In our case, such periods are assumed to be the four largest (and costliest) wars fought by the United States in the second half of the 20th century, namely World War II (WWII), the Korean War, the Vietnam War, and the War on Terror. Each month and over multi-year periods around the start of each of the wars, we form value-weighted portfolios of stocks with patriotic names. The abnormal returns that we find during the War on Terror, the Korean War, and most of WWII are as high as 6% per year and are therefore economically significant.
These tests reveal a pattern in the timing of abnormal returns: The stock price response of the "patriotic name" portfolios is not immediate, i.e. for WWII, the Korean War, and the War on Terror, we report the average number of patriotic stocks in our portfolios and size and book-to-market statistics for the duration of the war (below) and a 48-month period immediately preceding the war (before). Patriotic stock portfolios exhibit abnormal returns over several periods after 9/2001 and over a four-year period after the start of the Korean War.
All alphas (except one) are insignificant and suggest that there is no “patriotic name bias.”13 We suspect that there is a lack of “patriotic name bias” during the Vietnam War. If this intuition is correct, perhaps there could be a “patriotic name bias” during subperiods of World War II when the public was more optimistic about the outcome of the war. We therefore re-estimate model (1) over several periods after the Battle of Midway (i.e. excluding the first six months of the war) and until the end of the war.
In Table 10, we combine all prewar and wartime periods to determine the economic significance of the "patriotic name bias."
Cumulative Abnormal Returns
However, a simple examination of the proportion of IPOs with patriotic names to total IPOs does not reveal any wartime pattern19. Had it not been for foreign investors, the "patriotic name effect" would probably have been stronger. Those of equally weighted portfolios are generally more important than those of value-weighted portfolios.
The CAR of the value-weighted portfolio peaks at about 52% about a year and a half after 11/9/01 and then declines over the next two and a half years. This means that in years 3 and 4 there is a reversal and the portfolio of value-weighted patriot stocks earns negative abnormal returns. An equally weighted portfolio shows no reversal over these four years, probably because mispricing at a larger company is easier to detect and therefore more likely to be corrected sooner.
These cumulative abnormal returns are consistent with the Fama-French regression results of the zero net investment portfolio. We do this because the market model parameters - estimated over a period before the benchmark dates - become less relevant over longer time windows through the change in the stock return distribution. First, that there is no immediate price reaction of the patriotic name stocks after the benchmark dates.
Second, the bias appears to be more pronounced for smaller firms, as the relative performance of value and balanced portfolios indicates. For WWII and the Korean War, the equal-weighted portfolios outperform the value-weighted; for the War on Terror, the opposite is true for the first year after, but thereafter the value-weighted portfolio CARs begin to decline, while those of the equally-weighted portfolios continue to increase and are significant over longer event windows. In particular, using stock market data around and during WWII, the Korean War, and the War On Terror, we show that when investors are more likely to be enthusiastic about their country, firms with "patriotic" names earn positive abnormal returns.
The size of the effect varies depending on the war under study: for some specifications, the monthly abnormal return is as high as 0.8% and is significant at the 1% level. For example, for all three war value and equally weighted portfolios earn CARs of at least 24% (and as high as 69%) in the first year. To the extent that certain war periods are associated with stronger patriotic sentiment among investors, our results can be interpreted as evidence of a “patriotic name bias,” whereby patriotic sentiment causes investors to slowly gravitate toward stocks with patriotic-sounding names.
Cumulative abnormal returns
As before, the cumulative abnormal returns (CARs) over a window of x days after the event are given by:
Bosch Jean-Claude and Mark Hirschey, 1989, "The Valuation Effects of Corporate Name Changes", Financial Management, Vol. Each month we form value-weighted portfolios of stocks with patriotic names over a period of two, three and four years before and after December 1941, after June 1942 and until the end of the war (August 1945). Each month we form value-weighted portfolios of stocks with patriotic names over a period of two, three and four years before and after June 1950 and until the end of the war (July 1953).
Each month, we construct value-weighted portfolios of stocks with patriotic names in the two, three, and four years before and after that date. Each month we construct value-weighted portfolios of stocks with patriotic names for the two-, three-, and four-year periods before and after September 2001. We construct value-weighted portfolios of stocks with patriotic names each month and estimate a four-factor model ( 1 ) for each of the four-year time periods.
Because the exact start date of the Vietnam War is disputed or unclear, we provide for the purpose of comparison some key dates along with a brief description of the events that occurred on these dates (Source: Wikipedia). Each month of these periods we form value-weighted portfolios of stocks with patriotic names. The match sample consists of stocks that do not have a patriotic name but are in the same industry and have a similar size and book to the market.
Each month, we construct a value-weighted portfolio of stocks with patriotic names along with a matching portfolio that contains stocks that do not have patriotic names but are in the same industry and have a similar market size and book to the stocks with patriotic names. We regress return differences on the two-, three-, and four-year periods before and after September 2001. We also create a control portfolio that contains stocks that do not have patriotic names but are in the same industry and have a similar market size and book value as a patriotic name.
We then average the differences in sales growth rates between Patriot and control stocks each quarter over multi-year periods. Each month, we construct equally weighted portfolios of stocks with patriotic names over the two-, three-, and four-year periods before and after September 2001. CARs are calculated from control portfolios constructed by comparing each patriotic stock to one that it doesn't have a patriotic name, but it's in the same industry and has a similar size and book to the market.
Time Windows Cumulative Abnormal Returns War on Terror (trading days) Value-weighted portfolios Equal-weighted portfolios. CARs are calculated using the market model as a benchmark. t-statistics are in parentheses; t-statistics significant at ten percent are in bold. Time Windows Cumulative Abnormal Returns Korean War (trading days) Value-weighted portfolios Equal-weighted portfolios.
Examination of the articles revealed that this superiority is due to reporting on the presidential election campaign.