Why do companies exclude stock-based compensation from adjusted earnings?

Companies say it is a non-cash charge whose size depends on valuation models and the share price at grant, so leaving it out gives a cleaner view of the operating business. The effect is a higher earnings figure. The employees still did the work, and shareholders still bear the cost through new shares.

Where do you find stock-based compensation in a 10-Q?

Look in the cash flow statement, where it is added back as a non-cash item near the top of operating cash flow, and in the notes on equity compensation. If the company publishes adjusted earnings, the reconciliation table in the earnings release lists it as one of the items added back to GAAP net income.

Does stock-based compensation lower EPS?

It lowers GAAP EPS twice over time. The expense reduces net income in the year of the grant, and the shares issued when awards vest or options are exercised raise the share count. Adjusted EPS usually strips out the first effect, and the second only shows up as the share count creeps higher.

Stock-Based Pay Is a Real Cost, Whatever the Adjusted Figure Says

Paying staff in stock costs no cash. So why count it against earnings? That question is the whole case for the adjusted figure, and it sounds reasonable until you ask who ends up paying, because the answer is the shareholder, in one of two ways, and neither of them shows up in a number whose first step is to add the cost back.

The belief

Stock-based compensation is non-cash. No money leaves the bank when a company grants shares or options. So adjusted earnings, which drop the charge, show the true profit. On this view the GAAP figure is dragged down by a bookkeeping entry.

That’s the myth. Test it on numbers.

The gap on one hypothetical company

A 50% gap is large. The adjusted figure says the company earns 60 cents a share. GAAP says 40. Every valuation multiple you compute from the adjusted number makes the stock look a third cheaper than it is on GAAP earnings, since 40 cents is two thirds of 60, and that is before anyone asks what the $40 million was paying for. Set it beside a peer that pays its staff in cash and you are comparing a company that reports its whole wage bill with one that reports part of it.

The $40 million was paying for work, the same work a salary pays for. Pay the same $40 million in cash and everyone calls it an expense.

Where the cost arrives

Shares handed to employees come from somewhere. The company either issues new ones or buys old ones back.

Take the dilution route first. Suppose net income stays flat at $80 million. Awards vest, and the share count grows from 200 million to 206 million.

Each existing holder now owns a smaller slice of the same profit. Nothing left the bank account, and holders are poorer anyway.

The other route is cash. Some companies buy back enough stock to offset what they issue to staff. The share count holds steady and the dilution never shows. The money for those buybacks is real and it leaves the company, recorded in the financing section of the cash flow statement, which means the non-cash expense has become a cash one, just on a different page of the filing from the one that carries the adjusted earnings. The cash flow statement lesson shows where each line sits.

The objection: the models are guesses

The strongest case for excluding the charge is about measurement. The expense for options comes from a valuation model, the grant-date value can differ from what the shares are later worth, and some options expire worthless, with the expense already booked for vested ones staying on the books. Analysts using a fully diluted share count also worry about counting the cost twice.

Both points have something in them. Neither gets you to zero. A cost that is hard to measure is still a cost, and the reported figure is at least an estimate made under a published accounting standard. The double-count worry is about timing. The diluted count reflects grants already made. This year’s expense reflects grants that will swell next year’s count. Count each grant once and the worry goes away. Leaving out all $40 million answers the measurement question with zero, the one value certain to be wrong.

Reading the reconciliation

The reconciliation table is the one place the release has to show its work. Start with GAAP net income at the top. Read down the list of add-backs. Stock-based compensation is usually among the largest recurring ones for companies that pay heavily in equity, and a line that appears every single quarter is an operating cost by any sensible reading, however it is labeled. The answer on why adjusted earnings run higher than GAAP earnings goes through the other common add-backs, and Ten-Q Very Much covers the filings where they appear.

Then check the share count. Compare the diluted weighted average shares on the income statement with the same quarter a year earlier. Rising means you are paying in dilution. Flat, with a large repurchase line in the cash flow statement, means you are paying in cash. Either way, you are paying.

Use GAAP, or do the subtraction yourself

The call is simple. Use GAAP EPS when you compare companies. If you prefer the adjusted figure for its other exclusions, take the stock-based compensation line from the reconciliation and subtract it again: on the example, $0.60 less $40 million over 200 million shares, which is 20 cents, brings you back to $0.40. The Below the Beat quiz tests the habit on earnings headlines.

The call bends where stock pay is small. Then the gap between the two figures is a rounding matter, and arguing over which one to use costs more attention than it saves, so pick either and move on to the margin and the cash flow. It holds hardest for young companies that pay much of their staff in equity. There the adjusted number can describe a profitable business that GAAP says does not yet exist.