Why Reserve Ratio Matters More Than Entry Price in Applied Deep Tech

Andreessen Horowitz put roughly 17 percent of a fund into Skype. Sequoia Capital put 15 percent into WhatsApp. Thrive Capital put 20 percent into Stripe. GV put 22 percent into Uber. Founders Fund committed close to 30 percent of a fund to Anduril alone. None of those numbers describe an entry check. They describe what a fund was willing to do after the first check, once conviction had actually been earned, and in every case that concentration decision did more to build the eventual return than the terms of the original investment ever could.

Founders spend real negotiating energy on entry price. It is the number on the term sheet, the one a lawyer will redline. What almost never gets the same scrutiny is a far better predictor of outcome: whether the investor across the table is actually capable of, and willing to make, a concentrated bet like the ones above when the moment calls for it.

A reserve percentage on paper and a real concentrated bet are different things

In each of the cases above, reserves mattered not because they existed on a spreadsheet, but because they were decisive when it counted, concentrated into a single company rather than spread evenly across a portfolio to avoid picking favorites. Most institutional funds report holding somewhere between 40 and 60 percent of total capital in reserve, and nearly every LP asks about that figure before committing. The number is a weak signal on its own. A fund that reserves 50 percent but spreads it across two dozen companies to avoid confronting which ones are actually working will never produce a Skype or a Stripe sized outcome, no matter how large the reserve line reads on a fund model.

A well known early stage investor just reversed his own advice on this

Hunter Walk, the co-founder of Homebrew, publicly changed his position on reserves this year after more than a decade running an early stage fund. His argument centers on two problems with the conventional 20 to 50 percent reserve model: follow-on decisions made weeks or months after an early round rest on comparatively little real information, and crowded cap tables frequently pressure smaller funds away from taking their full pro rata even when they want it. Instead, Homebrew got to over 120 percent invested in each of its first two funds by recycling capital from early exits rather than holding a large passive reserve pool for the fund's entire life, a structural choice that expanded real capital availability without asking LPs for more money.

The failure mode nobody budgets for: the reserve runs out

There is a documented, recurring phenomenon in venture called reserves exhaustion, where a specific fund, not the whole firm, simply runs out of the capital set aside for follow-on rounds, typically because more portfolio companies needed bridge financing than the original model assumed. When that happens, a founder who assumed their existing investor's large reserve number meant automatic support in the next round discovers instead that a clear yes has quietly become a maybe, right as competition for whatever reserve capital remains intensifies across the fund's entire portfolio at once.

What this means for a founder comparing term sheets

A founder weighing two similar offers should weigh an investor's demonstrated willingness and capacity to concentrate real capital into a genuine winner as heavily as the entry valuation, because the valuation only describes today. Whether that investor has ever put a meaningful share of a fund behind one company, and whether their reserve capital is still intact rather than already stretched thin across bridge rounds elsewhere in the portfolio, describes what happens at the round that will actually matter.

Concentration is earned through visibility, not promised on a spreadsheet

Our approach treats reserve capital the way the funds behind Skype, WhatsApp, and Anduril treated theirs, as a decision to be earned by a company's trajectory rather than allocated evenly to avoid confronting which portfolio companies are actually working. Working Excellence's digital engineering strategy and BetterWorld Technology's cybersecurity services give us the operating visibility to make that call with real information rather than a board update, which is precisely the informational gap Hunter Walk pointed to as the weakness in reserve strategy generally. Our investment thesis treats concentrated, conviction driven follow-on capital as core to how we support applied deep tech companies, because the check that matters most is rarely the first one.