Some Thoughts on Equity

Equity underpins our entire industry, but we lack a good mental model for thinking about it. That won’t work anymore now that finance finally matters.

Certain things that “just don’t make sense” (like many of the late-stage fundraises) are downstream of Silicon Valley’s “foie gras” problem: the set of perverse incentives, shared across many market participants, to stuff great companies with far more equity than they can effectively deploy and chase vanity markups.

Those incentives are well-covered, so there’s no need to beat a dead horse. Incentives aside, many of the things that don’t make a lot of fundamental sense are simply downstream of our collective inability to think about the cost of and returns on capital, particularly with respect to equity.

This is somewhat ironic given how the postwar technology ecosystem evolved: Silicon Valley largely owes its entire existence to equity, both capital itself and employee ownership that engenders a long-term oriented spirit of value creation. It’s not totally surprising, though, given (i) the capital-light nature of many internet, mobile, and cloud winners over the last few decades (ii) the “permissionless” nature of many of those markets, particularly the former two, and the young, financially unsophisticated founder archetypes they favored and (iii) a decade-plus of low rates and fiscal profligacy that pushed investors far-out on the risk curve and made fundraising often feel easy. Nobody ever thought too hard about finance.

That culture is finally running up against a new reality: finance now matters. Frontier AI is ultra-capital intensive, with labs raising at deca-billion-dollar scale, and their entanglement with hyperscalers, lenders, and the government raising questions about whether they’re TBTF. New aerospace, defense, energy, and industrials businesses in El Segundo require capital structures that look nothing like those of software companies. “AI rollups” are trying to run the 1980s buyout playbook on steroids. “Straightforward” application-layer AI companies can no longer take high gross margins and zero marginal costs for granted. And even “legacy” software businesses must grapple with a world where the returns to R&D have fallen and distribution is destiny, with winners crowned depending on how effectively they capitalize their GTM engines.

How carefully you think about the equity you raise and deploy has never been more important.

A better model for the cost of equity

Equity isn’t free

Our capital allocation challenges are mostly a function of not understanding the cost of equity. Ironically, though the industry harbors a reflexive distrust of debt, there’s little confusion about how it works: the cost of debt is a lender’s required return. Its explicit qualities make the cost of debt easy to reason about and feel real.

Conceptually, the cost of equity is quite similar: an equity investor’s required return. But it’s a theoretical, forward-looking concept dependent on an investor’s assessment of a business’ prospects, risk, volatility, opportunity cost, etc. And it’s not explicit in the way the cost of debt is. Nobody spells it out, it’s not written down anywhere, it doesn’t show up in the P&L, and money doesn’t flow out of the bank every month. These dynamics understandably confuse people.

You may hear that because equity investors have no contractual right to a given return, equity is “free.” This perspective is odd when shared by founders, who surely realize that investors expect something in return. It’s downright insane when shared by other investors – I have yet to find a VC who believes they’re in the business of making donations. DPI doesn’t appear out of thin air. Someone is paying for it.

Founders intuit this reality to some extent when they discuss the “heavy expectations” that come with raising venture dollars: the requirement for generational outcomes, the fundraising treadmill, the doors that close, etc. They don’t consider these pressures in the context of their cost of equity, but they are inextricably linked. When capital is expensive, only a small number of future outcomes is acceptable to investors.

Don’t confuse cost and dilution

The confusion around the cost of equity is often most apparent in the way we talk about big fundraising events; many people make the mistake of myopically focusing on dilution rather than the cost of equity or even confusing the two concepts outright. They’re indeed related, because the number of shares you sell is a function of what price investors will pay, which depends on the return those investors require: the higher your cost of equity, the lower your valuation, and the more dilution you’ll incur for every dollar raised. But they’re decidedly not the same, and the fact that an equity financing for a late-stage business has “low dilution” doesn’t necessarily make it attractive. If dilution were truly the only thing that mattered, companies should just raise infinite debt, which comes with 0% dilution.

This is plainly ridiculous, because other factors matter. What does the per-share price investors pay imply about the per-share value of the business you’ve built? What opportunities do you have to deploy capital, how much can you deploy against them, and what are those returns in light of your cost? What risks does a certain kind of capital present?

Most of this would be much clearer if people could reason about the cost of equity the way they do about the cost of debt. If interest payments represent an annual return on a lender’s investment and a real dollar cost, what’s the equivalent for equity?

Consider the dollar cost

Traditional models for the cost of equity – CAPM, growth-adjusted earnings yield, etc. – are helpful for many companies but not particularly instructive for quickly scaling, unprofitable private companies. Fortunately, many VCs will simply tell you: “we look only for 10x opportunities” or “every deal must have the potential to return the fund” or “we underwrite to a 4x minimum.” Expected returns vary by stage and risk profile (and portfolio construction, power law dynamics, expected loss ratios, etc.), but the implied cost of growth dollars is often ~30%.

It’s most helpful to concretize the cost of equity by reasoning about this math in real dollars rather than theoretical returns. Consider a firm (with a 30% annual return target) that leads a $200M growth investment in a great business that ultimately has a successful IPO in five years, earning the firm a ~4x on its investment (30% IRR). That’s a $543M gain in dollar terms: over a half-billion dollars that would have otherwise gone to founders, employees, and earlier investors. These dollars don’t just appear out of thin air.

Because no dollars explicitly changed hands or showed up in the P&L, people don’t think of them as a cost. Hypothetically, what if we did take them out of the P&L? Consider a “debt-equivalent” example: a 30% annual return on $200M would have burdened the company’s P&L with an additional $60M each year. And realistically, because dollars coming out of the business couldn’t be reinvested at the same rate (in the same way that equity compounds), the annual dollar amount required to deliver a 30% compounded return would have been even higher. In any event, it’s an extraordinary amount of money.

This example is loosely illustrative only. Equity and debt are obviously different, and there are major practical limitations to burdening the P&L with hard, backwards-looking dollar costs to approximate an inherently theoretical, forward-looking metric like the cost of equity. But the important idea holds: when you raise growth equity and build a successful business, you’re taking real money out of your (and earlier shareholders’) pockets and handing them to new investors. That dollar cost is real, even if you don’t foot the bill until your IPO, and it’s worth it only if you do something special with the equity.

Return on equity matters

Companies of course need to raise equity to succeed, particularly at certain points in their lifecycle. And when this equity is deployed appropriately, the value transfer between shareholders is positive-sum; the cost is worth it. That’s the critical point: raising equity is worth it only when the returns on that equity justify its cost. Deploying new equity at a rate of return below its cost destroys value. Consider the obvious debt analogy: no company would borrow at 10% to invest in a project that yield 6%.

Ultimately, ROE – and incremental ROE on new equity invested – matters because the financial returns a business generates will approximate its fundamental ROE in the limit; this is true even if investors pay a very high entry price over time and experience multiple compression (or vice versa). Investors pay those very high prices only when they believe a business can (i) create value for its customers (ii) create that value in a differentiated, durable way that allows it to maintain pricing power and (iii) thus grow revenue and profits at a high rate for a very long time. In other words, they pay high prices for businesses with high ROEs that exceed their cost of capital.

This reality feels abstract at times, especially in venture, where calculating fundamental ROE is difficult because companies make large investments today but may not generate GAAP profits until many years in the future (under the traditional ROE calculation construct); the earlier a company is, the wider the confidence interval on its costs and returns will be. Still, especially as they scale and become predictable, companies should anchor their decisions in rigorous capital allocation principles. Are investors appropriately valuing the business built to date? Does the implied cost of equity make sense considering the ROE? And does raising equity make sense given the incremental return the new capital can generate?

How to think about these late-stage financings

One argument in favor of the late-stage equity financings is that these companies do need equity because they require compounding risk capital to pursue lots of risky, high-ROE endeavors. As I noted, this argument is uncompelling on its face: most deploy the majority of these dollars into predictable, range-bounded GTM rather than high-upside product and M&A initiatives with unbounded potential. This is directly value-destructive at worst or inefficient at best. The irony here is that these businesses could capitalize their predictable investments differently, giving themselves leverage on a smaller equity base and significantly increasing their ROE.

Herein lies an interesting twist. While the best companies often have the lowest theoretical cost of equity ex-ante and tend to be the most hand-wavy about the concept, the biggest winners often pay the highest price ex-post. Every dollar of equity raised delivers enormous sums of money to the new investors, at the expense of everyone else (given the way this equity is deployed). The higher your growth and bigger your success, the higher a price you ultimately pay for the unnecessary equity you raised along the way. Bad capital allocation punishes the best companies the most.

Another argument in favor of these financings is that founders are “selling high” to investors with irrational expectations, investing at valuations that imply a low or even negative cost of equity. It’s difficult to square this argument with the one that “dilution is low;” when capital is ultra-cheap, the rational choice is to maximize rather than minimize dilution given the massive universe of opportunities with returns that exceed the cost of capital. And if founders really believed that investors have an ultra-low or negative expected return, they’d sell far more secondaries and/or simply sell their business entirely; they could surely re-deploy the proceeds into higher-return opportunities elsewhere.

Occasionally, investors themselves will argue that the cost of equity in these rounds is cheap. This is stupefying, because it’s the equivalent of investors arguing that they expect to deliver low returns. Institutional LPs don’t pay 2 and 20 over ten years to earn treasury returns. Pack it up.

Others have argued that these financings have non-economic operating benefits. Will Manidis noted that raising a small round, even if the cost of capital is too high, might be attractive because it conflicts new investors out of investing in competitors; the cost of capital may be uneconomical in isolation, but incurring this cost is worth it because these conflicts impair competitors. That may be true at the margin, and it was often true historically, when investors were highly sensitive to conflicts. As the amount of capital directed at private companies has exploded, however, firms have become less sensitive to these competitive dynamics. And as the collaboration-oriented social contract in Silicon Valley (which Will has written compellingly about) comes under pressure and cedes way to more mercenary, one-shot gamesmanship, I don’t expect firms to really care.

Similarly, some people argue that higher valuations help attract great talent. Again, that may be true for some people, but I’ve often found the opposite to be the case: while higher valuations are a noisy leading indicator of success, they also raise questions, especially among the most ambitious operators, about the remaining upside for new hires.

If you have questions, comments, or feedback, please reach out: andrewziperski [at] gmail [dot] com.

The views expressed herein are the author’s own and are not the views of General Catalyst Group Management, LLC or its affiliates.