Research Tools

FactSet and Capital IQ are the two dominant financial data and research platforms used for company financials, comps screening, precedent transaction databases, and ownership data, most banks license one or both, and analysts are expected to become fluent navigating whichever their desk uses within the first weeks on the job. Bloomberg Terminal is more heavily weighted toward real-time market data, fixed income, and trading-desk functions, though it's also used in banking for market data, comps, and company research, particularly its widely used company-overview and comps functions.

Each platform has its own query language, export conventions, and quirks, and genuine fluency, knowing exactly which function pulls which data fastest, is built through repeated daily use rather than a single training session, which is why new analysts are expected to spend real time in their first weeks simply exploring the platform's functionality.

Every figure pulled from a data platform should ideally be cross-checked against the primary source (the company's own SEC filing or press release) for anything that will appear in a client-facing document, since data platforms occasionally have errors, lag on newly reported figures, or apply their own standardized adjustments that may not match the specific definition needed for a given analysis. Tracking exactly where each number came from, which platform, which specific filing or data point, is essential discipline, since a senior banker or client questioning a number needs a clear, immediate answer about its source.

This sourcing discipline connects directly to the error-checking habits covered elsewhere: a wrong or stale data pull is one of the most common sources of an embarrassing error in a live deck, and verifying against a primary source is the most reliable safeguard against it.

Rather than manually re-pulling the same data every time it's needed, experienced analysts build reusable templates and saved screens/searches within these platforms, a saved comps screen with the right filters already applied, an Excel plugin-linked template that refreshes financial data automatically, which dramatically speeds up recurring tasks like updating a comps page or screening for potential acquisition targets.

Most platforms offer a direct Excel plugin (allowing live-linked data pulls that update within a model rather than manual copy-paste), and learning to use that plugin fluently is one of the highest-leverage specific skills a new analyst can build in their first weeks, since it turns what would otherwise be a repetitive, error-prone manual task into a fast, reliable, repeatable one.

A screen is a filtered search across a data platform's universe of companies, built from specific criteria, industry classification, size (revenue or market cap range), geography, growth rate, margin profile, used to generate a target list of potential comps, acquisition targets, or buyers for a given mandate. Building a good screen starts from a clear question (what specifically is this list meant to identify) and iterates: an initial broad screen is almost always too wide, and the real skill is progressively narrowing the criteria based on judgment about what's actually relevant, not just mechanically applying filters.

A raw screen output should never be delivered as-is; it needs a manual review pass to remove companies that technically match the filter criteria but don't make business sense in context (a company classified in the right industry code but structurally different in practice), which is exactly the kind of judgment a purely mechanical screen can't apply on its own.

Benchmarking compares a company against a defined peer set across operating and financial metrics, margins, growth rates, capital intensity, leverage, to assess relative performance and identify what's driving any gaps, often used in a pitch to show a client where they stand versus competitors, or to support a valuation argument (why a company deserves a premium or discount multiple versus peers). A useful benchmarking analysis goes beyond just presenting the numbers side by side; it explains, briefly, why any notable gaps exist, a structural difference in business model, a temporary operational issue, a difference in scale.

Building this well requires the same peer-selection judgment as a comps analysis (discussed in the valuation lessons), since a benchmarking exercise built on a poorly chosen peer set produces a misleading comparison regardless of how carefully the underlying numbers were pulled.

The final step of most screening work is converting a raw data output into a clean, prioritized, client-ready list, typically ranked or bucketed by relevance (strategic fit, size, likely appetite for a transaction) with a brief rationale for each entry, not a raw, unannotated spreadsheet export. This is often one of the earliest substantive analytical deliverables a new analyst produces independently, and doing it well (clear prioritization, defensible logic, clean formatting) is a meaningful early signal of judgment to the staffer reviewing it.

Because screening work often directly feeds a buyer list or target list for a live process, accuracy and defensibility matter just as much here as in any valuation work, a client or senior banker will ask why a specific company is or isn't on the list, and the analyst needs a clear, immediate answer.