Robert Heaton

Software Engineer /
One-track lover / Down a two-way lane

Peeking through stock option exercise windows

06 Aug 2019

I wrote this post several years ago when I was doing some research into stock options, but never got round to publishing it. None of the research actually applies to me anymore, but I figured I’d publish the post anyway because it’s got some good gags and contains some useful information.

Note that while I am a careful driver and a delight at dinner parties, I am most emphatically not a lawyer or an accountant. You should not taking anything I say at face value.

The high-risk nature of privately-held technology companies is often masked by survivorship bias, boosterism, and good old-fashioned over-optimism. But this risk still makes life confusing and easy to screw up for anyone who receives a significant portion of their compensation in the form of stock options.

One of the most important mechanics responsible for this unwelcome intrigue is the exercise window. Not understanding the difference between long- and short-term capital gains can leave you with a larger tax bill than you expected, but not understanding exercise windows can cause you to lose everything.

What is an exercise window and why does it matter?

An exercise window on an option grant means that when you leave the company, you have a set period of time in which to exercise the options and purchase the underlying shares. If you don’t, you lose the options completely.

The reason that exercise windows are so important to understand is that exercising options in a private company can be prohibitively expensive, at least in the US. The IRS requires you to immediately pay tax on any appreciation in the value of the underlying shares between the time the grant was awarded and the time it was exercised. Shares in private companies are generally very illiquid - indeed, most option grants contain clauses prohibiting you from selling your shares at all. This means that you have to pay these taxes out of your own pocket, years before you are able to sell the shares, if indeed you ever are.

The industry default exercise window is 90 days. This means that a lot of people who leave their companies end up losing their options, since they can’t afford the taxes that would be required to exercise them. On the other side of the fence, a lot of would-be leavers end up unhappily marking time until their shares are liquidized by a sale of some kind.

There’s nothing intrinsically wrong with a company offering, or an employee signing, a contract saying that the employee will perform work in exchange for a combination of cash and stock options that have a 90-day exercise window. The employee should be sure that they understand that 90 days is not a long time. They should also consider the fact that liquidity events (like IPOs) are occurring later in companys’ lifetimes than they used to, and that there is a good chance that the employee will want to go and do something else before their company gets there. But nothing inherently bad is happening.

However, it’s still very easy to find yourself gazing through a 90-day exercise window without having fully internalized what this means. The impact of an exercise window on a grant holder is extremely context-dependent and doesn’t translate easily to numbers. If someone has a gun pointed at your head (a literal metal gun, not a metaphorical financial one) and is forcing you to stay at your company for the next twenty years, then you’re already locked in and an exercise window won’t make much difference to you. On the other hand, if you have to leave to take care of a sick relative, move across the country to support your spouse’s career, or just want to become a scuba-diving instructor, then it makes all the difference in the world.

If you’re somewhere in between then it’s very hard to know what you should think. No accountant can tell you, no matter their hourly rate.

Can’t we just get some bigger windows?

The current industry default window size is 90 days. If you have stock options and don’t know how long their window is, it’s 90 days. Happily, some companies have switched to plans with windows of up to 10 years (usually measured from the employee’s start date, rather than exit date, for legal reasons I haven’t attempted to understand). This allows employees to leave without forfeiting any of their vested options, at least until 10 years after their start date, at which point I imagine that all hell breaks loose.

Stock options with a long exercise window are intuitively worth more than those with short windows, even though exactly how much more is very unclear. This means that one should expect that if a company started offering long-window grants then it would start making those grants correspondingly smaller. It would then put effort into explaining to prospective hires why less is more. Paradoxically I suspect that most prospective hires would get confused and go with the bigger but more restrictive numbers, since that’s an easier decision to explain to your grandma. But, we can still see that by granting fewer, longer-windowed options, the company will still end up giving away the same amount of value, just in a different form. The effect on employees’ expected value should be exactly zero.

Despite this postulated no-op, in reality I believe that a broad shift to longer windows would be a good thing for employees. The worth of a long-windowed option is much easier to estimate than a short one, because it doesn’t require putting a dollar value on the ability to freely switch jobs. Companies will always be much better at evaluating and making deals than employees, because they have more experience and data. Anything that introduces additional complexity, like hard-to-evaluate financial instruments and salary secrecy, is probably a win for companies, at least in its first-order effects. This is because it gives companies an additional way to be better at deal-making than their employees. I have no idea exactly how much more valuable long-window options are than short ones, but I’m willing to bet that there are plenty of companies and VC firms who have or could put time and effort into making decent guesses.

This isn’t to say that everyone involved in contract negotiation is completely cynical and out to screw you. But unless your prospective employer sends you all of the market data they used to arrive at your offer and tells you the absolute, absolute, absolute maximum they would be willing to negotiate up to, then they aren’t entirely on your side at this point in time.

And of course I’m sure that the tech job market is a long, long way from being efficient. It seems unlikely that the 90 day default was produced by an equilibrium of market forces, and much more likely that it was an artifact of Silicon Valley history that no one has had much incentive to change. If ten year windows became the new normal then who knows what would actually happen

The future

I’m not sure how much of the current state of affairs is deliberate. There’s an easy-to-imagine narrative whereby at first a company’s founders don’t really know much about stock-based compensation and are just trying to get their legals done quickly and simply and the same way as everyone else. Then perhaps most early employees aren’t aware that there’s any alternative or aren’t in much of a mood to argue. Then it’s only when the company is relatively stable and people with vested stock start thinking that they might like to go and do something else that the question of how and why these option things actually work starts being asked in earnest. The higher-ups are just as surprised as everyone else at the answers.

On the other hand, whenever dollars and zeros are involved you should always at least consider the possibility that everything is the way it is because that’s how the people who currently own the dollars and zeros would prefer it.

An article on the Triplebyte blog suggests that new companies should offer ISOs that automatically convert to NSOs with a 10-year exercise window if not exercised within 90 days of leaving. The article also claims that existing companies with outstanding grants are entirely capable of extending the exercise windows on these grants, should they choose to do so. I have no idea if this is accurate, but I also have no reason or incentive to doubt it.

It feels easy to castigate companies that decline to extend their employees’ exercise windows. But I do have some intellectual sympathy for the (implicit or explicit) position that long-window grants are more valuable than those with short-windows, you should have read what you were signing, and why should we amend your contract to give you free money and not anyone else?

The conclusion

There are no doubt plenty of of people who have been saying all of this for years and can’t understand how everyone else is only just noticing it. I now personally don’t think I would join a company that only offered short-windowed stock options. I value flexibility very highly, and have no idea what price I should sell it for. So when looking for your next job you may want to negotiate a little harder, demand RSUs or death, and start asking more awkward questions.

I’ve written a lot more about stock options in general here.

Subscribe to my new work on programming, security, and a few other topics. Published a few times a month.
Follow me on Twitter ➜ RSS ➜