This site is not investment advice. It is a personal research portfolio published for educational purposes. Read the full disclosures before relying on anything here.
Shelf Life Research

Independent · Consumer sector · Unaffiliated

Ratings and analysis on the companies you actually buy from.

The list

Coverage universe

Every name I have published a view on, with a 12-month price target and a rating. Click any card for the full write-up.

Methodology

How the ratings work

The scale is absolute, not relative to a benchmark or to the rest of the coverage list. Ratings are reviewed at least quarterly and after any material company event.

Rating scale · 12-month horizon
RatingExpected total returnWhat it means
Strong Buy> +25%High conviction. The market is mispricing something specific and identifiable, and I can name the catalyst that closes the gap.
Buy+10% to +25%Attractive risk/reward, but either the upside is more modest or the path is less certain.
Hold–10% to +10%Fairly valued on my numbers. A good business can still be a Hold; price is a separate question from quality.
Sell–10% to –25%Expectations embedded in the price look too high relative to what the business can deliver.
Strong Sell< –25%Material downside risk — structural decline, balance-sheet stress, or a thesis I believe is broken.
Under Review—Rating and target suspended pending new information, typically after an acquisition, restatement, or major guidance change.
STEP 1

Read the filings, not the headlines

Every report starts with the last three 10-Ks and four 10-Qs, plus the proxy. Segment disclosures and the footnotes usually say more than the earnings release.

STEP 2

Build the operating model

Revenue is driven bottom-up where the company gives me the units: store count, same-store sales, traffic versus ticket, category mix. Margins are built off gross profit per unit, not a percentage assumption.

STEP 3

Check the model against the world

Consumer companies are unusually checkable. Store visits, app rankings, pricing on the shelf, job postings, and channel commentary all serve as a reality test on the model's assumptions.

STEP 4

Value it, then argue against it

The target is the probability-weighted output of an explicit bear, base and bull case, each built on the company's own disclosed drivers rather than a single point forecast. Peer multiples are context, not support — they rest on third-party estimates that are neither filed nor auditable, so they stay out of the target arithmetic. Before publishing, I write the strongest case for the opposite rating; if it is more persuasive than mine, the rating changes.

Conviction levels

  • High — the thesis rests on a small number of assumptions I can defend in detail.
  • Medium — the direction is clear, the magnitude and timing are not.
  • Low — an initial view published to be tested and revised in public.

What this is not

  • Not a model portfolio, a newsletter, or a trade-alert service.
  • No position sizing, no entry or exit prices, no stop losses.
  • No coverage initiated for payment, and no company reviews a report before it is published.

Scorecard

Track record

Every call scored against what the stock actually did after it was rated. Self-reported and unaudited. Price return only — dividends, fees, taxes and transaction costs are excluded.

Prices as entered

A call is scored correct when the stock moved the way the rating implied — up for a Buy, down for a Sell, within the ±10% band for a Hold. Returns run from the closing price on the initiation date to the latest price shown, and are not annualised. Calls inside their twelve-month window are still open and are marked as such; an open call can still go wrong. Past performance is not indicative of future results, a record this short is not evidence of skill, and scoring against the stock alone ignores what the market did over the same period.

About

Who writes this

Why the consumer sector

It is the one part of the market you can walk into. Store traffic, shelf pricing, packaging changes, app downloads, and menu tests are all observable before they show up in a quarterly print, which makes it the best sector to learn genuine fundamental research in rather than just reading someone else's model.

Get in touch

Corrections, counter-arguments, and "your model is wrong because…" emails are all welcome — being publicly wrong and then fixing it is the point of doing this in the open.

Email me

Legal

Disclosures & terms of use

By accessing this site you agree to the terms below. If you do not agree, please do not use the site.

The short version

I am not a financial adviser and nothing here is financial, investment, legal, accounting, or tax advice. These are one person's opinions, published to build and demonstrate research skill. They are not tailored to you, they may be wrong, and acting on them could lose you money. Do your own research and speak to a licensed professional before making any investment decision.