OKX Founder Star Xu: No Fund Outperforms Bitcoin, Bought First Bitcoin at About $10 in 2011 and Mined on Company Servers
Star Xu, founder and CEO of OKX, stated in a long interview aimed at Europe that in over a decade in the industry, he has not seen any fund outperform Bitcoin; the core position should be a foundational holding tied to the industry, which he suggests is Bitcoin, emphasizing to stay at the table.
He mentioned buying his first Bitcoin for about $10 in 2011 and mining on the company servers; from 2012 to 2013, he could place bets of 500 Bitcoins on the dice betting site Just Dice. After discovering the random algorithm was too simplistic, he wrote his own program, entering when the house was losing and exiting when it was winning, claiming to have gained about 10,000 Bitcoins from this. He compared the site's gameplay to today's Uniswap-style liquidity pools, which later closed due to compliance issues. He once called Ethereum a garbage coin to Vitalik's face, claiming it was the biggest misjudgment.
For industry practitioners, he reiterated his old stance: if you don’t have 10 Bitcoins, don’t come to the meeting; crypto practitioners should feel ashamed if they don’t have 10 Bitcoins. The portfolio should consist of a foundational holding tied to the industry and the remaining holdings; buying and selling may seem fifty-fifty, but the challenge lies in direction; principles are most easily changed when the chips on the table grow larger. The fifth year earns more than the previous four years, and the tenth year far exceeds the previous nine years—provided the direction is not wrong.
Regarding artificial intelligence, he stated that true intelligence has arrived, surpassing top experts, with over 90% of coding tasks already completed by models; however, models cannot automatically make money, as principles, risk limits, and algorithms still need to be defined by humans, with models merely executing more accurately. The world is not designed for models and needs upgrading to adapt; stablecoins are described as tokenized government bonds, indicating the world is upgrading itself to adapt to blockchain. He claimed to have possibly coined the terms "isolated margin" and "full margin" himself.
On entrepreneurship, he stated that luck is the most important factor for success, with 90% of entrepreneurs likely to fail; resilience is more important than cash, as one might encounter opportunities even after three months without money; if the direction is wrong, no amount of effort will help. If he could go back fifteen years, he would prefer to be a vice president or technical director rather than a boss, as the boss's equity is hard to sell, and wealth is mostly on paper, while employees can live off options. He mentioned he still works about 12 hours a day and eats the same boxed meals as his employees. Trading is a professional job, and one should not bet their entire fortune on it.
In terms of market mechanisms, this is the exchange founder rewriting Bitcoin from a trading pair to an industry entry certificate: what is bought is the consensus of "hold first, then discuss strategy," while what is sold is excess fund performance, altcoin rotation, and all-in bets. Beneficiaries are spot and long-term custodians; those under pressure are practitioners who rely on stories for fundraising but have no Bitcoin in their own warehouses. The Just Dice story rewrites early on-chain gambling into the origin of quant trading, also writing 10,000 Bitcoins simultaneously as results of mining, betting, and market making, which cannot be verified transaction by transaction on a public ledger.
Additionally, the European interview also discussed MiCA and product implementation; in recent years, he has publicly emphasized "always holding Bitcoin," has not seen any fund outperforming it, and that stablecoins will become financial infrastructure.
Source: Public Information
ABAB AI Insight
XU overlays personal and exchange balance sheets. The $10 purchase, 10,000 on Just Dice, and the 10-Bitcoin threshold represent the same set of professional ethics of "first becoming a holder, then becoming an intermediary." His rejection of Ethereum serves as tuition to prove that judgment can be wrong, but position structure cannot. Working 12 hours a day and eating boxed meals is a way to detach the boss identity from the narrative of cashing out, returning to the operations counter.
Capital flows within the OKX system towards spot, contracts, Web3 wallets, and stablecoin channels. He describes stablecoins as tokenized government bonds, connecting compliant fiat rails to on-chain settlement, rather than telling another public chain myth. With 90% of code handed to models, engineers shift from writing interfaces to defining principles; models cannot decide liquidation points for individuals. He claims the terms isolated margin and full margin, indicating that once product language becomes industry default, founders gain naming rights that create path dependency.
In contrast to Satoshi's absence, early miners treating coins as game tokens, and the FTX founder treating customer assets as risk budgets, Xu chooses to "first become a whale, then open a casino." The industry phase is a window where regulators acknowledge stablecoins, and funds still do not outperform spot: exchanges are no longer selling the next hundredfold token but ensuring people are not cleared from the table by the next hundredfold token.
Structure belongs to the return of pricing power from trading skills back to asset binding. The fee business requires people to stay; the cheapest way to keep them is to make practitioners their own Bitcoin bulls. The mechanism is: if funds want to collect management fees, they must prove they can beat the benchmark; when the benchmark is Bitcoin and has not been beaten for a long time, the rational approach is to hold the benchmark first, then use the remaining risk budget for trading. The money printer won't lend out because those who can truly print are already at the table.
ABAB News · Cognitive Laws
First become a holder, then qualify to discuss strategy
Principles are first changed when chips grow larger
Models can write code but cannot decide your liquidation points