Earlier this week, Peter Walker from Card common data on the use of assorted mechanisms in SAFE issued to U.S. startups in 2024: 61% used the cap, 30% used the cap and discount, 8% used only the discount, and 1% used neither.
In response to a query about the variety of unrestricted SAFE funds, or Simple Agreement for Future Equity, one investor replied, “That means at least 1/12 of VC funds have no idea what they’re doing.”
This is the legacy of the unrestricted SAFE, which is now synonymous with bubbles, irresponsible fundraising and investors get burned.
The Beginnings of Unlimited SAFE
In August 2021 Combiner Y (the creators of SAFE) have updated their SAFE templates to remove the version that contained each a limit and a discount, and have clarified this note:
“YC’s recommendation to founders was to either issue a capped safe or a discounted safe. We didn’t encounter any situations where the combo safe was the preferred choice. As such, we felt it was inconsistent to issue this safe.”
The query arises: why does YC see the point of introducing SAFE (only with a discount)~8% SAFE), but not the discount and the SAFE limit (~30% SAFE)?
The purpose of limits and discounts
Both mechanisms are designed to offer early investors with preferential access when their investment converts to shares in a future pricing round.
- Caps are intended for startups with a more practically predictable future, comparable to SaaS or e-commerce. The cap should reflect the projected growth at a given point in time, in addition to the additional risk SAFE investors have taken on in the meantime.
- The rebates are designed for corporations that have upfront risks with too much uncertainty to set a reasonable limit. Consider a startup that raises SAFE to prototype hardware, conduct clinical trials of a drug, or fund large-scale pilot projects.
YC understood it this manner: limits and discounts solve the risk problem from two different perspectives.
The practice of mixing the two, where a discount is applied if a startup raises capital at or below its cap, is punitive. In such a scenario, investors must be seeking to put startups in the strongest position for future growth, not dilute them even further.
SAFEs combined also distorted expectations about discount levels. median discount 20% reflects the role of a ceiling hedge and is much less meaningful as a standalone mechanism. This is not an attractive compensation for risk at this level, but founders may feel uncomfortable with investors suggesting something significantly higher than the median without a clear basis.
SAFE discount update
When YC originally created the SAFE rebate, it missed the way the fundraising process would evolve: over time, SAFE rebates would grow to be more common, stacked, and founders would wish an incentive to lift a funding round and convert.
Caps work well here as the company grows and the supposed dilution becomes more inconvenient. This is not as true for the (flat) discount, which makes SAFE discounts should compound over time. In fact, looking at required return on investment In the case of enterprise capital as the basis of calculation, the discount has a logical elegance, ensuring that investors are adequately compensated for the risk.
The results of this approach can be much larger discounts (over 50%), but applied without a cap—offering more flexibility for startups at the extreme end of the risk/reward spectrum. The dilution results are still inside a practical range, especially for startups on a path to rapid (but capex-intensive) growth.
This would breathe recent life into a helpful fundraising tool that had fallen into misuse, just in time for the hard tech founder renaissance of 2024.