Amy and I arrived in Homer today. Yippee. Everything in the house works except the hot water. Oops. Guess we’ll have to figure that out tomorrow. Following is the view from my living room at 10pm.
I was recently asked the following question by email by a reader of my blog. Rather than respond with a one-off email, I figured I’d post my answer here since it’s broadly applicable (ah – the joy of a cross country airplane ride when one is caught up on email.) The question follows:
At my company, we’re looking at recapitalizing from 3.5 million shares to 35 million (and contemplating 350 million). The reason is pretty straightforward (although we might be way off base here): we want to create “more shares” so that as we roll out our stock option plan there is some enhanced psychological value in “getting more shares,” for employees. Although the value once you do the math is the same, I personally believe that people would rather get 5,000 shares than 5 shares, regardless of the monetary value. So, I asked my attorney what his recommendation is, but I’d love a second opinion. Do we stay at 3.5m, do we go to 35m, do we go to 350m?
My general rule of thumb for a venture backed company is to try to establish a share base from the beginning so that you never have to do a forward or reverse stock split (referred to in the question as “recapitalizing from 3.5m shares to 35m shares – or a 10:1 split.)” A range of 10m to 50m shares – depending on what you think your exit value will be (the more optimistic you are, the more shares you should use) – is a good range to work with.
Now, the share count is heavily dependent on your financing history. If you have a successful company with ever increasing valuations with each financing, you can effectively manage your share count using my rule of thumb above. However, if you end up doing a “down round financing” (one at a lower price than a previous round) – especially if it is a recap or at a significantly lower price) this approach will quickly become irrelevant as you’ll end up with a huge share count. So – while it’s nice to plan in advance (and for success), recognize that circumstances will dictate where the share count goes.
To answer this question, let’s ignore the future financing dynamics for a second and do some math. Let’s start with the 3.5m shares we have in the question. Let’s also assume that company did a financing and is worth $10.5m post-money (e.g. $3 / share), that the financing was done with preferred stock, and the board determined that the fair market value (FMV) for the common stock is $0.30 / share (common stock in a venture-backed company is often valued at 10% – 25% of the preferred – I’ll leave that for a separate post.)
A typical VP level employee that joins this company will get between 1% and 2% of the company (possibly more depending on her role – let’s choose 1% to keep the math easy.) This will be an option grant – in this case of 35,000 shares at an exercise price of $0.30 / share. So – if the company is sold for $21m (2x the current value), each share will be worth $6 and each option will be worth $5.70 ($6 minus the exercise price of $0.30) and the VP will get $199,500.
Now – let’s do a 10:1 stock split. As a result, the company has 35m shares, instead of 3.5m. Each preferred share is worth $0.30 ($10.5m / 35m). Our newly minted VP gets 350,000 shares at a strike price of $0.03. If I’m the new VP, I “think” I like this better. However, when the company is acquired for $21m, each share is now worth $0.60 and each option is worth $0.57. The VP still gets $199,500.
You can see that another 10:1 split (to 350m shares) would make the math pretty impractical.
Now – let’s go the other way. Assume you do a down round financing that results in the share count increasing from 35m shares to 105m shares. The VP still has 350,000 shares. However, the board decides that 105m shares is hard to deal with and decides to do a 3:1 reverse split (3 shares become 1). So – the total share count goes back to 35m. However, the VP’s options now go to 116,667. And – most importantly, the strike price goes from $0.03 to $0.09. The VP now believes she has less options and they cost more (true – based on the dilution) and subsequently wants an option refresh grant (probably not unreasonable.)
However, let’s say this wasn’t a financing, but an IPO. Let’s assume the 105m share case, but this time let’s increase the company valuation so the VP has 3x what she used to have (1,050,000). Let’s leave the strike price at $0.03. The investment bankers work with the company and determine the target valuation for the IPO is a pre-money value of $105m (low, but let’s keep the math easy). As a result, the bankers ask the company to do a 10:1 reverse split. Our VP now has 105,000 shares at $0.30. Since she’s been previously thinking her shares will be worth $10 in an IPO (just ask around – I bet most of your employees think most IPOs happen at around $10 – $20 / share), she now thinks she has 10x LESS value (in this case – NOT true – he has exactly what he had before.) Bad karma ensues (ironically at a time everyone should be psyched because the company is going public.)
Another issue to consider when you figure out you share count is your annual franchise tax. If you are a Delaware corp (and many venture backed companies are), you have to pay an annual franchise tax – this is calculated either one of two ways:
- The authorized share method: a straight calculation as to the number of shares authorized based on the State’s rates.
- The assumed par value method: calculates a ratio of shares actually issued and outstanding (does not include the options not yet exercised) against total gross assets for the prior year against the total number of shares authorized.
Generally, the assumed par value method is the less expensive of the two approaches, however, if the company is profitable and has high total gross assets as compared to number of shares outstanding, the authorized share method may be cheaper. In either case, the maximum annual taxes ever owed is $165,000 (at least in 2004).
For perspective, using the 350m share count maxes out both of these calculations in most situations, so if you are a Delaware corp you are going to write a check for $165,000 for the privilege of having 350m shares. If you reduce the share count to 3.5m, your taxes under the authorized share method are approximately $22k and are only $5k under the assumed par value method. I’d personally much rather save the $160k and explain the 3.5m share count to my employees. Just for perspective, 35m shares maxes you out at $165k for the authorized share method and you end up around $30k for the assumed par value. So – a question you have to ask is whether you want to pay an extra $25k / year of franchise tax to go from 3.5m to 35m shares?
You mileage will vary if you are incorporated in other states – ask your attorney and tax account for advice.
Overall, I believe that increasing the share count in a company to create the perception that an employee is “getting more shares” is a mistake. I recommend you pick a realistic share count (again – my 10m – 50m range is a decent rule of thumb) so that – unless you have down round financings – you’ll never have to monkey with the share amounts in any scenario. Then – when you grant options to a new employee – explain clearly to them what they are getting.
I’m on the board of the Colorado Conservation Trust. I think it’s currently the most impactful environmental conservation based organization in Colorado – I am so enthusiastic about it that I merged a non-profit environmental organization that I helped start – the Front Range Alliance – into CCT last year and joined the CCT board. CCT is continuing the mission of the Front Range Alliance with the Front Range Mountain Backdrop initiative.
One of the really fun projects that CCT did last year (at the request of one of our major donors) was to hold a high school video essay contest. The winner – 10th grader Michael Beggs of Boulder High School – produced a remarkable video titled “Grasslands”. Second place was titled “Oil and Gas Development on the Roan Plateau” and third place was “Eurasian Milfoil – A Non-Native Grass that is Polluting Colorado’s Rivers and Streams.”
If you are conservation minded, you should be pleased that we are “growing them young” here in Colorado.
I’m at MIT all day at a symposium run by Eric von Hippel on Democratizing Innovation. It’s a classic “drink from a fire hose” type of day – short (15 minute) descriptions of 40 or so research projects over two days.
Most of what I’m interested in is the research around open source. However, given that I recently read FAB and have been thinking about a “personal fabrication machine”, I was totally jazzed to hear about eMachineShop – an online machine shop that allows a user to design, price, and order custom machined parts online.
The company’s tag line is “Why waste time traveling, calling, faxing or emailing to conventional machine shops – and waiting days for quotations? Reduce your total time up to 90%! Open doors to new products and projects, to inventing new things, to reducing the cost of parts and more. Quantity 1 to 1,000,000.”
The examples are great. Pricing is straightforward and easy to deal with. And – for people like me that barely know how to use a stapler (unless it’s a virtual one) – this is a brilliant example of the shift from physical to virtual, enabling me to create stuff using software that I’d previously never have a chance of even thinking about playing around with.
Matt Blumberg – CEO of Return Path – posted a job description for a new VP of Marketing on his blog. Jennifer Wilson – Return Path’s long standing VP Marketing is having a baby and going on maternity leave (congrats Jennifer!) – when she returns it’ll be on a part time basis. So – if you are (or know) and A+ VP of Marketing in the Denver or New York area, please drop Matt a note.
I hate Spam. Fortunately, one of our companies – Postini – does a remarkable job of eliminating it (turn on Postini, magic happens, Spam disappears.) Recently, I’ve been under attack from other Spam variants, including trackback/comment Spam and IM Spam (SPIM). Today, Postini announced a partnership with IMlogic to collaborate on a Spim blocking service. Dudes – IM there.
As I wander around in summer school, I’ve been working my way through Brian Harvey’s Computer Science Logo Style 2/e, Vol. 1: Symbolic Computing. So far I’m finding Harvey to be a fantastic writer and his approach to introducing computer science concepts to be very accessible (to me). In addition to being a good writer, Harvey has written a Logo interpreter that goes along with the book series.
For those of you that are sneering and thinking “Logo – that’s for kids”, Harvey has a fun four line program showing the power of Logo that he uses to challenge one to reproduce in four lines of Java.
I’m still too early in the book to have any thoughtful commentary about my “programming rebirth” other to say that it’s fun to stretch my brain on this stuff again after a long hiatus.
When I read James Frey’s first book A Million Little Pieces, I thought it was the most intense book I had read in a long time. Frey’s second book – My Friend Leonard, which continues where the first book leaves off, is even more intense.
This book starts at the bottom (like the first one does), starts to climb a little, and then turns you upside down and smashes your head into a concrete floor over and over again. I started it in the Seattle airport last night on my way to Boston on the red eye – I figured I’d just get started and then put it down when I settled in on the plane to fall asleep. Bad plan. An hour into the flight, I finally gave it up and went to sleep. I finished the book off after I woke up in Boston this morning. The roller coaster ended – finally – on page 357.
Frey is incredible on numerous dimensions. For starters, this dude is unbelievably stubborn. It works against him for a while, but it ultimately works for him in major ways. Next, he transitions from a devastating first 24 years of his life into a supremely articulate, yet rawly emotional, human. While he struggles, is afraid of himself, hates his fear, and suffers, he never gives up, especially since he knows that giving up will mean death for him.
His friend Leonard is a major part of this. Leonard has his own issues, but is an awesome friend when Frey keenly needed one. Leonard is unapologetic, unyielding, and knows life is meant to be lived. The interplay between Frey and Leonard is awesome, especially when you toss in Leonard’s bodyguard Snapper.
This book is brilliant. Do NOT read it without reading A Million Little Pieces first.
I’ve just agreed to be a judge for the finals of the Microsoft produced Imagine Cup Programming Competition. This is a college programming competition that now has over 15,000 students enter. I’m heading over to Yokohama, Japan at the end of July with Scott Maxell of Insight Venture Partners and Chris Pacitti of Austin Ventures who are my co-judges for the finals. We’re apparently going to do an American Idol like thing – nerd style. Scott insisted on being Simon and I offered up – much to the chagrin of Amy – that I’d be Paula. I think it’s awesome that Microsoft produces a worldwide college programming competition – it was easy to agree to judge when asked.