All correlations are based on past performance. In general, I think correlation data between asset classes is misleading and investors should never rely on it.
The only thing investors should be concerned with is how the assets they own correlate to the economy and not each other. Stocks for prosperity. Gold for inflation. Bonds for deflation (and prosperity). Cash to ride out recessions.
Stocks and bonds are not negatively correlated all the time. During the 1970s stocks and bonds both did horribly in real terms because inflation was so bad. The correlation of them to each other didn't matter, only how they correlated to the economy did.
Prior to 2008 investment gurus said that stocks and long term bonds were actually positively correlated and that both would suffer at the same time. I guess that's maybe true in mild recession or inflation, but not true in a deflation. They reached their conclusions because we hadn't seen deflation since the 1930s and they were just looking at the past few decades. Their data and economic analysis was incomplete and they were unpleasantly surprised by the outcome when stocks lost over 30% in value and long term Treasury bonds went up 30% in value. Prior to 2008, many investment gurus poo-poo'd treasuries and encouraged investors to take on credit risk in lower quality bonds for more return. That was also a bad bet. Junk bond funds for instance dove almost as much as stocks. Corporate bonds also did not fare as well as Treasuries. All the correlation data they had proved to be incorrect because they were looking at the wrong things.
Gold is a volatile commodity because there is so little of it in relation to dollars in circulation. If there is no serious inflation expected in the dollar (say 5% or less per year), then gold will not perform well. But if it gets over 5% a year (or it is anticipated to do so) investors will start thinking that gold is looking pretty good compared to holding dollars and will buy it running up the price.
But we should also remember that asset allocation strategy should avoid looking at any one asset in isolation. Only how all the assets perform in the entire package matters.
I can make a case that any asset is horrible to own. Stocks the last 10 years have been real stinkers but I still own them because the next 10 years they could be great. In the 1990s nobody wanted gold but by the time the tech bubble popped gold was poised to go on a tear the next 10 years to today. In the late 1970s nobody wanted bonds because inflation had killed them the prior decade. But once inflation came under control bond prices went through the roof and continued to turn in excellent performance the next 30 years!
So I strongly encourage investors to step away from the mindset of correlations and guessing what assets will do best and just spread your money around. It is counterintuitive, but by owning four major assets of stocks, bonds, cash and gold you are actually far safer risk-wise than concentrating your investments.
This question of which asset class will do best and whether to own Asset X or Asset Y comes up a lot. I wrote a post about it that includes a snippet from Browne explaining his experience in this as well that readers may enjoy:
The only thing investors should be concerned with is how the assets they own correlate to the economy and not each other. Stocks for prosperity. Gold for inflation. Bonds for deflation (and prosperity). Cash to ride out recessions.
Stocks and bonds are not negatively correlated all the time. During the 1970s stocks and bonds both did horribly in real terms because inflation was so bad. The correlation of them to each other didn't matter, only how they correlated to the economy did.
Prior to 2008 investment gurus said that stocks and long term bonds were actually positively correlated and that both would suffer at the same time. I guess that's maybe true in mild recession or inflation, but not true in a deflation. They reached their conclusions because we hadn't seen deflation since the 1930s and they were just looking at the past few decades. Their data and economic analysis was incomplete and they were unpleasantly surprised by the outcome when stocks lost over 30% in value and long term Treasury bonds went up 30% in value. Prior to 2008, many investment gurus poo-poo'd treasuries and encouraged investors to take on credit risk in lower quality bonds for more return. That was also a bad bet. Junk bond funds for instance dove almost as much as stocks. Corporate bonds also did not fare as well as Treasuries. All the correlation data they had proved to be incorrect because they were looking at the wrong things.
Gold is a volatile commodity because there is so little of it in relation to dollars in circulation. If there is no serious inflation expected in the dollar (say 5% or less per year), then gold will not perform well. But if it gets over 5% a year (or it is anticipated to do so) investors will start thinking that gold is looking pretty good compared to holding dollars and will buy it running up the price.
But we should also remember that asset allocation strategy should avoid looking at any one asset in isolation. Only how all the assets perform in the entire package matters.
I can make a case that any asset is horrible to own. Stocks the last 10 years have been real stinkers but I still own them because the next 10 years they could be great. In the 1990s nobody wanted gold but by the time the tech bubble popped gold was poised to go on a tear the next 10 years to today. In the late 1970s nobody wanted bonds because inflation had killed them the prior decade. But once inflation came under control bond prices went through the roof and continued to turn in excellent performance the next 30 years!
So I strongly encourage investors to step away from the mindset of correlations and guessing what assets will do best and just spread your money around. It is counterintuitive, but by owning four major assets of stocks, bonds, cash and gold you are actually far safer risk-wise than concentrating your investments.
This question of which asset class will do best and whether to own Asset X or Asset Y comes up a lot. I wrote a post about it that includes a snippet from Browne explaining his experience in this as well that readers may enjoy:
http://crawlingroad.com/blog/2010/01/12/what-asset-will-do-b...