The Vietnam War and the U.S. stock market The effect of war-related events on the stock market

The effect of war-related events on the stock

market

Faculty Economics, Management and Organization MSc. BA Finance

Supervisor:

The Vietnam War and the U.S. stock market

The effect of war-related events on the stock market

Abstract

This paper investigated the effect of positive and negative news-events related to the Vietnam War on the return of the stock market of the United States. This effect was measured by calculating the average CARs of the S&P 500, following the different war-related events.

Because these events were expected to influence the stock market in the short term, the effect was measured with relatively small event-windows. In total 38 positive and 38 negative events were found. The results indicated that all events do seem to influence the stock market when the CARs were stated in absolute values. Also, there was a weak indication that positive events related to the Vietnam War have a negative effect on the U.S. stock market when the average CARs were calculated two days after the event appeared in a medium (for the (-1,+2) and (-2,+2) event-window). Furthermore, the results indicated that negative events related to the Vietnam War have a negative effect when the CARs were calculated four days after the event appeared in a medium (for the (-1,+4) and (-4,+4) event-window and weak evidence for the (-5,+5) event-window). However, significance at the 5% level was also found for the (-4,+4) and (-2,+2) event-window when the events weren’t divided into positive and negative events. The results give an indication that different events during a war influence the stock markets of countries involved in that war. However, because of the limitations of this research, future research on this topic will be needed.

Keywords

Vietnam War, stock market, events, media, United States, CAR

Contact Details: Joël B. Vellinga

Kornoeljestraat 2B26 9741 JB Groningen

Table of contents

1. Introduction

2. Theory and Background

2.1. Literature Review 11

2.2. Why a Case-Study of News-Events During the Vietnam War 15

2.3. History of the Vietnam War 16

2.2. Expectations and Hypotheses 21

3. Data and Methodology

3.1. Data 23

3.1.1. Stock Market Data 23

3.1.2. Events Data 24

3.1.3 Weaknesses and Biases of the Data 24

3.2. Timeline Events 25

3.3. Research Method 27

4. Analysis and Empirical Results

4.1. Stock Market Return Analysis of All Events (CARs in Absolute Values) 31

4.2. Stock Market Return Analysis of Positive Events 32

4.3. Stock Market Return Analysis of Negative Events 33

4.4. Stock Market Return Analysis of All Events 35

4.5. Event-related Stock Market Return Analysis for Alternative

Benchmark-Measurement Periods 36

4.5.1. Robustness Tests All Events Analysis (CARs stated in

absolute values) 36

4.5.2. Robustness Tests Positive Events Analysis 36

4.5.3. Robustness Tests Negative Events Analysis 37

4.5.4. Robustness Tests All Events Analysis

5. Conclusions and Discussion

5.1. Summary and Conclusions 39

5.2. Biases and Weaknesses 41

References

Appendices

Appendix 1: Copy of Table 1.1., From the Book Written by Paul Post (2006) 48

Appendix 2: Timeline Events 49

Appendix 3: Classification of Events 62

Appendix 4: Positive and Negative Events Included in the Research 65

Appendix 5: Returns Positive and Negative Events 66

Appendix 6: Tables Robustness Tests on Benchmark-Period Returns

(With Absolute Values), All Events 69

Appendix 7: Tables Robustness Tests on Benchmark-Period Returns,

Positive Events 73

Appendix 8: Tables Robustness Tests on Benchmark-Period returns,

Negative Events 77

Appendix 9: Tables Robustness Tests on Benchmark-Period returns,

1. Introduction

It seems that during human history there have always been wars. It almost seems that human nature forces us to battle and kill each other. Even at the present date, a lot of news is still focused on wars and their consequences. It seems that there is always an excuse for one country to fight another country. Because these wars have an influence on the confidence of consumers and investors, it seems very probable that these wars also influence the stock markets. This thesis tries to answer the question if positive and negative events related to a war have an influence on the return of the stock market. When such an influence exits, this thesis tries to find out what this influence is. The events, that will be used in this research, have to be related to the war, and will be researched for a country that is involved in the war, but is not fighting the war on its own soil.

At first, the objective of this thesis was to research all the wars after 1900, in which Western or Western-oriented countries were fighting abroad. However, at a certain point in the research it was decided to perform a case study of one war. This war could then be researched more

extensively than would have been the case when more wars would have been taken into account. The war this thesis focuses on is the Vietnam War.

Since this paper was supposed to research different wars, the research was focused on the effect of the start of a war, the events during a war and the end of a war. Therefore, different articles were researched to come up with a theory about the effect of certain events on the value and return of the stock market. However, since the research is now focused on only one war, a representative analysis focused on different time-periods of a war (the beginning, during and the end of a war) is no longer relevant. Instead, partly based on the literature, this research will now focus on the influence of news about events that are dubbed positive or negative.

escalated in the year 1964 with the Gulf of Tonkin incident. The war ended officially on the 27th of January 1973 when the Paris Peace Accords were signed.

After this, it was researched if the stock market data was available. As mentioned before, the stock market that had to be researched was the stock market of the United States. Since data about the Dow Jones Industrial Average (DJIA) in the Thomson Financial DataStream dated back to 1951, this seemed to be the most obvious choice. However, the DJIA consists of only thirty companies. Although the DJIA is an adequate sentiment index, the aim of this research was to show what the events would do to the entire stock market of the United States. Therefore, it was researched if an index existed that consisted of more companies. Data was available for the 1964-1973 period for the S&P 500 index and therefore it was decided that this index would be used.

The Vietnam War was chosen because this is a relatively long war and this war seemed to have a big influence on public opinion in the United States (U.S.). Since investors have to get their news about events during a war from a medium, this thesis focuses on media coverage of events. This means that events that were announced in the media will be taken into account in this research. It also brings us to another advantage of the Vietnam War, namely that it seems that during this war, the media were less constrained by their governments to report on certain events during the war. Therefore, the coverage of this war was more extensive than that of other wars that could have been taken into account, like for example the Falklands War.

Graph 1. Reaction of the S&P 500 Following Positive Events During November ‘67 840 845 850 855 860 865 870 875 880 885 15-11 16-11 17-11 18-11 19-11 20-11 21-11 22-11 23-11 24-11 25-11 26-11 27-11 date v a lu e S & P 5 0 0

This graph shows the reaction of the S&P 500 following three events that were found in the literature. November 16, General Westmoreland asserted that the situation in Vietnam is ‘very very encouraging’. November 17, American troops won a ‘dramatic victory’ in the battle for Dak To.

November 21, General Westmoreland told news reporters that whereas in 1965 the enemy was winning today they are losing.

Graph 2. Reaction of the S&P 500 Following the Gulf of Tonkin Incident ‘64 810 815 820 825 830 835 840 845 29-7 30-7 31-7 1-8 2-8 3-8 4-8 5-8 6-8 7-8 Date V a lu e S & P 5 0 0

Graph 3. Reaction of the S&P 500 Following the Tet-Offensive ‘68 810 820 830 840 850 860 870 26-1 27-1 28-1 29-1 30-1 31-1 1-2 2-2 3-2 4-2 5-2 6-2 7-2 8-2 9-2 10-2 11-2 12-2 13-2 Date V a lu e S & P 5 0 0

This Graph shows the reaction of the S&P 500 following the Tet-offensive which started on January 30, 1968. The reaction o the S&P 500 is shown by the blue line. Although this became known as a military victory of the U.S., the press and public reacted shocked and saw it as a clear indication that the war was not under control and that it would not be over soon.

Graph 2 indicates that the S&P 500 immediately drops after the Gulf of Tonkin incident. Graph 3 does not show an immediate reaction but at some point the stock market drops very sharply, perhaps indicating that it took some time before the public took hold of the information they were getting.

It will be researched what kind of influence the events will have on the stock market of the United States. As a stock market index for the United States the Standard & Poors 500 (S&P 500) will be applied. This index is chosen because it gives a good indication of the U.S. stock market

The relationship this thesis expected to find is that positive war-related events have a positive influence on stock market performance and that negative events have a negative influence on stock market performance. This is because events about the war that give positive news for a country involved in the war are expected to boost investor confidence. Therefore investors will start to save less and invest more in the stock market which will make the stock market go up. Events that give negative news about the war are expected to decrease investor confidence. Therefore, investors will invest less in the stock market and therefore the stock market is expected to go down.

The Cumulative Abnormal Returns (CARs) were calculated for events that appeared and were covered by the media. Therefore, a benchmark return needed to be calculated. The benchmark return chosen in this article is the average return. Taking account of the results of previous articles, it seemed that 100 trading days would be appropriate. However, because it is not certain that 100 trading days is really the most appropriate measure, it was tested if the results changed when a different number of trading days were taken into account. The number of trading days, used to calculate the average return, ranged from 50 to 120 trading days, with intervals of 10 trading days.

The results of the research indicated that positive war-related events influence the return of the stock market, but that they do so in a negative manner. These results appeared when the event-window was (-1,+2) and (-2,+2). Furthermore, the results indicated that negative war-related events affect the stock market in a negative manner. The strongest results were found for the (-4,+4) window, and significant results were also found for the (-1,+4) and (-5,+5) event-window. Moreover, when the CARs of both the positive and negative war-related events were stated in absolute terms, the results indicated that events in general do influence the stock market for the different event-windows. These results were significant at the 1% significance level.

2. Theory and Background

This section will first give an overview of the information, found in the related literature, which describes the possible relationship between wars and the stock market, and the causes for these relationships. After this an explanation is given why it was decided to perform a case-study of the news-events during the Vietnam War. Furthermore, it also gives a brief overview of the history of the Vietnam War, and finally it shows the hypotheses.

2.1. Literature Review

In the financial magazines, a lot is written about the influences of wars on stock markets. The Economist (1991) wrote: “It seems that Wars can be good for business, as Wall Street recognized in 1939, when the Dow industrial average rose 16% in the 12 days after Hitler invaded Poland. But usually stock markets change their mind, once their own country is involved or likely to be.” This means that wars are likely to have a significant influence on the stock market. It also means that involvement in a war, or further involvement during a war, is seen by investors as a negative event.

Rigobon et al. (2003) indicate that the risk of war has a negative influence on equity prices. They use a heteroskedasticity-based estimation technique to come up with these results. The evidence in their article indicates that greater war risk has been associated with a shift by investors away from risky assets. This means than in the period before a conflict starts, the stock market reacts to news that indicates whether or not the risk of going to war is more or less likely. If the news indicates a greater risk of war, the stock market drops. If the news indicates a lessening in the treat of going to war, the stock market rises. Leigh et al. (2003) have analyzed how stock markets react to news about the probability of going to war with Iraq. They did this by studying an unusual financial instrument called the “Saddam security”. This instrument was used to track shifts in the probability of war. After this the regression between this security and a stock market index was researched. They found that the start of a war lowers the value of U.S. equities by around 15 percent (when the probability of a war rises from 0% to 100%). This effect was mainly concentrated in the consumer discretionary industries, airlines, finance and information

According to the author Paul Post (2006) there are two ways in which a war influences the economy. War has a psychological effect, in that it scares people. War also has a real effect, namely that it is expensive. The real effect entails a large and sustained increase in government military expenditures and the mobilization of physical and labor resources.

The psychological and real effect can also affect the stock market. The psychological effect could cause people to withhold investments during wartime, because they are scared and there is a lot of uncertainty. The real effect could be the cause of less money in the economy to invest in the stock market; because the government needs more money to invest in the war, less money becomes available to spend on public services and/or taxes need to increase to fund the different expenses of the war. This will mean that investors will have less money to invest in the stock market, which will also cause the stock market to drop.

It is expected that investor sentiment (the psychological effect) will show the most obvious results. Especially when the war begins it is hard to comprehend what a war will cost. Therefore, especially the fact that a war scares investors will have a profound impact on the stock market. If a war begins, investors will know that money must be spend on this war and also they expect the future to become less certain. Therefore, investors will divest their investments in risky financial instruments and move to more secure ones. Since stocks on the stock market are a relatively uncertain financial instrument, investors will invest less in this stock market. This will cause the stock market to drop.

Table 1.1., in the book of Paul Post (2006), shows the change of the Dow Jones Industrial Average index at the day when a certain crisis occurs. Furthermore, it also shows the effect 6 months and 1 year after the occurrence of a crisis. The table shows that on average these events have a negative impact on the Dow Jones Industrial Average index on the day when the event occurs. Therefore, it could also be expected that negative events, which could be seen as small crises, during wars will influence the daily performance of the stock market in a negative manner. The before mentioned table can be found in the first appendix.

war and the duration of the war. As long as the information about the war is of a positive nature (victory in war, few losses during war etc.), the stock market is expected to rise. When the opposite occurs (loss in war, more losses during war etc.), the stock market is expected to drop.

During a war, the investors will also learn more about the different costs of the war. Investors will notice these costs because the government can increase taxes to cover the funds needed for the war or because these costs are shown in the media. If investors notice that the war is relatively successful because of positive battle reports or political statements they know that the war will cost less than initially expected. Therefore, more money becomes available to invest on other services and therefore more money will flow into the economy. More availability of money will also lead to more investments, also in the stock market, which will cause the stock market to rise. If investors notice the opposite, that the war progresses worse than expected, the opposite is likely to occur. However, it is still expected that news about the results of the war, and therefore news that is expected to make investors less or more confident will have the biggest impact on the stock market. This is because it is proved that there is only a weak negative relationship between U.S. military expenditures and U.S. private investment. Furthermore, since only the short term stock market effects are measured in this article, it is probable that being scared or relieved by the events has a bigger impact than future costs of the war. However, the effect will be similar. Again, investors will look for signs that indicate how the war is going. If news-reports indicate that the war is going better than expected and/or that the end of a war comes into sight the investors will be relieved, will think they could be more certain about the future and will therefore invest more in risky assets like the stock market. However, news that indicates the opposite will again have the opposite effect.

It is expected that the stock market will react positively to news about the end of a war. From graphs of both the Vietnam war and Dessert Storm1, it seems that the end of a war has a positive influence on the prices on the stock market, whether the war is won or not. It seems to be that the negative aspects of a war are already taken into account before the war comes to an end.

Therefore, the stock market seems to show a sigh of relief, and the return on the stock market rises, once the war is finally over.

At the end of a war, the fact that investors are relieved that a war is over will likely have a bigger impact on the stock market than the effect that more money will become available for investors.

1

However, the effect of both influences will again be the same since both relief and the expected availability of more money are expected to make investors more confident and therefore will make them invest more in the stock market. If more money becomes available for the “general” public, they can invest more, and therefore they will probably also invest more in the stock market. The end of a war will cause investors to become more certain about the future. Therefore, they will invest more money in risky financial instruments and therefore the prices of these instruments, like the price of the stock market, will go up.

Berkman and Jacobsen (2005) study the stock market reactions to international crises. They used a database of 440 international political crises. They research whether world market returns depend on the number of crises in any given month and after this, they distinguish between start effects, effects during a crisis and end effects. The results are estimated with a GARCH (1,1) model. Their results indicate a strong and statistically negative relation between world market stock market returns and the number of international crises. Furthermore, the results indicate that the start of a crisis has a large negative impact on world market returns. They also indicate that this drop in market prices is followed by lower than average returns during the crisis and a price increase when the crisis ends. Moreover, they find that the more severe a crisis, the larger are the negative effects. Another conclusion is that investors in stock markets of countries involved in an international crisis suffer more: Stock markets in those countries drop by almost two percent when a crisis starts; lose an additional one percent for every month the crisis lasts; and this value loss is only partially recovered when the crisis ends.

The results of the research of Berkman and Jacobsen (2005) also show that the volatility of the stock market decreases when a crisis has come to an end, indicating that after a war, the stock market reactions are less sever as during a crisis. However, since the period after a war is not part of the research in this thesis, no further research has been done to come up with results that show what happens to the stock market after a crisis comes to an end.

with a general theory. Furthermore, the literature did not indicate whether or not a longer crisis has a more severe impact on the stock market than a shorter crisis. However, the literature did indicate that the stock markets lose value with every month a crisis lasts, so that stock markets lose more value when a crisis lasts longer. Furthermore it indicated that the more severe a crisis the higher the negative stock market reaction to this crisis would be.

What the literature did indicate was that the start of a crisis or war generally leads to a decline in the stock market and the end of a crisis or war generally leads to a rise in the stock market. Furthermore, the literature indicated that the way the news was presented probably has a big impact on the stock market reaction. If the news is presented in a positive fashion indicating that the war or conflict is going well, or that a loss is just a minor incident, the stock market seems to react accordingly. When the presenter shows pessimism about the news and therefore shows that the war or conflict is not going well, the stock market also react accordingly, it drops. However, it must be noted that these insights do not come from academic journals but from observations made in short articles in magazines.

It seems that media-reports about events that were brought positively and shows positive news related to the war will bring a rise in the stock market of a country involved in that war, while media-reports that were brought negatively and shows negative news related to the war will lead to a drop in the stock market. This research will try to come up with evidence that these

relationships exist and will therefore be an extension of the existing literature

2.2. Why a Case-Study of News-Events During the Vietnam War

As has been reported in the previous section, different articles were researched to come up with a theory about the influence of certain events on the value and the return of the stock market. Since this paper was supposed to research different wars, the research was focused on the effect of the start of a war, the events during a war and the end of a war. However, since the research is now focused on only one war, a representative analysis focused on different time-periods of a war (the beginning, during and the end of a war) is no longer relevant. Instead, partly based on the

literature, this research will now focus on the influence of news about events that are dubbed positive or negative.

became clear that the media were very constraint during the wars by their national governments, except during the Vietnam War. This reason, and the reason that media data about different events during the Vietnam War was available, made the Vietnam War the most obvious choice. It was decided to research the period from the 1st of January 1964 until the 31st of January 1973. This is because the Vietnam War doesn’t really have a clear starting point but the situation escalated in the year 1964 with the Gulf of Tonkin incident. The war ended officially on the 27th of January 1973 when the Paris Peace Accords were signed. A complete overview of the history of the Vietnam War will follow in the next section.

2.3. History of the Vietnam War2

From 110 BC to 939 AD, a large part of Vietnam, as we know it today, was part of China. After gaining independence, Vietnam went through a long period of resisting foreign aggressors. By 1802, centuries of internal battles between the Trinh and Nguyen lords ended when Emperor Gia Long unified Vietnam under the Nguyen dynasty. The French gained control of Indochina (this included Vietnam) during a series of colonial wars, from 1859 to 1885.

During the Second World War, the Vichy France government cooperated with the Japanese forces, which controlled Vietnam. Therefore, the French continued to serve as the day to day administrators. In 1941, the Communist-dominated national resistance group called the “League for the Independence of Vietnam” (better known as the Viet Minh) was formed. Ho Chi Minh assumed the leadership of the Viet Minh. The Viet Minh began to craft a strategy to seize control of the country when the war would come to an end.

During the Second World War the Viet Minh helped the U.S. Army. The Japanese overthrew the French administration in 1944. Emperor Bao Dai declared the independence within the Greater East Asia Co-Prosperity Sphere on March 11, 1945.

Following the Japanese surrender, Vietnamese nationalists, communists and other groups hoped to seize control of Vietnam. The Japanese army transferred the power to the Viet Minh. Ho Chi Minh declared independence from France on September 2, 1945, in what became known as the August Revolution. Because the U.S. opposed the revival of European colonialism, Ho Chi Minh expected that the U.S. would ally itself with the Vietnamese nationalist government, communist or otherwise. Power politics intervened, however, and the U.S. changed its position. It was

2

recognized that France would play a crucial role in deterring communist ambitions in continental Europe. Therefore, its colonial aspirations could not be ignored.

The new Vietnamese government only lasted a few days. At the Potsdam Conference, the allies decided that Vietnam would be occupied jointly by China and Britain. The French prevailed upon them to turn over control. French officials immediately tried to regain control. By agreeing to give up its concessions in China, the French persuaded the Chinese to allow them to return to the north and negotiate with the Viet Minh. After negotiations collapsed over the formation of a government within the new French Union, the French bombarded Haiphong. In December, 1946, the French re-occupied Hanoi. The Viet Minh requested U.S. support but got none. The Viet Minh fled into the mountains to start an insurgency, marking the beginning of the First Indochina War. During this war a long and bloody struggle ensued. Because the Soviet Union was now seen as a serious competitor, the U.S. saw the entire communist world as controlled by Moscow. Therefore, the U.S. denied Ho’s eloquent pleas for U.S. recognition.

In 1950, the Democratic Republic of Vietnam and China recognized each other diplomatically and the Soviet Union followed. The U.S. countered by recognizing the French government of Vietnam. The U.S. Military Assistance and Advisory Group (MAAG) arrived in 1950 to screen the French requests for aid, to advice on strategy and train Vietnamese soldiers. By 1954, the U.S. had supplied 300.000 small arms and spent approximately one billion dollars in support of the French military effort in Vietnam. The Viet Minh, on the other hand, received crucial support from the Soviet Union and the People’s Republic of China.

The Battle of Dien Bien Phu marked the end of the French involvement in Indochina. The Viet Minh defeated the French. The French forces surrendered on May 7, 1954. At the Geneva Conference the French negotiated a ceasefire agreement with the Viet Minh. Furthermore,

independence was granted to Vietnam, Laos and Cambodia. The Viet Minh established a socialist state, the Democratic Republic of Vietnam, in the north. In the south a non-communist state was established under the emperor Bao Dai. Ngo Dinh Diem became his prime minister.

counter communist expansion in the region. Ngo Dinh Diem was chosen by the U.S. to lead South Vietnam.

On October 26, 1955, Diem declared the new Republic of Vietnam with himself as president. Diem’s human rights abuses increasingly alienated the population. As opposition to Diem in South Vietnam grew, a low-level insurgency began to take shape in 1957.

Observing the increasing unpopularity of the Diem regime, Hanoi authorized the creation of the National Front for the Liberation of South Vietnam (NLF). The principle objective of the NLF was to seize political power and reunify Vietnam.

In 1960, John F. Kennedy became president of the United States. During his leadersip the number of U.S. military advisres rose. Also because the quality of the South Vietnamese army remained poor, some policy-makers in Washington began to conclude that Diem was incapable of defeating the communists. President Diem was overthrown and excecuted, along with his brother, on November 2, 1963. However, following the coupe chaos ensued. Hanoi took advantage of this and increased its support for insurgents. South Vietnam entered a period of extreme political turmoil as one military government was replaced by the other.

Kennedy was assisinated on November 22, 1963 and was succeeded by Lyndon B. Johnson. On August 2, 1964, the U.S.S Madox was allegedly attacked by torpedo boats in the Gulf of Tonkin. A second attack on the U.S.S. Turner Joy was reported two days later. The second attack led to retaliatory air strikes and prompted the congres to approve the Gulf of Tonkin Resolution. This gave the U.S. president power to conduct military operations in South East Asia withouth declaring war.

On March, 1965, following an attack on U.S. marine baracks at Pleiku, Operating Flaming Dart and Operation Rolling Thunder commenced. This bombing campaign, which would last three years, was intended to force North Vietnam to cease its support for the NLF. Furthermore, it was aimed at bolstering the morale of the South Vietnamese. However, North Vietnam continued to support the NLF throughout the bombings.

Operation Starlite was the first major ground operation by U.S. troops and proved highly succesful. Learnig from their defeat, the NLF began to engage in small-unit guerilla warfare. Under the command of Westmoreland, the U.S. increased its troop commitment to more than 553.000 by 1969. Washington encouraged its SEATO allies to contribute troops. Australia, New Zealand, the Repunlic of Korea, Thailand, and the Philippines all agreed to send trops. However, major allies declined. Meanwhile, the political situation in South Vietnam began to stabilize. The relative calm allowed the Army of the Republic of Vietnam (ARVN) to collaborate more

effectively with its allies.

In January, 1967, the People’s Army of Vietnam (PAVN) and NLF broke the truce that had traditionally accompanies the Lunar New Year (TET) holiday. They launched the surprise Tet Offensive in the hope of sparking a national uprising. Over 100 cities were atacked. Although the Tet offensive became a military victory for the South Vietnamese and U.S. forces, the U.S. public took it as a defeat. Because the war was expected to be under control, the Tet offensive came as a surprise and the public and press were shocked and confused. Before the Tet offensive the press reports about the war were mainly positive. After the Tet offensive the reports became

increasingly negative. In a large part because of this, the Tet offensive was the turning point in the U.S. involvement in the Vietnam War.

On May 10, 1968, peace talks began between the U.S. and the Democratic Republic of Vietnam. Negotiations stagnated for five months, untill Johnson ordered to halt the bombing of North Vietnam.

During the 1968 presidential election, Richard M. Nixon promised “peace with honor”. His plan was to build up the ARVN, so they could take over the defense of South Vietnam. This policy became known as Vietnamization. Nixon also pursued negotiations. Operations became smaller and the ties with China and the Soviet Union improved. In September, 1969, Ho Chi Minh died. Due in part to the negative press reports about the Vietnam War, the anti-war movement became stronger in the U.S.. Nixon appealed to the “Silent Majority” of Americans to support the war. However, revelations of the Mail Lai Massacre, in which U.S. forces went on a rampage and killed civilians, provoked national and international outrage.

sparked U.S. protests. These protests led to four students being killed by National Guardsmen at the Kent State University.

In 1971, the Pentagon Papers were leaked to the New York Times. The top-secret history of the U.S. involvement in Vietnam, commissioned by the Department of Defense, detailed a long series of public deceptions.

Also in 1971, Australia and New Zealand withdrew their soldiers. The U.S. troops were reduces to 196.000, with a deadline for further reduction of 45.000 troops by February 1972. ARVN troops came under heavy attack by North Vietnamese forces with the Easter Offensive of 1972, a massive invasion of South Vietnam. American airpower came to the rescue with Operation Linebacker and the offensive was halted. The last American ground troops were withdrawn in August.

Nixon’s National Security Adviser, Henry Kissinger, held secret negotiations with North Vietnam’s Le Duc Tho. In October 1972, they reached an agreement. However, South Vietnams president Thieu demanded massive changes to the peace accord. When North Vietnam went public with the agreement’s details, the Nixon administration claimed that the North tried to embarrass the president. The negotiations became deadlocked. To show support for South Vietnam and force Hanoi back to the negotiating table, Nixon ordered Operation Linebacker II, a massive bombing of Hanoi and Haiphong. Simultaneously, Nixon pressured Thieu to accept the terms of the agreement. On 15 January, 1973, Nixon announced the suspension of offensive action against North Vietnam. The Paris Peace accords on ‘Ending the War and Restoring Peace in North Vietnam’ were signed on 27 January 1973, officially ending U.S. involvement in the Vietnam War. A cease-fire was declared across South Vietnam, U.S. hostages were released and the territorial integrity of Vietnam was guaranteed. It also stipulated a sixty day period for the total withdrawal of U.S. forces.

2.3. Expectations and Hypotheses

The evidence in the literature indicates that wars will have an influence on the stock market. Therefore, events during a war that get attention from the media are also likely to influence the stock market. This means that it will be expected that media news about the Vietnam War will have an influence on the stock market. The first hypothesis that will be tested in this article is the following:

Hypothesis 1:

- H0) Media news about the Vietnam War will have an influence on the stock market of the United States.

- H1) Media news about the Vietnam War will not have an influence on the stock market of the United States.

The different events, which will be used for this research, will be categorized in positive and negative news-events. The literature shows that the stock market reacts to news that gives an indication about the risk of going to war and that the stock market reacts to the end of a war. Furthermore it showed that the stock market would show lower than average returns during a crisis and that manic swings are expected during war time. The expectation is that these swings are caused by news that is related to the war. These expectations are also formed because it was indicated that the stock market seemed to react to news that had an influence on the expectations about the war. Therefore, it is expected that both positive related events and negative war-related events have a certain influence on the stock market.

Furthermore, it is expected that positive war-related news influences the return of a countries stock market positively while negative war-related news is expected to have the opposite effect. This leads to the following hypotheses:

Hypothesis 2:

- H0) Positive news about the Vietnam War leads to a positive abnormal return on the U.S. stock market.

Hypothesis 3:

- H0) Negative news about the Vietnam War leads to a negative abnormal return on the U.S. stock market.

- H1) Negative news about the Vietnam War does not lead to a negative abnormal return on the U.S. stock market.

Of course there are also many other events that have an influence on the stock market. Financial markets differentiate between different kinds of information. However, the involvement of a country in a war is such a major event that it is likely to have a big impact on consumers and investors. Therefore, it is also likely to have a major impact on the stock market. Furthermore, the relative short event-windows exclude a lot of the possible confounding effects. To take account of other factors, influencing the stock market, the average return will be calculated over a period prior to the occurrence of the news-event in the media. This return is expected to show the normal return and the return, after the news-event occurred in the media, will be calculated as the

3. Data and Methodology

This section contains information about the data and the methodology. For this research, the data is collected by using a stock market performance database, the Thomson Financial DataStream. Furthermore, data about the different events during a war had to be found. This data can be found in different books written about the Vietnam War, articles that are written about this war and encyclopedias. The encyclopedia that will be used in this research is the online encyclopedia Wikipedia3. However, because the information of this online encyclopedia can be edited by everyone, the events found in this encyclopedia will be checked with other sources.

3.1. Data

For this research, data is needed on stock market performance and events during the Vietnam War (like the beginning and end of the war and the different events during the war).

3.1.1. Stock Market Data

To come up with the stock market performance data, Thomson Financial DataStream has been used. Since the research is focused on the war in Vietnam, the stock market data that will be used is that of the United States. This country was the main combatant against the North Vietnamese army and the Vietcong. Data about the Dow Jones Industrial Average (DJIA) in the Thomson Financial DataStream dated back to 1951, and therefore this seemed to be the most obvious choice. However, the DJIA consists of only thirty companies. Although the DJIA is an adequate sentiment index, the aim of this research was to show what the events would do to the entire stock market of the United States. An index that consisted of more companies seemed more appropriate for this purpose. Therefore, it was researched if an index existed that consisted of more

companies. Data was available for the 1964-1973 period for the S&P 500 index and therefore it was decided that this index would be used.

3

3.1.2. Events Data

The events that will be used are events that are mentioned in the different media, like newspapers and television broadcasts. These events can be found in different books about wars and the media. Furthermore encyclopedias, like the online encyclopedia Wikipedia, can be used to come up with certain events during the war. With the help of these sources a timeline is constructed that describes different events during the Vietnam War. The main source for this timeline is the book of Hallin (1989), which describes different events during the Vietnam War as written in the New

York Times (NYT) and as broadcasted by different television stations.

3.1.3. Weaknesses and Biases of the Data

Of course this data also has some biases and weaknesses. One of the most important weaknesses of the data is that the events-data is dependent on what the writers of books deem important. This means that the data could be biased to certain events, which are seen as important by the authors. In order to correct this bias other sources have been used that report about the media coverage in war periods, but the data that can be found by using these sources will also depend on the preferences of the writers. Therefore, it is likely that not all important war-related events that appeared in the media have been found. However, from the research done, it seems that the researched sources focus on a lot of the same events.

Another important weakness is that stock market data is influenced by other events than wars. To correct for this problem the values of the stock market indices can be benchmarked against a world index. However, if this benchmark will behave exactly the same as the stock market index of a country than no specific influence can ever be found. Because the United States was

expected to have the biggest influence on a world market index this problem was expected to be most prominent for this country. Therefore, the correlation between the S&P 500 index and the FTSE World Market index were researched. This research indicated a correlation of 98,7%4. This means that both indexes behave almost completely similar. Furthermore, Thomson Financial DataStream does not provide a world index that goes further back than the year 1969. Because the world market index seems to be influenced by the United States to a very large extent and because not enough data for a world index is available such an index will not be used as a

4

benchmark for the S&P 500. This weakness is also further reduced because of the event-study method used. Because this research is not focused on a single company and its stock return, but on an entire country and its return, the return used is already that of an index and therefore it is less prone to confounding effects. Furthermore, the event-windows are relatively small, so there are fewer possibilities in the data for confounding effects to occur.

Another weakness of the data is that it has to be decided by the researcher/author what events are seen as positive, negative or not relevant for this research. Therefore, it might be possible that certain events that are seen as positive or negative in this research are seen as the opposite by others.

In total seventy-six news-events were found. These events were equally divided between the positive and negative events. This means that the conclusions are based on thirty-eight positive news-events and thirty-eight negative news-events. It would have been better if there had been a larger number of events so more weight could be given to the conclusions of this paper. The relative small amount of events is therefore another weakness of the data.

3.2. Timeline Events

In order to come up with a classification of the different events, a timeline was constructed. This timeline shows the different events that are included in this research and ranges from the year 1964 until January of 1973. A complete overview of the events in the timeline can be found in appendix 2. The events are split into positive and negative events.

The following events are seen as positive: The end of the war, a decision to withdraw troops from the battlefield (this will indicate a lesser involvement in the war), an announcement about the start of peace-talks and signs of progress in these peace talks, a U.S. victory in a battle or other

positive battle reports, a statement by a government official that announces that the war will not be expanded and positive statements about the Vietnam War from government officials.

The way the events are divided can be found in appendix 3. The events that are marked as positive or negative and the amount of positive and negative events can be found in appendix 4. The way the different events are distributed over the different years can be found in graph 4.

Graph 4. Bar Graph distribution events from 1964-1973.

0 2 4 6 8 10 12 14 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 year n u m b e r o f e v e n ts positive negative

This graph shows the distribution of the events that were used in this research over the different years. The bar, indicated as positive, shows the positive events during a certain year. The bar, indicated as negative, shows the negative events that were found for a certain year.

The graph shows that the most events were found in 1965. Also, it shows that in earlier years more news about the war reaches the public than in later years. It is likely that this is because the U.S. policy changed during the war, from large scale battles to smaller scaled battles.

3.3. Research Method

In this thesis, it will be tested if the different events during the Vietnam War have an influence on the stock market and when this influence exists it tries to answer what this influence is. This means that this will be an events-study. The article of MacKinlay (1997) has been studied to come up with a proper research method. Since only one war will be researched, only the influence of positive and negative events related to this war will be looked upon.

Since the aim of this research is to come up with an indication that these events have an influence on the stock market on the day they are announced in the media, the event-window will be centered around the announcement day. At first the (-1, 0) event-window will be researched. This is done for both the positive and negative events. With this kind of research, it is possible to find out if these events have an influence on the stock market on the day when they are announced. The same is done for one day (-1,+1) and two days (-1,+2) after the announcement day, to take account of a possible delayed stock market reaction. The returns on the S&P 500 of these event-windows can be found in the first appendix. For robustness, the event-window will also be extended to three days (-1,+3), four days (-1,+4) and five days (-1,+5) after the announcement date. Furthermore, it will also be tested if the results change when the abnormal returns are calculated for extended event-windows because it could be that for example certain government decisions about the Vietnam War are already expected, or leaked, before they are officially announced. Furthermore, the news reports tend to be quite mixed and therefore the public could react differently to these different forms of news. Therefore, the average Cumulative Abnormal Returns will also be researched for the (-2,+2), (-3,+3), (-4,+4) and (-5,+5) event-windows.

Abnormal Returns (ARs) are defined as the difference between the Realized Return (R) and a Benchmark Return (BR).The formula that will be used to calculate the abnormal returns is:

AR

=R

–BR

research, not the returns of a firm but the returns of a country are researched. This means that the index that is being researched is already a market index. However it could be possible to relate this index to a world market index. The problem, as explained before, is that no appropriate world market index could be found that dated back far enough.

Another possibility is to use the average return as the appropriate benchmark or to use excess returns (returns above the risk-free rate) instead of nominal returns. If the average return is used as a benchmark the documented returns are defined as abnormal returns. The average return is the return over a period prior to the event-window. This method is based on the fact that returns on a relatively long period before an event should show a certain tendency which indicates the normal return. Returns above or below this return are Abnormal Returns.

In the article of Brown and Warner (1980 and 1985) it is indicated that the Average Return Model (or Mean Adjusted Returns Model) performs worse than the Market Model when event month clustering appears. Event month clustering means that the securities of a sample experience an event during the same calendar time period in which the influence of an event on the price of the security is being examined. However, since the influence on the U.S. stock market is researched in this study, and not the effect on an individual security, this effect is expected to be of minor importance. In other situations, they found that the Average Return Model performs the same as the Market Model. Furthermore, Dyckman et al. (1984) report that the Market Model performs significantly better than the Average Return Models, but that this difference does not appear to be important since the rejection percentages are quite close in all cases. Klein and Rosenfeld (1987) indicate that the Average Return Model, but not the Market Model, produces upwardly biased Abnormal Returns during a bull market and downwardly biased returns during a bear market. This is a reason for caution in using this model. Brenner (1979) and Brick et al. (1989) report statistically different returns between results using the Market Model and using a model that incorporates the risk free rate. Since it is reported in the different studies that the Market Model is the most reliable model this discrepancy causes some worries.

The average return will be calculated by averaging the S&P 500 100 trading days before an event appears in a medium. It will also be tested if different results appear when the average return is calculated by averaging the S&P 500 50, 60, 70, 80, 90, 110 and 120 days before an event appeared in one or more media. From previous research it became apparent that the estimation window to calculate the average returns should be between 100 and 300 days before the appearance of an event. However, since trading days are taken into account, 100 days already seems to be an extensive period to come up with the average return. Also, the literature didn’t give a clear explanation as to why exactly the amount of days needed to be between 100 and 300 days.

The formula that will be used to calculate the Cumulative Abnormal Returns ( s) is:

(2) This formula can be used to calculate the average Cumulative Abnormal Return ( ):

(3) Where Ndesignates the number of events and τ indicates the event-window (t1,t2).

To examine whether the average Cumulative Abnormal Returns significantly differ from zero a student t-test will be employed.

(4)

(5)

Where is the estimator of the standard deviation of the ARs for the S&P 500 computed over the estimation window (t1,t2):

(6)

Where

L

4. Analysis and Empirical Results

As mentioned before, the results of this research will be obtained by performing a student T-test. The different tables in this section show the results of these tests. Section 4.1. explains the results when the average Cumulative Abnormal Returns (in absolute values) are calculated for all news-events that are used in this study. The following two sections show the results for the positive and negative news-events that appear when the average return is calculated by averaging the nominal returns of the S&P 500, 100 trading days before the start of the event-window. Section 4.4. explains the results when the student T-tests are performed for all events. Finally, section 4.5. shows the results of the student T-tests with alternative benchmark-measurement periods. The different tables show the average Cumulative Abnormal Returns, the standard deviation of the average CARs and the results of the student T-test. The student T-tests are performed for the different event-windows.

4.1. Stock Market Return Analysis of All Events (CARs in Absolute Values)

Table 1. T-Tests All Events ( CARs in Absolute Values and AR 100 days). Event- window Average CAR σ _average CAR T-test (-5,+5) 0.0178 0.0019 9.1452*** (-4,+4) 0.0147 0.0017 8.4619*** (-3,+3) 0.0115 0.0015 7.5987*** (-2,+2) 0.0090 0.0012 7.2609*** (-1,+1) 0.0061 0.0009 6.9570*** (-1, 0) 0.0046 0.0006 7.4561*** (-1,+2) 0.0085 0.0011 7.9496*** (-1,+3) 0.0105 0.0012 8.5306*** (-1,+4) 0.0120 0.0014 8.7212*** (-1,+5) 0.0141 0.0015 9.3354***

This table indicates the student T-tests of the average Cumulative Abnormal Returns in absolute values (irrespective of whether they are positive or negative) for all news- events when the average returns is calculated over 100 trading days before the event-window of an event. * = significant at the 10% level, ** = significant at the 5% level and *** = significant at the 1% level.

4.2. Stock Market Return Analysis of Positive Events

Table 2 shows the results of the student T-tests for the positive events. The average returns are again calculated by averaging the nominal returns of the S&P 500 100 trading days before the start of the window of a news-event. The table shows the result for the different event-windows and shows the results when the average CAR is calculated with the nominal daily returns.

Table 2. T-Tests Positive Events (AR 100 days).

Event- window Average CAR σ _average CAR T-test (-5,+5) -0.0020 0.0031 -0.6535 (-4,+4) -0.0019 0.0027 -0.6809 (-3,+3) -0.0021 0.0024 -0.9004 (-2,+2) -0.0035 0.0019 -1.8349* (-1,+1) -0.0018 0.0014 -1.3522 (-1, 0) -0.0005 0.0010 -0.5337 (-1,+2) -0.0030 0.0017 -1.7662* (-1,+3) -0.0020 0.0019 -1.0217 (-1,+4) -0.0005 0.0022 -0.2262 (-1,+5) -0.0011 0.0024 -0.4732

Table 2 indicates that the average Cumulative Abnormal Returns have a negative sign for all the event-windows. For the (-1,+2) event-window, the results are significant at the 10% level, indicating that positive events have a negative effect on the S&P 500. Furthermore, the results also indicate at the 10% significance level that a negative relationship exist between the positive events and the S&P 500 for the (-2,+2) event-window.

This result appears to be quit surprising. First of all, it was expected that the positive events would have an immediate effect on the stock market and not a delayed effect. Therefore, it is surprising that the results indicated a delayed effect instead of an immediate effect. Secondly, it is found, that there is a weak indication that positive events, related to the Vietnam War, affect the U.S. stock market in a negative manner. This contradicts the expectation that positive events affect the stock market positively. However, since the results are only significant at the 10% level, the evidence is very weak. Therefore, it is not possible to come up with strong conclusions indicating this negative relationship

Because the results for the (-1,+2) and (-2,+2) event-window are only significant at the 10% level, it is not possible to be completely certain whether or not the hypothesis that states that the U.S. stock market will react to events that state positive news about the Vietnam War should be accepted or rejected. However, since the evidence did not indicate the possibility of a positive relationship between the positive war-related events and the S&P 500, the second hypothesis (that states that positive events, related to the Vietnam War, lead to a positive return on the U.S. stock market) can not be accepted.

4.3. Stock Market Return Analysis of Negative Events

Table 3. T-Tests Negative events (AR 100 days). Event- window Average CAR σ _average CAR T-test (-5,+5) -0.0044 0.0024 -1.8336* (-4,+4) -0.0058 0.0022 -2.6957** (-3,+3) -0.0022 0.0019 -1.1595 (-2,+2) -0.0018 0.0015 -1.1786 (-1,+1) -0.0001 0.0011 -0.0642 (-1, 0) -0.0004 0.0008 -0.4719 (-1,+2) -0.0005 0.0013 -0.3432 (-1,+3) -0.0017 0.0015 -1.1289 (-1,+4) -0.0037 0.0017 -2.1456** (-1,+5) -0.0029 0.0019 -1.5467

This table indicates the student T-tests of the average Cumulative Abnormal Returns of the negative news-events, related to the Vietnam War when the average returns is calculated over 100 trading days before the event-window of an event. *=significant at the 10% level, ** = significant at the 5% level and *** = significant at the 1% level.

Table 3 shows that the average Cumulative Abnormal Returns for the different event-windows have a negative sign. The results of the student T-tests are significant at the 10% level for the 5,+5) event-window. Moreover, the results are significant at the 5% significance level for the (-4,+4) and the (-1,+4) event-window. These results indicate that negative events, related to the Vietnam War, lead to negative returns on the U.S. stock market when the event-window is centered ten days around the event date. Furthermore, the evidence for this relationship is

4.4. Stock Market Return Analysis of All Events

It would have also been possible to conduct the research in another manner. It would have been possible to use a dummy-variable method or to calculate the Abnormal Returns on different days after an event. The first test that has been conducted in this chapter could also be done differently. Instead of stating the returns in absolute values, it could have been possible to state the returns in their nominal values. This means that the returns will be stated as either positive or negative and that these values will be summed. To give a better indication of the influence of all events on the U.S. stock market, this test has also been conducted.

Table 4 shows the tests for the different event-windows when the Cumulative Abnormal Returns are stated in nominal values.

Table 4. T-Tests all events (AR 100 days)

Event- window Average CAR σ _average CAR T-test (-5,+5) -0.0032 0.0019 -1.6484 (-4,+4) -0.0037 0.0017 -2.1462** (-3,+3) -0.0022 0.0015 -1.4266 (-2,+2) -0.0027 0.0012 -2.1694** (-1,+1) -0.0010 0.0009 -1.0927 (-1, 0) -0.0004 0.0006 -0.7095 (-1,+2) -0.0017 0.0011 -1.5902 (-1,+3) -0.0019 0.0012 -1.4968 (-1,+4) -0.0021 0.0014 -1.4987 (-1,+5) -0.0020 0.0015 -1.3060

This table indicates the student T-tests of the average Cumulative Abnormal Returns all events, related to the Vietnam War when the average return is calculated over 100 trading days before the event-window of an event. *=significant at the 10% level, ** = significant at the 5% level and *** = significant at the 1% level.

4.5. Event-related Stock Market Return Analysis for Alternative Benchmark-Measurement Periods

In this section, the remaining robustness tests are discussed. These are the student T-tests with all the events and the tests with both the positive and negative events with the alternative

benchmark-measurement windows. The complete tables of the different tests can be found in appendices 6, 7, 8 and 9. Appendix 6 shows the results of the student T-tests for all events, when the returns are stated in absolute values. Appendix 7 shows the results of the student T-tests for the positive war-related events and appendix 8 shows the results for the negative war-related events. Finally, Appendix 9 shows the results of the student T-tests of all events, when the returns are stated in their calculated values.

4.5.1. Robustness Tests All Events Analysis (CARs stated in absolute values)

As appendix 6 shows, there is an indication of a trend when more or less trading days are taken into account to calculate the average returns. This trend shows that the results of the student T-tests decrease in value when more trading days are taken into account and that the results of the student T-tests increase in value when less trading days are taken into account. Therefore, the significance of the student T-tests rises when less trading days are taken into account. However, there are a few examples that deviate from this trend, especially when the average return is calculated over 110 trading days. The most significant results are obtained when 50 trading days are taken into account to calculate the average return.

4.5.2. Robustness Tests Positive Events Analysis

Table 5. Robustness Tests Positive Events

Trading days T-test (-2,+2) T-test (-1,+2)

120 -1.7389* -1.6662 110 -1.7732* -1.6802* 100 -1.8349* -1.7662* 90 -1.9354* -1.8456* 80 -2.0401** -1.9039* 70 -2.1524** -2.0482** 60 -2.2335** -2.1388** 50 -2.1048** -2.0064*

This table indicates the results and significance of the student T-tests for the (-2,+2) and (-1,+2) event-window. These tests are performed for all the events and are performed with the average Cumulative Abnormal Returns. * = significant at the 10% level and ** = significant at the 5% level and *** =significant at the 1% level.

4.5.3. Robustness Tests Negative Events Analysis

As shown in table 6, when more or less trading days are taken into account to calculate the average returns for the negative events, there is also a visible trend for the 5,+5), 4,+4) and (-1,+4) event-windows. The significance of the student T-tests increases when more trading days are taken into account and the significance decreases when less trading days are taken into account. However, the test with 110 trading days for the (-4,+4) and (-1,+4) event-window and with 70 trading days for the (-4,+4) event-window differ from this trend. When the average returns are calculated over 120 trading days before the start of the event-windows the highest T-values and most significant results are obtained. With this amount of trading days, the results are significant at the 10% level for the 5,+5) event-window, significant at the 5% level for the (-1,+4) event-window and significant at the 1% level for the (-4,+4) event-window.

Table 6. Robustness Tests Negative Events

Trading days T-test (-5,+5) T-test (-4,+4) T-test (-1,+4)

120 -1.9626* -2.7528*** -2.1639** 110 -1.8851* -2.6830** -2.1080** 100 -1.8336* -2.6957** -2.1456** 90 -1.5936 -2.4501** -2.0181* 80 -1.4239 -2.2623** -1.9477* 70 -1.4126 -2.2806** -1.8801* 60 -1.4032 -2.2260** -1.7705* 50 -1.2640 -2.1288** -1.7578*

4.5.4. Robustness Tests All Events Analysis

Table 7 shows the robustness tests of all events for the event-windows in which one or more significant results were found.

Table 7. Robustness Tests All Events (nominal returns)

Trading days T-test (-5,+5) T-test (-4,+4) T-test (-2,+2) T-test (-1,+2)

120 -1.6459 -2.1504** -2.1366** -1.5722 110 -1.6097 -2.1104** -2.1424** -1.5415 100 -1.6484 -2.1462** -2.1694** -1.5902 90 -1.5655 -2.1193** -2.1665** -1.5780 80 -1.6426 -2.1682** -2.2004** -1.5812 70 -1.6867* -2.2369** -2.2545** -1.6666* 60 -1.8305* -2.3081** -2.2602** -1.6465 50 -1.5807 -2.1514** -2.1279** -1.5452

This table indicates the results and significance of the student T-tests for the (-5,+5), (-4,+4), (-2,+2), and (-1,+2) event-window. These tests are performed for all the events and are performed with the nominal Cumulative Abnormal Returns. * = significant at the 10% level, ** = significant at the 5% level, *** = significant at the 1% level.

5. Conclusions and Discussion

In this section the results of the different tests are discussed and conclusions are drawn.

Furthermore, it summarizes the thesis, mentions again the weaknesses and biases of the research and gives some possible explanations for the results and recommendations for future research.

5.1. Summary and Conclusions

This thesis investigated the effect of positive and negative news-events during the Vietnam War on the stock market of the United States. As an indicator for the U.S. stock market, the S&P 500 index was chosen. Expectations were formed form the related literature and it was decided which events would be seen as positive events and which would be seen as negative events. Student T-tests were performed using average Cumulative Abnormal Returns to see if the results bore any significance.

The different tests lead to the following conclusions: A clear indication was found that events do influence the S&P 500 when the average Cumulative Abnormal Returns were stated as absolute values. The results were significant at the 1% level. The stock market return analysis, with alternative benchmark-measurement periods, showed a downward trend when more trading days were taken into account to calculate the average returns. The most significant results were obtained with 50 trading days. However, because this research focused on the daily returns it is easier to find proof for such a relationship than when longer returns, like monthly returns are taken into account. Because the stock market can fluctuate strongly in certain periods, it is easier to find a deviation from a certain trend when the returns are calculated over one or a few days than when more extensive periods are taken into account.

Furthermore, there was a weak indication that the positive news-events, related to the Vietnam War, have a certain influence on the U.S. stock market. When the average Cumulative Abnormal Returns were tested for the (-1,+2) and the (-2,+2) event-windows, there was proof at the 10% significance level that these events have a negative influence on the S&P 500. The stock market return analysis with alternative benchmark-measurement periods also showed a downward trend, for the significant results, when more trading days were taken into account. However, taking account of 50 trading days decreased the value of the student T-tests and therefore, the most significant results were obtained with 60 trading days to calculate the average return. The results for the (-1,+2) and (-2,+2) event-windows were significant at the 5% level with this amount of trading days.

The student T-tests performed with the negative war-related news-events indicated a relationship between these news-events and the U.S. stock market. The results were significant at the 10% level for the (-5,+5) event-window, and at the 5 % significance level for the (-4,+4) and (-1,+4) event-window when the average returns were calculated over 100 trading days. The tests

performed for the other event-windows did not lead to significant results. The stock market return analysis with alternative benchmark-measurement periods showed an upward trend, for the significant results, when more trading days were taken into account to calculate the average returns. Therefore, the most significant results were obtained with 120 trading days to calculate the average return. The results with 120 trading days were significant at the 10% level for the 5,+5) event-window, at the 5% level for the 1,+4) event-window and at the 1% level for the (-4,+4) event-window with this amount of trading days.

of the student T-tests when less trading days were taken into account). The most significant results were obtained when 60 trading days were taken into account to calculate the average returns. However, the results for the (-2,+2) and (-4,+4) event-windows are sill significant at the 5% level.

The results of the different student T-tests do not indicate that positive or negative events related to the Vietnam War have an immediate influence on the U.S. stock market. However, with extended event-windows, some significant results appeared. The surprising thing is that there is an indication that positive war-related events seem to influence the stock market two days after an event appeared in the media, while the negative war-related events seem to influence the stock market four days after their media appearance. However, the indication for the positive events is weak and similar results, for the (-4,+4) event-window taking account of the negative events and for the (-2,+2) event-window taking account of the positive events, are also found whit the student T-tests with the CARs of all events. Furthermore, the different sorts of events seem to influence the stock market in a negative manner. Therefore, the only conclusion that can be drawn is that the different news-events seem to influence the stock market negatively.

5.2. Biases and Weaknesses

One of the most important weaknesses of the data is that the events-data is dependent on what the writers of books deem important. This means that the data could be biased to certain events, which are seen as important by the authors. In order to correct this bias other sources have been used that report about the media coverage in war periods, but the data that can be found by using these sources will also depend on the preferences of the writers. However, the research done to showed that the researched sources focus on a lot of the same events.

such an index will not be used as a benchmark for the S&P 500. This weakness is also further reduced because this research is not focused on a single company and its stock return, but on an entire country and its return, the return used is already that of an index and therefore it is less prone to confounding effects. Furthermore, the event-windows are relatively short, so there are fewer opportunities for confounding effects to occur.

Another weakness of the data is that it has to be decided by the researcher/author what events are seen as positive, negative or not relevant for this research. Therefore, it might be possible that certain events that are seen as positive or negative in this research are seen as the opposite by others. The relative small amount of events is another weakness of the data.

5.3. Possible Explanations and Recommendations for Future Research

A possible explanation for the negative return of the positive events is the public distrust about statements made by the media and by the governments. Another possible explanation is the high expectations of the public. Before, the Vietnam War, the U.S. war machine had fought a number of wars abroad. In most of these wars the U.S. army had been proved successful in defeating the enemy. Furthermore, the Vietnamese army was expected to be much smaller than for example the German and the Japanese army. Therefore, the public might have expected that the U.S. army would win this war easily. Because of this, it could be that the news about positive events could not really satisfy the public since they expected better results. Because of this, the public could see this news as an indication that the war would last longer and therefore they could decide to invest in more secure financial instruments than the stock market.