The profit that belongs to your share
After a company has paid for materials, staff, borrowing costs and tax, whatever remains is profit. Earnings per share asks a simple follow-up question: if that profit were split evenly across every share in existence, how much would land on yours?
This is the number that moves share prices four times a year, because it is the closest thing to a scoreboard the market has. When commentators say a company "beat expectations", they almost always mean this figure came in above what analysts had guessed.
Apple, step by step
Take roughly $97 billion of annual net profit and about 15.2 billion Apple shares in existence. The arithmetic is 97,000,000,000 divided by 15,200,000,000, which gives approximately $6.38 of earnings per share.
That single figure now lets you do something you could not do before: compare Apple to a company a hundredth its size. Total profit would be a meaningless comparison. Profit per share puts both businesses on the same footing, one share against one share.
Figures here are rounded recent full-year approximations, shown so you can follow the method rather than memorise the result.
The buyback illusion
Apple is the clearest example of the trap in this number. The company spends heavily buying back its own shares, which shrinks the denominator in that division. Hold profit perfectly flat, retire 5% of the shares, and earnings per share climbs by about 5.3% anyway.
A headline reading "earnings per share up 5%" therefore does not always mean the company earned more. Sometimes it means the same profit is being divided among fewer owners. That is genuinely useful to you as a shareholder, but it is a financial manoeuvre rather than a better business, and the two deserve different levels of enthusiasm.
The check takes ten seconds: look at whether total net profit grew as well. If profit grew 5% and earnings per share grew 5%, the business improved. If profit was flat and only the per-share figure rose, the share count did the work.
One good year proves very little
Profit is the most easily distorted line in company accounts. Selling a building, winning a lawsuit, or writing down the value of a failed acquisition can swing a single year’s earnings per share dramatically without telling you anything about how the business trades day to day.
The defence is to read five years rather than one. A company whose earnings per share has climbed steadily through a recession is telling you something real. A company with one spectacular year surrounded by mediocre ones is telling you about an event, not a trend.
Making it a rule inside Stax
The rule worth writing is rarely "high earnings per share". A large figure often just means the company has relatively few shares in issue. The rule worth writing is consistent growth in earnings per share over several years, which describes a business that keeps getting better at keeping profit.
Once you have a list of companies that pass, the honest next step is a test. Stax replays your plan across five or more years of real prices with fees deducted from every simulated trade, and eight independent checks audit the run before you see the result. That tells you how the rule behaved, including in the years it struggled.