How Safe are SAFEs?

The joys and pitfalls of everyone's favorite convertible instrument.

Picture this:

It’s a Friday morning. It’s summer, it’s a gorgeous day out. You are in your favorite coffee shop, minding your own business, but not really. You start eavesdropping on the people chatting at the table next to you. 

They seem business-y. But they also seem like they might like computers. You focus more on what they are saying. They definitely like computers. The conversation is animated, you decide to pay attention, hoping maybe there’s some good workplace gossip. There’s a pause. You hear one say to the other: “how safe do you think a safe is?”. WTF? What kind of weird riddle is this? Is this the next generation of interview teaser questions? Everyone knows the warm lightbulb is the one that was just turned off. 

To contextualize, a close friend was raising money for their startup. They asked me how safe I think SAFEs are. To the random people overhearing us, without the benefit of capitalization and startup jargon, I apologize. For anyone who has founded a venture backed company, there’s a decent chance this acronym has been on your radar. Thankfully, this post is for everyone!

What is a SAFE and why should I care?:

Let’s start with definitions and background. Maybe skip this part if being a CEO or VC is your day job! 

SAFE stands for Simple Agreement for Future Equity, introduced by the accelerator Y Combinator (YC). It’s a popular way for startups to raise their first institutional financing (i.e. professional investors at funds rather than friends & family). 

How do founders raise money? I am likely (definitely!) oversimplifying, but I see two main ways founders raise money at the early (pre-seed / seed) stages: 

1) priced equity rounds (investors give the company money in exchange for equity today), and 2) convertible instruments (investors give the company money today, which converts to equity upon a triggering event later). SAFEs fall into this second category. 

What are the pros and cons of each?

Priced Rounds

Priced equity rounds clearly define: the valuation of the company, who owns what % of the company, and the company’s governance structure / board. Important stuff! But, they are comparatively slow and expensive. After the founder and investor agree on terms, it’s at least a couple of weeks until the money is in the startup’s bank, and a month (or two) is more likely. The startup can also expect to spend ~$50-75K in legal fees for even a straightforward process. As we know, speed and cost matter to all founders. 

Convertible Instruments

Convertible instruments somewhat flip this. Investors give the company money and their investment converts to equity upon a triggering event: typically a priced round or a sale of the company. Convertible instruments do not set a valuation, but often have either a valuation cap (a ceiling valuation at which the instrument will convert) and / or a discount, giving these investors a favorable price per share when they convert. The downside? Uncertainty for both founders and investors around ownership and dilution. 

Historically, convertible instruments were debt (convertible notes), but SAFEs improved on this by removing interest and maturity dates, simplifying terms, and focusing negotiations on valuation. Introduced in 2013 (making them really old in startup years?), I think SAFEs are now widely preferred over convertible notes in early-stage financings. I can’t remember the last time I saw a note. 

Cracking the SAFE:

Before I offer too many opinions - one quick disclaimer: I’m not a lawyer and this is not legal advice! Founders should get a lawyer and consult them as they think through financing options. I’m happy to recommend a few! 

With that out the way (am I burying the lede?), I am generally positive on SAFEs for seed rounds, i.e. raising less than ~$5M at valuations under ~$20M. 

In those instances, my experience is they are a fast and cost-effective way for seed-stage companies to streamline the fundraising process. They mostly limit negotiation to the most impactful terms: valuation cap and discount. It is easy to add side letters or clauses for major investor or pro-rata rights. They remove the interest and maturity problem that convertible notes create. Ask me sometime about when I had to extend the maturity date on a note three times…or actually maybe don’t. (Definitely don't ask my prior firm’s finance team, not sure if they have forgiven me.) At larger round sizes and higher valuations, I think it’s better to pursue a priced round to set the company up for Series A and beyond. 

Let’s come back to my founder friend in the coffee shop. They had two major questions. First, as a founder, they wanted to know how to think through the dilution they had effectively taken that was not reflected on their cap table. Second, they wanted to understand how investors approached investor protections, governance and ownership with a SAFE.

The ownership & dilution question is a more straightforward one to answer as it is mathematical. Any founder and investor participating in a SAFE can (and should?) model out their likely ownership upon conversion:

• Carta has plenty of good benchmark median valuation and round size data which can inform the model. 

• For those who are not fans of Excel, Carta also have a calculator which automates the modeling process, as do my friends at 1984 Ventures. 

The protections and governance question is a more nuanced one. SAFEs provide investors with many of the desired features of an equity-like investment - liquidation preferences, company & investor representations etc. But there is no board seat or veto rights for a certain share class. Potential hot take, but I don’t know that this bothers me too much at this stage. To be clear, I am a big fan of both accountability and good governance! But most seed-stage companies still have founder-controlled boards even after priced rounds. And the bigger picture is that a huge component of investing at the earliest stages is investing in people and trusting them to have good judgment. If you are really having to lean heavily on governance and protective provisions to drive the desired impact at the earliest stages, then something has probably already gone wrong. 

When SAFEs go wrong: 

Lest I be accused of being a one-sided fanboy - I do not think SAFEs are perfect. Like many aspects of the fundraising process, not everything is always understood or obvious. 

The most common “gotcha” is around ownership on conversion: “I own less than I expected / I took more dilution than I wanted”. Real issue, and very much not a fun one. 

A couple of ways this can manifest: 

• Stacking SAFEs. When multiple SAFEs convert into the same round, the dilution can add up quickly and lead to surprises. Take this scenario - a founder raises their first $1M on a SAFE at a $7.5M valuation cap. The team is performing well, word gets out about what they are doing, and investors want in. They raise another $2M on a SAFE with a $15M post-money cap. Later, the company raises a $10M Series A at a $50M post-money valuation. When the SAFEs and Series A convert, the founder’s stake drops from 100% to about 50%. While this scenario is a success, the cumulative dilution from multiple SAFEs can be a shock if not carefully planned for. Imagine how messy it can get with three or four SAFEs converting at once! 

• Converting below the valuation cap. This is a tough one. Say a founder was able to raise money at a $20M cap during “good times” but the market turns, or the space they are in falls out of favor. After a grueling fundraise, they ultimately raise equity financing at $10M post. The founding team takes twice the dilution they were expecting to from the SAFE, as it converts at the $10M valuation, not the $20M cap they had planned for. Oof. 

• Option pool dynamics. Since SAFE investors are not yet on the cap table, the founders take the dilution burden of option pool expansion prior to SAFEs converting. This is usually not a big deal - but there can be weird edge cases (e.g. making an early non-founder executive hire that requires a big option grant) that require option pool expansion prior to the converting priced round, and founders take proportionately more dilution than they would have if the investors were already on the cap table.

• Pre vs. Post-Money SAFE. When the SAFE was first introduced, it was pre-money, creating a shifting valuation cap depending on how much money the company raised. This made it impossible to calculate ownership / dilution at the time the SAFE was issued. YC changed this (thank goodness) by making the SAFE templates post-money in 2018.¹ I think this is pretty much a solved issue now (if you go to YC today all the templates are post-money), but worth being aware of. 

Again, with a bit of modeling and some forethought, I think these issues are navigable…and often when this results in surprises it is because of a challenging or complex fundraise. 

To the person at the back who has been holding their hand up dying to ask a question - yes, there are indeed situations where investors will participate in an uncapped SAFE with no discount. I think most investors who do this kind of thing see it as a call option that gives them access to the founder(s) and business. This can make sense for deep-pocketed investors at multi-stage funds, but is probably a *questionable* choice for angel / seed investors. Personally, I avoid these.

I have also seen misunderstandings around when pro-rata rights, if applicable, take effect: in the converting round or only afterward. I likewise think this is a solved problem since the SAFE became post-money as standard. YC changed the generic pro-rata side letter to make it clear the rights applied into the converting round.² Since pro-rata rights are usually only granted to larger investors via a side letter, I think it is good practice for both founders and investors to clarify what their intentions are to make sure they are aligned before signing the side letter.  

In Lehmann’s Terms:

So to answer my friend’s question directly: when used properly in a seed round, I think SAFEs are pretty safe. 

As an investor, I feel as though the SAFE does what it says on the tin: a simple agreement for future ownership in a business. I can't say I own X% of the company today, but there are guardrails around what my ownership will be on conversion from the valuation cap / discount. I can get comfortable understanding if something is in my strike zone and make a judgment on relative risk and reward. It codifies what happens upon financing / sale, liquidation preferences, and provides company & investor representations. 

The variety of templates you can download from the YC website reflect the variables that people have tweaked over the years, but it generally feels as though the instrument has come to a point where it’s familiar and parties know what they want to negotiate on.

They are not perfect, but I’m also not sure you will find a group of lawyers who will provide consensus on any document being ideal (I’ll report back if I hear differently!). I’m also sure some people have horror stories too and will never use them again (and please tell me what happened if so). 

If you made it this far, thanks for reading! Hope you enjoyed me breaking it down, so you don’t have to.

Will 

1. More context from the YC site here: Our first safe was a “pre-money” safe, because at the time of its introduction, startups were raising smaller amounts of money in advance of raising a priced round of financing (typically, a Series A Preferred Stock round). The safe was a simple and fast way to get that first money into the company, and the concept was that holders of safes were merely early investors in that future priced round. But early stage fundraising evolved in the years following the introduction of the original safe, and now startups are raising much larger amounts of money as a first “seed” round of financing. While safes are being used for these seed rounds, these rounds are really better considered as wholly separate financings, rather than “bridges” into later priced rounds. In 2018 we released the “post-money” safe. By “post-money,” we mean that safe holder ownership is measured after (post) all the safe money is accounted for - which is its own round now - but still before (pre) the new money in the priced round that converts and dilutes the safes (usually the Series A, but sometimes Series Seed). The post-money safe has what we think is a huge advantage for both founders and investors - the ability to calculate immediately and precisely how much ownership of the company has been sold. It’s critically important for founders to understand how much dilution is caused by each safe they sell, just as it is fair for investors to know how much ownership of the company they have purchased.

2. From the YC Safe User Guide: In the post-money safe, we removed the pro rata right that existed as a default option in the original safe. That pro rata right applied to the financing after the round in which the original safe converted (e.g. if the original safe converted in the Series A, the pro rata right applied to the Series B). Instead, we created a standard side letter with pro rata rights that apply to the round in which the safe converts (e.g. if the safe converts in the Series A, the pro rata right applies to the Series A), which can be used if and when the parties agree to it. Although our original goal was to create a universal standard for pro rata rights for all start-up companies, our experience was that we couldn’t do this in a way that made sense for all parties. For example, a company raising $500,000 from 10 angels investing $50,000 often has significantly different considerations than one raising $2,000,000 from a single institutional investor. Also, the pro rata rights contained in the original safe were often misunderstood by both founders and investors as applying to the round in which the original safe converted (the Series A), rather than the round after the one in which the original safe converted (the Series B). So the standard pro rata side letter is an acknowledgment that this right is best handled case by case, and that the prior form of the right often wasn’t what founders and investors were expecting.