• Ingen resultater fundet

4. EMPIRICAL METHODOLOGY

4.1 D ATA

In order to answer our research questions a rich set of data has been collected. Furthermore, due to the scale of the study, secondary data has been applied in the empirical analysis. Using secondary data give us the opportunity to conduct a longitudinal analysis, i.e. using concentrated samples over a longer time period, thereby allowing us to test for the profitability of momentum investment strategies and for the persistence of the January Effect. Moreover, secondary data has been used extensively by previous existing literature for their empirical analysis.

4.1.1 Data Sources and collection

The data collection process can be divided into two steps. In the first step, S&P Capital IQ, a comprehensive market intelligence platform containing financial information about both private and public companies, was used to collect monthly information on market capitalization and dividend- and stock split- adjusted closing prices. Data was collected for companies that as of 31st of January 2021 were listed on the relevant stock exchanges. Please see section 4.2 for a more in-depth explanation of the stock exchanges included in the analysis.

In the second step, information about companies that had been delisted from the stock exchanges examined during our sample period was collected as follows. Firstly, to identify the companies which had been delisted we reviewed Nasdaq’s Nordic Surveillance Reports from 2006 to 2020. Once all the delisted companies were identified, S&P Capital IQ was used again to collect their historical data, thus enabling us to conduct a more thorough empirical analysis and reduce survivorship bias. If we were to not include delisted companies in our sample this could potentially distort the results of our findings, as the sample would not provide a complete representation of the historical market situation and equity universe. We refer to appendix 1 for a complete list of companies and their tickers included in the data sample.

4.1.2 Data Variables

The following section presents the variables employed in the analysis. The data sample consists of monthly dividend and stock split adjusted closing stock prices for each company included. In order to gain a better understanding of the overall value of the stocks and make informed investment decisions adjusted stock prices have been applied. Adjusted stock prices includes the impact of dividends, stock splits, seasoned equity offerings, etc. Thus, using adjusted stock prices have allowed us to conduct a more accurate analysis of historical performance as the impact of beforementioned factors are excluded.

Moreover, for the purpose of portfolio formation and conducting sub-analyses we have collected the market capitalization of each company. All market capitalizations have been converted to DKK through the S&P Capital IQ database applying the appropriate historical spot rate. This has been done to easily compare the size of companies across multiple stock exchanges. The codes applied in the S&P Capital IQ excel plugin to collect historical data are as follows:

- Adjusted closing prices = IQ_CLOSEPRICE_ADJ - Market capitalization = IQ_MARKETCAP

4.1.3 Data Intervals

Our analysis has been conducted based on-end-of month adjusted stock prices. Using monthly data points is in line with a large part of previous literature and allows us to compare our results with previous findings. Alternatively, we could have used daily or weekly data points in our analysis, as this would have given us a more comprehensive collection of data and consequently provide us with

a basis for a more nuanced and exhaustive analysis. However, it can be argued that in practice investors will most likely not develop an investment strategy based on daily or weekly rebalancing of the portfolio due to transaction costs, time availability, etc., why an empirical analysis based on monthly data points is deemed adequate by the authors of this study.

4.1.4 Data Adjustments

A number of selection criteria has been applied in our data collection strategy. In line with previous studies, we have excluded listed mutual and investment funds from our data sample. These have been excluded from our study due to their strong correlation with other stocks applied in our analysis.

Another important factor to consider when employing a momentum investment strategy is share liquidity, i.e. how easily stocks can be bought or sold. Thus, the next criteria in our data collection strategy was in relation to the share class included in the analysis. For each company we have included only one share class, since they are often highly correlated. Previous studies have excluded Class A shares in their analysis as these are argued to be less liquid than Class B shares. However, we found that for some companies Class A shares had a higher trading volume than Class B shares. Hence, in order to ensure that our data sample only includes shares with the highest liquidity, we have examined the trading volume of each company’s stocks currently listed in the last full calendar year, i.e. 2020, to identify the most liquid stocks. In terms of delisted companies, we examined each company’s last year of trading to identify the share type with the highest liquidity. In summary, for companies with multiple share classes and companies listed on more than one stock exchange, e.g. Nordea which is listed in both Denmark, Sweden and Finland, the share with the highest trading volume was included in the final data sample.

In addition to A- and B-shares, some companies have preference shares. Preference shares may be defined as a hybrid of common stocks and bonds, since they have different rights than common shares, e.g. preference shareholders receive dividends before common shareholders. Furthermore, in most cases preference shares only comprise a small percentage of total shares outstanding, why they are typically less liquid than common shares. As a result of these factors, preference shares have been excluded from the data sample.

RELATEREDE DOKUMENTER