z-of-a.

Spectral · Project #1

The astrology was wrong.
The tool wasn't.

It found a 788-day cycle in the credit market — p = 0.021, replicated, and Wall Street isn't watching it.

Phase Clock · 2026-04-10
tighteningeasingtroughpeak
Quadrant Q4_easing · envelope 0.0681 · phase 314.2°

We took astrology seriously enough to test it.

The astrology died. The test found a cycle nobody else sees.

That's what we do at z-of-a — we take ideas Wall Street won't touch and build the tools to test them. Sometimes the wrong tool finds the right thing. Spectral is what fell out the first time. There will be more.

788
day cycle, BAA-10y spread
0.021
p-value, replicated out-of-sample
25
topology nodes, k-means clustered
analog regimes since 1950

Who this is for

  • Allocators and PMs who already trust their own judgment and want one more honest signal.
  • Quants who notice when the consensus model has nothing to say about the current regime.
  • People who'd rather read a one-page chart with a real p-value than a 30-page deck with none.

Who this isn't for

  • Anyone looking for a daytrading signal. The cycle is 788 days. The horizon is months, not minutes.
  • Anyone who needs a backstory dressed up in tasteful editorial prose. We don't do that here.
  • Anyone who can't handle being early.

Get this Friday's Signal — free.

One email, Friday afternoon, with the current Phase Clock and a one-paragraph read on what it's doing.

No spam. One email a week. Unsubscribe anytime.

Or: skip the wait. Buy this month's report.

Each monthly report is one chart, one regime call, one page of methodology, and the analog dates we're watching. $20. No subscription required.

If you'd rather not think about it again, the monthly subscription is $100. Yes, that's more than four reports. It's priced that way on purpose. We'd rather have a hundred subscribers who actually want this than a thousand who are price-shopping research.

P.S. The first thing we tested was whether Saturn-in-Taurus periods showed elevated equity drawdowns. The astrology didn't survive. But the spectral tool we built to test it found something else: a 788-day cycle in the BAA-Treasury spread, p=0.021, that nobody else seems to be watching. We're publishing the chart every Friday until somebody notices.
P.P.S. If you're a quant at a fund and you think you can break this in five minutes — please try. Reply to any Friday email and tell us what you found. We mean that.