larger image.

Source: Bloomberg

TheStreet Recommends

We can rearrange the data in the chart above to a long-term map of correlation of returns between the two markets.

The present three-month correlation of returns is one of the most negative on record, surpassed only by the first Persian Gulf War and approached a few times thereafter.

But take a look at how many times the rolling three-month correlation of returns is positive; it includes the March-June 2008 period following the

Bear Stearns

panic low. More than 45% of the daily rolling three-month correlations of returns in the chart at right are positive.

Many Unhappy Returns

Now let's steal a page from the "commodities as a financial investment" crowd and shift the focus from the spot price of crude oil to its total return. This includes the roll yield, which can be negative as well as positive, of selling one future or set of futures and buying another, along with the return earned on the collateral deposits. We can compare this series as calculated by Dow Jones-AIG (idle thought: Why do I get much of my index information from firms that regularly lose $5 billion to $10 billion a quarter?) to the total return of the S&P 500 (second idle thought: Why do I get the rest of my index information from firms that slap AAA ratings on things with the word "pile" in the description?).

Is This a Long-TermDeterministic Relationship?

Click here for larger image.

Source: Bloomberg

The long-term relationship between the two total return series is highly unstable but weakly positive. The largest feature on the S&P 500 line since the data began in 1991 is the tech boom-and-bust between 1995 and 2002, and that occurred independently of crude oil. The largest features on the crude oil line have been the two rallies of 2002-2006 and August 2007-July 2008, and as noted above, only the second one is coincident with a downturn in stocks. Negative causality from crude oil to stocks cannot be demonstrated.

Industry Group Impact

Finally, let's return to an analysis first introduced in

February 2005

on assessing the impact of factor prices on S&P industry groups, and add the twist introduced in

November 2006

on weighting these factors by the groups' representation in the index. We can construct a table of groups both helped and hurt by rising crude oil prices at a 90% confidence interval.

Crude Oil Beta-Weighted Impacton S&P 500

Click here for larger image.

Source: Bloomberg

There are 69 industry groups accounting for 68.6% of the S&P 500's market capitalization with statistically significant negative relationships to crude oil prices. These groups are concentrated in the consumer-related and financial sectors. If we multiply their weights by the betas relative to crude oil, we get a negative impact of -3.56%.

There are only 16 industry groups accounting for 18.84% of the S&P 500's market capitalization with statistically significant


relationships to crude oil prices; these are concentrated, obviously, in the energy and utility sectors. However, their betas are much higher, so when we multiply them by their weights, we get a positive impact of 5.08%.

The net impact across industry groups is 1.51%. Every 1% rise in crude oil prices should lead to a 0.0151% rise in the S&P 500, all else held equal. That is correct: The net partial contribution of higher crude oil prices to stocks is positive, not negative; just glance at the chart at right and tell me otherwise.

Experience is a great teacher. You have to know what to forget and when to forget it, but the one thing you never should forget is to look at the data before offering conclusions based on short-term thinking.

Howard L. Simons is president of Simons Research, a strategist for Bianco Research, a trading consultant and the author of

The Dynamic Option Selection System

. Under no circumstances does the information in this column represent a recommendation to buy or sell securities. While Simons cannot provide investment advice or recommendations, he appreciates your feedback;

click here

to send him an email. has a revenue-sharing relationship with Trader's Library under which it receives a portion of the revenue from purchases by customers directed there from