For a while now, I have been intrigued about the strong co-movement of the S&P 500 versus the relative price of cyclical and non-cyclical stocks. I have remarked about this relationship in past posts. For example here, here, here, or here.
In some ways, I think the relative movement of cyclical stocks versus non-cyclical stocks was capturing investors' risk appetite. As the apparent risk appetite rose or fell, so did the equity markets. I found the same dynamic to hold true for other risk versus non- or less risk asset classes. For example, the relative price of the Russell 2000 versus the Russell 3000, the price of copper versus gold, or the price of junks bonds versus investment grade corporate bonds.
All of these relative price indices have tracked the price of the S&P 500, to various degrees, in the recent past. However, there were and are nuances in the data for each- be it trend, directional changes at inflection points, volatility, etc. Instead of looking at each individually or taking a simple average, I wanted a more robust combination of the signals in each. The below graph represents my first stab at this combination.
The blue line in above graph represents the price trend of the S&P 500 since the early 2000's. The pink line- called Risk- is derived from four relative price indices to which are applied various weighting schemes. To derive these relative price indices, I used a number of different financial data series including the Morgan Stanley Cyclical Index, the Dow Jones U.S. Consumer Goods Fund, Gold, Copper, the Russell 2000, the Russell 3000, the Barclays High Yield Bond Fund, and the iShares Investment Grade Corporate Bond Fund.
What I find interesting about the the risk measure is that it appears to have turned before the market did in the three major moves over the last 12 years, turning up before the market in both 2002 and 2008/2009 and down before the market in 2007. The same also appears to be the case for a number of the counter-trend moves within the larger trends.
However, I am keenly aware that what we think we see may just be an illusion and what appears to be a signal may just be noise. There is more research to do here and over the weeks and months ahead I hope to expand on this analysis.
In some ways, I think the relative movement of cyclical stocks versus non-cyclical stocks was capturing investors' risk appetite. As the apparent risk appetite rose or fell, so did the equity markets. I found the same dynamic to hold true for other risk versus non- or less risk asset classes. For example, the relative price of the Russell 2000 versus the Russell 3000, the price of copper versus gold, or the price of junks bonds versus investment grade corporate bonds.
All of these relative price indices have tracked the price of the S&P 500, to various degrees, in the recent past. However, there were and are nuances in the data for each- be it trend, directional changes at inflection points, volatility, etc. Instead of looking at each individually or taking a simple average, I wanted a more robust combination of the signals in each. The below graph represents my first stab at this combination.
The blue line in above graph represents the price trend of the S&P 500 since the early 2000's. The pink line- called Risk- is derived from four relative price indices to which are applied various weighting schemes. To derive these relative price indices, I used a number of different financial data series including the Morgan Stanley Cyclical Index, the Dow Jones U.S. Consumer Goods Fund, Gold, Copper, the Russell 2000, the Russell 3000, the Barclays High Yield Bond Fund, and the iShares Investment Grade Corporate Bond Fund.
What I find interesting about the the risk measure is that it appears to have turned before the market did in the three major moves over the last 12 years, turning up before the market in both 2002 and 2008/2009 and down before the market in 2007. The same also appears to be the case for a number of the counter-trend moves within the larger trends.
However, I am keenly aware that what we think we see may just be an illusion and what appears to be a signal may just be noise. There is more research to do here and over the weeks and months ahead I hope to expand on this analysis.
No comments:
Post a Comment