I was reading this column
yesterday, which described a number of items you may have not known
about gold. That got me thinking. What can the price of gold tell us
about the broader market or the economy? As I said in my comments in the post, I thought I would test the thesis that the relative price of the Dow Jones Industrial Average to gold can predict growth in the economy.
Similar to the outcome I found in Fun with Gold Part 1 and despite a better looking chart analysis, the explanatory power of the the DJIA/gold ratio on the economy is not much better than gold's predictive power for the course of the market. Looking at the chart below, I initially thought the DJIA/gold ratio would be a descent measure to track economic growth. The chart shows the year-over-year return of the S&P 500 and the DIJA/gold ratio.
Initially, the ratio looked like it could help predict the course of the economy, as the two statistics show signs of co-movement. However, the statistics do not bear this out. The real kicker in my mind here is the analysis of the R-squared of the data sets. First, the coincident R-squared is just 11.4%. Lagging the data just makes the explanatory power even worse, as the R-squared falls to 5.9% with a one-quarter lag and just 0.1% for a 4-quarter lag.
I do not want to say that there is no usable information in this relationship, as graphically there appears to be some co-movement, but the statistics suggests you cannot rely solely on signals from the DJIA/gold ratio to predict the economy.
Similar to the outcome I found in Fun with Gold Part 1 and despite a better looking chart analysis, the explanatory power of the the DJIA/gold ratio on the economy is not much better than gold's predictive power for the course of the market. Looking at the chart below, I initially thought the DJIA/gold ratio would be a descent measure to track economic growth. The chart shows the year-over-year return of the S&P 500 and the DIJA/gold ratio.
Initially, the ratio looked like it could help predict the course of the economy, as the two statistics show signs of co-movement. However, the statistics do not bear this out. The real kicker in my mind here is the analysis of the R-squared of the data sets. First, the coincident R-squared is just 11.4%. Lagging the data just makes the explanatory power even worse, as the R-squared falls to 5.9% with a one-quarter lag and just 0.1% for a 4-quarter lag.
I do not want to say that there is no usable information in this relationship, as graphically there appears to be some co-movement, but the statistics suggests you cannot rely solely on signals from the DJIA/gold ratio to predict the economy.
No comments:
Post a Comment