Sisyphus Market

“So much to lose in exchange for so little to gain.”

That’s the thesis in one sentence. Everything else is footnotes.

Let’s do the temporal math again, stripped of narrative.

Starting point: CAPE ~40+. Historically, forward 10-year real returns from this zone tend to be low—call it 0–3% annualized—before you price in credibility erosion, policy volatility, tariff drag, and AI disruption risk.

The best case from here, using a century-plus of data, is that equities roughly match what cash is paying right now—with vastly more path dependency.

Your window: 10–15 years of peak agency. Call it 12.

Push the boulder uphill at these prices and gravity rolls it back. That’s not a market. That’s a myth.

“The bull market is effectively over for now” is the honest framing for rational long-horizon investors. The “uptrend over time” premise is broken at current prices, under current institutional quality, on any meaningful horizon. It will be unbroken at different prices, different institutional quality—or both.

Our job until then is the hardest job in investing: doing nothing, correctly, for a long time. Rolling in cash with absolute optionality.

The permabulls aren’t wrong about the market. They’re wrong about you.

A 30-year-old with a 40-year horizon? Fine. Ride it out. Time can heal drawdowns. The math works when time is abundant and cheap.

But time isn't cheap anymore once you're getting along in years. It's your scarcest asset—more scarce than capital. And the permabull playbook treats time as infinite, which is the one assumption guaranteed to be wrong for every human being.

I’m not afraid of being wrong. I’ve been wrong before and recovered. I’m afraid of being right about the market but wrong about the timing—deploying capital at what looks like a reasonable entry, watching it grind lower for 3–4 years, and realizing I burned a third of my remaining window waiting for a thesis to play out.

They’ll call it fear. It isn’t fear. It’s math: a finite horizon meeting one of the most expensive markets in modern history during accelerating institutional decay.

I arrived here a long time ago. A defined scout tranche is the admission I might be wrong on timing. Two opposing tranches are the hedge against being wrong about AI’s impact. The rest is cash optionality.

So what am I doing every day—stress-testing this conviction against multiple frontier models?

Making sure the reasoning holds from every angle.

So far:

Valuation: holds.
Credibility risk: holds.
Policy chaos: holds.
Temporal math: holds.

None of them has seriously challenged it yet.

The fortress holds.