Jul

5

Time for another look at war's effect on the stock market (yes this is mercantile, but someone's got to pay for the BBQ), using DJIA monthly returns (w/o div), checked wartime against peacetime. Dates used for US involvement might be contestable (they’re from web timelines):

WWII 12/41-8-45
Korea 6/50-7/53
Vietnam 8/64-4/75
Iraq II 3/03-present (or until Democrats rescue us, whichever comes first)

First, I lined up monthly returns for the four wars starting at their first month through the 38th (the shortest was Korea at 38 months), and regressed monthly return vs. number of months into war:

Regression Analysis: ret versus Month

The regression equation is ret = 0.0093 - 0.000123 Month

Predictor     Coef    SE Coef       T          P
Constant   0.0093    0.005     1.85    0.066
Month      -0.00013  0.0002   -0.57   0.567

S = 0.0305157   R-Sq = 0.2%   R-Sq(adj) = 0.0%

There was no trend in returns as wars progress from the start. Next compare mean monthly returns in and out of war (peace back to 2/41):

Two-sample T for war vs peace

                  N      Mean     StDev     SE Mean
war            264  0.0050   0.0370   0.0023  T=-0.9
peace        533   0.0076   0.0419   0.0018

War months just a tick lower, but not significantly. What about volatility?

Test for Equal Variances: war, peace

95% Bonferroni confidence intervals for standard deviations 

             N      Lower      StDev      Upper
war      264  0.03370  0.03701  0.04099
peace   533  0.03921  0.04191  0.04499

F-Test (normal distribution)
Test statistic = 0.78, p-value = 0.022

Interestingly, war was less volatile than peace (maybe investor focus on war news or patriotism distracted them from looking too hard at the economy?). With war’s end possibly just an election away, traders pining for those volatile months may get their wish. 


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search