Kyle Samani on Tokens, Network Effects and Crypto Investing

Kyle Samani says his previous startup, Pristine, built Google Glass software for surgeons. When Google discontinued Glass, the company pivoted and was later acquired for what he describes as a modest sum. Looking for a new direction, he discovered Ethereum in 2016 and became interested in the flexibility of smart contracts and permissionless finance.

After months of study, Samani and Tushar Jain agreed to start Multicoin Capital in 2017. Its liquid fund began that October, and its first venture fund followed in July 2018. By this interview in July 2022, Samani says the firm was on its third venture fund.

One painful lesson was that the firm could not excel simultaneously at trading, fundamental research and venture. After trading mistakes, Multicoin stopped trying to trade the market and focused on forming long-term investment theses. Samani says its intended holding period is indefinite, though it can still sell.

For Samani, high performance starts with an organization and culture that let the team execute. He also stresses cognitive performance and personal health habits, including sleep, diet and exercise.

Samani describes tokens as assets whose value can be considered in different ways. Whether a discounted cash flow model will be necessary or appropriate for each crypto asset remains an open question in his view. His firm focuses much of its attention on assets it believes may support such a model in the future.

When a startup offers both equity and tokens, Samani argues that value accruing to both can create conflicts, with one claim gaining at the other's expense. Multicoin generally prefers tokens. He notes that deals can be structured as SAFEs or equity rounds with token warrants, but says the firm often regards the token as the primary asset.

He argues that tokens make sense when a project needs to coordinate and incentivize public participation without requiring permission. His example is Helium, a wireless network that rewards participants for deploying hotspots. Multicoin had a position in Helium, giving Samani an investment interest in the example.

Samani estimates that Helium had roughly 800,000 to 900,000 hotspots at the time. He contrasts distributed installation with the costs of a centrally planned carrier network and argues that the model can remove much of that cost. The estimate and cost claims are his, not independently verified results.

A wireless network needs some minimum coverage before customers find it useful, Samani argues, though he does not know the threshold. In his description, token issuance rewards the earliest hotspot operators for taking greater risk before it is clear the network will reach critical mass. As more hotspots join, rewards per hotspot decline under the distribution mechanism.

The same incentive mechanism that helps a network grow may also make it easier to bootstrap a competitor, Samani says. Stebbings challenges him, pointing to existing hardware and token commitments as sources of lock-in. Samani replies that the prospect of large returns could motivate participants to try a new network; those returns are speculative and not guaranteed.

Asked whether tokens concentrate or distribute wealth, Samani argues they can broaden participation beyond employees and investors to customers and suppliers. This is his prediction about the mechanism, not evidence that every token distribution is equitable. He also believes existing capital-market norms make such participation harder.

Token issuance also creates operational risks. Samani identifies potential employee trading on inside information and the difficulty of engaging thousands of token holders as challenges for a small startup. He argues that a community with capital at stake may nevertheless contribute effort as well as money.

Liquidity does not change Multicoin's initial investment decision, Samani says: the firm invests only when its thesis supports an intended long-term hold, while retaining the ability to sell if circumstances change. He values quicker feedback from liquid markets, although market prices and returns can move sharply.

Asked how he handles volatile prices, Samani jokes that prices do not matter if he does not look at them. That is his personal approach to investor psychology, not a reduction in the financial risk of holding the assets.

Samani says Multicoin begins investment committee discussions with written memos and concurrent comments identifying assumptions, gaps and logical errors. The group then debates details at length. He believes active disagreement helps the team challenge its biases and improve its understanding.

Stebbings asks whether the process must feel so fierce and whether junior colleagues might be discouraged from speaking. Samani says the criticism is directed at statements rather than people, but acknowledges the process requires fortitude. Applicants begin by writing an essay, which he says he reads before looking at their names or backgrounds.

On whether crypto investing is collaborative, Samani says he thinks it had become less so as fund sizes increased. He does not claim certainty about whether decentralized ownership can make outcomes sufficiently larger to offset smaller ownership stakes.

Multicoin has few formal portfolio-construction rules, according to Samani, and does not set aside conventional reserves for follow-on investments. He opposes fixed sector allocations because they might have excluded unexpected opportunities such as Helium. He says the firm invests across stages and cited tickets from $1 million to $300 million; this is the firm's strategy, not a general rule.

Stebbings emphasizes founders in early-stage decisions. Samani says Multicoin spends most of its discussion on markets, while checking that a founder's background fits the task. He believes careful analysis of certain markets can support concentrated positions, although he acknowledges that conviction is difficult in many markets.

The team asks where returns to scale come from if a product begins to work: whether each additional unit of revenue or usage gets easier to achieve. It also considers how competitors might respond and what initial wedge could bring the product to a self-reinforcing point.

Samani distinguishes the shape of a network effect as users are added: sublinear, linear or superlinear. He believes most are sublinear, but says the curve's shape and its changes over time inform his conviction and position size. He also considers inclusionary versus exclusionary effects a separate distinction.

He says Multicoin generally avoids SaaS tools because he believes relatively few have meaningful network effects, though some tools can develop valuable networks across customers. This is an investment preference, not a claim that SaaS businesses cannot succeed.

Samani admits to persistent fear of missing out. He says he devotes three to four hours a day to reading, including newsletters and social media, and fits it around exercise, meals and other commitments. He believes consuming a broad range of information can improve judgment even though the process is inefficient.

Stebbings asks whether crypto bull markets result from token distribution innovations or leverage. Samani argues that recent cycles were driven by new forms of capital formation, while acknowledging explicit and hidden leverage, including derivatives and lending. His description of the then-recent Three Arrows Capital failure and its scale was an estimate made in 2022.

On whether volatility might deter institutional investors, Samani says he does not know. Multicoin had raised funds before the market turned, and he says the firm had not tested the market for a new fund. He wants the firm to scale its research process but doubts an investment team larger than ten can maintain the same depth of discussion.

He says Multicoin does not target a particular deployment pace: it invests when it loves an opportunity. Its quarterly decision reviews revisit earlier investments to ask whether missed risks were knowable and whether positions were sized appropriately. He estimates, from memory, that roughly half of meaningful problems in a mature subset of the portfolio were knowable in advance.

Samani claims crypto loss rates can be lower because token liquidity may allow a firm to exit at a gain even after its investment thesis proves wrong. He says selling can signal negatively and describes discussing progress with founders and executing sales over extended periods. Liquidity, price and exit opportunities are not assured, and his statement is an observation about his firm's experience and opportunities at that time.

Asked whether Bitcoin hedges inflation, Samani says no: he prefers productive assets and suggests some businesses may be inflation resistant. This is his opinion in 2022, not a settled conclusion or a guarantee about Bitcoin or the businesses he mentions.

Samani favors Solana over Ethereum and argues that Ethereum lacked a convincing scaling plan at the time. He predicts it will not have one in the next 12 months and emphasizes that developers need a credible, specific roadmap. Multicoin had a position in Solana, so his comparison carries a direct investment interest; the claim and forecast are his own.

Looking back, he wishes he had been more open to ideas that might be interesting in the short term even if they later seemed foolish. He also calls for clearer, relatively permissive rules for crypto capital formation and says that achieving them would require congressional action or changes to SEC rules.

Samani identifies Delphia as a recent publicly announced Multicoin investment. He describes its proposed model: participants export and contribute data, which the company hopes to aggregate to inform trades in US equities. He presents superior forecasting and returns as a theory and hope, not an established result. The investment gives Multicoin a financial interest in that assessment.