My workstack to build an investment thesis
This is what I do with an idea. Every investment thesis I work on goes through some version of the stack below. The full version is what a position deserves before real money goes behind it. But time is the scarce resource in this job, and most ideas do not survive contact with the work, so I keep three versions: a full stack to build conviction, a shorter one to progress or kill an idea cheaply, and a minimal one just to get up to speed on a name.
1. The full work stack
Per investment thesis, before a position. Roughly 60 to 70 hours of work.
| Step | What I am doing | Hours |
|---|---|---|
| 01 | Read at least one annual report, cover to cover. | 6 |
| 02 | Read or listen to the three to four latest earnings call transcripts. | 3 |
| 03 | Read, listen to or watch the latest investor day. | 3 |
| 04 | Two sell-side initiation reports. | 4 |
| 05 | A handful of sell-side reports with structural work, plus one or two earnings updates to see what the market is focused on. | 3 |
| 06 | Read or listen to earnings transcripts of competitors for the read-through, plus the big sell-side sector reports. | 3 |
| 07 | Check whether any respected podcasts or investors have done work on the name (Acquired and the like) and go through it. | 2 |
| 08 | Full financial model (breakdown below). | 20 |
| 09 | Speak to the sell-side analysts with the best written reports, on both the long and the short side, to test and confirm the thesis. Ask what the market and their other clients think, to get the investor view on the stock. | 4 |
| 10 | If the view is differentiated from the market: test and confirm it through product testing and expert calls (see below). | 10-15 |
| 11 | Run the differentiated view by investors I respect, to gauge sentiment on the name and to find where I might be wrong. | 2 |
Step 08: the full financial model
- Identify the three key drivers of the business.
- Reverse DCF to see what is baked into the price.
- Forecast the key drivers and create sensitivities.
- Construct base, bull and bear cases.
- Understand the narrative.
- Create an initial upside/downside assessment.
Steps 10 and 11: testing a differentiated view
This step only happens if I think we have a view that is different to the market, and one that will drive share price alpha: under- or over-appreciated earnings, or a mispriced risk or opportunity that would drive a PER re-rating. If that bar is met, the thesis gets tested and confirmed through product testing and by speaking to experts: people in the supply chain, ex-employees of the company, current employees of competitors, customers, and so on. Once the work stands up, I also run the thesis by a few investors I respect: partly to gauge sentiment on the name, and partly because they are the fastest way to find out where I am wrong.
2. The progress-or-kill stack
Enough work to decide whether an idea deserves the full stack. Roughly 13 to 16 hours.
| Step | What I am doing | Hours |
|---|---|---|
| 01 | Latest annual report, targeted read: business description, segments, MD&A. | 3 |
| 02 | Two most recent earnings calls. | 1.5 |
| 03 | One sell-side initiation report. | 2 |
| 04 | Read one or two expert reports on the sector or the company. | 1.5 |
| 05 | Rough model only: the three key drivers and a reverse DCF for what is priced in. | 4 |
| 06 | One or two sell-side calls to hear the pushback. | 1.5 |
| 07 | Decision: progress to the full stack, park it, or kill it. | 1 |
The kill decision is the whole point of this tier. If the reverse DCF says the market is already paying for my best case, or the drivers are not forecastable with any edge, the idea dies here at a cost of two days instead of two weeks.
3. The up-to-speed stack
Not a thesis, just literacy. Enough to hold a sensible conversation about a name and know whether it is worth a second look. Three to four hours.
| Step | What I am doing | Hours |
|---|---|---|
| 01 | Latest investor presentation or capital markets day deck. | 1 |
| 02 | Most recent earnings call. | 0.75 |
| 03 | Skim the annual report business section and the key metrics. | 1 |
| 04 | One recent sell-side note, purely for the consensus framing. | 0.5 |
A closing note on why the tiers exist. The full stack is where conviction comes from, but conviction is expensive. If every idea got the full treatment, I would cover a handful of names a year and most of that work would end in a pass. The shorter stacks are how ideas earn the right to the deep work: get literate for a few hours, spend two days deciding whether there is any chance the market is wrong, and only then commit the weeks. The 80% of effort that nobody wants to spend still gets spent. Just only where it might matter.
