Gold Remains a Poor Inflation Gauge

Gold has long been considered an inflation hedge, but the relationship between the two may be mostly a one-way street. Historically, the value of the precious metal turns out to be a weak indicator of inflationary pressures, unless prices are climbing rapidly. Even then, the link is tenuous. Moderating inflation, however, may give clues to the direction of the yellow metal's price.

Broadly speaking, inflation is nothing more than an increase in prices, with the Commerce Department's GDP price deflator and the Labor Department's Consumer Price Index [CPI] the key measures. Focusing on the headline CPI figure, which is a broad measure of consumer prices, and the core number, which strips out volatile food and energy, we compared the data, adjusted [SA] and unadjusted [NSA] for seasonal factors, with the monthly change in gold prices.

Correlation is a simple yet useful tool for examining a relationship between variables. This does not tell us about causality or the direction of the relationship, only that a relationship exists. Variables that move together perfectly have a correlation of 1.00, while variables that are perfectly inversely related have a correlation of -1.00. Variable movements that are not related at all have a correlation of 0.

For the entire period we examined - March 1968 through September 2006 - the correlation between the monthly change in the price of gold and the monthly change in the seasonally adjusted headline CPI (CPI-SA) was 0.15. The unadjusted CPI (CPI-NSA) had a correlation of 0.14 with the price of gold. As indicated below, this is better than the correlation between the change in the price of gold and the monthly inflation rates as measured by the seasonally adjusted and not seasonally adjusted core CPIs. (Click here for an Excel sheet of the data and highlights of our analysis.)

Correlation Results of Monthly Changes Price of Gold Headline CPI [SA] 0.15 Headline CPI [NSA] 0.14 Core CPI [SA] 0.08 Core CPI [NSA] 0.04

SA: Seasonally Adjusted NSA: Not Seasonally Adjusted

At first blush, a correlation of 0.15 does not seem all that strong, particularly if gold is supposed to be a hedge against inflation. If a strong correlation exists, it shouldn't be a stretch to expect the price of gold to shoot up when inflation picks up, and then climb at a slower clip when inflation eases. If that were the case, you would expect the relationship to be a lot stronger than just 0.15. Why isn't it?

For starters, there is the very nature of our comparison - we analyzed the monthly change in gold spot rates versus the monthly change in the CPI. A potential problem with this is that the CPI for a given month is released the following month. While investors know what gold prices are for a given period during that time frame, they don't know what inflation for that period is until the next month. Arguably, then, we are looking at the correlation of inflation with how well gold investors' forecast of inflation is reflected in gold prices. This is a subtle difference, but important just the same.

Because of the timing of the CPI releases, our next step was to lag the CPI by different periods. We started with a month and went back up to 12 months. Because the strongest relationship was between gold and CPI-SA, we used this index for determining the optimal lag.

We found the correlation between the change in gold prices and inflation was the strongest when we compared same-month information, i.e., a specific month's change in the price of gold with the inflation rate for that same month. The relationship between these two variables trended lower the more we lagged the CPI.

Correlation Results For Different Lags Correlation No Lag 0.15 1 Month 0.09 2 Months 0.10 3 Months 0.14 4 Months 0.05 5 Months 0.06 6 Months 0.10 7 Months 0.06 8 Months 0.00 9 Months 0.03 10 Months 0.01 11 Months 0.04 12 Months -0.05

This is where a little bit of history might help shed some light on the relationship between gold and inflation. The early years of the period we examined were influenced by the currency-exchange rate system in place at the time. Under the terms of the Bretton Woods agreement, which was initiated after World War II and existed until August 1971, currencies were effectively fixed at specified rates and the U.S. dollar was backed by gold at the rate of $35/ounce. The industrialized countries at the time worked to keep some type of fixed-exchanged rate system in place, but ultimately gave up, and in 1973, this gold-exchange system was completely abandoned and currencies were left to float freely.

With that in mind, we looked at the correlation of the monthly change in the price of gold to the monthly change in inflation from January 1974 through September 2006. Here, the relationship between monthly changes in the price of gold and the CPI-SA jumped to 0.18, while the relationship with the CPI-NSA climbed to 0.17. Although this is an improvement, the relationship still seems tenuous. Again, we turned to history to provide some insight into the relationship between gold prices and inflation.

We can break down the last 30 years or so into three distinct periods based on the price of gold:

• Period 1: August 1976 through June 1982. The period is characterized by a rally and sell-off in gold. During this period, the price of gold climbed from just over $105/ounce to a high closing price of nearly $667 before dropping to just below $318.

• Period 2: June 1982 through April 2001. During this period, gold was in a trading range with a lower limit of about $255 and an upper limit of about $500.

• Period 3: April 2001 through September 2006. This is the latest rally, during which gold has climbed from about $339 to about $600.

We then calculated the correlations for the changes in the price of gold and the headline CPI-SA and CPI-NSA for these periods. As indicated below, the relationship between gold and inflation changed substantially over these periods.

Correlation Results of Monthly Changes Dates & Correlations Period 1: 8/76 - 6/82 Headline CPI [SA] 0.17 Headline CPI [NSA] 0.18 Period 2: 6/82 - 4/01 Headline CPI [SA] 0.07 Headline CPI [NSA] 0.03 Period 3: 4/01 - 9/06 Headline CPI [SA] 0.23 Headline CPI [NSA] 0.16

SA: Seasonally Adjusted NSA: Not Seasonally Adjusted

The correlation of gold and inflation during Period 1 is not too different from the reading for the period from January 1974 through September 2006. But the relationship disintegrates during the middle years of June 1982 through April 2001, a period that is characterized by relatively stable prices when compared with Period 1.

Now focus on the results for Period 3. There is a relatively strong correlation between the monthly change in gold prices and the inflation rate, with the correlation for the CPI-SA at 0.23. During Period 3, gold prices only trended higher. This led us to go back to Period 1 and ignore the sell-off and just focus on the rally in gold prices that spanned from August 1976 through September 1980, a time frame that we called Period 1A. We found that gold prices better reflected the change in prices than they did when we examined all of Period 1, as indicated below.

Correlation Results of Monthly Changes Date & Correlations Period 1A: 8/76 - 9/80 Headline CPI [SA] 0.22 Headline CPI [NSA] 0.20

SA: Seasonally Adjusted NSA: Not Seasonally Adjusted

Meanwhile, the correlation between gold prices and inflation drops in the latter portion of the Period 1, which we call Period 1B. During the period September 1980 through June 1982, consumer prices continued to climb, but at a slower pace. The price of gold, however, dropped precipitously and the correlation between the two also fell.

Correlation Results of Monthly Changes Date & Correlations Period 1B: 9/80 - 6/82 Headline CPI [SA] -0.07 Headline CPI [NSA] -0.01

SA: Seasonally Adjusted NSA: Not Seasonally Adjusted

The data suggest that the yellow metal reflects inflation -- although not very strongly -- mostly when prices are rising at a rapid pace. The relationship weakens significantly during times of moderate inflation. Indeed, if history is any guide, then an easing in inflationary pressure could signal the start of a protracted decline in gold prices.

Disclosure: At the time of publication, Erik Dellith did not directly own gold. He may be an owner, albeit indirectly, as an investor in a mutual fund or an Exchange Traded Fund.

Note: This is independent investment and analysis from the Reuters.com investment channel, and is not connected with Reuters News. The opinions and views expressed herein are those of the author and are not endorsed by Reuters.com.