No matter how many years have passed since I wrote my last bit of production code, I will always stay a software developer at heart. A software developer is not a profession, it’s a state of mind, and real devs tend to view the world in a certain way. This is why the topic of the “SAFE” and how it has been abused in the Israeli startup industry is so painful for me.
Good software developers are lazy. This is a well known fact. When they run into a problem, their minds immediately attempt to generalize the problem, parameterize the problem, and ensure the solution to the problem is reusable.
As a software developer turned Venture Capitalist, the idea that investment documents were drafted from scratch each and every time pained me very much, and still does. Specifically, convertible loans, which until 2013 were the industry standard for facilitating startup funding between priced equity rounds — I could not, for the life of me, understand why they were not generalized, parameterized and reused. And then they were:
The 2013 Original Pre-Money SAFE
In 2013, Y Combinator introduced the original SAFE, also known today as the Pre-Money SAFE. While the use of convertible loans for startup financing is basically a legal/accounting hack (“ugh!” said the programmer in me) which uses a debt instrument to express a non-debt transaction, the Pre-Money SAFE was engineered as a non-debt instrument. It was designed to be a simple and fast way to get that first money into the company, and an expression of the concept that holders of SAFEs were merely early investors in a future priced round, with the hidden assumption that it will be used to raise a relatively small amount of money — essentially “bridges” into later priced rounds.
While originally designed as a bridging mechanism, the SAFE documents were quickly adapted in Silicon Valley as the most common financing method for Pre-Seed, and when dynamics changed and startups started raising higher and higher amounts at a faster velocity, SAFEs were often used, or mis-used in these rounds as well. This led to complicated situations in which companies had SAFEs on top of SAFEs, and the actual ownership of shareholders was not clear for long periods of time: it was harder for founders to calculate precisely how they were being diluted, and harder for investors to calculate the ownership stake they were buying. In the words of the YC folks:
The answer to “how much of the company are we selling” was dependent on a recursive loop of how much was raised on other SAFEs, plus a hypothetical assumption about the Series A option pool increase that would be negotiated years later.
Note: these documents are somewhat hard to find today. We’ve provided an unmodified copy of the original YC docs here.
Meet the new 2018 Post-Money SAFE
In September 2018, the folks at Y Combinator created a new set of SAFE documents. While retaining the simplicity and clarity of the previous version, the new SAFE is fundamentally different from the original one as it is “Post-Money”, allowing dilution and ownership to be calculated in a simple manner, much like a priced equity round.
Important note to founders: The side effect of this approach is that each additional SAFE raised effectively dilutes just the current stockholders, which is often the founders and early employees. However, since the calculation is now simpler and more deterministic, it does allow all stakeholders to make rational decisions with the rules of the game being known and clear in advance, which is always a plus.
The 2018 version of the SAFE has one additional, important change. It contains the following language:
This Safe is one of the forms available at http://ycombinator.com/documents and the Company and the Investor agree that neither one has modified the form, except to fill in blanks and bracketed terms.
In other words, this new version of SAFE is not meant to be modified in any way, and should be used as is. If respected, this should lead to lower legal fees, shorter negotiations, and faster money-in-the-bank. Essentially, the startup and the investors only need to agree on the terms (discount no/yes-how-much, cap no/yes-how-much, Pro Rata rights yes/no), fill in the blanks, sign, and get back to focusing on the important stuff.
Comparing pre-money SAFE with post-money SAFE
The impact of the different SAFE types on shareholders dilution is important, so I’ll demonstrate it with an example.
Let’s simulate the following scenario:
- You are raising $250K using SAFE with a cap of $750K pre / $1M post, no discount.
- You then raise $1M in an equity round with a pre-money valuation of $9M (so the SAFE converts at cap value).
To simplify, we will assume you hold all the shares prior to this, and that no option pool is allocated at any point.
In this scenario, regardless of the type of safe used (2013 with a cap of $750K pre, or 2017 with a cap of $1M post), the SAFE holders will own 25% of the shares after conversion, and 22.5% after the round. You will retain 67.5% ownership.
Now let’s assume that before raising the equity round, you decide to raise another $250K using SAFE, on the same terms. So it’s a $250K SAFE, followed by another $250K SAFE, and only then the equity round.
If both SAFEs are the 2013 Pre-Money kind, you’ll end up with the SAFE holders owning each 20% of the shares after conversion and 18% after the round, and you will retain 54% ownership.
However, if both SAFEs are the 2018 Post-Money kind, you’ll end up with the SAFE holders owning each 25% of the shares after conversion, and 22.5% after the round, and you will retain 45% ownership.
So be careful not to pile (post-money) SAFEs on top of each other without fully understanding the impact of doing so on all stakeholders.
SAFE Makes an Aliya, Oy Vey
So much for Silicon Valley drama. What about Israel you ask? Good question.
Being a seed-focused fund, it has always been very important for us to increase the fundability of the startups we invest in, and that includes reliance on standard, Silicon Valley-compatible terms and legal documentation.
Therefore, for years we’ve tried to get our portfolio companies, or more precisely, their legal counselors, to use SAFEs when appropriate — and have met resistance. Each firm had their own CLA (“convertible loan agreement”) version that evolved over many years of use, adapting to the always changing rules and regulations of Israeli Tax Authority, tried and true — they were reluctant to give up the comfort of the familiar tools (and perhaps also the legal fees associated with them).
In 2017 this started to change, and more and more deals were getting done using SAFEs — sort of. Old habits die hard, and the legal teams kept modifying the SAFE templates, sometimes in a rather liberal manner. Each and every SAFE-based transaction still needed to be reviewed carefully to spot the changes and analyze their impact.
Once the new 2018 SAFE was introduced, things got even more complicated. News of the New SAFE were slow to reach the Israeli ecosystem, and we often found out that the term SAFE in a rather liberal manner, that could mean either the original 2013 SAFE, or some Israeli-modified variant of it, or the 2018 SAFE, or even *gasp* a modified version of the non-modifiable 2018 SAFE.
Only recently did we run into a case where such a modification caused the discount to be reversed (i.e. 80% instead of 20%). Good thing we caught this in time before it became the New Israeli standard SAFE :)
Let’s Make SAFE safe again
This is a call to the Israeli startups ecosystem: let’s clean up our act and settle on the 2018 Post-Money SAFE as a standard. The benefits are tremendous: fast and efficient transactions, reduced legal fees, clarity for all stakeholders — founders and investors, and — compatibility with Silicon Valley standards, resulting in shorter, simpler Due Diligence processes with American investors.
You can find links to the most recent unmodified (duh) Y Combinator 2018 SAFE documents here: