Anti-Dilution: Full Ratchet Versus Weighted Average

There are two principal approaches to anti-dilution provisions: the "Full Ratchet" and the "Weighted Average." These concepts tend to be explained by reference to computational formulas, which work off a variable called the "conversion price". Forget that. For two reasons. First, conversion price doesn't refer to currency but instead to an arbitrary ratio (usually 1 or .3 but it can be anything), so it's confusing. Second, the computation of the conversion price/ratio, while a necessary accounting tool, is a distraction from the operational result: the reallocation of stock from the founders to the angels.

Both approaches are based on the same predicate concern: what happens to the value of the angel's investment (in the form of preferred stock (that converts into common stock upon IPO/merger)) in the event that new stock is issued (sold) to someone else at a lower price (relative to that paid by the angel)?FN1

In most investment scenarios, like real estate, for example, if the post purchase market value is lower than the price paid, so it goes. That's the risk of real estate investment. The seller wouldn't retroactively kick in extra acreage or have a guest house built on the property to make the investor feel better.

Start-up financing does not operate like that. In the event the company's value goes down post purchase, anti-dilution provisions work to re-allocate shares from the existing common stock pool to the angel's preferred stock pool to make the angels feel better. The only question is, how many shares are re-allocated?

According to the Full Ratchet approach, the answer is: a lot. In fact, full ratchet does not even care about the number of the shares per se. It just cares about the bargain given to the new investors. How much are the new investors paying per share? That's deal the angel's now want.

An analogy would be you bought 10 pants at Banana Republic for $100 then someone else comes along and buys 10 pants for $10. Under FR, you get to say, 'Give me another 90 pairs of pants.' 

It's called "Full Ratchet" because as a technical matter it demands that the ratio by which their preferred shares will be converted into common shares be retroactively decreased - "ratcheted down" - to the effective conversion ratio the new buyer is getting.FN2 The salient feature is that this occurs no matter how large or small the subsequent financing is (if - this is an extreme example - a company issues just 100 shares to subsequent investors at a $1.00 per share and the angels have 5,000 shares (for which they paid $10 per share), then under application of FR the angels gain (and the founders lose (the compensation has to come from somewhere)) an additional 45,000 shares.FN3

Weighted Average is a more measured approach. Although WA fixates on the deal the new buyers are getting, it takes into account the magnitude of the financing - how much new stock has been issued. The weight of the better deal price is balanced against the weight of the size of the purchase (similarly, the extent to which going 3 for 4 affects one's batting average depends on how late in the season the game is).FN4

There's a number of different formulas used for the calculation but as noted they aren't very intuitive so let's ignore them since the mathematical concept is straightforward. Take the post-money monetary value of the company under the old deal (just multiple the total number of company shares by the share price paid by the angel (in this case, $10 X the total number of shares (10,000 (assume 5,000 for angels, 5,000 for founders) = $100,000. This is our baseline company value under the old deal. Add to that the value being added by the new buyer - in this case $100 (100 shares at $1.00 per share). Divide (average out) that total value - $100,100 - by the total number of shares after the new buyer buys in: 10,100.

The number you get - $9.91 in this case - is going to be lower than the price paid by the angel - $10. It's an average between the old price and new price dependent on the aggregate number of shares being issued at the new price.FN2 How many shares does the angel's $50,000 investment get him now? Not 5,000 ($50,000/$10pershare) but 5,045 ($50,000/$9.91pershare).

FN1. "Original investor" and "new investor" are the two players here; "angels" is just shorthand for the former.

FN2. As noted, "conversion price" is misleading. The confusion becomes manifest in most examples you'll see - which will use $1 for the angel investment and then like fifty cents for the new investment and refer to these as conversion prices even though that is conflating two separate concepts. The conversion price is just the ratio by which preferred shares are converted into common shares, and often is 1:1 or .5:1. As a matter of fact, any investor will give a company a certain amount of money in exchange for a certain amount of preferred shares (which will have a certain price per share value that no one cares about) and THEN those shares will be deemed to convert to common shares according to a certain ratio.

FN3. Note that in these examples (and any such examples you'll see involving formulas) the calculation of the shares given to the new and old investors will be with respect to COMMON shares that the investors will eventually own. As indicated above no one really cares about how many preferred shares are being issued.

FN4. Under Weighted Average, the pants analogy works as such: There's 20 pairs of pants valued at $200 (you bought 10 for $100 and assume the founders also have 10). Add to the $200 valuation the $10 paid by the new buyers (for 10 pants). Divide $210 by the new total number of pants: 30, which equals $7. This becomes the retroactive price you paid. How many pants does $100 get you if it's $7 rather than $10 per pant? 14.28 instead of 10. Fashion for a fortnight.