doug444 wrote:Here's a chart I'd like to see -- the price changes for industrial metals, percentage-wise, over the past 5 years; all in one chart, aluminum, copper, zinc, nickel, chromium, iron. I couldn't find such a chart with Google, but I quit after 10 minutes. Now add the price of oil (WTI) in a different color, and see if oil prices are a good predictor of copper (and other) prices.
Even if oil's price was a good predictor relative to copper over the past five years, what about the last one, three, ten and twenty years? By assuming that something that worked once will always work, you will suffer the perils of those who use small sample sizes.
It is just like some stock enthusiast who touts something like a high volume breakout being significant and shows you a chart of a stock that legitimately doubled after a high volume breakout. I find myself thinking, "Ok, so it worked for that one stock, but I'd like to see an exhaustive study of at least several hundred."