Linda Xie: Six of 50 Angel Bets Above Cost, All Crypto

Six of the 50 personal angel investments Linda Xie has disclosed returned more than their original cost, and every one of them backed a crypto company, according to her Nov. 8 post reviewing a portfolio built over nearly a decade.

Xie, a co-founder of Scalar Capital and a former product lead at Coinbase, published the review of 50 investments she made personally as an angel investor. Spread across almost ten years, that averages roughly five investments a year.

The portfolio's shape explains part of the result. Crypto accounted for 60% of the companies she backed, AI for 10%, with the remainder distributed across other industries. On those percentages, crypto covers roughly 30 of the 50 positions and AI about five. The concentration in crypto therefore does more than explain why crypto won; it means crypto was the only category in the sample where any position cleared its cost, because all six winners sit in that sector.

How did the six winners clear their cost?

• Five came through token subscription rights, rights to buy tokens rather than shares.
• One was an equity investment.
• No single investment returned as much as the total amount invested across the portfolio.

That last line is the review's most important qualifier. The aggregate return from the best-performing position stayed below the aggregate capital deployed across all 50 investments, so the portfolio has not produced one outsized payoff offsetting the write-offs elsewhere.

What does the token-rights structure change for an angel investor?

The practical question for anyone writing a private check today is whether the instrument still matters more than the sector. Xie's own numbers suggest it does: within the crypto cohort, the five above-cost outcomes arrived through rights to buy tokens rather than through share ownership, while one equity position also cleared cost.

The distinction is mechanical. A subscription right can be marked against a token price without any sale of the underlying business, and it carries no voting or dividend claim on that business. An equity position has no such reference price, and it converts to cash only when a buyer appears. For an investor, paper gains in the token-rights bucket can be realized by selling a listed asset, while an equity gain requires a liquidity event.

The cost of that route is concentration. Returns depend on a token market existing at a price the issuer can support, rather than on an acquirer or a public listing. Xie's data does not settle whether that tradeoff is favorable in general, because the review reports outcomes across a mixed 50-deal sample rather than a controlled comparison.

What do the 12 exits show?

Twelve of the companies she backed shut down or exited through acqui-hires, and those exits provide the review's only realized return data:

• Seven returned nothing.
• Four recovered 80% to 90% of their initial cost.
• One outcome remained pending.

Scaling that realized band to a household-sized check: for every $100 committed, $80 to $90 came back, a loss of $10 to $20 on deals that found a buyer. Against that, seven of twelve, more than half of the realized outcomes, returned nothing at all.

The other 38 positions in the portfolio have no realized outcome, which places a ceiling on how much can be read from the six winners.

How far do the numbers reach?

Xie's accounting covers her personal equity investments. It excludes her funds, including Scalar Capital, and it excludes her direct holdings of Bitcoin and Ether. The crypto winners therefore reflect equity or token-rights exposure to crypto companies, not a bet on the tokens themselves, and the review does not measure her overall portfolio.

The portfolio's highest paper return rate belongs to an AI company, well above every other position. If monetized, Xie wrote, its value could theoretically cover the full amount invested across the entire portfolio, with a single unrealized mark standing in for the whole book.

For now the read-through is structural rather than directional. One pending acqui-hire remains to resolve, and the next cohort of private rounds, priced through token subscription rights rather than equity, will test whether a pattern drawn from 50 deals over a decade holds up.