The Halvren Read
One number, computed in the open.
The Halvren Read reduces ten checklist verdicts to a single number for legibility. It is the same arithmetic for every operator in coverage. No analyst override, no thumb on the scale, no separate model. The verdicts already live on each operator page; the score is what those verdicts add up to under a fixed weighting.
The weighting
Each of the ten Halvren Checklist questions resolves to a verdict: green, amber, or red. Each verdict maps to a point value. The points sum within three pillars, and the pillars are weighted into the final score.
Pillar I · The business
Questions 1 – 4
Up to 40 points
- Q1. Full-cycle free cash flow
- Q2. Unit economics at the worst price of the last decade
- Q3. Balance sheet at trough pricing
- Q4. ROIC on incremental capital
Pillar II · The people
Questions 5 – 8
Up to 30 points
- Q5. Insider ownership, bought not granted
- Q6. Behaviour in 2015 and 2020
- Q7. Compensation tied to per-share value
- Q8. Visible succession
Pillar III · The cycle
Questions 9 – 10
Up to 30 points
- Q9. Position on the cost curve that matters
- Q10. The decade test — what a normal year looks like in 2035
The math
Green = 10 points per question, amber = 5, red = 0. Within each pillar, the raw point total is rescaled to the pillar cap. The three pillar contributions sum to the Halvren Read, capped 0–100.
An operator who clears every question lands at 100. An operator who fails every question lands at 0. The score is a clean projection of the checklist; it is not derived from price, sentiment, or a separate model.
Bands
Colour-coded for the operator cards and the watchlist spread. The bands are not verdicts; they are a reading aid.
- 75 – 100 (green). The desk has very few reservations. Either three full pillars line up, or one weakness is well-understood.
- 50 – 74 (amber). The business is fine; one of the pillars is unresolved. A close-watch operator.
- 0 – 49 (red). Multiple pillars are not yet clean. The desk is reading, not buying.
What the score is not
This is not a rating. It is not a buy or sell signal. It is the desk's view of operator quality reduced to one number for legibility. The full reasoning lives in the writeup.
Two operators with the same Halvren Read can be very different businesses. A 91 in oil & gas reads differently than a 91 in midstream. The number rewards passing the same ten questions; it does not normalise across cycles, capital intensities, or commodity exposures. A reader who only knows the number knows less than a reader who reads the writeup.
The score is recomputed every time the underlying checklist verdicts change. Verdicts change after a new quarter, a CEO change, an insider buy, a debt issuance, a balance-sheet pivot. The score is downstream of the desk's read; the read does not chase the score.
For operators not in coverage
The Halvren Read is computed only for the 20 operators in formal coverage. Checklist Live applies the same ten questions to any ticker on demand, but does not produce a 0–100 score for off-universe names — the desk's review is the part that earns the number. The streaming read is the desk's checklist applied; not the desk's verdict.
See the full Halvren Checklist for the framework, the Coverage page for the Halvren Spread, and the research archive for the writeups behind each score.