The Venture Capital Stack, Explained

Ask a new VC associate what the "capital stack" is and you will usually get the startup answer: SAFEs at the bottom of the ledger, preferred stock on top of common, an option pool somewhere in the middle. That is half the picture. A venture firm sits between two capital structures — the one that funds the fund, and the one the fund invests into. Every dollar a GP deploys travels through both.

Understanding the two-layer view is the single most useful mental model for anyone building a firm, because most of the strategic decisions a GP makes — fund size, ownership targets, reserves, when to sell — are really decisions about how these two stacks interact.

Layer 1 — Capital into the fund

Before a GP can wire a dollar to a founder, that dollar has to be raised into a blind-pool fund. This layer has its own internal ordering:

• LP commitments — the equity of the fund itself. Endowments, foundations, pensions, funds-of-funds, family offices, and individuals commit capital that is drawn down over the investment period, usually three to four years, via capital calls.
• GP commitment — the manager's own money in the fund, conventionally 1–2% of the total (often more for emerging managers who want to signal conviction). It is the alignment mechanism LPs look at first.
• Fund-level facilities — subscription lines that bridge capital calls, and (rarely, later in a fund's life) NAV-based borrowing. These sit above the LP equity at the fund level and are covered in depth in our fund finance guide.

The economics of this layer are set by the limited partnership agreement (LPA): a management fee (the classic 2%, often stepping down after the investment period), carried interest (typically 20% of profits), and the distribution waterfall that determines when the GP actually gets paid. Unlike buyout funds, most venture funds have no preferred return hurdle — a reflection of the power-law return profile LPs are underwriting.

One structural feature deserves emphasis: LP capital is contractually patient but reputationally impatient. The fund term is ten years plus extensions, and LPs cannot redeem. But a GP who wants to raise Fund II and Fund III is effectively re-underwritten every two to three years, and interim marks — TVPI, and increasingly DPI — are the currency of that re-underwriting.

Layer 2 — The cap table the fund invests into

Once the fund holds committed capital, the GP deploys it into a second structure: the startup's cap table. This is where venture differs most sharply from other private asset classes. There is no senior debt tranche to negotiate, no EBITDA to lever. The instruments are few and standardized:

SAFEs and convertible notes

At pre-seed and seed, most first checks are written on a SAFE (Simple Agreement for Future Equity) or, less commonly now, a convertible note. These are not equity yet — they are contracts that convert into preferred stock at the next priced round, with a valuation cap and sometimes a discount setting the conversion price. The post-money SAFE, now the market standard, fixes the investor's ownership at conversion, which pushes dilution onto founders and earlier holders. A GP who does not model SAFE stacking carefully will systematically overestimate ownership.

Preferred stock

Priced rounds — Series Seed, A, B, and onward — issue preferred stock. Venture preferred carries a bundle of rights that common does not:

• Liquidation preference — almost always 1x non-participating in healthy markets: in a sale, the investor takes the greater of their money back or their as-converted common value. Participating preferred and multiples above 1x appear in distressed or overheated moments and are a red flag either way.
• Anti-dilution protection — typically broad-based weighted average, adjusting the conversion price if the company later raises a down round.
• Protective provisions — veto rights over sale, new senior stock, debt above a threshold, and other major actions.
• Pro rata rights — the right to maintain ownership in future rounds, which is where reserves get spent.

Each new series usually sits pari passu or senior to earlier series in the preference stack. In a strong exit, none of this matters — everyone converts to common. In a mediocre one, the preference stack decides who gets paid, and seed investors discover exactly how junior they are.

Option pool and common

Below the preferred sits the employee option pool (typically 10–20% of the cap table, and note who funds the expansion — the pre-money "option pool shuffle" is a real cost to whoever isn't paying attention) and founder common stock, the residual claim.

The same dollar, two frames

Here is the key insight the two-layer model produces: fund capital is the most senior money LPs have committed, and among the most junior claims on any given cap table. Your Series A preferred beats common in a downside — but there is usually no debt above you and no collateral beneath you, and in a zero the whole position is a zero. Venture equity has no coupon, no amortization, no covenant. Its only protection is price, ownership, and the preference.

This is why venture underwriting looks nothing like credit underwriting. There is no fixed return to solve for. A GP underwrites each investment to the question: can this single position return a meaningful fraction of the fund? The math is unforgiving — at 10% ownership, returning a $50M fund requires a $500M outcome before dilution, and dilution is certain. That is the power law doing its work, and it is why ownership targets, entry price, and follow-on reserves are the three levers of portfolio construction.

How GPs think about the mix

In buyouts, structuring a deal means choosing debt tranches. In venture, "structuring" happens at the portfolio level, in the fund's own construction:

• Fund size sets strategy. A $20M fund writing $500K checks and a $500M fund leading Series Bs are different businesses with different LP bases, ownership math, and exit requirements.
• Ownership targets determine how concentrated the fund's outcome distribution will be. Higher ownership means fewer positions and more idiosyncratic risk — and more upside per winner.
• Reserves (commonly 30–50% of the fund for follow-ons) are the venture equivalent of dry powder: capital held back to exercise pro rata in the companies that are working. Reserve discipline — following on into signal, not into loyalty — separates good funds from average ones with identical initial portfolios.
• Recycling — reinvesting early exit proceeds during the investment period — lets a fund put more than 100% of committed capital to work, offsetting fee drag.

The takeaway

The venture capital stack is two structures bolted together. Layer 1 — LP commitments, the GP commit, the LPA — determines your economics, your timeline, and who you answer to. Layer 2 — SAFEs, preferred, the pool, common — determines what you actually own and what it is worth in each exit scenario. GPs who reason across both layers at once make better decisions at every step: how big a fund to raise, what ownership to demand, when to follow on, and when to sell. Master the two-layer view and the rest of fund strategy becomes arithmetic.