Saturday, October 27, 2012

Stock Prices of Major Defense Contractors since 1990

Posted by Glenn Alpert
In an effort to better understand drivers behind the stock prices of major defense contractors, I undertook a study correlating their stock prices with major economic indicators.

IHS Jane's A&D Analyst Matthew Bell (09/2011)

In my research, I took the Defense budget, January IPI and M3 monthly data, military sales values, and wartime/peacetime status over the last 21 years, from 1990 – 2010. Over 21 observations of each of these data sets, I also recorded the historical stock price of 4 major diversified defense companies. 

These companies included Lockheed Martin, General Dynamics, Northrop Grumman, and Raytheon. Since 1990, Lockheed merged with Martin Marietta to become Lockheed Martin, and Northrop merged with Grumman to become Northrop Grumman. When searching for historical stock prices for these two companies, I was able to find records dating back to 1990, but I assume that the stock price which predates the mergers should reflect the stock price of the acquiring company.

I also chose to test the regression for years where the US was not considered to be at war; this would date from 1993 (conclusion of Gulf War 1) to 2001 (beginning of the global war on terror).

Stock Price = b0 + b1DEF + b2JIPI + b3JM3 + b4MS + b5W+ μ

The dependent variable represents the stock price of a given Defense company at the beginning of January for a given year.
DEF – Defense spending (yearly budget) – b1
JIPI – January Industrial Production Index for defense capital goods – b2
JM3 – January M3 Survey for new orders of defense capital goods – b3
MS – Military sales to foreign countries – b4
W – War – b5



There is not much change in the stock price during war or peace (although Tolstoy would be a good read). One could infer that GD makes more “support” equipment and maintenance services, rather products or services consumed or used exclusively during conflicts.

During wartime, [8.6(W)], Lockheed’s stock can be expected to rise significantly.
During peacetime, [-.593(DEF)], the government will contribute fewer dollars to Lockheed’s earnings (in other words, the government will buy less from Lockheed). U.S. Foreign Military Sales are key to Lockheed’s bottom line.


War [23.25(W)] has a massive impact on Northrop Grumman’s stock price. 
Northrop Grumman does not appear to run its own large scale manufacturing operations [-2.3(JIPI)]. During peacetime, Northrop Grumman concentrates more on Foreign Military Sales [.001(MS)] than wartime [.0005(MS)]. Another way of looking at this is that during a period of time where the US government is buying NG products, it shifts its business development efforts overseas.


War actually decreases demand for Raytheon products [-4.3(W)] (it would make more sense to build and install massive radar and satellite systems before you need them rather than in the middle of a conflict). Raytheon runs its own robust manufacturing operations [2.88(JIPI)].



There are some considerations to keep in mind regarding the results of these regressions. First of all, I highly suspect that the “Independent variables” are not completely independent of each other. The overlap would cause some distortion in the equation results.

Secondly, the stock price data was for the first recorded date in January of the given year, and the M3 and IPI data was end-of-month data for January. In retrospect, it may have been a good idea to use IPI and M3 data from December of the previous year, but the change should not be drastic, since manufacturing trends tend to increase or decrease gradually from month to month the majority of the time.

Thirdly, the Defense budget data calculated on a yearly basis, so this could skew the correlation. Monthly DoD spending data would add accuracy to the equations.

In addition, another idea for this project was to test the data for the recession years to see if there was a significant impact on the other variables. The data includes three recession years, but it would be difficult to test recession vs. non-recession years with only three data points (‘08, ‘09, and ’10). One would to leave out Defense spending from the equation, then just measure manufacturing data (stock price = Bo + IPI + M3)

This exercise provided additional insight into the stock price of US defense companies. While the data did provide some insight, we should remember that other factors affect stock prices, and this is reflected by R-squared figures ranging from the mid-50% to mid-90% range. Of course, further study is necessary.

No comments:

Post a Comment