I think we are getting semantics in the way of value computations.
But there is some common ground and I think we can find it and stand on it. From which wobbly turf perhaps find some resolution.
Let's say I promise your lawn will be mowed and thus you promise to pay $20. Then I get my son to promise me to mow it. Then he mows it. We can argue for a long time who actually promised to mow the lawn (did my son promise me, or you transitively through me...). The fact of the matter is that the lawn is mowed and he did the work, as promised. So pay up.
Same with stock options and the company. There's a promise (by somebody) to employees that they will get some booty if the stock price goes up. So somebody has a debt to employees.
We can say that the debt born by the whoeveritises was incurred by the company (which issued the options) or by the whoeveritises themselves. The important thing is not to lose sight of the cost to the whoeveritises. Which cost has a dollar value.
We can compute this dollar value by any number of ways. One is the expected value given by the dubiously applicable Black Scholes equation. Or the expected value given by the more dubiously applicable Shannon Computation.
What remains are really two variables. Who are the whoeveritises and what ways can we use to estimate this cost that they bear.
I maintain that the whoeveritises are the shareholders. Not only by process of elimination (not employees, not company, not IRS, not customer, not supplier... ), but by precise economic rigor of funds flow.
I also maintain that one of the ways to estimate this cost is to estimate the value to the employees. Cost not equal to value implies missing money. Or extra money. And the only folks who can create US Dollars are chaired by Alan Greenspan (who is well practiced at this), and they don't get involved here.
If anybody has cogent arguments to refute either of these two simple claims, please feel free to advance them.
There is agreement here, perhaps more than you know. When the company issues a share to an employee ( love your reference to insider, sounds ominous) the existing shareholder suffers dilution in ownership equal to 1/(shares outstanding) and he is no longer the only shareholder. The value the company receives for the issued share = strike price + ( % of the employees services attributable to that compensation), ...
Presumably the "% of the employees services" bit is reflected somewhere in the sum of future EPS. And you will note I already allowed for it. Hence my reference to $27 or $35 in the original example. So actually, we agree.
...When our hypothetical shareholder purchases the share from the employee the dilution has already taken place and should be reflected in the stock price, he is buying a diluted share.
This is what doesn't happen. When was the last time you heard ANYBODY on this thread (besides maybe me and mucho) compute the dilution-effective PE of QCOM? When was the last time you didn't feel obliged to post some witty but pithy reply to the fool who would propose such thinking? <ggg> Just kidding around. Anyway, the effect is gradual as it occurs, but large in cumulative effect. One is better off to compute the cumulative effect on sum of EPS and determine a terminal value, than trying to buy a share at a price plus or minus two tenths of a penny based on how many options were exercised this week!
Now as for My problem is with you taking appreciation after the stock option grant and calling it a cost to shareholders. The potential cost to shareholders happens at the time of grant and is equal the discount in strike price to the value of the stock.
Not true. Not true at all. Tantamount to claiming that there is no cost to shareholders when a stock option is issued at 0 discount.
An actual cost to shareholders does occur when the option is exercised. Here is why.
Immediately before stock option exercise, the shareholders own an asset. It is worth something. This worth is not changed by virtue of issuing shares or canceling shares. Or the price on the market at the time. While an option holder holds an option, the holder has no claim on these assets. It can be cancelled or retracted. Or maybe never even exercised.
On the instant of stock option exercise, the number of shares outstanding goes up. The asset value that each shareholder holds goes down. Just a tiny bit, unnoticeable in the millions of shares already sloshing around, but when these millionths are added up across those millions of shares held, the sum is precisely equal to the asset value held by the new shareholder, minus of course whatever additional asset value this new shareholder contributed. The algebra is fairly straight forward, and I encourage anyone who is interested to try and figure it out for themselves. There is a Before (assets / # shares = value per share before) and an After (assets + contribution / # shares + options = value per share after). Existing shareholders have the same number of shares, and the difference in their value (as a group) is computable. Although it's messier than most people's algebra skills can handle. Correctly, that is.
Anyway, it is inescapable that at the moment of exercise THERE IS A COST, measurable as a reduction in value. And thus, at the moment of grant, we know there WILL BE a cost. Only we aren't sure what it will be or when it will be incurred. If at all. The best we can do is estimate it.
Let's turn attention back to when the option is granted. Shareholders (or at least smart ones) expect that their holdings will be diluted in the future by some exercise event in the future, exactly as described. They expect to incur a cost. How much is an open question and hotly debated. But it is not zero. It depends on the stock price in the future.
But they (gladly) give up this cost, in return for expecting something else. Like that the person will work hard and make them wealthy. That still doesn't mean that the cost is zero. It just means they are OK with whatever it is. Often they are OK with it just because they don't really know what it is and they figure that since everyone else is OK that some smarter person has figured this out so they are safe. Sheep theory. Safety in numbers. Fess up, when was the last time you tried to compute this cost and see whether or not Qualcomm deserved its stock price?
It's easier to claim that the cost is zero, thus allowing you to make no effort (why compute zero). And through the wonders of circular logic, end up convinced that the cost is zero.
Even though we know that it is not. For real value as determined by apportioning the same asset base amongst more shares is a reduction in value. A cost by definition.
Now, this cost changes depending on when the option is exercised. If it is exercised when the strike price is equal to the market price, then there is no cost. As the share price goes up, the cost goes up. Thinking for a few moments, nobody in their right mind will exercise a stock option for parity. Nor hold a growth stock they think will have a flat stock price forever. So really it's naive to assume that a stock option will have zero cost. It will be exercised at a time when the stock price is higher than the grant price. And thus, shareholders will incur a cost.
They can use the Black Scholes equation if they like. Or back in 1998 they might have said "if the price of the stock goes from the current $6.00 to the phenomenal value of $32 then the $2.1 B$ that these options would represent, spread out amongst all 400 Million of us ($5/share) is a drop in the comparative bucket." Which is how the thinking usually goes.
But this is thinking like owners of a pyramid scheme. Not owners of the company. Frankly, most investors in public companies that produce no dividend are only owners of pyramid schemes, so this thinking is perfectly OK. It just has to be followed very carefully by "but what will the guys in the future be willing to pay for my shares"?
And sooner or later someone will be adding up highly diluted future Earnings Per Share and getting a value per share that will send many options underwater. Hopefully one is no longer a shareholder when this happens. Or so went prudent thinkers back in '98. So sorry for the folks who hopped on the bus in 2000 (but thank's for the option income folks, it's gone a long way to improve my lifestyle).
But back to the "what is a share worth" question.
Which question is answered by adding up a cost/benefit equation. The first part (the benefit) is profit. The risk-discounted present value of the sum of annual profits into perpetuity should be pretty close to market cap. In a fair world.
That's what shareholders plan to split up. The next part is the negative contribution by option holders. The dilution. Subtract this cost from market cap, and divide the total sum by all of the parties who will be at the table on that day in the infinite future (so include most, if not all, option holders).
That factor is what a present value share should be worth.
And it isn't just 3% lower 'cause dilution was 3% so EPS is reduced by 3%. Not when you do the real math. It's the cumulative effect of 3% forever, which can add up to a lot over not so many years.
An alternative approach that also works is to ignore all this business and pretend that the option strike price is zero. Add up the present value of eps and divide by as many shares as you figure will have been issued and exercised.
This is what most people do. But it understates the actual value, because along the way the company will be accreting assets from the issuing of these shares. Which never made it into that adding up.
So if you just use dilution, it's over-estimating the cost. Feel free to do that if you like. It requires a lot less work.
John |
