VC portfolio power-law simulator

• Money invested: fund size × share invested after fees. It is split equally across the companies, including any follow-on money, so £50m × 80% ÷ 30 companies = about £1.33m each. The other 20% pays management fees.
• Outcomes: each company is randomly given one outcome, with chances set by your outcome mix. If the mix does not add up to 100%, we scale it so it does. To make results look more like real life, each multiple is nudged by a random factor between 0.7 and 1.4 (so a “20x” company returns somewhere between 14x and 28x). Companies that lose everything stay at 0x.
• Proceeds = the sum of what every company returns (money invested in it × its multiple). Gross multiple = proceeds ÷ money invested.
• Carry = carry % × (proceeds − fund size), if that is positive. The fund’s investors get their full fund back, fees included, before the managers share in profit. There is no hurdle (preferred return), which many funds do have; a hurdle would only change things for funds just above 1x.
• Net multiple = (proceeds − carry) ÷ fund size. This is what the fund’s investors actually get back for each £1 they put in.
• Lost money means a company returned less than was invested in it (below 1x).
• 2,000 runs: the same settings are run 2,000 times with fresh random draws. We count how often the fund returns at least 1x and 3x net, take the middle (median) result, and count how often the best company alone returned more than all the others put together.

Assumptions and limits: the outcome mix is an illustrative rule of thumb, not data from real funds. The model ignores timing (a 3x fund over 12 years is far less impressive than over 6), follow-on strategy (real funds put more money into their winners), recycling of early proceeds, dilution, and the way outcomes cluster in good and bad years. Treat it as a way to see the shape of venture returns, not to forecast a real fund.