Stock Analysis - InterContinental Hotels Groups Stock Price Reaction - Essay Example

?Introduction: The relationship between special information, event announcement, and stock return behavior is most specifically concerned with an increasing number of investors, portfolio managers, regulators, and researchers. This report is directed towards helping the reader/analyst to examine the InterContinental Hotels Group’s (IHG) stock price reaction with regards to its disclosure of company specific events. In order to conduct such a level of investigation, the relationship, method of event study, and other key factors will be utilized within this process. Before delving into the research, it is necessary to integrate an understanding of key background information. This includes, but is not limited to, the background of IHG and a brief statement of four special informational events which will be analyzed in depth within this brief response. Accordingly, a literature review will be shown which helps to underscore and elaborate upon many of the specific theoretical interpretations of the methods utilized within this report as well as relate pertinent information with regards to how such a process has been traditionally understood. Additionally, a demonstration of development of an event study is included and defined. Furthermore, the third part of the piece will include a methodological review. The explanations of approaches which have been utilized to conduct this report will be given as well as some assumptions and critical common factors that are associated with this methodology. Likewise, a research design and results section will comprise the fourth part of this analysis. Ultimately, the issues which will be explored and/or discussed will be analyzed based upon an event study methodology. This part will begin with defining the date of events and then choose the estimated period and test period. Following this, marker models will be chosen as a specific benchmark for abnormal returns. Similarly, a linear regression for the market will be chosen to shed further light on the theoretical interpretation and key data presented. In such a way, ordinary least square assumptions will be necessary to estimate the stability or whether or not it exists to a discernible degree at all. Based on the results of linear regression, the abnormal return will be calculated. Next, the Cumulative abnormal return, or CAR will be utilized as a means of indicating the effect of events specific to those which have already been related and identified. Further, a T-test by CAR will be presented to show the significant of these events. A final portion of the research will necessarily be the conclusion of this report. Within this conclusion, it is the intent of this researcher to show the basic information of this report and summarize it according to the inference which has been drawn. Background information: InterContinental Hotels Group (IHG) is global hotel group which operates nine hotel brands; which in turn comprise a total of 4,600 hotels in more than 100 countries and regions, more than 676,000 rooms. IHG was established in 1946 and is the world's largest and most widely distributed network of professionally managed hotels. With regards to the business model that IHG exhibits, this can be understood in three distinct ways. First, IHG operates as a franchisor, second as a management accessory, and thirdly it operates upon a rental basis. In addition, the franchise has been recognized as exhibiting an effective business philosophy by many leaders within the industry. This report chose four specific events to explore the relationship with the IHG stock price volatility for a period time. These four events are shown below: (1) On the 18th April 2012: IHG shows its support for small business owners, entrepreneurs and Road Warriors. (2) On the 7th August 2012: IHG reported interim results from the 30th June 2012 and announced $1bn return of capital. (3) On the 8th October 2012: IHG reported the results of general meeting about Special Dividend and Share Consolidation. (4) On the 12th November 2012: IHG announced that they will have a new Vice President of Development within the European market. Literature review Efficient Markets Hypothesis (EMH): The Efficient Markets Hypothesis (EMH)originally came to prominence in 1970; promoted and defined by Eugene Fama. Moreover, the Efficient market hypothesis was originally developed by Louis Bachelier (1900); who represented the perspective of the stochastic process. Bachelier studied the movement of Brownian motion, stock price changes, and recognized that the effectiveness of the market information, present events, and even the discounted value of future events, are reflected in market prices. As such, the efficient market theory relates that the market price already contains all the available information. Based on the past information and price changes in last period, it is not possible to make money. EMH contains three underlying assumptions. Firstly, investors in the stock market can depend on fundamental analysis to forecast the future rate of return and carefully delineate between risks and return trade-offs. Secondly, the price of the stock reflects the balance of supply and demand of these rational people who want to buy exactly equal to what they want to sell. In other words, arbitrageurs will immediately move towards buying or selling stock when the price is overvalued or undervalued. Thus, the stock price changes rapidly to enable the two to become equal once again; not unlike the supply and demand curves always seek a point of equilibrium. Thirdly, the stock price can fully reflect all available information of the asset. This in turn is termed "information efficiency". The price of the stock will then notice a change due to these changes of information. Thus, as positive news or negative news spreads, the price of the stock begins to change; then the stock price either rises or falls as it is known to the investor. On the other hand, these assumptions can be difficult to set up. The investors must consider some costs, such as transaction costs, tax, and opportunity costs. Based on the nature of market efficiency, Fama raised three forms of EMH. (1) Weak-Form Market Efficiency: the price of stock fully reflects all the past history of the prices of securities, including stock trading price, trading volume, and the amount of financing. In this form, technical analysis does not work anymore. This is to say that a fundamental analysis may earn more profits as compared to a non-fundamental analysis. (2) Semi-Strong-Form Market Efficiency: The information published about the business prospects of the Company fully reflect on the price. Technical analysis and fundamental analysis both are inefficient, only the insider message may raise excessive profits. (3) Strong-Form Market Efficiency: it includes not only all public information, but also includes all the insider information. No methods to help investors earning a high profit are included; nor even insider information.  The Role of Statistics in Business and finance In business applications, Statistical techniques are regarded as a necessary tool. The data that is derived from statistical techniques will be shown on the financial statements and some analytical reports. Based on the results, corporations can make a quick and accurate adjustment for their strategy and investors can make informed decisions. The importance of statistics is never to be underestimated as researchers and managers work to develop and learn the statistics of techniques; especially with regards to stock prices. Comparative to efficiency markets, the response of stock prices is also quite different. Scholars in the field use statistical data to analyze the relationship between stock price movement and other factors. In 1973, economist Jordan published an article entitled “An empirical investigation of the adjustment of stock prices to new quarterly earnings information”. This paper shows that new earnings information have an effect on the stock price. Moreover, the companies which experience high growth find it more difficult to adjust the new earning information than those companies with low growth. Tumarkin and Whitelaw (2001) examined the effects of the postings in Internet financial forums on stock prices. The article demonstrates that there is a relationship between Internet message board activity and abnormal returns; especially in the companies that belong to the Internet services sectors. Charels(2010) investigated the relationship between the stock market performance and environmental news discloses from corporations. This article is dedicated to Green rankings and Scores and seeks to state that these have little to no influence on the market price behavior. In addition, this paper shows that there is a negative effect by Overall Newsweek Green studies for all of the stock prices for the companies in this study. Event study: An event study is a study which is performed in a short period of time before and after the announcement of special informational events. As such, this oftentimes portends a change in stock returns due to the changes of investor’s behavior in this period. Based on the analysis of statistics, this method is to judge the impact of the event on the shareholders' wealth and enterprise value. The event study depends on the rational market models, which indicate market efficiency. The estimation of a typical event study needs to be conducted and test the hypothesis; which is the effect of special events after announcement on value of firm. The event study is widely used in the field of finance and accounting. It was first used by Dore (Dolley, 1933) who adopted the method to study the impacts of stock splits on stock price changes. After that, the method has a wide range of application for securities market research. Ball and Brown (1968) introduced the event study methodology currently used to study the information content of earnings. Further, Fama (1969) studied the effect of stock dividends by utilizing Brown’s methodology. Brown and Warner (1980) helped to further improve the event study methodology by revising several statistics. The event study method came to be known as the "semi-strong form validation method”. The event may be the date of the company's major events, such as company merger announcement date, issue date, earnings announcement day, the stock dividend payment date, it also can be the general economic events, such as inflation, trade deficit occurred. This can be broadly understood up until the time that Kholar'i and Warner (2006) stated the two main changes in event studies. The first is that the daily measurement is more prevalent than the monthly one and thereby has more accuracy for the abnormal return. And the second is that the methods that are to measure the abnormal returns and test statistics became more perfect. Methodology Review: Under different forms of efficiency markets the response of stock price will necessarily be different. Sometimes the effects of events may be early or delayed; other times they might overreact or lack a response entirely. Event study is regarded as a statistical methodology to investigate how, from an economic standpoint, the event affects the value of corporation. Thus, for this research the relationship between special events and stock price performance of IHG, the event study would be utilized. The level and scale of interval measurement: The length of the interval may have different effect on the result of calculating returns. Mostly, monthly, weekly and daily intervals are used to estimate the stock price. According to Brzeszczynski et al. (2011), a short-length may provide an accuracy result for returns. So it comes without question that a short interval will be better for estimating effect of information events. Estimation period and test period: In order to analyze the effect of events, the data must be collected both ex-ante and ex-post. The researcher must then decide the length of the event window. There are two period in the event window, estimation period (EP) and testing period (TP). The estimation period is a period without events. This period helps researches to indicate the normal stock return without events. Mostly, the period before events is used as estimation period. The testing period includes the date of event and the period before and after the event. It is a necessary period to demonstrate the predicted abnormal return. Most importantly, the EP and TP cannot overlap. Effects of event will be shown by comparing two periods. Two common forms are shown below: The estimation period is from t=-T to t=-s while t=-s to t=t via t=0 is the test period. Some researches design the estimation period on the both sides of test period. Others have one estimation period before the testing period. The empirical models (benchmark): There are several empirical models to calculate the abnormal return:Mean-adjusted returns model, Market-adjusted returns model ,Risk-adjusted returns model and market model benchmark. Mean-adjusted returns model: This model maybe the simplest model, it assumes that the normal return is constant and the normal return of a stock and the mean return of estimation window are equal as follow: And then the difference between the normal return and obvious return is regarded as the abnormal return as below: = It shows that the abnormal return by using this model depends on the previous mean return within the estimation period. This model must be systematically biased within a bull and/or bear market. Market- adjusted model: The main assumption of market- adjusted model is that the expected return of a stock is equal to the market return. It is shown below: This model just applies to estimate the market wide movements that take place within the same time of the event for the company. Based on this model, the abnormal results are shown below: Risk-adjusted return model The advantage of risk-adjusted return model is that it considers both market-wide and firm specific factors of each stock. This model assumes that a particular phase of the expected rate of return of individual securities in the event period is same as the market rate of return as follow: , E= If return rate under the influence of event is not equal to the return rate without effect, the abnormal return can be calculated as follow: , E= The market model benchmark The assumption of this model is that asset returns are jointly multivariate normal and distributed independently and identically by time. In addition, it assumes that there a linear relationship between the stock return and the return of a market portfolio or index. This model is widely used to estimate influence of event in practice. The return is based upon this model shown below: The stock return is the dependent and the market return or index is the independent variance. Where is independent and indicative of random variables with a mean of zero. This report will be under the unsystematic component. In this model, both and should be estimated, and the ordinary least squares is necessary to estimate. So, the abnormal return is calculated by: According to Beaver (1981), the reasons for popularly employing the market model benchmark are: 1. This model could provide smaller variances of abnormal returns. 2. The market model will lead to more powerful statistical tests. 3. This model can produce smaller correlations through stock abnormal returns and provide a closer consistency with standard statistical tests. The benefits of employing the MM model may depend on the R square of the MM model regression. R square has a positive correlation with variance reduction of abnormal returns; thus a higher R square will lead to a larger gain through a greater variance reduction. Calculation of stock returns Arithmetic returns and logarithmic returns are underlying methods used to calculate stock returns. The arithmetic returns method is as follow: The logarithmic returns method is as below: where is donated to the price of the end of period t; shows the price of the period t-1; is equal to the dividends paid at period t. In fact, the logarithmic returns method is widely used to calculate the returns. The reason for this is that this method is better for both theoretical and empirical. In aspect of theoretically, this method is good at capturing the compound effect of rate of return. And it is more likely to be normally distributed in empirical aspect. Hypotheses test: For this type of study, the most common form of null and alternative hypotheses are showed below (Gonedes 1975): HN: f (|yi) – f () = 0 for all yi HA: f (|yi) – f () ? 0 for at least one yi Where is the return of stock j in an event period of interest; Yi is a signal from information structure ? announced in the event period that potentially affects stock j. f (Rj|yi) is the distribution of Rj conditional on the information signal yi from the information structure ?. f(Rj) is the marginal distribution of Rj. There are a variety of dimensions with calculation of abnormal returns. Cumulative abnormal returns (CAR): Estimation effect of event information disclosed is most used in the abnormal return over an estimation period in financial research. Mostly, researchers investigate the movement of stock price in a close estimation by applying cumulative abnormal returns. The wide use of cumulative abnormal returns is due to it can fully show the effects of events on stock price and it also can solve the uncertainty over the specific event date. Most researchers call it CAR and calculate as follow: In continuous time, the CAR demonstrates the abnormal return on a portfolio which is rebalanced every time period to provide an equal weighting in each stocks. Research design and results 1.Four specific events of IHG This report is dedicated to investigate the relationship of the movement of stock return and the announcement of specific information events. The nature of four specific information events chosen will be demonstrated. The first event is on 18th April 2012, IHG announced that it will support for small business owners, entrepreneurs and road warriors. The company operate a promotion strategy offering around a $200 cash reward for travelers who stay on weekdays. Based on a recent survey, it found that travel services are among the top purchases that small business owners plan to make in 2012, with more than half of survey respondents planning to book a hotel stay for their travel needs. At the same time, IHG announced that they would support for small business owners. And the data of first event is 18th April 2012. IHG reported interim results to 30th June 2012 and announced $1bn return of capital on 7th August 2012. This is the second event. Interim results reported that IHG had a solid performance during this period. The total revenue is $10.3bn, up 7.3%. The interim dividend increased by 31% reflecting these results. The company is confidence in the future prospects. They stated that they would continue to strengthen both existing and new brands. At the same time, IHG decided that they would return $1bn to shareholders. So, 7th August 2012 is the date of the second event. The third event is that IHG reported the results of general meeting about Special Dividend and Share Consolidation on 8th October 2012. The company state that the share consolidation is US$1.72 under the special dividend context. The special dividend should to be paid on 22nd October 2012. And the date of the third event is 8th October 2012 because they disclosed the event on this day. The last special event is that of IHG’s announcement of a new Vice President of Development in Europe on the 12th November 2012. Philippe Bijaoui will be the new Vice President of Development in Europe. He will lead all IHG brands throughout Europe. He has over 18 years of international experience in the hotel sector and real estate sectors. IHG believed the new vice president will bring some positive influences for IHG brand as a whole. A future 86 hotel will be operated by Philippe in Europe. The date of this last special event was 12th November 2012. 2. Decision of EP and TP The estimation period of this report continues for two years from 1st April 2010 to 1st April 2012. IHG is a company of FTSE 100 index, so, the FTSE is used as an estimation index for this report. The date of four special events is as follow: first event (18th April 2012), second event (7th August 2012), third event (8th October 2012) and fourth event (12th November 2012). Thus, the test period of this report is from April to November 2012. 3. Decision of model benchmark 3.1 calculate the returns The logarithmic method is chosen to calculate the returns in this report due to the advantages of this method. The adjusted close price is used to calculate the returns. The adjusted close price has considered the dividend factor; thus the formula should be as below: In order to investigate the effect of four special events, aspects of model benchmark and the market model benchmark have been chosen to estimate the expected returns. It is because the several advantages of this model. And the expect returns could be calculated by the formula below: In this report, where is showed the expected returns of IHG stock. denotes the FTSE 100 index return. is a systematic component linearly related to market return and ? and ? must be estimated here. is a unsystematic component uncorrelated with market return. So under the special events, the abnormal return could be calculated within the formula as follow: In order to insure the correct of the formula, ? and ? must be estimated here. For estimating ? and ?, a model called linear regression model is presented as: Yi = ?0 + ?1*Xi + ?i Where: 1. Yi is a dependent variable (or the LHS variable or regressand) 2. Xi is a independent variable (or the RHS variable or regressor) 3. ?0 denotes the intercept and ?1 denotes the slope , both ?0 and ?1 are parameters for the model and unobservable. (?0 + ?1*Xi) is the regression function. 4. ?i is the random error which is not from observation. Therefore the relationship between Y and X is not deterministic. 5. Y and X are standard writing in the model but we often use Rt or Rm to present them. The most important part is estimating the parameters or coefficients of linear regression model. Essentially, we need to find the straight line which fits the given data. For the reason of existence of vertical error, minimizing the total errors is of considerable significant to getting the best fit. This equation denotes the sum of square of errors , we need to fit and to minimize the total squared errors. So we should use the ordinary least squares estimators which are presented as follows: After minimizing the errors then we have this: Where X bar and Y bar are average values of X and Y respectively There are some issues in these formulas that the reader should pay attention to: (1) All the observations are equally weighted in this minimization upon the condition of all the observations are independent. In other instances where some observations are deemed better than others we do weighted least squares. (2)What we have done is to fit a straight line through the data points. (3)The slope is the sample covariance divided by the sample variance. (4)The estimated line goes through the average values. Regression of MM model Based on the market model benchmark formula: In order to regress and estimate the ? and ?, Daily stock price of IHG and daily FTSE 100 are used in to the formula from 1st April 2010 to 1st April 2012. The continuous two year daily stock prices are entered Excel by using regression analysis to get the value of ? and ? as exhibited within Table 3. The result of this regression is shown as follows: Table 1 Table 2 Table 3 Depending on the summary out of regression, ? is T and the positive slope coefficient ? (T). It is showed that T observations used in this regression. The value of t-statistic is calculated as T, it means that slope coefficient is statistically significantly different from zero. And the value of R2 (T) shows that the regression line does not fit the data well enough. The result of abnormal returns A new formula generated by entering the ? and ? and the abnormal returns is calculated by this new formula as below: = Rjt –(T+ TRmt ) (Where denotes the abnormal return) Thus, the abnormal return can be calculated. The abnormal return of IHG’s four special events on the event date show below in Table 4: Table 4 In the abnormal table, there are positive and negative abnormal returns in these four special events. T1 and T2 is the positive abnormal return and T3 and T4 represents the negative abnormal return. It means that the first and second event bring a favorable mention on the stock price of IHG. And the third and fourth events have an unfavorable mention on the IHG stock. Based on Table 4, the data point to the fact that special events have an influence on the stock price of IHG. In order to indicate the effect, the value of CAR will be calculated in next step. Result of CAR Insure the accuracy of the result, 10 days before and after event date are use as a test period of a special event (t=0 is the date of event, t?-10…0…10 ). Based on the CAR formula (), the CAR of four event could be calculated as in the table below: TABLE 5 The table 5 show the CAR of each event, CAR (first event)= T, CAR (second event)= -0.141453524 , CAR (third event)= -0.051273266 , CAR ( fourth event)= -0.179375249. the result of CAR show, event 1 and 2 have a positive effect on stock price and event 3 and 4 bring a negative effect on the stock price of IHG. Base on the result of abnormal return and cumulative abnormal return, it mean that the four special events have effect on the stock price of IHG. It does not indicate how a significance of these abnormal returns related to the four events. A t-test needed to demonstrate that these abnormal returns are statistically significantly different from zero or not. T-test Null and alternative hypotheses assumed with a n-2 degree of freedom in a two tailed result as shown below: This hypothesis will indicate that whether the four specific events disclosed have an impact on abnormal returns or not. The formula of t-test shows as below: (Where is the abnormal return of day t; denotes the standard deviation of abnormal returns with n as the sample size) Through the calculation by Excel, the value of t of each event is as follow: Table 6 The table shows that: In specific event 1, t=T, based on the degree of freedom is T. So the t518,0.05= T. Under a two-tailed test, if -1.96 Read More