Process

A draft, written in public. I expect it to change as the research record grows and the reasoning gets tested.

1. Two lenses, one decision

The CMT Program taught me to read the tape — trend, breadth, rotation, relative strength. The CFA Program teaches what something is worth. Most analysts pick a side; I work both. Fundamental gives me the what and why; technical gives me the when and how much.

2. Top-down, then bottom-up

Macro sets the opportunity set. Sector and factor rotation say which areas of the market are being rewarded right now. Bottom-up security work happens inside that frame — not in spite of it. Starting bottom-up alone means under-pricing regime risk.

3. Technicals as risk management, not prediction

Technical signals are most useful for one job: telling me when something I've fundamentally underwritten is being rejected by the market — and conversely, when a thesis is starting to be confirmed. Rotation, trend, and breadth are inputs to sizing and timing, not substitutes for valuation work.

4. Process > outcomes (over any single quarter)

Outcomes are a noisy signal of process quality at short horizons. The fix is to make the process inspectable: written theses, dated entries, and an honest review when a call goes wrong. The research that was wrong stays up, and the next note says why.

5. Concentration with discipline

Diversify by factor, not by ticker count. A 20-25 name book of high-conviction ideas across uncorrelated drivers beats a 60-name book where the marginal positions exist out of fear of being wrong.

6. Valuation matters — eventually

The market spends years not caring about valuation and then a quarter caring about nothing else. Sizing should reflect the gap between intrinsic value and price; technicals reflect whether the path to closing that gap has begun.

7. Sequence risk deserves more weight than it usually gets

Where capital has a spending path — the problem wealth advisory exists to solve — the right portfolio depends on the path, not just the destination. Mean-variance is a starting point; risk parity, drawdown sensitivity, and goals-based bucketing all live downstream. This is the area I most want to work in, and the one I have the least practice in.

8. Models are accountability

Every model behind this research is open math I built myself. If a thesis depends on a margin assumption I can't justify with a sensitivity table, or a setup that doesn't survive a different lookback window, the thesis is weaker than I think.

Last revised — draft v0.2. Some of these will change with experience. The site will reflect that.