Small-cap stock analysis requires the same intellectual framework as large-cap analysis — rigorous business evaluation, conservative valuation, and clear thinking about risk. What differs is the context: less available information, lower liquidity, greater price volatility, and the absence of professional coverage that would otherwise provide a market reference point. Each of these differences creates both a challenge and an opportunity.
The following five-step framework describes the process used in every Sifter Research report. It is not a formula. It is a sequence of questions, each of which must be answered before moving to the next.
Step 01 — Screening
The starting point is a large universe of potential candidates — ideally, all listed companies in a given set of markets — filtered down using quantitative criteria that eliminate the most obvious failures quickly. Typical screening criteria include minimum return on equity over multiple years, low or no financial debt, consistent free cash flow generation, and a market cap below a threshold that excludes institutional interest.
The goal of screening is speed, not precision. A company that fails basic financial quality metrics at the screening stage does not warrant deeper analysis. Most will fail in under two minutes. That is the point — a large funnel at the top produces a small, high-quality set of candidates for serious work. Binjiang Service Group (3316.HK) passed the initial screen with sub-10× P/E, 37.6% ROE, zero financial debt, and positive free cash flow across every reported period. That is a relatively rare combination — it signals something worth understanding before a valuation line is drawn.
AI tools have substantially improved screening capacity. What previously required building and maintaining a proprietary database of global financial data can now be approached through AI-assisted financial data extraction, allowing a solo analyst to screen across 40+ global exchanges without a Bloomberg terminal or a team of data engineers. The structural reasons why so many of these companies remain mispriced despite their quality are covered in Why Global Small-Caps Are Systematically Mispriced.
Step 02 — Business quality assessment
Once a candidate passes the quantitative screen, the real analysis begins: understanding what the business actually does, how it makes money, and whether its competitive position is durable.
The core question is always: why does this business earn more than its cost of capital, and will it continue to do so? The answer must come from primary sources — annual reports, earnings transcripts, regulatory filings, and competitor data. No secondhand summary is sufficient.
The specific factors to evaluate include:
- The source of competitive advantage (cost structure, switching costs, network effects, intangible assets, regulatory moat)
- Industry dynamics — concentration, growth rate, pricing power across the value chain
- Customer and supplier relationships — are they stable or fragile?
- Geographic market position — is the business dominant in a niche, or marginal in a large market?
- Technology exposure — is the business's competitive position vulnerable to digital disruption or AI displacement?
For global small-caps, business quality assessment often requires reading documents in foreign languages and understanding regulatory environments in unfamiliar jurisdictions. This is one of the primary barriers that keeps professional analysts away — and therefore one of the primary sources of edge for those willing to do the work. The specific factors evaluated at each step are formalized in Sifter Research's 82-question Multibagger Checklist.
"The question is always the same: why does this business earn more than its cost of capital, and will it continue to do so? The answer must come from primary sources."
Step 03 — Management and incentive evaluation
A business with a strong competitive moat and mediocre capital allocation can destroy a decade of compounding potential. Management evaluation is therefore not optional — it is the second most important step after business quality.
The primary question is alignment: does management think and act like an owner? The evidence comes from compensation structure, insider ownership levels, the history of capital allocation decisions (buybacks vs. dividends vs. acquisitions), and how management has communicated with shareholders over time. A CEO who owns 20% of the company and has never issued stock options behaves differently from one who collects annual bonuses regardless of performance.
Track record matters more than stated intention. Look at how capital was deployed in the five years prior to your analysis: Were acquisitions value-creative or empire-building? When business was slow, did management cut costs or preserve the balance sheet? Did they buy back stock at sensible prices or issue shares at the top? The pattern is almost always consistent with future behaviour.
For small-caps, management quality is often the swing factor. A mediocre business run by a genuinely exceptional allocator can still generate excellent returns. An outstanding business run by an extractive manager will eventually disappoint.
Step 04 — Scenario-based valuation
Once business quality and management alignment have been established, the question shifts to price. The standard mistake in equity analysis is to build a single model with a single set of assumptions and call the output a "fair value." This creates false precision and ignores the fundamental uncertainty of forecasting.
A more honest and useful approach is to build three explicit scenarios — bear, base, and bull — with different assumptions in each, and evaluate whether the margin of safety is sufficient in the base case to justify investment even if the bear case materialises.
Each scenario should specify:
- Revenue growth assumptions over 5 years, with justification
- Margin trajectory — expansion, compression, or stability, and why
- Capital intensity and reinvestment rate
- Terminal value methodology and assumptions
- Discount rate rationale (not just "WACC from a calculator")
The output is a range of intrinsic value estimates. The investment is only attractive if the current price offers a meaningful discount to the base case, with enough protection that the bear case does not result in permanent capital loss. Margin of safety before upside imagination. Always. One concrete illustration: FILA S.p.A. (FILA.MI) holds a controlling stake in a listed Indian subsidiary (Camlin) currently valued at approximately €348m. FILA's total enterprise value is €470m. A standard DCF applied to the consolidated business misses this entirely — the sum-of-parts analysis changes the picture completely, implying the European stationery operations are being purchased at a fraction of replacement cost.
Step 05 — Risk assessment
The final step is to invert the investment case: assume the investment fails, and identify the most likely reasons why. This is more useful than a standard "risk factors" section because it forces specificity about what the thesis actually depends on.
The key question is: what has to go right for this investment to fail? If the answer is "the entire business model has to collapse," the risk profile is different from "the management team has to make a single major acquisition error." Understanding the failure modes also helps calibrate position sizing and monitoring priorities after the investment is made.
Common risk categories in global small-caps include:
- Currency risk — especially for businesses with revenues in one currency and costs in another
- Minority shareholder risk — in controlled companies, does management have incentives to tunnel value to the controlling shareholder?
- Accounting quality — are the financial statements genuinely transparent, or do they require adjustment to reflect economic reality?
- Liquidity risk — can the position be exited if the thesis changes? And over what time horizon?
- Concentration risk — is the business highly dependent on a single customer, geography, or product?
This five-step framework is applied in every Sifter Research report. The published reports on FILA.MI and 3316.HK show the full process documented, with all assumptions visible and a final judgment that carries the author's name.
