Page 1 of 1

commodities = stocks?

Posted: Wed Aug 11, 2010 12:00 am
by techtrade89
Here's the story on Bloomberg.com:

http://www.bloomberg.com/news/2010-08-1 ... lysis.html

The website doesn't have the pic, I think, but I've seen it on the Bloomberg terminal. Basically after economic collapse started, the commodity index tracks movements in the S&P 500 remarkably closely.

Supposed to be independent, no?

Scary weird stuff . . .

Posted: Wed Aug 11, 2010 12:54 pm
by sluggo
It's not easy to find historical data for the GSCI cash index, so I just used the prices of the nearest futures contract as a proxy.

I multiplied the GSCI by 1.963 so both series can appear on the same scale.

(1.963 is the ratio of the geometric mean returns of the two series)

EDIT - uploaded new revision of spreadsheet, which calculates daily returns and displays them on a scatter plot

Posted: Wed Aug 11, 2010 10:35 pm
by techtrade89
thanks - that's the pic.

But what does it mean?

I'm not sure.

With a touch of fundamentals, here's a guess: Commodities trade based on supply and demand - ultimately. Psychology, other traders, systems, over the long enough haul it all comes out in the wash. But . . . this assumes a functional economic system underlying it all so that the commodities have value within that system. Presumably the economic system is a given nearly all the time, so it doesn't impact prices. However if you start to worry about the economy itself in a truly fundamental way, then the values of the commodities take on a new variable in the form of whether there will be a robust enough underlying society to make use of (and hence give value to) commodities.

In other words, commodities tracking the S&P like this is scary --- no? Maybe it's why trend following is having a tough time now, until things settle down and decouple, we are living during very "interesting" times . . . ?

Posted: Fri Aug 13, 2010 8:26 am
by sluggo
One tracks the other "remarkably closely"? I edited the spreadsheet above to calculate daily returns, then plotted them on a scatter diagram. The correlation doesn't look "remarkable" to me.