Changpeng Zhao explains his 2025 focus on BNB Chain, stablecoin 2.0, prediction markets and AI agents in a post-pardon, builder-first reset.
Summary
- Changpeng Zhao shifts from exchange operator to mentor investor, focusing on Giggle Academy, YZi Labs and BNB Chain builders.
- He advocates a “stablecoin 2.0” thesis: native, highly liquid, yield-bearing assets on chains like BNB, going beyond USDT’s 1.0 model.
- YZi Labs supports multiple prediction markets and RWA comes into play as BNB Chain scales throughput, users, and native liquidity.
Changpeng Zhao’s 2025 could have been a silent fadeout. Instead, Binance’s pardoned founder spends his time scaling an education app, building early-stage projects through YZi Labs, and quietly lobbying governments from coffee shops in Dubai. “Whether you call it freedom, closure or finally ‘turning the page,’ at its core it’s about the feeling of being able to move forward unencumbered again,” he said, emphasizing that his daily routine remains anchored at home – and at the gym – rather than on the trading floor.
Changpeng Zhao doubles down on Binance
The four pillars of that new routine are now clearly: Giggle Academy, YZi Labs, the BNB Chain ecosystem and direct consulting work with policymakers from Pakistan to the UAE. Giggle Academy, a free education platform with “more than 90,000 kids already using it,” has grown into what Zhao calls a long-term commitment, supported by a team of about 60 people that iterates on the product weekly.
YZi Labs, meanwhile, serves as its primary investment and mentoring vehicle. “I’m more of a mentor and coach: I work with founders and developers and help them grow,” Zhao said, noting that the company has evaluated more than 1,000 projects and completed nearly 70 investments by 2025, many of which build directly on BNB Chain and contribute to the EASY Residency program. That builder-first approach now includes a dedicated $1 billion Builder Fund focused on DeFi, AI, real-world assets and biotech on BNB Chain – a pool arriving as the network heads toward “CEX-like” confirmation times and record user numbers.
On-chain, BNB (BNB) Chain has quietly shifted from “undervalued and overlooked” to one of the busiest settlement tiers in the industry. The number of daily active addresses hovers around 2 million, with previously published figures around 2.4 million, while transaction volume on the chain grew by approximately 600% year-on-year, placing BNB Chain among the top networks in terms of throughput. At the same time, BNB’s spot and derivatives markets have turned into a magnet for volatility traders: after an October crash wiped out nearly $19 billion of leveraged positions in the market, BNB fell only about 10% before recovering, trading in a volatile range from $1,100 to $1,340 and hitting new all-time highs above $1,330. Order books on major exchanges now show fat resting bids just below $1,100, and a persistent sell wall in the $1,330 to $1,370 range – a structure that tells you exactly where the next liquidation cascade or tightness will begin. If BNB loses that $1,100 shelf, this rally will die quickly; on the other hand, a clear break above $1,370 will almost certainly trigger new momentum accounts targeting $1,450-$1,600.
Yet Zhao spends more time talking about stablecoins than price action. He is blunt: “What we see today is still largely ‘stablecoin 1.0,’ and real 2.0 is only just beginning.” The traditional model (depositing dollars in a bank, issuing tokens like Tether’s USDT on a chain) remains dominant, with network effects allowing USDT to lead the way even though it offers little return. “USDT is not particularly competitive when it comes to yield, and that creates an opening for other stablecoins,” he said, pointing to newer designs like Ethena, where “yield mechanisms are built into the design” and where YZi Labs has taken a stake.
BNB and YZi Labs, where are they going?
On BNB Chain specifically, he views the stablecoin competition as an “open garden” rather than a horse race. USDT circulates on BNB Chain as a packaged asset, but Zhao argues that the real story is the arrival of true native options: USD1 as a US-backed collateral model, previous experiments like FUSD that stalled due to clumsy issuance, and newer projects like the $U project, which he says “also has some potential.” According to him, the end game is simple but hard to implement: a stablecoin that is easy to trade, widely listed and yet offers a sustainable return. Ignore any design that can’t hit all three.
Prediction markets are the other story he doesn’t want to dismiss as a fad. Zhao credits Kalshi and Polymarket with dragging this category into the mainstream during the last US election cycle, where Polymarket saw betting volume of more than $3.6 billion and effectively led traditional polls with sharper odds. “In many cases, the results were even more accurate than traditional polls because participants were putting real money on the line,” he said. YZi Labs has backed multiple early-stage forecasting platforms, including BNB Chain-based projects such as Probable and the recently launched Opinion, where Zhao confirmed the company is only a minority investor.
Crucially, he emphasizes that this is not a winner-takes-all card. “This is not a winner-takes-all race.” In any market, multiple players usually coexist,” Zhao said, drawing a line from exchanges to stablecoins to prediction markets. That logic also underlies his skepticism toward subscription-based AI trading agents; the more a profitable strategy is sold, the faster it breaks down. Platform companies – exchanges, prediction markets and even RWA tokenization rails – can scale based on fees and spreads, but shared alpha is, in his words, a self-extinguishing product.
If there’s one through line in Zhao’s 2025 playbook, it’s patience. “Success takes time,” he said, comparing BNB Chain’s trajectory to Nvidia’s decades-long climb and describing the build as a “marathon mixed with a boxing match.” They are mission-driven founders, not tourists chasing the next memecoin meta, the ones he wants to fund – and the ones he expects to still be around in 2027 when today’s stories have been revised, recycled or revealed as pure exit liquidity.
