Stock Splits

Chapter 5

Empirical Results

In the following chapter we initially present a summary of the data statistics we have applied throughout our analyses in section 5.1. Next, we present and comment on the empirical results from our event-time and calendar-time studies insection 5.2. We finish the chapter by elaborating on relevant robustness tests insection 5.3.

5.1 Summary Statistics

This section includes summary statistics concerning stock splits, insider trading actions, and lastly stock splits with insider trading actions.

60 Empirical Results Figure 5.1: Annual Frequency of Stock Splits and Stock Dividends

The figure shows the historical development in stock splits and stock dividends. The numbers are reported as a fraction of the total number of listed stocks at the end of a year. Thus, the number of stock splits or stock dividends are normalised compared to the total number of listed stocks in our sample (NYSE, AMEX, and NASDAQ stocks).

Figure 5.1 shows that the annual frequency of stock splits and stock dividends have de-veloped somewhat similar throughout the sample period, albeit at clearly different levels.

Stock dividends have throughout the sample period moved within the interval between 0.0%

and 0.3%, whereas stock splits have shown a higher volatility, namely in the range between 0.0% and 1.5%. The data shows a large peak in stock splits between the 1980s and 1990s, followed by a large drop in the percentage of stock splits. Overall, stock splits and stock dividends have become less frequent in the past 30 years, enduring high volatility in the annual frequency of stock splits over the sample period. Most recently, the frequency of stock splits has significantly dropped following the recession in the beginning of the 2000s in the US and later during the the financial crisis in 2007 to 2008. Thus, when considering the annual frequency of stock distributions infigure 5.1, stock splits and stock dividends may have become less relevant compared to historical levels. However, managers and board of directors may stille be cautious as the markets have been through challenging periods since the beginning of the 2000s with recurring financial crises. In addition, firms are entering a new level of low market growth for the coming years, which has moved a potential stock split lower down on the corporate agenda. Lately, managers have instead favoured stock buybacks as the preferred corporate action. We still regard stock splits as important in the financial markets and hypothesise that stock splits will return as the markets normalise.

Next, table 5.1 shows that 83% of the stock splits during the entire sample period

Summary Statistics 61 is derived from companies that split multiple times. Thus, a significant part of the com-panies engaging in stock splits will be recurring throughout the sample period. Krieger and Peterson (2009) argue that firms that have already conducted a stock split are more likely to split again, and that the prior stock split’s announcement effect is a useful predic-tor of future splits. This implies that the market will partially incorporate the information implied by a stock split based on the abnormal returns evident from the previous stock split.

Table 5.1: Frequency of Splits and Number of Firms

The table shows the sample firm’s frequency of splits during the period 1926-2015. Throughout the period, a total of 8,905 firms conducted 22,638 splits with 83% of the stock splits stemming from firms splitting multiple times.

Next, we present the development in the split factor in relation to the average FACPR value as well as the size of the stock split compared to the general market.

Figure 5.2: Average Size of Split Factor (FACPR Value) and Normalised Split value

The two figures show the average FACPR value as well as a measure of the value of the stock distributions. The right hand side of the figure shows the percentage size of annual stock splits in dollar value divided by the end of year total market capitalisation. The figures show a higher FACPR value when less firm split their stocks. Similarly, there is a higher average split size relative to the market when there is a fewer number of firms splitting.

Figure 5.2shows the historical average size of the factor to adjust price (FACPR).⁶² The

62Everything else equal, a higher split factor increases the number of outstanding shares issued when a firm undertakes a stock split.

62 Empirical Results peak in the beginning of the period is prone to outliers, as the number of splitting firms was very low. Similarly, we see a slight increase in the factor to adjust price after the decrease in the number of splits per year (cf. figure 5.1and5.2). Thus, we argue that the evidence suggest a higher split factor when less firms are undertaking stocks splits.

However, by just looking at the factor to adjust price in the stock split we can infer limited information about the size of the stock splits themselves. We can only obtain information about the size of the split by looking at the size of the firms that undertake a stock split. Infigure 5.2 we also present the market value of the stock splits compared to the market value of the entire market capitalisation. Thus, we multiply the factor to adjust price with the stock’s market capitalisation at the day of the stock split. We then aggregate the annual market capitalisation stemming from stock splits and divide the dollar amount (value) with the total market capitalisation at the end of the year.

Figure 5.2 shows that the relative size of a stock split (market value of additional stocks issued at the stock split) compared to the market is between 0 and 1.5% of the total market capitalisation throughout the sample period. However, in recent years, the market value of the stock splits has been more volatile and peaked twice towards the 2010s. This may reflect the fact that a few large corporations has undertaken large stock splits. E.g., Apple engaged in a 7-for-1 stock split in 2014. Because of the size of the stock splits in recent years (2008 to 2015), we still find it very relevant to study stock splits.