All In on Risk, Cash Spent — the Fed Pipe Behind the Bet Is Still Idle

All In on Risk, Cash Spent — the Fed Pipe Behind the Bet Is Still Idle

I run a crypto family office called Maelstrom, and I've been telling anyone who will listen that the book is at "maximum risk": near-zero stablecoin cash, Bitcoin at the core, Ethereum as the largest non-Bitcoin position, and a call for ETH to roughly double from about $2,500 to $5,000 by year-end. Headlines like that read like a green light to chase. Before anyone mistakes my portfolio for a roadmap, it's worth seeing what the book actually is — because this is not a trade on Ethereum at all.

Start with what the book is. Maelstrom is my own money, a family office, not a public fund with redemptions breathing down its neck. In January I described the book as "almost maximum risk", deep in risk assets, with a dollar-stable holding "very low", and Bitcoin itself bought with spare cash thrown off by financing trades. In June I dumped the HYPE and NEAR positions outright to de-risk through early Q3, citing higher energy prices from the Iran war, an AI IPO pipeline, and politics. By late August I was back to "maximum risk", telling Crypto Banter it would be "foolish" to avoid risk assets and warning that the next phase brings higher volatility. The man moves — but here the man is me, and the current all-in has a shelf life. The interesting question is what flips the book next.

The first useful read is what the trade is not. Ethereum's roadmap, layer-2s, and tokenomics? Not the trade. Liquidity and positioning are. In August I named the exact channels in my essays.

The first channel is the U.S. Treasury. This month the Treasury said it would at least double its long-end debt buybacks starting in September, and Bitcoin jumped as much as 9% in a day on the news, with more than $1 billion in crypto short positions liquidated. The same idea in plumbing terms: when the Treasury spends into the market — repurchasing its own bonds, or running down what I put at about $1 trillion sitting idle in its General Account at the Fed — those dollars have to land somewhere, and BTC is first in line.

The second channel is the Fed's FIMA repo facility, jargon for a window where foreign central banks can temporarily swap their U.S. Treasury holdings for dollars. Right now the window has a $60 billion per-counterparty cap and, per the latest Fed release, zero foreign offices are actually using it. The pipe is plumbed and dry. I wrote a whole essay, "Yen-quake", on this: if the Fed raises that cap, Japan can defend a falling yen by borrowing dollars against as much as $1.37 trillion of Treasury collateral instead of selling the bonds — and every dollar borrowed is fresh dollar liquidity that sloshes toward Bitcoin, Ether, and gold. No cap change, no yen event, no fresh money.

The third channel is the long arc, and it's where I get cinematic. In "Situationship" I argue the AI data-center boom is a credit story, not an earnings story: the Fed holds short rates below nominal growth, real rates stay negative, and the Treasury and Fed coordinate — T-bills issued, "reserve management purchases" capping the long end — to keep money cheap without calling it QE. When AI capex growth eventually decelerates while credit keeps expanding, the excess liquidity has nowhere to go but risk assets; and if the AI bubble cracks, I expect a panic-print bigger than 2008, with emergency vehicles buying AI equity directly, which carries Bitcoin to a million and beyond.

Now the piece that matters for Ether. I call Ethereum the "most hated" large cap — the asset every fund is underexposed to — and that is partly checkable. ETH's 52-week high is about $4,764, still below its 2021 record near $4,800; Bitcoin, by contrast, printed an all-time high near $125,000 within the past year. Ethereum's share of crypto market cap sits around 11%, and the altcoin-season index around 32, meaning by that technical measure the rotation into alts hasn't even started. Clean the narrative out, liquidity coming in behind it, and my claim is that a rally above $3,000 turns reflexive — underweight funds forced back in, momentum feeding itself — and the target is $5,000 by New Year's. That is a positioning trade, not a valuation trade.

There is an entry behind this stance worth staring at. Fully invested means no buffer. If the Treasury doesn't run down the TGA, the buybacks run their course, and the FIMA cap stays at $60 billion of capacity and zero usage, then the liquidity I'm counting on has not arrived — and a near-cashless book has nothing between it and a drawdown but the exit button. My own record is "profitable but uneven": HYPE and PENDLE paid, a token called PUMP didn't. And "all in" spans more than spot positions: Maelstrom is also mid-way through raising a $250 million private-equity fund to buy profitable off-chain trading-infrastructure and analytics firms, and seeded a chain-abstraction stablecoin called River. That money is committed for years.

None of this tells you to buy or sell ETH. It tells you where my conviction is anchored, and what decides it: Treasury buybacks that actually swell the dollar pool, a TGA balance actually spent, a Fed window that moves from zero usage to something flowing. I invented the perpetual swap, the most-used product in crypto's history, and I'm handing you the exact accounting entry I'm watching — the same pipe that generated this month's rally. Track that, not my cash position, and you'll know before most whether an all-in book is a bull signal or the next supply. A fully invested, cashless fund is only conviction while the liquidity it bet on is actually arriving.