Sep

25

The attached is a plot of log SPX vs contemporaneous yield curve (10Y-2Y), monthly 1976-August 2011. Dates are not shown, but the plot is a continuous (albeit mathematically not one-to-one) and shows various regimes between log SPX and 10Y-2Y.

The series starts in 1976 at the red dot near the bottom. Stocks (vertical axis) made little headway while 10Y-2Y varied above and below zero, making a series low about -2 in 1980. From the early 1980's to 1990's, stocks moved sharply upward, with 10Y-2Y varying between slightly negative and +1.5. From 1990 to 1992 (~Iraq I), stocks moved up while 10Y-2Y widened. From 1992-94, stocks went up while 10Y-2Y narrowed. From 1994-2000, stocks rose strongly while 10Y-2Y remained in a tight range (activist FED?), and 10Y-2Y went negative when stocks peaked in 2000. From 2000 - 2011 (green dot), 10Y-2Y varied considerably from -0.4 to +2.8 while log SPX was range-bound with considerable variation.

The recent picture - range-bound stocks with varying yield curve - resembles the late 1970's, and the overall pattern suggests no consistent relationship between stock levels and 10Y-2Y.

Steve Ellison writes: 

Attached is a regression graph of 1981-2010 S&P 500 annual returns vs. the difference between the 10-year yield and 3-month yield at the beginning of the year. N0, t=1.06, p=0.30, Rsq=0.04

The results I posted earlier that showed significance were the second year's S&P 500 returns regressed against the 10-year/3-month yield differential. That result seems suspicious–why should year-old data be more predictive than current data? 2011 so far is not going according to form.

George Zachar adds: 

Because between Volcker's rise and Lehman's fall, the main way the curve steepened was when the Fed lowered rates in the front end, stimulating the economy and stocks, months down the road.


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search