The earnings per share metric is a valuable tool for investors to use when investigating the profitability of a company against its outstanding equity. From there, the components that come together to make this metric are as useful to investors as the output that they create, because it allows us to understand how it is that different earnings and dilution scenarios would change a single share’s claim against the company’s overall earnings, depending on the nature of the securities that are being used to finance this dilution.
Calculating earnings per share is simple, you take a company’s net incomes, you subtract out preferred share dividends, and then you divide the result by the average number of shares outstanding over the course of the year. The preferred dividends are subtracted out from the profits to show only those earnings that are available to common shareholders, and therefore give us an understanding of what an investor’s nominal claim to the company is. However, if the company changes the volume of shares it has issued throughout the year, it will also change the way in which we calculate actual number of shares used in the equation.
There are three main ways for a company to change the average number of shares outstanding during the year: a buyback, an issuance, or a split/reverse-split. In a split/reverse-split situation, the net effect on proportionate shareholder ownership remains constant, because everyone still owns the same ratio of shares to one-another. This means that a split doesn’t change an individual investor’s EPS claim, but instead impacts the EPS that an incremental purchase would have access to (because you’d need to buy a greater volume of shares to have a similar claim).
In those instances where a company chooses to issue or buy-back shares part-way through a year, it is important to take into account the weighted impact of such an event. This is accomplished by multiplying the change in share volume by the proportionate date of the change. For example, if a company bought back 1,000 shares in March, we’d multiply 1,000x(3/12), because March is the third month out of 12. This shows us the weighted number of shares bought back over the course of the year, and ensures that we are calculating a final result that accurately represents the time value of shares outstanding.