The venture industry got comfortable with valuation “rules of thumb” in the follow-on rounds using the typical model of Series A, Series B etc.  Bullpen plays in the “rational B” follow-on round , and we see both priced seed rounds coming to us as well as unpriced convertible notes.  We have been on both sides – in capped convertible notes (“capped converts”) and in priced rounds going into rounds that follow us.  What we have noticed is vast confusion between pricing a follow-on to a priced round and a follow-on to a convertible note.

The confusion seems to center on what is the meaning of “pre” money.  This is all that counts:

Pre =  post-money less new cash

When debt is rolled up, it raises the effective pre since it is not new cash.  When all sorts of discount shares are added on top, due to the cap in the capped convertible notes, the post rises quite a bit.

The headline is that VCs are over-valuing the follow-on to capped converts by apply traditional rules-of-thumb for valuation without appreciating the impact on the post-money value.  When they run their spreadsheets (often AFTER setting price) they are shocked, shocked to find that valuations are much higher than expected.  I think this has caused an inadvertent valuation creep upwards that has been noticed across the board recently.

It means capped converts are better deals for founders, and worse for follow-on VCs.  The angels end up in about the same position either way .

Let me explain with a rather simple example, below the fold.  

Let’s say the first round is $3m on $4 pre, or with a capped convert, $3m with a $4m cap.  What happens with a $3m second round priced at $12m pre?  As you shall see, the outcomes are remarkably different between a capped convert and a priced seed round.

Priced Rounds

In the traditional venture priced rounds, the outcomes are straightforward.  Assume in both cases the stock option pool is set to 20% post each financing.

The first round prices out as $3m on $4 pre, which is $7m post, so the investors own 3/7 or 43%. The cap table:

37% – founders
20% – pool
43% – angels / seed VC

The second round buys 20% ($3m on $12 pre or 3/15 post).  The pool needs to get refreshed to be 20% post. It started at 20% after the first round, and now needs to be increased again since it will be smaller than 20% after this second round.  The new money will shrink it by 20% to 16%, and the new pool refresh (to bring it back to 20% post the second round) actually dilutes that old pool slightly, so the actual math is more complex than 20% dilution of the original 20%; in this case it works out to a 25% dilution and the pool refresh is 5%.  (Most VCs do this in a spreadsheet and plug in the numbers.)  The pool is calculated on the post money value of $15M, but the dilution from the additional 5% pool refresh is pushed in to the pre-money, so it shrinks the prior investors and founders stake.  The pre starts at $12M but shrinks by that pool refresh to $11.25M (calculated as $12M less 5% of the post-money of $15M, or $12M less $0.75M value), before the pool refresh is added back.

It is usually easier to see this working with shares and stock prices and not percents.  Let’s say the initial round comes with 10M shares, so the ownership in the pre starts as:

3,714,286 shares – founders
2,000,000 shares – pool
4,285,714 shares  - angels / seed VC

You can calculate the price per share of the VC at 70c for their $3m.

After the next round, the cap table increases from 10M shares to a little over 13.3M, the additional 3.3M being the new 20% ownership by the new investor, plus the pool refresh:

3,714,286 shares – founders (same as before)
2,666,667 shares – pool (adds 666,667  shares = pool refresh)
4,285,714 shares  - angels / seed VC (same as before)
2,666,667 shares – new VC

You can calculate the price per share at  $1.125, a nice increase from the prior round.

The percents end up as follows:

28% – founders
20% – pool
32% – angels / seed VC
20% – new VC

Yo can see how the dilution of the new round is 25% – 20% from the money and 5% from the pool refresh.

Capped Convert Rounds

Now let’s compare that to an initial capped convert of $3M with a $4M cap.  Typically it accrues some interest; let’s say it was set at 8% per annum and has accrued over 10 months for an additional $0.2M of value.  When the VC prices the follow-on round at $3M on $12M pre, expecting $15M post, funky things happen.

First, the capped convert rolls in to the round, so the actual round is not $3M on $12M pre, but $3m new money + $3.2M convertible debt or now $6.2M on $12M pre, for a “preliminary” post of $18.2M.

Second, the cap is applied, giving the capped convert holders additional shares.  The typical calculation is to price the shares to the debt holders as if the round were done at the cap, or in this case $4M pre, so it calculates as $6.2M on $4M pre for a $10.2M post.  The capped convert get 3.2/10.2 or 31.4% of the company.  This compares with getting 3.2/18.2 or 17.6% without the cap or a discount – an additional 13.8%.  (if this had been an uncapped convert with a 20% discount, they would have gotten 25% more shares, or gone from 13.8% ownership to 17.25%, an additional 3.45%; the cap adds over 10% more ownership!) The discount shares of the cap pushes the post-money to $21M.

Third, the pool is created at 20% post, which can push the actual post-money up by 25% to $26.25M unless it is specifically put in the pre-money value (as is typical in priced rounds – pool shares are typically loaded in to the pre-money shares even though they are calculated off the post money shares).  The calculation is to add the pool (or pool refresh if there is a preexisting pool) to the founders shares, since often there are no investor shares yet in the company. In our example, the pool is put in to the pre, so the post remains at $21M.

These calculations are inter-related – the calculation of the discount shares is done assuming a fully-diluted round including the pool.  When a stack of notes sit in front of a round, with different caps, interest calculations & rates, the calculations can get very complex. Break out the spreadsheet!  Typically the spreadsheet is used to figure out the post-money price and number of shares, and this is applied against the caps to get the discount.  The formula is the discount price per share is the capped-amount over the total pre-money fully-diluted shares, including the pool refresh.

Our simple example should end up as follows:

35% – founders
20% – pool
31% – angels (capped convert investors)
14% – new VC

Summary

Compare the two cap tables:

In one, the VC believes it is putting in $3M for 20%, and it gets that.  In the other, it believes it is offering the same sort of deal, but it ends top with only 14% – a huge difference.  The earlier investors end up in about the same position, whether a round at $4M pre or a capped convert at a $4M cap.  The founders make out much better.

What is the pre?  In the priced round case, it is as stated: $12M.  The post is $15M and the effective pre is that $15M less new cash of $3m.

In the capped convert case, it is much higher than $12M.  The post-money in our example turned out to be $21M, so the effective pre is $18M.  The rollup of debt added $3.2M to the stated $12M pre, and the discount shares added another $2.9M worth of shares – rounded, $6M higher than expected.

What to do? I suppose one could propose rounds in post-money terms only, but in the example above the stated pre had the benefit to the new investor of rolling the pool into the pre-money shares.  Stating a round in post-money terms only could also make a difference in how the capped convert discount shares are calculated, depending how the notes are drawn up.  Another option (which we have followed) is to put the discount shares into the pre-money; it is then much more straightforward to determine the effective pre.  To get back to the same pricing structure as a priced round, both the discount shares and debt would have to be rolled into the pre.

In the founder-friendly spirit of capped converts and super-angels, I do not expect a sudden shift to green eyeshade super-angel driving capped converts down like this. Indeed, I think they appreciate that their looser approach is contributing to grade inflation in the follow-on rounds – and trying to put it onto the backs of the traditional VCs.  We shall see how the Empire strikes back.

In the meantime, Caveat Inaugurator.