LP/GP structure, management fees, carried interest, and how the business of venture capital actually works - explained from the ground up.
A venture capital fund is a pool of money managed by General Partners (GPs) and funded by Limited Partners (LPs). GPs are the investors you see on Twitter - they source deals, lead rounds, and sit on boards. LPs are the people behind the money: pension funds, endowments, family offices, fund-of-funds, and high-net-worth individuals. LPs commit capital to the fund, and GPs decide where it goes. The legal structure is almost always a limited partnership - LPs have limited liability (they can only lose what they put in), while GPs have full management responsibility.
Fund size determines everything: check sizes, target ownership, portfolio construction, and the kind of returns a fund needs to generate. A $30M seed fund writes $500K–$1M checks and needs a few big winners. A $500M growth fund writes $15M–$30M checks and needs massive outcomes to move the needle. Fund size is a strategy decision - it dictates which companies a VC can invest in, how many bets they make, and what a "good" return looks like. A 10x return on a $50M fund is $500M. A 10x on a $1B fund is $10B - a very different bar.
GPs charge an annual management fee - typically 2% of committed capital during the investment period (usually the first 3–5 years), then 2% of invested capital for the remainder of the fund's life. On a $100M fund, that's $2M per year to cover salaries, office space, travel, legal costs, and operations. For smaller funds, the economics are tighter - a $30M fund generates $600K/year in fees, which doesn't go far once you account for a small team and overhead. This is why many emerging managers run lean.
Carry is how GPs make real money. It's the share of profits the GP keeps after returning all capital to LPs. The industry standard is 20% - meaning if a fund returns $300M on $100M invested, the $200M in profit is split 80/20: $160M to LPs, $40M to the GP team. But carry only kicks in after LPs get their money back (and sometimes a preferred return, called a "hurdle rate," typically 8%). This structure aligns incentives: GPs only do well when LPs do well.
A typical VC fund has a 10-year life, split into two phases. The investment period (years 1–4) is when the fund actively deploys capital into new companies. The harvest period (years 5–10) is when the fund manages its portfolio, supports follow-on rounds, and waits for exits (acquisitions or IPOs). In practice, many funds extend beyond 10 years - startups don't operate on clean timelines. GPs usually start raising their next fund around year 3–4 of the current one, using early results and portfolio momentum to attract LPs for the next cycle.
Portfolio construction is how a fund allocates its capital across investments. Key decisions include: how many companies to invest in, how much to reserve for follow-on rounds, and what ownership target to aim for. A typical seed fund might invest in 25–40 companies, reserving 30–50% of the fund for follow-ons into winners. The math is driven by power law - a small number of investments will generate the vast majority of returns. Everything else is portfolio construction designed to make sure you're in those winners.
VC returns follow a power law: a single company often returns more than the rest of the portfolio combined. The industry benchmarks are 3x net (returning $3 for every $1 invested) for a "good" fund, and 5x+ for a top-quartile fund. But averages are misleading - the median VC fund barely returns capital. Returns are measured as TVPI (Total Value to Paid-In capital, including unrealized gains), DPI (Distributions to Paid-In capital - actual cash returned), and IRR (Internal Rate of Return - time-weighted annualized return). LPs care most about DPI, because paper gains don't pay pensions.
VC funds raise capital from LPs in a process that mirrors startup fundraising - but on a longer timeline. GPs pitch their track record, strategy, and team to LPs over 6–18 months. First-time fund managers ("emerging managers") have the hardest time - they're raising on a thesis and reputation without a proven portfolio. Established firms raise on track record and existing LP relationships. Once a fund hits its target (or a "first close" threshold), it begins deploying. Most firms operate multiple funds in sequence - Fund I, Fund II, Fund III - each building on the last.