I recently closed my third angel investment and the first one I led from beginning to end. Leading made me think a lot about deal terms. We ended up using slightly modified Series Seed docs, which I plan (and would love) to continue to use.
The one term we slightly modified from the model term sheet was the liquidation preference. As I seem to be at odds in this change with people much more experienced than myself, e.g. Fred Wilson and Chris Dixon, I want to lay out my thinking and hopefully get some feedback.
First, why would I ever differ with these guys? I think it's simply because I don't have as much money to invest, and as a consequence, my angel investing strategy is different.
These guys are so-called super angels. And if I had more money I'd probably end up being one too. But right now, I have a limited pool of angel money. It's enough to do a bunch of deals and I'm fine with losing it all, but it isn't enough to do the YC/Conway (super angel) strategy.
Here's my basic argument. Suppose you have a "hit" rate of 10%. I realize it's more complicated than that and I've developed a spreadsheet to model it better, but for sake of simplification just go with me for a moment.
At this rate, every deal you do has a 90% chance of not being a hit. People then think intuitively, OK, I need to do 10 deals then. Nope--the default intuition is wrong. What's the probability that you do 10 deals and don't get a hit? (9/10)^(10) =~ 35%!!! That's a massive %. At 20 deals you're down to 12%, but that's still pretty high.
So basically you have to do a lot of deals if you're at all serious about making money off of angel investing. A recent Monte Carlo simulation by Jeff Miller confirms this thinking. He concludes 20 deals is a sweet spot.
If you're a new/small angel like myself with a fixed investment pool, what can you do to maximize the # of deals? Two things really. First, make smaller investments--though this isn't great because there's a lower limit where you'll get shut out of deals ($25K is probably the realistic minimum) and, when you do hit, you won't get that much back.
Second, liquidation preference. In my previous post on angel investment scenario planning, I noted that small returns (1-3x) don't influence overall deal pool IRR that much. And that's true in a static scenario. But the real world is dynamic and if you can do more deals you can increase the chances of hits, as noted above. Getting money back through liquidation preference won't move the IRR needle, but it will enable you to do more deals. And to the small angel (me) that is the most important thing.
If you have no idea what liquidation preference is (and you want to know!), check out Brad Feld's two great intro posts on the subject. The default term from the Series Seed Term Sheet is essentially "One times the Original Issue Price." That means in an exit event, the investors have a choice to a) get their money back or b) get their equity %. It's an economic choice. If it's a small exit and the equity % doesn't yield as much money as was put in, you'd opt to get your money back; otherwise, you'd opt to get your %.
Tweaks on this model are:
- Get multiple times your money back (e.g. 1.5, 2 or 3 times) in choice a.
- Get your money back first, and then also participate in the distribution of what's left based on your %, which is called participating preferred.
- Participation, but with a cap. In this one you participate up to a limit, like 3x. So you get your money back, and then your % allocation, but if that exceeds 3x your money you have the choice whether to get that or your straight %.
Anyway, entrepreneurs seem to hate this stuff, especially the participation part. Some people call it "double dipping" because you get money back and then you double dip and get more money back out of what's left. I think this is somewhat of an understandable knee-jerk reaction, but everyone should look into different realistic scenarios to see what the actual implications might be.
I woke up today to read Fred Wilson's evolved view of liquidation preference, which is that he used to insist on participating preferred, but now only does (and with a cap) in certain specific situations.
Apparently Chris Dixon agrees with this view, and goes further saying "[a]nything more than that (participating preferred, multiple liquidation preferences) divide incentives of investors and the entrepreneurs. Also, this sort of crud tends to get amplified in follow on rounds."
My current thinking is that a 2x liquidation preference is a sweet spot. Let's look at a concrete example. Here is a spreadsheet I made where you can play around with it and other scenarios at your leisure.
Suppose you put in 50K in a 500K round on a pre-money of 2.5M (post-money 3M). You end up with 1.67%. Let's say the company doesn't take any more investment (for simplicity sake) and in year five it exits/liquidates.
In scenario 1, you got the normal 1x liquidation preference. In scenario 2, you get 2x. And in scenario 3, you get 1x, participating. Here's how those scenarios play out in terms of IRR at different exit sizes.
1M - 0.%, 15%, 6.%
2M - 0.%, 15%, 13%
5M - 11%, 15%, 22%
10M - 27%, 27%, 34%
20M - 46%, 46%, 50%
30M - 58%, 58%, 61%
At high exits, the term doesn't matter much--they converge pretty quickly. It clearly matters <=5M (at least in terms of IRR). With 2x liquidation preference, your IRR jumps from 0 to 15%, a healthy % that you can use to re-invest (and do two more deals at the same size). And interestingly, you can see the participation right does worse than 2x at small exits, does better in the middle, and smooths out at the high end.
Now let's take the cons in turn.
- Splits incentives. When you get crazy preferences, this can certainly happen. And if you have participating with a cap, as Brad notes in his post there is a plateau range in the middle at which the investors are indifferent to exit size, which is weird. But at 2x, non-participating, in a smallish round (<=500K) I don't see much issue with regards to incentives. The preference only kicks in for low exits (in this case sub 6M). The entrepreneur should not be getting financing if they would be happy with an exit lower than that.
- Mucks up later rounds. There is a risk here in that, at the very least, you can expect later investors to get similar preferences. Brad points out scenarios where you can actually lose out as an investor by having these rights (if the next investors get them too). Again though, if they get limited to 2x, non-participating, I think that is a manageable risk.
- They'll get thrown out anyway. I've heard this, but don't have enough experience to know if this happens or not. Ideally you'd also get some control rights on the next financing so at least you can be sure you won't get screwed in that next round.
- IRR doesn't matter for low exits. It matters to me :). In particular, as stated above, if it did return the preference amount, this would yield the ability to do 2 more deals instead of 1. A few of those, and an initial pool of money allocated to 10 investments may easily stretch to 20.
Update: some good comments on HN.
Update 2: I made a new post explaining more of the background on this post and addressing the comments.