Perhaps we should require the companies to list the options effects this way... as a "dividend not paid because the cash was used to offset option dilution."
I suggest to you that "Earnings" is supposed to be the total amount of a dividend that could be paid. So whether you report "dividend that can't get paid" or "reduced earnings", it's the same thing in the end.
But your proposal lends itself to some interesting thinking. Let's look further.
In the end, you need to compute the actual (or expected) dilutive effect of option exercise on a year-by-year basis.
Let's say I'm Cisco. I issue a stock option to an employee for $20.00. The employee exercises this stock option when the price is $22.00. Cisco then buys back the share on the open market for $19.00, what is the cost of that option?
Well, the actual effect on the company's ability to pay dividends was $2.00 to the worse. Not $19 or $20 or $22 or even $1.00 to the good! It's $2.00 Pricing the effect based only on the cost of buyback isn't quite appropriate. We have to factor in the cash flow component based on the strike price paid by the employee. And we also have to consider that the company issued that share ultimately to somebody who paid $22 in the first place and bought it back for $19 for a net increase in dividend potential of $3.00 for which it actually only realized $1.00 (because of the employee intermediary) for a net of cost of potential dividend reduction of $2.00
Now that's a High PITA factor (PITA is not a bread, it's a Pain In The A$$) speculative computation if I ever saw one!
So for sure, anybody who is calculating the impact of stock options wants the COMPANY (which has all the data) to do the work. Not guess based on information that the company doesn't report!!!
Anyway, the next issue is whether we want an accounting package that is backwards looking, or forwards looking. Let's pick backwards looking for the sake of argument: we'll record the actual cost of the actual exercise/buyback transaction based on what ends up as historical facts.
Blended with some fiction, of course to make it an interesting story.
Let's pretend Hueyone is suddenly appointed CEO of Cisco (this is fiction, after all). And immediately he bans issuing stock options. Well, right away the employees will be (a) mighty upset, and (b) demand higher cash compensation. So Huey, being a highly principled guy would increase employee cash compensation by some affordable amount and then use any subsequent whining as a detection mechanism for jobs that are due to be outsourced (thus solving two problems at the same time).
But meanwhile back at the nerd ranch, there are still all those unvested options out there, granted on Chambers' watch before Hueyone came on board. And every exercise depletes the ability of the company to pay a dividend. So every exercise represents one of those costs you're trying to figure out.
But even if we figure it out, if we account for the cost of stock options based on the actual cost at the time of exercise, then poor Huey gets hit with a double whammy when he reports to the board. In the first place, he has to report higher wage costs (which actually reduces his ability to pay a dividend). In the second place he has to report a further reduction in the dividend potential based on the dilution from exercise based on prior grants! Yeuch!
And so the company would under-state the ability of the business to pay a dividend as Huey is running it. He gets hit with Chambers' costs AND his own costs. And probably ends up being hated and, eventually replaced.
Maybe by some guy Mindmeld who knows how to play the seams comes in afterwards just as the option overhang is removed and looks like he's engineered a turnaround and increased profitability, when all he had to do was keep a steady hand on the helm. Mindmeld will be perceived as a better manager than Huey (and maybe get more options!) for doing the same job as Huey.
Accountants hate this kind of scenario. Although mindmeld might not mind.
Anyway, folks who like things reported where there is some sort of cause-and-effect accountability would lobby the lobbiests to record the cost of options issued on Chambers' watch as a cost to Chambers' results. So that Huey doesn't get dinged for Chambers' decisions and Mindmeld can't ride Huey's coat tails.
But this means figuring out in advance how much the difference between strike and exercise is likely to be if and when each granted option is exercised. Because now it's the prospective reduction in future dividends payable that we want to be mindful of. Ohmigosh.
Thankfully, this mind-bendingly difficult task was vastly (over) simplified some decades ago by two guys whose names have profaned this thread many many many times. Einstein said "All models are broken, some are merely useful", and these guys came up with such a broken but useful model that they won a Nobel prize.
Appropriate application of which takes us approximately close to where the current option-expensing crowd is thinking. Or the unthinkable, as certain other self-serving members of the high-tech community would rather label it.
Although we started from your premise (with which I agree) - anything that reduces the ability of the company to pay a dividend is not something that should be recorded on the books as having zero cost.
In my simple minded mind, I lump all possible dividends payable into the term "earnings". Any operating activity that reduces the amount of dividends that I could pay would be something that reduces earnings, and anything that increase the amount of dividends I could pay would increase earnings.
John
Names changed to protect the innocent. Or at least to get a rise out of a few of the guilty :o) |
