Anatomy of a Governance 'Heist'

BTC and crypto assets popped this week, recovering from the late-June crash to outperform the broader market. The S&P 500 and Gold were up over the past week, but moving further out the risk curve in equities there’s a different story to be told.

The Nasdaq 100, measured here by QQQ performance, is down 1.8% so far in July, and its latest addition might not steady the ship. Elon Musk's SpaceX was approved to enter the index this week, just as the newly public company suffered a double digit decline over the past 10 days. That's remarkable vol for one of the largest companies in the world, especially as price targets point to a $10trn valuation.

The cooling is visible on crypto venues too, as daily volumes in SpaceX perps have fallen more than $9bn from their late June highs, the speculative appetite fading alongside the price.

Crypto was a relative oasis of calm, at least on the surface, as BTC traded above $63k early in the week, and most of the top 10 are in the green. The mildly hawkish Fed minutes on Wednesday did nothing to deter traders, as there was little of note and relatively little new info. The central bank is in a wait and see mode, not ready to pull the trigger on a rate hike just yet.

Looking beneath the hood trading volumes were quite muted compared to last week. BTC was naturally slow over the weekend and built modestly into Monday before paring off again. Meanwhile, ETH lagged the recovery — although price doesn’t tell the whole tail here, and we’ll get into the different behaviour between the two majors later.

BTC traded roughly $18.8bn in spot this week against ETH's $9.0bn, and the direction of that flow is the story. BTC's cumulative net taker flow climbed all week to nearly $280mn.

ETH went the other way though, with CVD peaking above $100mn last Friday, before sellers moved in over the weekend. While there was a similar spike in volume on Monday, more than double the previous day, the flow didn't hold.

While that flow was leaving, the leverage was arriving in perps. ETH open interest rose 7.4% on the week against BTC's modest 1.4%. Longs building into fading flow is the kind of setup options desks charge for, and they are (more to come on that).

The other half of the tape was forced. ETH liquidated $154mn of shorts this week against $66mn of longs, and the biggest hour ran $24.7mn at 21:00 UTC Monday, 99% of it shorts, and the squeeze top-ticked the week at $1,814. Short liquidations are buy flow, but the kind that exhausts rather than accumulates: the squeeze marked the high, and by Tuesday the CVD bleed was doing the rest. BTC's buyers this week chose to buy. A share of ETH's had no choice.

Two weeks ago I wrote about the options market breaking, dealers pulled quotes as prices whipsawed (common enough in crypto options where liquidity can be patchy). That can be a deterrent for some, and might go some way towards explaining the relative immaturity of the crypto options market versus other derivatives products.

However, the market has shown considerable resilience in the past couple of weeks, more than recovering. The whole curve out to September now sits below that pre-crash 41%, the front by 6 to 10 vol points, and only December is above it at 42%. BTC realized 38% this week while grinding quietly higher, which puts the front expiries 3 to 7 points under trailing realised. The weekend is the extreme end of it, ATM IV for those expiries drops from 35% on Saturday to nearly 32% on Sunday. It snaps straight back to 35% by the following Friday.

One interesting observation, analysis of the June crash found the curve flattened and cheapened in the days before the 24th. One instance isn’t a trend and I’m not calling a repeat, but the shape is back, and it is cheaper now than it was then.

ETH, meanwhile, holds about 1.35x BTC vol at every tenor, 12 to 14 points depending where you look. That premium is structural, not news, but this week's tape is a live demonstration of what it pays for. Leverage built on flow that didn't stick is exactly the uncertainty it prices, and nobody watching this week's ETH tape would be rushing to sell it.

Skew adds a slight wrinkle though. BTC's front is heavily put-skewed while ETH's front is call-skewed, on the same days. Near-term BTC protection and near-term ETH upside, both bid. Both flatten out the curve.

Last week brought an interesting “heist” of sorts on the BONK DAO — the decentralised governance body managing the treasury. On June 30 an anonymous wallet submitted a proposal asking for over 4 trillion BONK, about $20mn, to be sent to a wallet it controlled. The proposal sat there in public for several days.

Over the holiday weekend in the US, a wallet bought more than 880 billion BONK, or $4.4mn, on Binance and Bybit. The quorum for the vote the anon wallet proposed was 879.95 billion. On Monday the vote closed, seven wallets voted out of more than 18,000 members, a 2.9% turnout (token weighted), with 99% in favour. The transfer executed automatically, BONK sold off and the internet debated whether this was bad governance or a heist. We’re not too interested in that kerfuffle, the interesting details are in the orderbook data.

The combined Binance and Bybit tape over the two accumulation days ran about $19mn. The attacker bought $4.4mn, or 23% of total volume, in one of the least liquid memecoins on the market during the summer months on a holiday weekend in the US. How? Because BONK has quite an interesting liquidity profile.

On gross flow it's one of the most fragile names in crypto, with Amihud illiquidity of 791bps per $1mn traded (i.e. trading $1mn moves the price by about 7.91%). Slam a $4.4mn market order in and it gaps.

But net flow barely registers, in fact net taker flow explains essentially none of BONK's hourly price moves this week (R² of roughly zero). It's a market that punishes impatience and completely ignores patience. The tape agrees, price went flat to slightly down while the buying ran.

We reconstructed the Binance order book for the accumulation weekend and set it against a matched control weekend to show how the anon did it. Resting bids were refilled about 1.6x more often per unit of book traffic, sustained across both days, at flat price, into 5x the notional. Bids getting hit and topped straight back up, over and over, is what a patient accumulator sitting on the bid looks like.

Bybit shows a weaker, front-loaded version of the same pattern, so this reads as strongest on Binance rather than uniform across venues.

One caveat worth stating plainly. The trade tape over the window actually leans buy-initiated, so the passive-bid read rests on the book, where a maker is visible, not on taker prints, where by definition it isn't.

Two things had to line up here, only one is a clear failure. A treasury anyone could buy their way into, and a market where patient size leaves no footprint. Both were visible in the data before the money moved.

World Cup knockouts will keep prediction markets busy. June CPI lands on the 15th and earnings season is rolling, with Coinbase and Strategy set to share earnings at the end of July or early August. Either can shake the front end, and I'll be watching whether the front of the curve stays this cheap.