Attack the Week (ATW)

Cosmic Fault Lines / FX Spreadsheet / Calendar / Dashboard / Allocation Update

Sunday Thoughts

I hope those of you in Europe got to see the eclipse this week. Spain, Iceland and Greenland got the full show; most of the continent got 90% or more, and I got nothing because I couldn’t be bothered to buy the glasses in time. Lesson learned, and I shall be prepared for the next one, which is arriving in 2081. Yikes, very unlikely that I will be around then!

You know, I’m a sucker for cosmic scale. The facts that don’t fit in your head no matter how hard you push.

Here’s one I heard on a podcast. You can fit 1.3 million Earths inside the Sun. Fine, we’ve all seen the graphic. So shrink it down. Make the Sun the dot on top of an “i” in a normal book. At that scale, the whole Earth is a single bacterium drifting four centimetres from the dot on the page. And that dot is one star. The book holds more stars than there are grains of sand on every beach on Earth.

Out there the facts are all like this, quiet and absurd and enormous. Down here we potter about treating our little world as perfectly ordinary. We build our axioms, we call things facts, and we forget that complex systems don’t behave nicely all the time and were never built to.

Think of a fault line. Two plates lock, stress builds for years, nothing moves, everything looks calm. Then one small slip lets go and it all releases at once. Could be a tremor. Could level a city. You can map where the faults run; you just can’t call the day.

Boom and bust is that slip. As normal as weather. Yet we keep telling ourselves we’ve grown cleverer, that this time the good stuff simply rolls on.

Only 6 years ago people were locked indoors calling the end of the world. Then rates got jacked to put the inflation genie back, and we lived through a bond bear market of a generation.

I’ve been reading the latest Myrmikan Research, which is always a joy. Their latest post on AI runs the argument straight through the plumbing, and the plumbing is where things always break. The bit that stuck: the oldest trick in finance is making risk disappear by tucking it into a pocket that’s meant to be safe. It never disappears. It moves to where nobody’s looking until it’s too late.

Follow the money. The hyperscalers are spending more than all their cash flow on AI and dressing it up. Rather than build the data centres and take the capex hit up front, they sign take-or-pay contracts with the neoclouds, so the commitment sits off-balance-sheet as a future operating expense. Goldman reckons a trillion dollars of these lease commitments never show on the books.

Then it gets creative. The AI debt, bonds secured against Nvidia chips used to buy more Nvidia chips, gets placed with private equity. And PE has spent a decade buying life insurers outright, because an insurer’s liabilities run decades out and get topped up by monthly premiums. Permanent capital, near enough. The perfect home for illiquid, mark-to-model paper.

So the risk ends up in the annuity book of a life insurer, often reinsured through a captive in Bermuda where the disclosures are thinner. The whole US life industry runs on a 4% equity cushion against $11 trillion of liabilities. That’s the safe pocket. That’s where the fragility sits now, in plain sight, under the pension savings of people who’ve never heard the word neocloud.

Stack margin loans on inflated stock, then policy loans on top, and you get leverage built on leverage. We have seen that script many times before.

Which brings me back to the fault line. The slip that starts it won’t be big. A single flawed assumption, one rating that gets questioned, one insurer reaching for cash too fast and spooking the rest. Senator Root said it in 1913, arguing against the Fed before it existed: eventually one man of poor judgment breaks, and as he falls he knocks the next brick, then another, and down comes the structure. Cioffi’s fund at Bear Stearns went in June 2007. Lehman held on another fourteen months. The slip and the quake are never the same size.

We will always be uncertain about timing, but rarely about the reaction function. When the redemptions start, and insurers sell what’s liquid first, the government does what it always does: protect the banking system over the currency, every time. Freeze the policies, blame the free market, lean on the Fed to fund the bailout. The QE wagon will be driving again.

Until then, we do what we do best: put our ears on the ground, stay open-minded, and be prepared for whatever might come.

Let’s now read Macro D’s latest thoughts and look at his Macro FX book of trades before we briefly scan next week’s macro calendar, analyse the latest output of our macro dashboard and check what asset allocation the model has in store for us as we enter yet another trading week.

Let’s roll.