I think YC takes mid single digit percentages in some form early on? So you figure they get diluted down over time and maybe they have, as a firm, 1-2 percent of the total, so, maybe, 150-250mm? And then PG certainly has a big chunk of that - a quarter? A third? He was already pretty rich because he got Yahoo stock (in substantial amounts!) for selling Viaweb, and then Yahoo stock went on to increase by a factor of more than 10 to its peak - if he had 20% of Viaweb at sale (10mm on the day of the transaction) and got out at a reasonable time he had, say, 75mm by the early part of last decade...
All of which is to say, it's true that YC has been very successful for all parties, and PG is probably much richer now than he was when he started it, but I suspect he'd say the money wasn't the important thing to him and based on these numbers I'd be inclined to believe him.
Edit: incidentally, I think PG has probably been underpaid for his YC work - I suspect most investment firms do not add substantial value to the companies they invest in, but the YC brand is clearly enormously influential and, given the enormous amount of noise in startup-land, I bet the brand - which PG has worked on very carefully for years - is responsible for a meaningful percentage of the value of the portfolio.
I would suspect that YC has done very well as a fund.
But I also suspect that the partners have all made FAR more by investing individually into YC companies after they've been in YC. Here are links to some of the investments YC partners have made. (I suspect there are more... these are what are shown in Crunchbase.)
> The difference between me and the other partners is mostly that I got to be publicly known by writing.
I wonder what the rate of Y Combinator founders' blogging (either personally or placing importance on their official business' blogs) vs those outside of it, if such a thing could be quantified.
My understanding is that these "valuations" are calculated by taking the price per share of the most senior tranche of preferred stock and multiplying that by the total number of shares outstanding (+ the option pool) of all classes.
This kind of "valuation" may be useful and the way business is done, so if everybody is wise to the game, it's not clear anybody who matters is being deceived. But as far as an outsider or accounting geek is concerned, this is not an appropriate way to value the equity in a company with multiple share classes. Common shares and less-senior preferred shares are simply not as valuable as the most senior preferred stock. You can't apply the same per-share value to everything.
These same companies are not claiming $11.5 billion of "valuation" when they set a per-share value for 409(a) purposes. In broad strokes, these companies claim to have low valuations to the IRS (which helps their employees when they grant options) and high valuations to the press (which helps them look good and attract employees) at the same time.
Which may all be fine, but the press (and prospective employees) should be as wise to the game as the investors and company are.
When preferred stock investors calculate the value of a company, they do so with reference to a cap table showing the fully diluted capitalization of the company (that is, all classes of stock plus options) assuming that all stock is converted to common and all options exercised.
In this way, a lead investor investing, say, $4M in an A round with a $6M pre-money company valuation and paying $1/sh against a model of 10M authorized would own 40% of a $10M company (post-money valuation), would own 4M shares of Series A preferred stock valued at $1/sh, and would do so in a context where there may be, say, 4M common shares held by founders and 2M common shares held in a pool for equity incentives. When that same company goes to do its 409A valuation following that preferred round, it retains an outside independent valuation company to calculate the price of its common stock for purposes of valuing options granted to employees (409A is solely designed as a compliance provision requiring that options and other forms of deferred compensation granted to executives and other employees not be underpriced as of the date of grant). When it does that valuation, the outside company will measure the value of the common stock by looking at the arms-length deal done by the Series A investors by which they paid $1/sh that gave them each 1 share of preferred stock that is typically convertible 1:1 into common stock on the happening of certain events. As of the date of the Series A investment, that common stock clearly does not have the same value as the preferred stock because the preferred stock includes a series of valuable preferences that the common stock lacks (e.g., a liquidation preference giving it first preference in being paid the proceeds of any acquisition, subject to various limitations; a right to class approval of major corporate actions; a right to anti-dilution protection in the event of a down round, etc.). Any common stockholder who has been "wiped out" in a merger where the preferred holders take everything and the common holders get nothing can attest to the value of these preferences. This is why the stock is called "preferred" in the first place. In any case, in the days before 409A, a company's board of directors would routinely use a 10:1 ratio to value the preferred stock in relation to the common owing to the value of the preferences (that is, in our example above, the preferred stock would be priced at $1/sh and the common at $.10/sh). Since the enactment of 409A, that ratio has shrunk considerably and, today, the outside companies who do the valuations will at their most aggressive use a 5:1 ratio and will much more typically use more like a 3:1 ratio in accounting for the valuation difference (again, in our example, resulting in common stock prices of $.20/sh and $.33/sh respectively). The common stock valuation so determined will then be used by the company for valuing its stock options granted to employees. After such options are granted, should the company hit the skids and go down (or get acquired in a fire sale), the common stock holders would typically get nothing while the preferred holders may get some or all of their money back. On the other hand, if the company does very well and is ultimately acquired at a premium, then the preferred holders (in a typical case where the preferred stock is non-participating and only gets its money back on the liquidation preference without participating in any further payout) will convert to common in order to get the benefits of the acquisition and will share equally (in accordance with their percentage interest in the company) with all other common holders in those proceeds. In this way, in the end, the discounted value of the common stock as used for 409A purposes becomes irrelevant to the value of the company and what really counts is the arms-length price that someone will pay for the acquisition, as to which the common will share equally with the preferred. In other words, there was a reason for discounting the value of the common at the time of the A round...
I agree with your description of the mechanics, but not with your conclusion that "there was a reason for discounting the value of the common at the time of the A round but that reason goes away once the acquisition occurs" and that "it is legitimate for company valuations to be based on the latest preferred pricing."
That analysis assumes that the acquisition will occur at a price that makes it worthwhile for the preferred investors to convert their shares to common, and therefore that the preference will be irrelevant.
Of course if there is a tender at a high-enough price, then the common shares will be worth the same as the preferred. But until that occurs, the preference is valuable. After all, the point of the preference is that it's a hedge against an uncertain outcome -- in this case, an acquisition that takes place at less than the original price of the preferred shares.
That's why you can't properly value the equity in a company by multiplying the price of the most valuable kind of shares by the total number of shares outstanding. In the presence of uncertainty about the value of a future buyout, the shares that don't have the preference really are worth less.
Fortunately, the company does not really claim this fake-valuation as its value anywhere that matters -- only when boasting to the press and to prospective employees.
14 comments
[ 3.0 ms ] story [ 47.3 ms ] threadAll of which is to say, it's true that YC has been very successful for all parties, and PG is probably much richer now than he was when he started it, but I suspect he'd say the money wasn't the important thing to him and based on these numbers I'd be inclined to believe him.
Edit: incidentally, I think PG has probably been underpaid for his YC work - I suspect most investment firms do not add substantial value to the companies they invest in, but the YC brand is clearly enormously influential and, given the enormous amount of noise in startup-land, I bet the brand - which PG has worked on very carefully for years - is responsible for a meaningful percentage of the value of the portfolio.
But I also suspect that the partners have all made FAR more by investing individually into YC companies after they've been in YC. Here are links to some of the investments YC partners have made. (I suspect there are more... these are what are shown in Crunchbase.)
Paul Buchheit: http://www.seed-db.com/investorgraph/investorview?investorid... Garry Tan: http://www.seed-db.com/investorgraph/investorview?investorid...
So in a way... Paul's stake in YC is kind of irrelevant. :) (Also, I think he's having a lot of fun with YC, which really makes it irrelevant.)
I wonder what the rate of Y Combinator founders' blogging (either personally or placing importance on their official business' blogs) vs those outside of it, if such a thing could be quantified.
This kind of "valuation" may be useful and the way business is done, so if everybody is wise to the game, it's not clear anybody who matters is being deceived. But as far as an outsider or accounting geek is concerned, this is not an appropriate way to value the equity in a company with multiple share classes. Common shares and less-senior preferred shares are simply not as valuable as the most senior preferred stock. You can't apply the same per-share value to everything.
These same companies are not claiming $11.5 billion of "valuation" when they set a per-share value for 409(a) purposes. In broad strokes, these companies claim to have low valuations to the IRS (which helps their employees when they grant options) and high valuations to the press (which helps them look good and attract employees) at the same time.
Which may all be fine, but the press (and prospective employees) should be as wise to the game as the investors and company are.
In this way, a lead investor investing, say, $4M in an A round with a $6M pre-money company valuation and paying $1/sh against a model of 10M authorized would own 40% of a $10M company (post-money valuation), would own 4M shares of Series A preferred stock valued at $1/sh, and would do so in a context where there may be, say, 4M common shares held by founders and 2M common shares held in a pool for equity incentives. When that same company goes to do its 409A valuation following that preferred round, it retains an outside independent valuation company to calculate the price of its common stock for purposes of valuing options granted to employees (409A is solely designed as a compliance provision requiring that options and other forms of deferred compensation granted to executives and other employees not be underpriced as of the date of grant). When it does that valuation, the outside company will measure the value of the common stock by looking at the arms-length deal done by the Series A investors by which they paid $1/sh that gave them each 1 share of preferred stock that is typically convertible 1:1 into common stock on the happening of certain events. As of the date of the Series A investment, that common stock clearly does not have the same value as the preferred stock because the preferred stock includes a series of valuable preferences that the common stock lacks (e.g., a liquidation preference giving it first preference in being paid the proceeds of any acquisition, subject to various limitations; a right to class approval of major corporate actions; a right to anti-dilution protection in the event of a down round, etc.). Any common stockholder who has been "wiped out" in a merger where the preferred holders take everything and the common holders get nothing can attest to the value of these preferences. This is why the stock is called "preferred" in the first place. In any case, in the days before 409A, a company's board of directors would routinely use a 10:1 ratio to value the preferred stock in relation to the common owing to the value of the preferences (that is, in our example above, the preferred stock would be priced at $1/sh and the common at $.10/sh). Since the enactment of 409A, that ratio has shrunk considerably and, today, the outside companies who do the valuations will at their most aggressive use a 5:1 ratio and will much more typically use more like a 3:1 ratio in accounting for the valuation difference (again, in our example, resulting in common stock prices of $.20/sh and $.33/sh respectively). The common stock valuation so determined will then be used by the company for valuing its stock options granted to employees. After such options are granted, should the company hit the skids and go down (or get acquired in a fire sale), the common stock holders would typically get nothing while the preferred holders may get some or all of their money back. On the other hand, if the company does very well and is ultimately acquired at a premium, then the preferred holders (in a typical case where the preferred stock is non-participating and only gets its money back on the liquidation preference without participating in any further payout) will convert to common in order to get the benefits of the acquisition and will share equally (in accordance with their percentage interest in the company) with all other common holders in those proceeds. In this way, in the end, the discounted value of the common stock as used for 409A purposes becomes irrelevant to the value of the company and what really counts is the arms-length price that someone will pay for the acquisition, as to which the common will share equally with the preferred. In other words, there was a reason for discounting the value of the common at the time of the A round...
That analysis assumes that the acquisition will occur at a price that makes it worthwhile for the preferred investors to convert their shares to common, and therefore that the preference will be irrelevant.
Of course if there is a tender at a high-enough price, then the common shares will be worth the same as the preferred. But until that occurs, the preference is valuable. After all, the point of the preference is that it's a hedge against an uncertain outcome -- in this case, an acquisition that takes place at less than the original price of the preferred shares.
That's why you can't properly value the equity in a company by multiplying the price of the most valuable kind of shares by the total number of shares outstanding. In the presence of uncertainty about the value of a future buyout, the shares that don't have the preference really are worth less.
Fortunately, the company does not really claim this fake-valuation as its value anywhere that matters -- only when boasting to the press and to prospective employees.
I think it would be bad for YC (and to a lesser extent HN), if not disastrous.
511 ( startups ) 11(.)5 (B)
for it to be one " startups " , decimal and B have to be parsed out ;-)