IOSG partner Jocy Lin used a keynote at Hong Kong Money Frontier 2026 to argue that crypto bear markets tend to produce the industry’s best investment opportunities, not because conditions look comfortable, but because pricing and fundamentals drift far apart.
Speaking on a theme she framed as contrarian investing and “frontier betting,” Lin said this is the fourth bear market she has lived through. She added that IOSG is still investing and does not want to become “the last gatekeeper among Asian crypto VCs.” Instead, she said, the firm wants more investors to enter Web3 and help more Asian founders build.
Bitcoin: the four-year cycle is still overpowering broader narratives
Lin began with Bitcoin. She said the recovery from the 2022 low was driven by several catalysts specific to Bitcoin: the collapse of Silicon Valley Bank in 2022, spot ETF speculation that pushed the price to $60,000, and Donald Trump’s election, which she said lifted it to about $120,000. By the end of 2025, most of those catalysts had been fully priced in.
After that, she said, the setup changed. The AI cycle started after 2025, and AI-linked assets, gold and Nvidia all strengthened. Under that backdrop, Bitcoin should have benefited, but it stopped responding and was down 29% year to date.
Lin cited a view from Bens Community figure “Bite Shenli,” saying Bitcoin’s four-year cycle had once again overpowered more complicated narratives. In her telling, that framework had pointed to a top in October 2025, and the market then entered a bear phase from that same month.
She said many people dismiss the cycle as coincidence, but argued that it can become self-reinforcing. If enough crypto veterans expect the market to peak at a certain stage, they tend to reduce exposure at roughly the same time. That collective behavior can then help create the cycle they expected. In her view, that also explains why Bitcoin turned lower even when the macro backdrop looked supportive.
For the next few months, Lin said the more important question is not why Bitcoin has lagged, but whether the next accumulation phase has already started.
She also shared IOSG’s internal cycle research. Looking at past market phases, the firm found that each upcycle since 2015 lasted roughly 1,060 days. Based on that work, IOSG outlined several scenarios:
- Scenario one: Bitcoin bottoms in a $45,000 to $60,000 range, possibly by late October 2026.
- Scenario two: Bitcoin falls to $40,000 to $55,000 in Q1 to Q2 of 2027.
- Scenario three: other versions of the bottoming path.
Lin said there are many ways to model the cycle, but IOSG’s internal research points more closely to the earlier version she described.
Stablecoins: Lin compares 2026 to 1975 and 2001
Lin then turned to stablecoins and called this year a major one for the sector.
She used two historical comparisons. The first was May 1, 1975, when the U.S. Securities and Exchange Commission abolished Wall Street’s fixed commission system after 183 years. In her account, that change turned stock trading from an activity associated with the wealthy into one open to ordinary participants and eventually helped create a $140 trillion asset-management industry, with firms such as Fidelity and Vanguard emerging afterward.
The second comparison was China’s entry into the World Trade Organization in 2001, which she described as one of the most important events in the country’s economic development.
Lin said she sees a similar turning point now in stablecoins. She pointed to the GENIUS Act becoming law and said that means banks can now legally issue dollars onchain. In her view, the passage of the bill marks a shift for crypto from a high-risk asset category toward lawful financial infrastructure. Looking back, she said, 2026 may be remembered the way 2001 is remembered.
Ethereum: strengths remain, but leadership is the harder question
On Ethereum, Lin split the discussion into strengths and weaknesses.
She listed four strengths. Ethereum remains the main settlement layer for the largest stablecoins and real-world assets, or RWA. Most of DeFi’s security and yield remain on Ethereum. Layer 2 scaling still settles finality back to Ethereum across the broader rollup stack. And Ethereum retains deeper developer infrastructure than any other Layer 1, she said.
She also described the challenges plainly. Layer 2 networks have captured fees that would otherwise have gone to the Layer 1. ETH/BTC has clearly underperformed in this cycle. Ethereum is also facing competition from newer chains such as Hyperliquid as well as offchain alternatives.
Even so, Lin said those are not the deepest issue. She has written extensively on X about Ethereum and said the more important question is leadership. Comparing Vitalik Buterin with Elon Musk, she said many of Vitalik’s statements only become fully understandable 10 years later, once people see why he said them and how they can be realized. At the same time, she said the market wants to see him closer to the front line in the way Musk often is.
Lin noted that Vitalik has said he wants Ethereum to become a “smaller ship.” But from Tomasz’s departure to the creation of a new Ethereum institution, she said the same question keeps resurfacing: who will provide leadership, governance and management, and which founder can stand at the front line and help this decentralized, open organization become strong again.
She also drew a distinction in stablecoin usage by chain. Ethereum, she said, is more closely tied to institutional capital and developed markets, while Tron is centered more on USDT and serves developing markets and cross-border dollar flows.
How IOSG is allocating in the bear market
Lin said IOSG has changed its portfolio mix during the downturn. The firm has reduced the share allocated to the primary market, while OTC, secondary-market activity and incubation now account for a larger share.
She said crypto is evolving from a trading market into internet-native financial infrastructure and broke that transition into four layers.
- Better Money and Better Rails: stablecoins have already shown that better money and better settlement rails can become internet-native ahead of the traditional system.
- Internet Capital Markets: RWA and tokenization can give onchain dollars more explicit real yield and collateral backing.
- Mass applications: Lin said she believes crypto, like the internet industry before it, can still produce the next ByteDance or the next Pinduoduo.
- AI Agent x Crypto: as financial infrastructure becomes internet-native, the end users will not only be humans. Agents will also need wallets, payments, identity and programmable ownership.
Her conclusion was that money will move onchain, assets will move onchain, and value will concentrate more heavily in the application and interface layers, with AI agents and humans sharing the same underlying stack.
Why bear markets produce the best investments
Lin said the best trades and investments often show up in bear markets because quality is frequently mispriced while real revenue becomes easier to test.
She summarized that idea with three points:
- Valuation dislocation: when sentiment is at its worst, strong projects can trade below intrinsic value, and the best entry prices often appear only after others leave.
- Real revenue survives: IOSG prefers businesses where cash flow can be checked instead of narratives alone. If a project can survive a bear market with a clear product and clear customers, that tends to compound in the next bull cycle.
- The model must stand up to scrutiny: the firm only backs projects that can clearly explain who pays, why they pay and how much they pay. Revenue that can be verified externally is what Lin considers real revenue.
She said genuinely profitable crypto projects in this cycle range from a few hundred million dollars in scale to about $5.8 billion and are spread across DeFi and infrastructure.
Area one: stablecoins and payments
Circle
Lin broke Circle’s business into three parts.
The first is reserve income. She said roughly $73 billion in reserves are invested in short-dated debt instruments, producing about $2.6 billion a year at a 3.5% annualized rate, with that figure moving alongside Federal Reserve policy.
The second is distribution cost. According to Lin, Circle shares about 62% of its main revenue stream with partners such as Binance and Coinbase, leaving it with a gross margin of 38%.
The third is long-term equity value at the infrastructure layer. Circle has built its own chain, the Cross-Chain Transfer Protocol, or CCTP, and developer APIs, though those products currently contribute only about 6% of revenue.
Based on those variables, IOSG built a discounted cash flow model for Circle: start from stablecoin float, convert that into reserve income, apply the 38% gross margin to derive stablecoin gross profit, assign a 15x multiple in year three, then discount the result by 20% to reach present value per share.
fun.xyz
Lin described fun.xyz as the crypto version of a Stripe checkout layer. It provides a universal deposit address, or UDA, and she said all money flowing into Polymarket passes through that entry point. IOSG is also an investor in the company, she added.
RedotPay
Lin called RedotPay the most popular crypto payment card in the market today. She said the card can be used anywhere that accepts Visa or Mastercard, bringing crypto into real spending environments.
She gave several operating figures: more than 5 million cumulative cards issued, global availability, about $3 billion in TPV, a market lead of 4x, and about $150 million in annualized revenue.
Her broader conclusion was that products built at the intersection of payments and stablecoins are now moving toward the mass market, and that RedotPay has already earned meaningful adoption.
Area two: prediction markets
Lin said prediction markets are moving into the mainstream. In her framing, niche products migrate toward mass products, and “the future of the niche is the future of the mass,” making this a useful cold-start path for crypto adoption.
She highlighted two leaders. Polymarket recorded $26.2 billion in volume in the first quarter of 2026, up 90% from the previous quarter. During the World Cup period, from June 11 to July 19, volume exceeded $15 billion. Lin said its edge lies in global reach and non-custodial design, making it one of the clearest examples of next-generation crypto applications.
Kalshi has taken a compliance-first route. Lin said it secured a federal license, recorded about $32.1 billion in first-quarter 2026 volume, and distributes through Robinhood and Interactive Brokers. Founded in 2018, Kalshi spent years facing skepticism, but survived the bear market, obtained its license and built one of the steepest revenue curves she has seen in U.S. fintech.
Area three: AI and crypto
On AI and crypto, Lin said the market spent the last five years focusing on model quality, asking whose Transformer architecture is better or whose RLHF stack is smarter. She said the more basic questions now are where compute comes from, where data comes from and how money moves in AI.
On compute, she said crypto has already shown that open networks can coordinate hardware resources globally. In Ethereum’s proof-of-work era, before the 2022 Merge, the GPU power aggregated by the network was comparable in scale to a frontier training cluster. She was careful to say that does not mean miners’ GPUs could simply be repurposed to train frontier models, because the metrics are not directly comparable. What it does show, she said, is that token incentives can aggregate idle hardware around the world. Applied properly to AI compute markets, that is a real opportunity.
She then turned to data and agent banking and mentioned several examples.
Grass
Grass has about 8.5 million users, according to Lin. Users share spare bandwidth through a browser extension and app in exchange for points. The network layer distributes scraping jobs from AI labs to those nodes, then cleans and structures web data into enterprise-grade datasets.
Lin said the key architectural feature is that Grass can route demand from AI buyers to onchain users and reward them with tokens or revenue. She named companies such as OpenAI and Anthropic as examples of firms willing to pay for training data.
Grass now holds more than 250 PB of data, she said. On revenue, she put 2025 at about $17 million and said 2026 is expected to exceed $70 million. Her takeaway was that tokenized crypto projects can now look very real as businesses, with revenue, cash flow and clear customers. Going to business users first and then to consumers is one of IOSG’s investment approaches, she said.
Hyperbolic
Another IOSG portfolio company, Hyperbolic, focuses on inference and GPU compute markets. Lin said more than 250,000 developers are building on the platform and that its customers include several frontier AI labs.
Nous Research
She also cited Nous Research as a representative example of a project moving from crypto toward AI, adding that users are already familiar with Hermes. She did not go deeper into the technical stack because of time limits during the talk.
Area four: onchain trading and credit
Collector Crypt
Lin described Collector Crypt as an onchain “pawn shop plus card shop.” By revenue, she said, it is already one of the top two applications on Solana.
She laid out the model in three steps. First, the supply side acquires and custody-holds physical cards in a secure vault. Second, the cards are tokenized onchain so each one becomes tradable and redeemable at any time. Third comes monetization and exit.
Collector Crypt has processed more than $1 billion in cumulative volume, has 4 million monthly active users, and ranks among the largest Solana applications by daily revenue, Lin said. She also gave protocol revenue figures of about $7.2 million in April, $9 million in May, $15 million in June and $12 million in July. At that scale, she said, annual revenue could exceed $200 million this year.
For Lin, that makes it a project with actual revenue, a token and clear user demand all at once.
Hyperliquid
Lin said Hyperliquid is another case worth watching and explained why some market participants say it is “eating Binance.” In simple terms, she called it a decentralized futures-focused Binance with a self-reinforcing buyback flywheel.
Each trade generates fees, and about 97% of those fees are used to buy back the token, she said. More trading leads to more fees, more fees lead to more buybacks, and the cycle repeats.
Why IOSG thinks it has an edge
Near the end of the talk, Lin explained IOSG’s own positioning. She said the firm’s advantage lies partly in access to opportunities across regions. Most of its portfolio spans both East and West, with exposure across North America and Asia. She also said IOSG has been able to identify the dominant paradigm of each cycle early.
The foundation for that work, in her view, is research. Deep research is the firm’s investment moat, she said.
The closing point: bear markets are where positioning starts
Lin closed by returning to the core claim of the talk: the best opportunities are often hidden inside the worst sentiment. She highlighted one more time the view from “Bite Shenli” that Bitcoin could reach a low point around the end of October this year, meaning October 2026.
That, she said, is also a reason for investors who shifted toward AI or U.S. equities to revisit crypto and look for new opportunities.
She grouped the opportunities IOSG is watching into three categories: projects with real revenue, especially in stablecoins, payments and AI-crypto intersections; businesses that were wrongly punished during this crypto winter despite solid fundamentals; and the need to identify likely winners before the next bull market arrives.
Her last line was simple: bear markets are where deployment begins.