Happy Friday! 

It’s risk-off across the board with BTC down 1.9% and gold’s weakness extending to seven consecutive sessions, as Brent crude surges above $110 on Iranian escalation. AI was the only sector in the green, largely driven by TAO's continued breakout. 

Below, we break down why Solana’s perps market still can’t compete with Hyperliquid, and how Bulk is trying to rearchitect the execution layer from the ground up.

Despite major tech benchmarks closing flat on the day, both BTC and gold saw notable drawdowns of −1.9% and −4.1%, respectively.

Gold’s recent weakness is particularly striking. After a strong start to 2026, seven consecutive down days have erased a large portion of its gains, leaving it up just 7.45% YTD. The shift is being driven by rising oil prices and renewed inflation concerns. Brent crude has surged above $110 following Iranian attacks on energy infrastructure, raising fears of a prolonged conflict. In response, markets are pricing in a higher probability of rates staying elevated for longer, weighing on gold. Polymarket currently implies a 33% chance of a US-Iran ceasefire by April 30.

Across crypto sectors, it was broadly a risk-off session, with AI standing out as the only sector in the green. This was largely driven by TAO, which is up 3% on the day and has rallied 47% over the past month after dipping below $200.

Under the hood, the strength is even more pronounced. Subnets have significantly outperformed over the last 30 days, with many up between 30% and 270%. This aligns with the thesis from our December Bittensor report. The TAO emission halving has reduced yields on the root subnet while increasing relative rewards for alpha subnets, incentivizing capital to rotate into higher-yielding subnet opportunities. The continued rise in the share of TAO staked in subnets reinforces this trend.

In other news, in the world of prediction markets, regulatory pressure is beginning to build again. Kalshi and Polymarket are facing scrutiny from state officials over whether their March Madness markets bypass gambling laws, particularly in states where sports betting is restricted. March Madness is the NCAA men’s college basketball tournament and remains one of the largest betting events globally, with over $3.3B expected to be wagered through traditional channels this year.

In just four days, Polymarket and Kalshi have already processed $212M in NCAA-related volume. At this pace, volumes could reach $1.2B by the end of the tournament. This would already represent close to 35% of the $3.3B expected across traditional sportsbooks, and you can see why the potential loss of this revenue is becoming a point of contention for regulators.

If you want to go deeper into how prediction markets are navigating these regulatory challenges and where the next phase of growth comes from, I will be hosting a panel at the Digital Asset Summit 2026 next week with industry leaders in this space. So grab your tickets now!

— Kunal

Bulk’s bet on Solana perps

The problem with perps on Solana is structural. HyperCore is purpose-built to do one thing; Solana is a general-purpose state machine where perps compete for blockspace alongside everything else, and the chain isn’t optimized for makers. That tradeoff shows up directly in maker economics: Execution uncertainty keeps spreads wide, wide spreads keep books thin, and market makers who need reliable fills have little reason to quote aggressively onchain. 

Bulk’s thesis is that you cannot fix this within L1. Their sidecar model pulls the entire trading path (matching, risk, consensus) into a dedicated parallel network run by Solana validators. Meanwhile, deposits, withdrawals, and settlement remain on the base layer. The sidecar is additive: Validators run Bulk alongside their existing client stack rather than replacing it. Around 12 validators representing ~5−6% of Solana stake are live today, with 12.5% USDC share of exchange fees for participation.

A key differentiator is the risk engine. Bulk is building something closer to CME SPAN than anything in crypto today: portfolio-aware margining that prices cross-asset correlation rather than treating each position independently.

A trader long $100K BTC and short $100K ETH at 0.85 correlation would face ~$55K in effective exposure rather than $200K gross, with margin requirements potentially 70%+ lower on hedged books. 

Several of Bulk’s core design choices are also explicitly aimed at improving maker economics and, by extension, tightening spreads and deepening book quality: gasless trading, ~20ms consensus/tick speed, and FIFO with cancel priority. 

Most perp venues require moving collateral into the exchange. On Bulk, Solana-native positions like Kamino or Loopscale collateral will be able to stay yield-bearing and still back margin through delegate access. 

In short, Bulk’s bet is that decentralization still matters, but only if it can be paired with real performance and native composability. 

Bulk is targeting mainnet before July, with unified margin accounts, dynamic risk engines, and external collateral integrations such as Kamino/Loopscale targeted by year-end. A key risk is complexity: Bulk is shipping a parallel consensus, unique risk engine, and composable collateral techniques, while Solana L1 upgrades like Alpenglow and MCP may make the case for a separate execution layer unnecessary. 

— Sam

1. Derive: The Case for Onchain Options

Blockworks Research finds that Derive is finally gaining real traction as onchain options activity matures, with record options volume, stronger RFQ flow, and improving BTC share against Deribit. Perps help support hedging and exchange activity, but options remain the more profitable business. New asset listings are also emerging as a key growth lever, with HYPE already becoming a meaningful options market and helping bring users into BTC options. 

The report argues the next phase depends less on one new product and more on better distribution, deeper integrations, and continued execution. DRV remains the only way to get exposure to that upside, though the token still carries meaningful risks from supply overhang, thin liquidity, and a wide range of valuation outcomes.

2. Inside Circle’s Stablecoin Economics

The report finds that Circle is still mainly a reserve-income business, not yet a scaled software or payments platform. Most of its revenue still comes from earning yield on USDC reserves, while newer products like CCTP, CPN and USYC remain strategically important but financially small. The report argues the main variables for 2026 are USDC balance growth, lower short-term rates, and how much reserve income Circle keeps after heavy partner payouts, especially to Coinbase. 

Circle’s strategic footprint is widening, and regulation is becoming more supportive, but the investment case still depends far more on float growth and rate sensitivity than on diversified platform monetization.