State of The Market, The Prediction Market Wars & Kraken Pursues Aave

Summary

• Santiago Santos reiterates his start-of-year call that equities would peak once the SpaceX IPO happened, and says he does not think crypto equities or tokens can recover without some AI-market rollover. He keeps that as a hedge, not a certainty. He also says a near-term AI rollover seems unlikely: H100 spot rates were down about 40% month over month, but long-term contracts continued rising, showing corporates remain committed to compute.
• Crypto venture is a barbell: 147 total deals so far in Q2 2026 is the lowest quarterly count since Q4 2020, while later-stage payments, stablecoin and tokenized-asset activity and M&A are improving. Rob says fintechs are increasingly exploring international expansion through stablecoins, non-custodial wallets and cards; early-stage crypto remains very difficult and token trading is down 40% year over year.
• Headline stablecoin metrics are noisy because most transfer volume is still bots trading against each other on-chain. Better signals include adjusted-volume work from Allium, Visa and Artemis, stablecoin-backed card-swipe data, and Rob’s pipeline of about 30 public companies that have begun RFPs or implementation. Yanowitz says inbound RFPs for on-chain data, asset intelligence and OFAC compliance have never been higher, including from companies he thought were years away from crypto.
• Canton has two possible futures: a strong enterprise database and SaaS product, or a broader institutional hub-and-spoke ecosystem. Rob says Broadridge mainly uses it as enterprise infrastructure and has no clear plan for open cross-Canton trading. Santos says Canton’s growth has been heavily subsidized with token rewards, while his team has struggled to validate adoption beyond proofs of concept and to reconcile token-versus-equity structures.
• The Journal’s Polymarket story is framed as part of a wider competitor-oppo cycle, not an isolated practice. Santos says the conduct surfaced in the article should not have happened but is common across prediction markets, exchanges and consumer trading products. He also says Meta entering the market would be a mistake and that retail financial-marketing standards need improvement.
• Kalshi’s potential $40B raise reflects rapid growth but remains only a reported possibility. Santos says Kalshi is currently about 1.8–2x Polymarket’s size and generates more revenue and volume, while warning of fee compression. He sees an important future in institutional clearing and notes compute block trades on Polymarket and, possibly, Kalshi. Both platforms reportedly did more actual volume last week than DraftKings’ stated $3.4B annualized volume.
• The reported Kraken–Aave terms are wrong in their details, according to Santos, who expects any transaction to be more equity-like than token-like. The broader strategic rationale is for exchanges to deepen relationships with on-chain lenders and offer DeFi/CeFi products. Santos says Aave does not need funding, and the discussion is strategic rather than a normal fund investment.
• Kraken’s secondary pricing is disputed: Yanowitz estimates roughly $10–12B, Rob cites a Deutsche Börse transaction implying about $13B, and the speakers disagree on whether the prior round was $15B or closer to $19–20B. FOMO raised $75M from Index, USV and Benchmark on a social-trading thesis spanning leaderboards, chat, published theses, perps, spot and RWAs; Rob invested at pre-seed and says the pitch was broader than crypto.

Deep dive

1. Equities, AI and the crypto recovery

• During a broad selloff, Yanowitz said SpaceX’s decline “triggered, I feel like,” OpenAI pulling its IPO. Santos then restated his start-of-year call: “the equities market would top once the SpaceX IPO happened.”
• Santos also says, with a hedge, “I don’t think crypto markets—crypto equities and crypto tokens—can come back without some sort of rollover in the AI market,” because AI has absorbed attention and capital across markets. He does not require a bubble burst; even a modest AI pullback could help crypto.
• His concern is that each quarter brings questions about capex returns, GDP acceleration and demand for compute, yet the market keeps working: P/E ratios remain mostly fine, earnings continue to grow, and markets often perform well after the anxiety. Volatility and uncertainty, however, are high.
• Santos cites SemiAnalysis data with some uncertainty around the exact product: short-term H100 rates were down about 40% month over month, while long-term contracts continued to rise. Spot buyers are showing less willingness to pay, but large corporates are still committing to compute for the long term. His conclusion is that a meaningful near-term AI rollover seems unlikely, so anyone waiting for one to revive crypto may be “a little bit cooked” for now.

2. Crypto venture’s barbell

• Yanowitz cites Blockworks data showing 147 crypto deals so far in Q2 2026, the lowest quarterly count since Q4 2020. M&A is rising, fundraising bottomed in February at its lowest monthly level since September 2024, and Yanowitz sees more activity in Series B, C and D than at pre-seed and seed.
• Rob says the areas showing product-market fit are payments, stablecoins and tokenized assets. Rain is growing month over month regardless of crypto prices, and he is receiving calls from fintechs that want to expand internationally through stablecoins, non-custodial wallets and cards.
• The catalyst is often either a board-level question—“what is your stablecoin strategy?”—or a founder watching another domestic or UK business expand globally faster through new infrastructure. Stablecoins can reduce the time and cost of onboarding banks and can create new revenue, although regulatory treatment remains gray in some places.
• Rob says he is also seeing interest from traditional fintechs and a little from AI companies, while crypto trading is down 40% on the year. Exchanges are responding with real-world assets, equities, prediction markets and pre-IPO products. Early-stage crypto remains especially weak, with founder quality low relative to the prior five or six years.

3. Measuring real stablecoin adoption

• Rob warns that “the vast majority of stablecoin transfer volume is still just bots trading against each other on-chain.” He makes a similar point about much of the Solana activity being discussed around tokenized equities: roughly 90% is still bots trading on low-fee AMMs.
• Adjusted-volume efforts from Allium, Visa and Artemis are useful but remain works in progress. On the retail side, Rob points to Payments Chad and stablecoin-backed credit-card swipe volumes as better indicators of actual fintech activity. Publicly announced programs involving MoneyGram, Western Union and Slash may eventually become measurable as well.
• Rob also has a privileged pipeline view: roughly 30 public companies are planning stablecoin businesses and have already done an RFP or begun implementation. Yanowitz sees the second-order effects through inbound requests for on-chain data, asset intelligence, user provenance and OFAC compliance. Those requests are coming from companies outside Blockworks’ usual target list, including slow-moving European and South Korean firms he would have expected to be years away from crypto.

4. Canton’s two possible futures

• Asked why Canton is attracting attention, Rob lays out two frames. In the negative frame, Broadridge uses Canton as a better database and a better way to operate its repo market, with no clear plan to open cross-Canton trading beyond its client base. Canton could therefore become a strong enterprise software business without becoming a broader financial network.
• The positive frame is that institutional change is slow. Over time, a hub-and-spoke Canton ecosystem could let institutions interact with one another more broadly. Rob says Canton is an eight- to ten-year-old project that has recently gained meaningful traction, with Goldman Sachs and other large institutions involved as investors or clients.
• Santos adds that the product has been heavily subsidized: the project has paid substantial token rewards or fees to encourage activity. He does not treat subsidies as automatically negative—venture-backed companies often subsidize growth—but says Canton is at a make-or-break point between enterprise proof of concept and broad adoption.
• Santos says his team has discussed the last three raises but struggled to validate that customers want to move from an interesting proof of concept or enterprise database to a much broader network. He also says the token-versus-equity structure is difficult for financial investors even though strategics may care less about the financial framing. He explicitly leaves room for being wrong.

5. Prediction-market competition and marketing

• Jason says Rob had previously predicted that negative Polymarket stories would appear. Santos frames the current Journal article as part of a wider competitive environment: people at companies and investors are actively trying to place negative stories about competitors. He calls that a statement of fact, not opinion.
• Santos says the conduct described in the article probably should not have happened, but argues it is not isolated to Polymarket. Yanowitz gives a similar account from a consumer-founder group chat, where the main reaction to the story was asking which firm had been hired and whether others could hire it too. He compares the behavior with tactics used in other high-stakes rivalries, including Uber and Lyft.
• The discussion also covers influencer claims of trading success. Paper trading is common across prediction markets, crypto exchanges, FanDuel and DraftKings. Santos says retail marketing for financial products needs better standards because “98% of the time” a social-media claim about a trading gain is not real.
• On Meta’s reported prediction-market product, Santos says he is extremely bearish on Meta building its own exchange. He still expects strong growth in the category: Polymarket, Kalshi, Robinhood, Susquehanna and J.P. Morgan are all expected to have very strong June activity.

6. Kalshi, Polymarket and the clearing opportunity

• Santos says Kalshi is currently about 1.8–2x Polymarket’s size, with more revenue and volume, so a valuation around twice Polymarket’s could make sense if the reported raise occurs. He emphasizes that the $40B figure is not confirmed. Kalshi remains more than 90% sports when parlays are included, while Polymarket is roughly 55–60% sports around the World Cup.
• Santos, who says he wrote his first Polymarket check about two and a half to three years ago, remains very positive on the category. Polymarket has expanded through exchange integrations, went live on Telegram, and is doing more institutional clearing and backend infrastructure, while Kalshi is further ahead in that institutional-access product.
• Institutional participation is still mostly market making against retail taker flow. However, Santos points to a compute block trade cleared by FalconX on Polymarket and says he thinks Galaxy did a similar compute trade on Kalshi. He is also hearing early conversations about insurance-linked and other contracts arranged through RFQ. The potential clearing market, he argues, is large.
• Kalshi’s take rate is roughly 1.5–2x Polymarket’s, but Santos expects fee compression and says investors should update their models. He does not view the businesses as indefensible: liquidity attracts more liquidity, and Polymarket has a strong retail brand.
• Yanowitz compares the category with DraftKings, which announced $3.4B of annualized volume. Santos says Polymarket and Kalshi each did more actual volume the prior week. His distinction is that DraftKings is primarily a direct-to-consumer sports product, while prediction markets could become modern digital exchanges spanning new contracts, institutional clearing and additional asset categories. Santos also references Coinbase launching what he describes as the first U.S. perp.

7. Aave, Kraken and FOMO

• Yanowitz relays a CoinDesk report that Kraken might acquire 15% of Aave at a $385M valuation, involving 35,000 ETH, 250,000 AAVE tokens and a 15% common-equity stake in Aave Labs. That would represent roughly a 70% discount to AAVE’s stated $1.4B market cap. Stani rejects the reported discount and says all Aave protocol and GHO revenue goes to the AAVE token, with no protocol or product revenue going to Aave Labs.
• Santos says the article is wrong in its details and expects any real transaction to be more equity-like than token-like. He says strategic interest in Aave, Morpho and similar on-chain lenders is real: exchanges and fintechs want deeper relationships that can support a DeFi/CeFi experience, vaults and eventually securities lending against tokenized assets.
• The broader race is for integrations. “If Morpho powers Revolut or Aave powers Kraken, that is the dream,” Santos says. He views the effort as strategic rather than fund-related because Aave does not need capital. He also says the matter is separate from the Gauntlet situation.
• On corporate structure, the discussion says the investment would be made through Payward Asset Management, which would be its first such deal as the group develops beyond Kraken. Yanowitz says most of Payward’s value is still Kraken.
• Kraken’s secondary price is uncertain. Yanowitz estimates that shares may be available around $10–12B. Rob cites a Deutsche Börse secondary announced around March that implied roughly $13B, while Rob initially described the prior round as about $19–20B and Yanowitz said he thought it was closer to $15B. Crypto trading is down 40% for everyone this year, adding to the discount.
• FOMO raised $75M in a Series B from Index, USV and Benchmark. Rob invested personally in its pre-seed angel round, when funds were not allowed in, and says Dragonfly reviewed the Benchmark-led Series A after seeing strong traction and retention. His guess is that Index was underwriting a broader social-trading thesis—not simply another crypto exchange—built around leaderboards, chat and published trading theses. The product currently includes perps, spot and crypto RWAs, with the stated ambition to become a broader trading app.