Bottom-up, repeatable, and run the same way, week after week, by every member of the team. Macro-aware but never macro-led. Our systematic screen narrows a vast universe to a manageable set of promising candidates, and that is where the research begins. We then build a concentrated portfolio one high-conviction company at a time.
The global small-cap universe consists of over 30,000 companies between $20 million and $2 billion in market cap.
Screening is the backbone of our discovery process. To cover it thoroughly, we run our screen on a scheduled basis. Every two weeks by industry and geography. Each PM and analyst owns specific sectors and regions, ensuring that every company in our universe is screened twice a year.
Our screen was built in-house to quantitatively reflect our investment philosophy. It puts the largest weight on return on assets, and looks for sustainable growth in sales, earnings, and cash flow, together with a steady-to-improving margin profile and a strong balance sheet.
What the screen surfaces, experience contextualizes. Over the years we have built an internal database of thousands of companies and many thousands more interactions — a shared memory the whole team draws on every time a name reemerges. In a market with minimal coverage, this is an asset that only immense experience and time can replicate.
From each screen, an analyst narrows the list to the most promising names and gets to work: reading filings, transcripts, and presentations, and building a financial model to understand the company.
In a typical year, our team conducts hundreds of direct meetings with company management. In addition to interactions with the new names coming out of screening, we aim for a touchpoint with our portfolio companies once a quarter. Most of these businesses carry little or no sell-side coverage, so there is often nothing published on them worth reading. Our notes become the record: what a company actually does, what drives the business, and what has to go right for it to grow. Thus, when a name resurfaces on a screen years later, we are not starting from zero. And when we want to know which companies we have met in a given end market, the answer sits in our own files.
Every company we touch gets a written recommendation: buy, watch, add, trim, or sell. This is distributed to the whole research team on a standard template.
Because we’re a small, experienced team, that thesis gets pressure-tested face to face. Every name is looked at by more than one set of eyes, and disagreement is welcome. We resolve it through deeper research and, where honest disagreement remains, through position sizing. A controversial name earns a small allocation.
The portfolio is assembled bottom-up, with each of the 50 to 70 holdings standing on its own merit. There is no optimizer and no tracking-error target. Position size reflects conviction, balancing our knowledge of the industry and management with the size of the opportunity, the business’ resilience, balance-sheet strength, and valuation. Positions often start small and scale up as our conviction grows.
Sector and country weights are a byproduct of stock-by-stock selection, not a target. We keep self-imposed limits deliberately wide so they never get in the way of good stock-picking: a maximum of 10% in any position, 40% in any country, and 40% in any GICS sector.
We define risk first as the permanent loss of capital, and second as underperformance versus our benchmark. Our single biggest risk-management tool is extensive due diligence. Beyond that, our controls are common-sense and rule-based: a cash buffer, position, sector, and country limits mentioned above; and a cap on the smallest, least-liquid names. We embrace company-specific risk, but avoid macro and geopolitical risks that can overwhelm a good business.
On selling, our logic reduces to two reasons: either the thesis is broken, or a better opportunity has come along and we’re recycling capital into it. Because quality compounders have the potential to re-rate over time, we don’t set price targets. And because we have no market-cap cutoff that forces us out of a winner, we can let our winners run.