# Stock Monthly Return Autocorrelation, from Kim Zussman

December 28, 2009 |

Using DJIA monthly returns, at the end of each December and end of decade (i.e., 12/ XXX9), checked correlation of this month's return with prior month's. Here are the correlations along with the mean return for months within each decade:

Date                  correl 120  av 120
12/1/1999       -0.120  0.013
12/1/1959       -0.022  0.011
12/3/1979        0.012  0.001
12/1/1949        0.019  0.003
12/1/1969        0.020  0.002
12/1/1989        0.054  0.011
11/2/2009        0.111  0.000
12/1/1939        0.119  0.001
12/1/1999       -0.120  0.013
12/1/1959       -0.022  0.011
12/3/1979        0.012  0.001
12/1/1949        0.019  0.003
12/1/1969        0.020  0.002
12/1/1989        0.054  0.011
11/2/2009        0.111  0.000
12/1/1939        0.119  0.001

Looks like returns were higher for decades with monthly correlation more negative, which is verified by regression:

Regression Analysis: av 120 versus correl 120

The regression equation is av 120 = 0.00647 - 0.0481 correl 120

Predictor Coef SE Coef T P
Constant 0.0065 0.00153 4.23 0.006
correl 120 -0.048 0.02036 -2.36 0.056

S = 0.00409714 R-Sq = 48.2% R-Sq(adj) = 39.6%

////////

Note the recent decade most resembled the one ending in 1940. In honor thereof, here is some good stuff by a former neighbor.