Convertible and SAFE notes
avc.com
avc.com
Of course it's still possible to get more dilution than expected if you raise at unrealistically high caps (or uncapped!) and convert at a lower valuation down the road, but even that is probably less painful than a conventional down round.
The biggest benefits of the note structure to us were a) Rolling close b) Not wasting time debating "valuation" when neither us nor the angels were capable of estimating a number with limited to no data.
In what way is the cap not a "valuation"?
I'd defer to Paul's point that choosing a too high (or non-existent) cap makes no sense. We did have to negotiate our cap and we set it so that we had a realistic shot of hitting 2-3x that in Series A pre money which in my mind is why an angel should be investing. I'd personally consider it a disappointing outcome if I raised at my cap and I'd hope my investors would as well.
A cap is not a valuation and I agree that everyone thinking it is, is a problem.
1) For a VC: more performance info / better access to the A.
2) For an angel: most angels won't have access to the A, so seed rounds are their chance to invest early in a potentially VC track company.
+100 on this. This is particularly problematic in India where most early stage investing is priced. The opportunity to bully founders to accept weird valuation terms is limitless.
notes are a trade-off of current capped valuation for future capped dilution. Guess which one gets misused by investors ?
Mysteriously, the only people who oppose SAFE notes are investors who would like you to instead enter a complex negotiation with them!
The "bad case" alternative to the second/third round of notes is often a down round or running out of money -- way worse.
The "good case" alternative of a second higher cap note batch is fine.
Founders and early investors can get screwed even if they raised as equity early on, too.
I do think the analysis is both incorrect in conclusion (SAFE is the best choice in many/most cases for both sophisticated founders and new founders), and the case badly supported in the blog post.
But it probably isn't someone doing the equivalent of pushing a self-serving agenda, which was my instinctive reaction and thus defensiveness.
He's an example of where having many notes can hurt founders:
- Raise $2m at an $8m cap, 15% discount.
- Raise $2m at an $18m cap, 15% discount.
- Raise $2m at a $28m cap, 15% discount.
- Sell 20% of company for $X in the Series A.
Caps are "sort of" like pre-money valuations, so the founder might expect that their dilution from the 3 notes is approximately 2/10 + 2/20 + 2/30 = 36.7%
Here's what actually happens in 4 different scenarios:
1) Raise $10m at a $40m pre. Dilution from the notes is 20.0 + 8.9 + 5.7 = 34.6%
2) Raise $7.5m at a $30m pre. Dilution from the notes is 20.0 + 8.9 + 6.3 = 35.2%
3) Raise $5.0m at a $20m pre. Dilution from the notes is 20.0 + 9.4 + 9.4 = 38.8%
4) Raise $4.0m at a $16m pre. Dilution from the notes is 20.0 + 11.8 + 11.8 = 43.6%
For the founder, #1 and #2 are better than expected, #3 and #4 are worse. It's not hard to get into situation #3 or #4 if you raise money from strategic investors or angels at high caps, and then Series A investors drive your price down below your most recent caps because that's where the market is. The 9% dilution difference between the first and last scenarios is fairly dramatic. When you combine that with the 20% dilution from the Series A, it's the difference between founders and employees having ~45% of the company vs ~36% of the company.
(Source of calculations: https://captable.io/company/8/convertible-notes/calculator)
If you're just raising a small seed in order to get something going, what's more important than moving fast and getting the product to market?
But notes/SAFEs are great for the earliest employees since we can grant them real shares instead of options. Plus, we can do it at par value.
Shares granted before a priced round are typically identical to the shares a founder has, with all the same tax advantages.
Doing priced rounds early means having to worry about a 409a early, and probably giving employees worse quality equity (or creating income tax problems for them).
And if you don't, that makes situation a lot worse for the employees, as they end up owning taxes at each vesting event.
The concern I think you have is that FMV of the shares might have increased after company formation? That is what 409a is for, but you generally have a decent window between reassessments, especially early on. (6mo? Is what I've seen)
83(b) clock starts when the restricted grant is issued to employee. The price is set at that time.
Priced rounds are great for VCs because they remove all their risk. But they don't remove any founder risk. Do a down-round after a priced round and you'll wish you were just taking more dilution from a SAFE.
> "1. They defer the issue of dilution until a later date".
When the company is doing well, notes allow founders get less dilution, sometimes way less. This is great for founders, and is just about the only time in the whole startup thing where things go better than expected for founders.
If the company is doing badly, I'll take a lower conversion rate over having to raise a down-round any day.
> "2. They obfuscate the amount of dilution the founder(s) is taking. I think many investors actually like this"
He sets up a moral imperative: you're an irresponsible founder if you don't know what dilution you're taking. But I know what dilution I'm taking: it's X _or less_!! (Again, in a down-round I'm fucked either way, so the downside of notes doesn't matter)
> "3. I cannot tell you how many angry pissed off angel investors [..] they own a LOT less than they thought they did."
And here we get to the crux of the issue, investors own less. If investors own less, who owns more? Founders do.
> "4. Most importantly, we cannot and should not continue to allow founders to issue notes to investors and not understand how much dilution they are taking on each time they do it. This is WRONG."
"Silly inexperienced founders accidentally give investors a bad deal. This is morally WRONG". Needless to say, I'm unsympathetic to this view.
Look I understand why VCs don't like notes. But to pretend that this is a founder-friendly thing is beyond irritating (note: he's far from the only VC to do this!). And this is one of the reasons I trust YC far more than any other investors - they're in it to help people make startups and they argue for founder-friendly terms.
(PS there is a real issue in that inexperienced founders might raise too much thinking its free money, and then they get a lower cap than expected. But my only bad experience with raising with notes is that VCs treat it as a lower implied valuation when raising the next round.)
This is a very disingenuous argument by Fred Wilson, which is a shame because his analysis is usually so intelligent. But when it comes to convincing founders to give VCs more equity, he's driven by his own incentives just like any VC.
99% of first time entrepreneurs have a simple question - "how do you value an idea, an early stage startup": It completely eliminates this question.
First, we start with evaluating ideas. Gee, guys, the research community does that many times daily. A researcher who pursues dumb ideas in a few years is known as a poor researcher. So, good researchers know how to evaluate ideas.
The researcher's superiors, the proposal reviewers, and the funding agencies also all know how to evaluate ideas and where here the ideas are just research directions with the ideas not really formed yet.
Moreover, for publication in peer reviewed journals of original research, the standard criteria for evaluating the ideas in a paper are "new, correct, and significant", and these criteria get applied with good accuracy.
Moreover, for a Ph.D., one way to evaluate the work is "an original contribution to knowledge worth of publication." And, before a Ph.D. student starts their research, they and usually also their faculty advisers have to evaluate the research direction, even before a solid idea.
The US NSF, NIH, DARPA, ONR, etc. are also quite good at evaluating research directions well before a finished idea.
Of course we can evaluate ideas, especially ideas already polished and ready for publication and also finished, polished ideas already implemented in solid, running code.
Then, sure, that evaluation does not necessarily lead to business success, that is, traction, revenue, earnings, money in the bank, an a good ROI and exit for investors. Right.
But, wait, there's more!
Second, here's a severe way to evaluate such a polished idea ready or nearly so to go for first traction: Or, with the running software, can see that it works. Now the severe question is, "Is that software a must have solution with enough users/customers and revenue per each to make the big bucks or is that software just a nice to have and/or for just a niche?" And we will want to know about barriers to entry, scalability, etc. Doing well on these criteria should be darned good signs.
Apparently instead the VC community wants to judge based almost entirely on traction and rate of growth in traction. Yes, that's hard headed; it's also dumb headed, ignoring important information.
Some businesses do quite well with new ideas:
Xerox did well with the research direction of electro-static photocopying -- from just the research direction through creating a major company.
At each smaller step in line width, Intel does well. Who would have believed 1 billion transistors on a postage stamp with line width 14 nm?
The pharmaceutical companies that do original research make their money on picking good research directions and then doing good research.
And may I have the envelope, please [drum roll]! The all-time, unchallenged, unique world-class, grand champion of going from mere research directions to world changing results is the US DoD. Second and third places are occupied by the NSF and NIH. Intel, some of the pharmaceutical companies, and more are also on the award stage. VCs need to start learning up from those examples.
The VC approach would have been, "You build and test the first one, and we will chip in for the gasoline for the Enola Gay."
Instead, IIRC from books of Richard Rhodes, the Manhattan Project cost ballpark $3 billion in 1940's dollars.
Okay, big, expensive waste just from war time, right? Maybe not: Let's divide $3 billion by one million -- right, even without a super-computer, that's $1000.
And the relevance? The estimates were that a US invasion of the home islands of Japan would cost the US 1 million casualties. Given the experience at Okinawa, Iwo Jima, etc., the 1 million was believable. So, the Manhattan Project saved expected 1 million casualties at, right, from the arithmetic, $3000 per casualty. Grand bargain just from the money alone! If your father, son, brother, uncle, etc. were one of those casualties, then you would have regarded the $3000 as a grand bargain.
And the Manhattan project spent money like it was worthless -- cost no object or concern. Good ideas are like that.
I doubt you are a VC but maybe you should start your own firm.
>And here we get to the crux of the issue, investors own less. If investors own less, who owns more? Founders do.
I think in this case there are just "more investors than the earlier investors expected to exist".
In any case, if there's a subsequent convertible round after the round you were in, of course there will be more dilution for the angel! And honestly if the subsequent noted were raised at higher caps comparable priced rounds, then the dilution to the earlier angels should be less anyways. This just sounds like optimistic angels not understanding the equations they are a part of.
Frankly said angry Angel should recognize that any subsequent round dilutes them and the founders in similar fashion, and any instrument that is "founder friendly" is existing investor friendly as well by definition.
I agree. I edited out of my rant the fact that investors are supposed to be sophisticated. Everyone knows what they're getting into here. Angels are typically way more sophisticated than startups (if you've done 1 deal that's usually 1 deal more than a founder has). If you can't do the math on this you shouldn't be investing in startups.
For context, I'm a long-time fan of your writings and REALLY appreciate how much you've given to the community (with the only VC blog worth reading IMO). Just want to fully understand the issue at hand here.
I don't have any agenda or ascribe any to you. All I think is that to the untrained eye, this post comes across as founder-unfriendly.
Among the asserted merits of SAFE's when the came into the YC ecosystem was that it removed some of the distraction that trying not to get screwed over created for founders at 'demo day' or earlier stage companies. At the time, an investor's willingness to use a SAFE might have had an inverse correlation to their willingness to take undue advantage of unsophisticated founders.
They'll only want to invest when 'I know this is the winner in this category', but I'll call it Series A and Series A has always been done at valuation $x.
FU! You won't fund a deal till it's profitable and all the risk is taken off. But, I need to get money from somewhere! So, I'll get it from small angels and individuals. And since I have to collect it 25-50K at a time, I can't do a priced round.
You set this board up, now don't complain that you have to deal with the mess.
[Edit] And if you'll only be dealing with a startup once it has gotten some level of success, you have no option but to deal with the mess, if you want the deal.
"Founders should insist that their lawyers publish, to them and the angel/seed investors, a “pro-forma” cap table at the closing of the note that shows how much of the company each of them would own if the note converted immediately at different prices. This “pro-forma” cap table should be updated each and every time another note is isssued. Most importantly, we cannot and should not continue to allow founders to issue notes to investors and not understand how much dilution they are taking on each time they do it."
Requiring that a cap-table be published is somewhat different than ensuring that the founders understand their dilution. It's quite likely there is a ton of overwhelming information/events going on during a close, and it's easy (though it shouldn't be) at this stage for a founders eyes to glaze over yet another-table. But taking the extra 60 minutes to walk through the implications of dilution, and ensuring that they actually get it is something a responsible investor/VC will do.
I wonder which situation is more common?
Looking at you, Uber.
That's what the cap if for. If the founder is 19 years old, fine, read a book on it (I recommend this: https://www.amazon.com/Funded-Entrepreneurs-Guide-Raising-Fi... but basically any will do) but angel investors that don't know what caps are are dumb angels.
> They obfuscate the amount of dilution the founder(s) is taking.
Also what a cap is for. Unless we're talking about it from the founders perspective, in which case it's their decision to raise another round or not. If the dilution doesn't make sense don't raise the round. If you're desperate and were banking on hitting the cap maybe the company wasn't worth what you thought it was worth when you raised your Angel round.
> They can build up, like a house of cards, on top of each other
Here I 100% agree. Early family and friends money or angel investments or small seeds should be convertible, but subsequent rounds should be priced, or a down round is going to eviscerate the cap table.
> They put the founder in the difficult position of promising an amount of ownership to an angel/seed investor that they cannot actually deliver down the round when the notes convert.
What? How? Only if the founder is lying or doesn't put in a cap can I possibly see this happening.
> The company has been around for a few years and has financed itself along with way with all sorts of various notes at various caps (or no cap) and finally the whole fucking mess is resolved and nobody owns anywhere near as much as they had thought.
False dilemma. The choice isn't "use convertible and get a fucking mess" or "do priced rounds for equity and keep things clean". If you keep your capped convertible round simple and follow it with a priced round for normal equity anyone with two brain cells can figure out what the outcome will be given different scenarios. I've seen some really, really fucked up priced rounds because some dumb angel investor took a dump on the shareholder agreement and it took hundreds of thousands in legal fees and payoffs to fix by the time they were going for what they ended up calling their seed (they renamed their small seed their angel round).
> It can easily be done for less than $5k in a few days and we do that quite often.
That isn't the issue. The issue is that a bunch of regulatory issues come in once a company is selling stock that a two person startup doesn't want to waste time dealing with.
> The first convertible or SAFE note issued in a company should have a cap on the total amount of notes than can be issued. A number like $1mm or max $2mm sounds right to me.
What? Is this at the idea phase or something further along? Because I don't know anyone that has a $1M cap if there is any serious amount of work or team.
> Don’t do multiple rounds of notes with multiple caps. It always ends badly for everyone, including the founder.
I agree. Only do this if you're desperate or if you raised a tiny angel round (<5% after cap).
> [...] a “pro-forma” cap table at the closing of the note
I agree, though more as a CYA thing to keep relationships good than something I think would be necessary if the founder set up a sane round in the first place.
I'm personally not a fan of SAFEs because they can get hairy when you have lots of angel investors, but convertible is great.
As a partial aside, this whole "strong views weakly held" (see http://avc.com/2016/06/strong-views-weakly-held/ for more details) is kinda annoying to debate with because it frequently boils down to me having to colour in the nuance. It's like a less extreme version of arguing with Trump supporters.
My style is:
Strong views strongly held, until they're no longer strong views because they've been eroded away by building, credible counter-arguments. I generally stop talking about the view during this period or when I do talk about it I'll ask questions. I think of this as weak views weakly held. Then I re-investigate the view's premises and:
1. If finding my original assessment still broadly correct, I evaulate the counter evidence and modify the view to include a greater degree of depth, since I now understand a further degree of the complexities. The view then becomes a "nuance aware-strong view, strongly held".
2. If finding my original assessment false, build a new strong view (strongly held) from the premises of others.
Strong support carries information "this information should be trusted" and I don't like broadcasting that signal if my view is weakly held.
By "cap" in this context, I believe he's referring to the aggregate amount of convertible notes.