Weekly Top Stories - 8/22/25

In this week's newsletter, Marc Hochstein contrasts a Federal Reserve official’s solicitous speech to a blockchain crowd in Wyoming with her predecessor’s standoffish stance toward the industry; Will Owens explains why HeavenDEX is suddenly the belle of the ball in Solana-land; and Thad Pinakiewicz gets to the bottom of the alleged 51% attack on Monero.

Fed’s Bowman Moves to Normalize Crypto

Federal Reserve Vice Chair for Supervision Michelle Bowman struck a markedly different tone from her predecessor on crypto, regulation, and banking in a speech Tuesday at the Wyoming Blockchain Symposium.

Perhaps most strikingly, Bowman said the Fed should consider allowing staff members to hold small amounts of cryptocurrency so they can learn hands-on how the technology works. “I certainly wouldn't trust someone to teach me to ski if they'd never put on skis, regardless of how many books and articles they have read, or even wrote, about it,” she quipped.

Bowman, whom President Trump nominated to the Fed board in his first term and elevated to vice chair this year, also made pointed remarks about the debanking of crypto and other legal but politically disfavored businesses. Although she didn’t mention the Obama administration’s Operation Choke Point or its unofficial revival under Biden, she suggested (in a roundabout way) that debanking was downstream of policymakers’ fixation on “reputational risk.”

“Exams or reviews focused on reputational risk have often lacked a sufficient nexus to financial risk and safety and soundness considerations that are the appropriate focus of our supervisory activities,” she said. Regulators “must allow and encourage banks to provide banking products and services to any legal business, without disfavoring any particular viewpoints, businesses, or industries.” (Emphasis added; debanking of individuals and businesses was often said to be politically motivated.)

Previously, the Fed and other banking regulators, with the encouragement of President Trump, had stated that they would no longer consider reputation risk when examining banks. On Tuesday, Bowman said she is additionally considering “whether we need a regulatory change to provide greater transparency and certainty about this approach,” without elaborating on what that change would be.

OUR TAKE:

This is a very important speech from a key Fed official signaling a historic pivot. (The Vice Chair for Supervision is the Fed’s point person for bank oversight and testifies before Congress twice a year.) Nothing happens in a vacuum. Let’s compare Bowman’s remarks to those of her predecessor, former Fed Vice Chair for Supervision Michael Barr.

In February, shortly before leaving the Fed, the Biden appointee denied that the Fed had pressured banks to dump or avoid crypto or other legal businesses as clients. “We don't tell a bank to do business with particular clients, and we don't tell a bank not to do business with a particular client," he reportedly said in an appearance at Georgetown Law. In a later speech, he reiterated: “The Federal Reserve neither prohibits nor discourages banking organizations from providing banking services to customers of any specific class or type, as permitted by law or regulation.”

As Chico Marx said, “who you gonna believe, me or your own eyes?”

In that light, Bowman’s comments, though understated (she is a Fed official, after all), were a remarkable acknowledgment from a top regulator that past policies indeed contributed to crypto firms’ struggles to obtain or keep bank accounts. Yes, it happened. It will be interesting to see what regulatory changes Bowman might propose to codify the administration’s removal of the nebulous “reputation risk” from bank examiners’ assessments.

To be fair, Barr (a former Ripple advisor) was less dismissive of crypto than some of his fellow Democrats in Washington. But in 2023, the year after FTX’s collapse, he told banks to be “careful and cautious” in dealing with the industry and said it would be “unsafe and unsound” for them to hold crypto assets on their balance sheets. That same year, he launched the Fed’s novel activities program, which subjected firms that deal with crypto and fintech to "enhance[d] supervision,” with special experts supplementing (not replacing) the usual supervisors.

By contrast, Bowman in her Wyoming speech sounded more bullish on blockchains than some of the crypto industry’s jaded veterans. Right off the bat, she said the technology has “the potential to fundamentally transform the way we live, work, and interact in society.” She urged banks and regulators to move past “an overly cautious mindset,” called embracing technology a "necessity” for TradFi, and identified a need to establish “regulatory certainty” and tailor rules to specific use cases rather than “applying expectations designed for an imaginary ‘worst-case’ scenario.’” (At the same Wyoming conference, Fed Governor Christoper Waller made similarly optimistic and open-minded remarks about stablecoins’ potential to upgrade payment systems.)

In one more departure from Barr’s playbook, Bowman is reintegrating the Fed’s novel supervision work back into the regular examination staff. This may explain why she’s keen for employees to play around with crypto wallets. Crypto is being normalized, so understanding how the technology works won't just be for specialists; it’ll be part of a regular bank examiner’s job, too. – Marc Hochstein

HeavenDEX Rises, But Can It Last?

The balance of power among Solana launchpads continues to shift. HeavenDEX, a new Solana launchpad and automated market maker (AMM), raised over $25 million onchain in just 11 hours through an initial coin offering (ICO) before launching its native token, LIGHT. In a market saturated with new launchpad products, Heaven is trying to stand out with its so-called “god flywheel.” This mechanism uses 100% of protocol fees to buy back and burn LIGHT.

At the time of writing, Heaven had launched 5,335 tokens in the previous 24 hours. That puts it well behind market leader pump.fun at 20,294, but far ahead of LetsBonk (724).

On revenue, the story gets more interesting. Heaven is now a close second behind pump.fun in Solana memecoin launchpad fees. On Aug. 20, pump.fun generated $1.35 million in revenue while Heaven pulled in $988,238. Since launch, Heaven has surpassed $290 million in cumulative volume and $3 million in cumulative revenue. To date, more than $3.4 million in LIGHT has been bought back and burned, shrinking total supply by 3.37%. Because fully diluted value (FDV) is roughly double the market cap, the burned supply equals more than 6% of circulating supply.

OUR TAKE:

Solana’s onchain culture remains hyper-focused on launchpads, memecoins, and speculative activity. Since early 2024, pump.fun has been the undisputed leader in that space (aside from a brief period when LetsBonk overtook it). Yet users frequently criticize pump.fun for being “extractive” to the ecosystem. HeavenDEX offers a different thesis: every dollar of protocol revenue is recycled into LIGHT, forever.

In a memecoin market where hold times are measured in seconds and scalping quick profits seems like the only way for traders to not lose money, it’s refreshing to see something that incentivizes longer-term behavior. But reflexivity cuts both ways. Liquidity is still relatively thin, and ongoing token unlocks add steady selling pressure. The same mechanics that amplify upside can just as quickly accelerate drawdowns.

Whether that ethos attracts developers who share the same values remains to be seen. LetsBonk’s quick rise and fall shows just how fast sentiment can rotate in Solana’s launchpad wars.

With pump.fun’s PUMP token still trading below its ICO level, the open question is whether Heaven’s model can sustain attention beyond the initial hype cycle. Will users continue to gravitate toward it, or is it just the “hot new thing”? – Will Owens

Monero Raided by Privateering Hashpower

Over the past several weeks, Qubic, an AI-and-“useful proof-of-work” layer-1, mobilized its miners to dominate Monero’s hashpower and publicly framed it as a live “51% takeover demo.” Qubic claims it briefly controlled >51% of hashpower, causing a six-block reorg and ~60 orphaned blocks; Kraken paused XMR deposits, then re-enabled them with a steep 720-block confirmation requirement. Monero contributors quickly floated ideas to harden the network. Meanwhile, independent researchers challenged the “>51%” claim and attributed the reorg largely to selfish-mining dynamics rather than sustained majority control.

A 51% takeover, or attack, is when one group of miners controls more than half of a proof-of-work blockchain’s mining power. With that majority, they can win the race to make blocks, rewrite the last few minutes or hours of history, reverse their own payments (double-spending coins), and temporarily block others’ transactions. It is a core risk of running a PoW blockchain: that coordination among miners, malicious or otherwise, can compromise the integrity of the network and ledger.

Qubic is an L1 that routes generalized compute toward tasks it calls “useful PoW.” As part of an “economic demonstration,” Qubic tweaked its protocol incentives to encourage CPU mining of Monero, alongside the GPU-based mining the network was conducting beforehand. Qubic further incentivized its own miners to make the change via the XMR rewards they would earn from CPU mining and attracted existing Monero miners to its mining pool through selfish mining techniques that increase a pool’s share of block rewards at the expense of all others’. Selfish mining reduces the efficiency of honest miners by invalidating some of their work. The net effect is that the selfish pool’s reward share > hashrate share, so it looks like it has more hashpower than it really does.

Qubic ramped up its attack and on Aug 11-12 published its “historic takeover” posts and press release, asserting it had crossed the 51% hashpower threshold and triggered a six-block reorg after a month-long hash war. This claim was challenged by several researchers, and confirmed to be a semantic argument by Qubic’s founder, who acknowledged it should be called a “34% attack” not a “51% attack” because of the selfish mining technique. Qubic did produce more than 50% of the blocks during short periods of the attack, but it is estimated that it only had ~33% of the total hashpower of the Monero network, not 51%+.

OUR TAKE:

Qubic’s move is a textbook stress-test of PoW economics, harsh but valid under existing incentives. Whether or not it truly cleared 51% of hashpower every minute is secondary to the demonstrated ability to coordinate enough hash, time private chains, and exploit miner payoffs to produce reorgs and extract outsized rewards.

Qubic didn’t look to simply economically deploy its available compute power to earn extra yield; it became a strict economic optimizer, willing to crack another network to extract value. Qubic became privateering hashpower, attacking another chain under the colors of more profitable mining, a not unexpected outcome in the untamed sea that is crypto. Qubic attacked a brittle edge of ASIC-resistant PoW: if it’s cheap to redirect general-purpose compute and profitable to do so (especially with an additional subsidy), the network that can pay miners more, directly or indirectly, wins block production.

That said, the way Qubic approached this evolution in its network wasn’t collaborative. Branding it a “51% takeover,” running selfish-mining strategies, and publicly polling which chain to “hit” next (Dogecoin “won” the honor) burns goodwill and imposes costs on honest network contributors.

Still, pressure can be productive. The Monero proposals – ChainLocks-like finality overlays, mining decentralization via P2Pool, or other consensus tweaks – are exactly the sort of pragmatic evolution PoW networks need. While Qubic’s approach is distasteful to many, it serves as a poignant reminder that security budgets do matter in digital assets and that thoughtfully constructed economic incentives are just as important as cryptographic security.

Perhaps this is the first death knell of small PoW chains, and we are about to see a collapse in that ecosystem, as Qubic goes from chain to chain extracting value. Or, maybe a solution like EigenLayer’s rented economic security can find a productive home securing PoW chains. – Thaddeus Pinakiewicz

Charts of the Week

Liquid staking tokens (LSTs) hit a record total value locked (TVL) of $86 billion last week before easing with the broader market. As of Aug. 21, LST TVL was $78 billion and accounted for 52% of the $148 billion DeFi market. We view this as a healthy signal for the DeFi staking ecosystem and evidence of investors' appetite for yield-generating products.

At $37.7 billion, Lido’s stETH [MH1] [CR2] captures about 48.5% of the LST market, followed by Binance’s staked ETH product at $13 billion, or roughly 17.5%. Lido’s TVL has risen about 185% since early April and about 65% since early July, when it was near $13 billion and $21 billion, respectively.

In a related development, Ethereum has recently seen a huge spike in both exit wait time and amount of ETH queued to exit. We believe this is driven by a few important factors that originated in the LST market, including [MH1] the recent depeg of stETH on Lido. In late July, roughly $1.7 billion of ETH left Aave, a move widely attributed by market watchers to wallets linked to Tron founder Justin Sun. The liquidity drop pushed Aave ETH borrow rates toward 10%, making popular stETH-to-[MH2] ETH leverage loop trades unprofitable. Traders unwound, selling stETH or submitting redemptions at Lido, and the validator exit queue swelled to more than 600,000 ETH with wait times of about 10 days. Longer waits tie up capital and reduce the effective return from arbitraging the stETH discount, so a small depeg can persist until the queue clears.

The mechanism linking Aave to Lido is straightforward. Aave is where many loopers post Lido’s stETH as collateral to borrow ETH and buy more stETH. When ETH borrowing becomes expensive, those loops break and positions are unwound, which sends redemptions to Lido. When redemptions exceed the buffer, Lido must turn off some validators to free the ETH they secure. Ethereum limits how many validators can exit in each time window, and that cap is shared across the network. When a large operator like Lido submits many exits at once, it uses up more of those slots, leaving fewer for others and lengthening the queue for everyone. During July’s price appreciation, new staking inflows lagged, so Lido’s ETH buffer did not refill quickly. Combined with some profit taking and redemptions, that kept pressure on exits and the peg. Because Lido is the largest staking provider, its exits consume a sizable share of the network’s exit capacity, extending wait times during periods of stress. — Christopher Rosa

Other News

⚖️ 'Writing code without ill intent not a crime,' says DOJ official after Tornado Cash case

⛓️‍💥 U.S. banking regulator OCC lifts enforcement order from Anchorage Digital

🚀 Solana handles 100k transactions per second in test run

🪪 U.S. Treasury seeks comments on digital ID verification for DeFi and Genius Act

₿ Treasury company Nakamoto purchases 5,744 BTC

🎭 MetaMask confirms stablecoin plan

❌ Nasdaq to delist BNB treasury company whose stock fell below $1