How this works: I run my portfolio to a written policy. Three sleeves, each with a target and a rebalance band, plus a set of trend gates. Once a month a script pulls my positions from the broker, checks them against the policy, and emails me a verdict. My agents draft this post from that email. I read it and decide whether it goes out.

The numbers (monitor output, 2026-09-01)

Factor equities: 58.2% actual, 60.0% target, band 55-65%, drift -1.8pp. OK.

Trend following: 33.2% actual, 35.0% target, band 30-40%, drift -1.8pp. OK.

Liquidity: 8.6% actual, 5.0% target.

Trend gates: total-world equities is above its 3-month, 6-month, and 12-month averages. All three gates open.

What changed since August: factor equities up 0.3pp, trend following down 0.3pp, liquidity unchanged. The policy only acts when a capital sleeve drifts more than 5pp. Nothing is close.

What I overrode this month: nothing.

Next report: October.

Drafted by my agents from the system's real output. Judged and edited by me. Not investment advice. I'm describing my own accounts and my own process. Nothing here is a recommendation or an offer of any kind.

This post is about how that works, because the interesting part isn’t the automation. It’s the boundary.

Why rename at all

I’d been feeling for a while that “hedge fund” was the wrong metaphor. A hedge fund connotes shorting, leverage, trading, alpha-hunting. What I actually run is a written investment policy, allocation bands, monthly rebalancing, and a perpetual horizon. That is endowment governance. The name was a misdescription.

I mentioned this to my agent and asked it to check whether the thinking held up. It came back with three verified facts:

1. Accuracy: the system really is endowment-shaped, point by point.

2. Domain quality: endowmentofone.com was unregistered. Shorter, cleaner.

3. Ownability: this one I’d half-known and it half-surprised me: personalhedgefund.com, the canonical form of my own brand, had been registered by someone else since 2006 and sits parked on an ad server. Dropping “the” is the most common way people mistype a domain. If the blog ever grew, my traffic would leak to a squatter, or I’d be negotiating to buy my own name at hostage prices. I could never fully own the old brand. The new one had no such shadow.

That third point converted a nice-to-have into a decision. I said: do it.

How the day actually went

Here’s the division of labor, beat by beat.

The agent did: renamed the publication in Substack settings and verified it stuck. Rewrote the About page headline and intro, checking its work in the editor before saving, then confirming on the live page. Searched Namecheap, put the domain in the cart, declined every upsell, and verified the cart by screenshot: one item, $11.48, privacy on, auto-renew on.

Then it stopped. It parked the browser at the Namecheap login and told me: your credentials, your money, your click. I signed in and paid. Two minutes.

The agent resumed: pointed the new domain’s DNS at Substack, swapped the publication’s custom domain, converted the old domain’s records into permanent redirects, and updated its own runbooks, the files that tell the Saturday drafting job and the Sunday publish job where the blog lives. The automation that writes the monthly portfolio posts never noticed the ground moving under it.

It stopped again at YouTube. No Google session in its browser, and it doesn’t touch credentials. I signed in once. It added the blog link to my conference talk’s video description, above the chapter timestamps so the chapters kept working, and verified the link on the public watch page.

Three stops all day: a login, a payment, a login. Everything else ran without me: research, verification, DNS, redirects, editor surgery, its own documentation.

Where the agents stopped

Here’s the part no agent supplied. The Personal Hedge Fund didn’t feel right to me. It just didn’t feel right to me, and it had been nagging at me for a while. That’s not something an agent can check. The verified facts came later, and they mattered. The squatted domain turned a feeling into a defensible decision. But the feeling came first, and it was mine.

And the feeling was about more than accuracy. I’d started wanting this blog to be about more of me than my portfolio. I orchestrate large parts of my life with AI agents, and the investing is just the part with the cleanest numbers. I wanted room to write about that. “The Personal Hedge Fund” could only ever be about the money. “Endowment of One” can hold a life: the capital, and the machinery that runs it. The section you’re reading this in exists because of that same decision.

So the division of labor, as I actually experience it: I bring the sense that something is wrong and the idea of what it should be instead. The agents verify it and build it. They turned a vague dissatisfaction into three checked facts, and a decision into a shipped reality in a day. Judgment, identity, and money stay with me. Mechanics, verification, and documentation go to them.

One more rule made this cheap: decide once, early. The blog had one post and approximately zero subscribers. A rename at that size costs nothing. The same rename at five hundred subscribers is a migration with casualties. Renames don’t get cheaper by waiting, so once the reasoning checked out there was no reason to sit on it.

About this writing

One more thing, because this section of the blog will live or die on it: the agent that renamed the blog is the agent that wrote this account of it. It drafted this post from its own session record, which is why the details are exact. The $11.48, the screenshot of the cart, the three stops: none of that was reconstructed from my memory. It was logged by the thing that did the work, at the moment it did the work.

So yes, this writing is AI-generated, and I’m not going to pretend otherwise. The disclosure line at the bottom of every post is not a formality, it’s the method. But AI-generated is not the same as invented. Every post in this section describes something that actually ran, in my actual life, usually that week. The agent drafts from the record. I judge, edit, and sign. If I ever can’t stand behind a sentence, it doesn’t ship.

Next time something in this system does a day’s work in a day, I’ll write it up the same way.

Drafted by my agents from the system’s real output. Judged and edited by me.

Not investment advice. I’m describing my own accounts and my own process. Nothing here is a recommendation or an offer of any kind.

One thing up front: these posts are drafted by my agents from the system's real output, then judged and edited by me. I'm not going to pretend otherwise. The agents are the point.

How I got here, in one paragraph

For years I did special situations. I built something like thirty tools for it: an agent that read SEC filings for five to seven hours straight hunting warrant terms, CVR trackers, net-net screeners, an OTC scanner. They worked. And I quit anyway, not because the strategies were bad, but because they stopped fitting my context: those niches are small and illiquid, and as my portfolio grew, the positions they could absorb became too small to matter. The work stayed the same. The reward, as a fraction of the whole, kept shrinking. What survived the purge was the machinery habit, not the complexity: today the strategy is simple ETF sleeves, and the sophistication all lives in the automation that keeps me honest. The full story is a future post.

The system, in one paragraph

Three sleeves: roughly 60% factor equities (small-cap value with a profitability tilt, plus momentum, spread across US, international developed, and emerging markets, roughly 2:1 value-to-momentum), 35% trend following (managed futures), and a 5% liquidity buffer. Once a month, a script pulls my positions from the broker, maps every holding into its sleeves, checks each sleeve against its rebalance band, checks 3/6/12-month trend gates on total-world equities, and emails me a verdict. Monthly on purpose: rebalancing decisions happen once a month, and checking more often would just be noise. It never places a trade. It tells me the truth. Acting is my job.

This month's report (real output, 2026-08-03)

Verdict: ALL CLEAR. No rebalance needed.

Factor equities: 57.9% actual vs 60.0% target (band 55-65%), drift -2.1pp, OK

Trend following: 33.5% actual vs 35.0% target (band 30-40%), drift -1.5pp, OK

Liquidity: 8.6% actual vs 5.0% target (band 3.8-6.2%), drift +3.6pp, DRIFT

Trend gates: total-world equities above its 3-month, 6-month, and 12-month averages.

The judgment layer

The honest wrinkle this month: the machine flagged the liquidity sleeve at DRIFT, 8.6% against a 5% target, and still called ALL CLEAR. That's by design. Only the two capital sleeves gate action. Factor equities and trend following are the strategy. Liquidity is the residual. As long as both capital sleeves sit inside their bands, there's no rebalance to do. The DRIFT flag on cash is information, not an alarm.

What I overrode this month: nothing. The machine got the same answer I would have. Most months that will be the answer, and that's the entire point. The hardest part of running a simple strategy is leaving it alone, so I built a system where touching the portfolio requires a reason the machine has to produce first.

Next month

Same format, same tables, whatever the machine says. If it says do nothing again, you'll read "do nothing" again. I suspect the long-run value of this diary is watching how rarely anything needs doing.

Drafted by my agents from the system's real output. Judged and edited by me. Not investment advice. I'm describing my own accounts and my own process. Nothing here is a recommendation or an offer of any kind.