Edge Capital's weekly assessment of geopolitical risk, capital flows, protocol developments, and market structure across digital assets.
According to Wu Blockchain News and Tokenomist, total unlock value for the upcoming 7 days exceeds $330 million.
Polymarket has announced perpetual futures trading, while Kalshi is reportedly preparing a similar launch beginning with crypto perpetuals. Both platforms are extending beyond event-based contracts into leveraged trading across BTC, equities, and commodities, introducing instruments without fixed expiry. This marks a structural expansion in the scope of prediction market platforms, moving them closer to full-spectrum trading venues.
The addition of perpetual futures shifts prediction markets into direct competition with established crypto exchanges. Perpetuals are the dominant instrument in digital asset markets, driving the majority of volume and open interest across both centralised and decentralised platforms. By integrating them, Polymarket and Kalshi are no longer operating as adjacent products — they are competing for the same speculative demand and liquidity flows that sustain existing venues.
This convergence, in our view, reflects a broader trend in crypto market structure where product boundaries are dissolving. Prediction markets, spot exchanges, derivatives platforms, and lending protocols are increasingly overlapping in functionality, with differentiation shifting toward user experience, regulatory standing, and distribution rather than asset class.
This expansion is occurring alongside a notable shift in U.S. regulatory posture. The CFTC has indicated openness to regulated perpetual futures in the United States, and Kalshi has already secured a margin trading licence, positioning it to operate within a compliant framework. At the same time, prediction markets remain subject to state-level scrutiny around gambling classification, creating overlapping regulatory regimes that introduce both opportunity and risk.
The regulatory environment will likely determine which platforms can scale. Venues that secure federal derivatives licences while maintaining prediction market functionality may capture a disproportionate share of U.S. retail and institutional flow. However, regulatory clarity remains incomplete, and the classification of certain contract types could shift as state and federal regulators establish their respective jurisdictions.
The integration of perpetual futures into prediction platforms reflects a broader convergence in market structure. User flows, liquidity, and trading behaviour are increasingly shared across prediction markets and traditional trading venues, with both competing for the same speculative demand. As product sets align, differentiation will likely shift toward liquidity depth, regulatory clarity, and distribution, rather than underlying asset class. The platforms that combine compliant perps with prediction market distribution may define the next phase of retail-accessible derivatives infrastructure.
On April 18, an attacker exploited KelpDAO's LayerZero bridge to mint approximately 116,500 rsETH — roughly $292 million — without depositing underlying ETH. The attacker then used this unbacked collateral on Aave to borrow approximately $190 million in WETH, triggering the largest DeFi exploit of 2026 and a cascading response across the lending ecosystem. The incident has been attributed to North Korea's Lazarus Group, specifically its TraderTraitor subunit, which has now been linked to roughly $575 million in DeFi exploits in April alone. The exploit did not involve a smart contract bug, social engineering, or key theft. It targeted off-chain infrastructure that connected two systems, exposing a category of risk that traditional security audits may not fully capture.
Aave froze rsETH and WETH markets across all deployments within 77 minutes of the drain, limiting additional borrowing and preventing broader cross-asset contagion. Other protocols responded in parallel: SparkLend, Fluid, and Upshift paused rsETH exposure, while Morpho's isolated vault architecture confined potential losses to specific markets. Ethena, Ether.fi, Curve, and over 20 additional protocols paused LayerZero OFT bridges as a precautionary measure.
The balance sheet impact is now the central focus. Aave has guided to $124 million to $230 million of potential bad debt depending on final loss allocation methodology. Stablecoin markets on Aave reached 100% utilisation as users withdrew into stables, effectively freezing approximately $5.1 billion of stablecoin deposits across the protocol. Total DeFi TVL declined by roughly $13 billion in the 48 hours following the exploit, with Aave absorbing the majority of outflows — its TVL fell from $26.4 billion to $17.9 billion.
The post-exploit capital flows suggest reallocation within DeFi in addition to systemic withdrawal from the sector. SparkLend absorbed over $1.4 billion of inflows and saw an increase in active borrowing, benefiting from its conservative risk policy — the protocol had deprecated rsETH as collateral in January and maintained ample ETH withdrawal liquidity throughout the crisis. Morpho similarly demonstrated architectural resilience, with its CEO disclosing minimal exposure due to isolated market design.
Stablecoin flows reinforced the rotation narrative. Net stablecoin supply grew by $1.8 billion on the week, with USDT absorbing $3 billion — the largest weekly print of the year. RWA-backed stablecoins including USYC, USDG, PYUSD, and BUIDL gained over $600 million combined, suggesting a flight to perceived safety within digital assets rather than an exit from the ecosystem.
The "DeFi United" initiative has mobilised contributions from Aave, Mantle, Lido, Ethena, EtherFi, and others to restore rsETH backing. Key measures include a proposed 30,000 ETH credit facility from Mantle, a 25,000 ETH contribution under discussion at Aave DAO, and the Arbitrum Security Council's emergency action to freeze 30,766 ETH of the attacker's downstream funds. The remaining funding gap is estimated at 75,000 to 90,000 ETH, depending on assumptions around liquidation recoveries and token freezes.
Execution will depend on coordination across protocols, governance timelines, bridge reopening, and the release of frozen assets — some of which face up to 49-day governance processes on Arbitrum. However, the speed and scale of the coordinated response represents an institutional maturation of DeFi's crisis management capability. The incident underscores, in our view, that cross-chain infrastructure security — not smart contract risk — has emerged as the sector's most consequential vulnerability, and that architectural choices around verification, isolation, and collateral policy will increasingly differentiate protocol resilience.
Disclaimer
This communication is for information purposes only and is not an advertisement, an offer, invitation or a solicitation to buy or sell securities or investment products, an official confirmation of any kind and is not intended as investment advice or recommendation. Before making an investment decision, investors should ensure they have sufficient information to ascertain the legal, financial, tax and regulatory consequences of an investment to enable them to make an informed investment decision. The information in this communication is subject to change without notice. No warranty is made as to the completeness or accuracy of the information contained in this communication, and the information in this email may be erroneous, invalid and/or unsubstantiated. The sender therefore does not accept liability for any errors, omissions or adverse consequences in the contents of this message which arise as a result of e-mail transmission or for any other reason.
The performance and value of any financial product may fluctuate and may be subject to sudden and large movements that could result in a loss equal to or in excess of the amount invested. Past performance is not a guide to future performance, future returns are not guaranteed, and a loss of original capital may occur. The presented figures are based on estimates, assumptions, models and third-party data, any or all of which may prove to be inaccurate.
