← Back to all projects

The Investing System

Completed
Investing Personal Finance Automation
Read the philosophy →

After a couple of years of arguing with myself about how to invest, I built a single system that takes me out of the equation. The whole thing runs on a 70/20/10 split.

70% Core: Broad U.S. and international index funds on full autopilot. Payroll deductions and recurring buys. No logging in to tinker.

20% Conviction: A small, capped sleeve for the few things I actually have a thesis on. Rule-based, and if it grows past its cap I trim back to target and the trimmings go straight into the core.

10% Dry Powder: Cash kept separate from the emergency fund, with pre-written rules for deploying into a real correction so I never have to make the call in the moment.

The investing page now lists the actual funds and their expense ratios, which is the part I would have wanted to see ten years ago.

Status check, October 2026: done. The machine has run five months without me touching it. Same five buys on the same day every month, the HSA finally moved out of cash, and the one-off stock bets are being worked out of the portfolio on their own written exit rules. I am not adding positions. I am not picking companies. The system is the product.

What I added in October: the Dip Ladder. The old dry-powder rule was two lines, deploy half at a 20% drop and the rest at 30%. That covers a crash and ignores the ordinary 10% correction that shows up about once a year. The ladder is three lines below the S&P 500's 52-week high, at 10%, 20%, and 30%. Cross a line and the monthly buy steps up, 1.5x at the first and 2x at the deeper two, with a one-time lump at the deeper two. Close back above a line for five straight days and the buy steps back down. The reserve has a floor so the rule ends the bump, not my nerves. The base never shrinks to fund any of this, because the cash was already sitting there for exactly this job.

The part that makes it real. A scheduled check runs every weekday after the close, compares the S&P to its 52-week high, and stays silent unless a line was crossed. When one is, I get an email and a Slack message with the exact two numbers to change in Fidelity. One tap. Nobody sells "if the index drops 10%, raise my recurring buy" as a product, so the human stays in the loop by design. The two big lumps are pre-staged as conditional orders so they fire even if I am on a plane.

What I will not pretend. Buying dips is not a return booster. The honest backtests show it beats plain monthly investing about one time in four. What it buys is behavior: a system that actually buys the crash instead of watching it, with every decision made months before the headline. That is the same trade the cash cushion already made. Call it what it is.

This is my personal household strategy, separate from any other venture. The goal isn't to be the smartest investor in the room. It's to be the most consistent one.

Ready to use it?

Read the philosophy →