Credit Analysis — Notes
Core mental model
•A credit analyst's job: decide/recommend whether a borrower is likely to repay, and on what terms — a specific analytical skill, distinct from general banking/regulatory knowledge (covered separately in this platform's Banking & RBI content).
•The 5 Cs (Character, Capacity, Capital, Collateral, Conditions) must be weighed TOGETHER, not checked independently — a borrower can be strong on some and weak on others, and the lending decision depends on the combined picture.
•Read the three financial statements as ONE connected story, not three separate checklists — income statement growth should be cross-checked against balance sheet (receivables) and cash flow (actual cash conversion), since revenue growth sitting in receivables rather than converting to cash is a real red flag only visible when reading them together.
DSCR and cash flow quality
•DSCR = Cash Available for Debt Service ÷ Total Debt Service. Below 1.0 means operating cash flow alone doesn't cover obligations — a genuinely more severe situation than "tight" cash flow, not just a lower number on the same scale.
•Adjust operating cash flow for NECESSARY maintenance capex (not growth capex) before using it in DSCR — using raw operating cash flow without this adjustment overstates genuine debt-service capacity, a common calculation trap.
•Two businesses with identical headline operating cash flow can have very different quality — examine the COMPOSITION (genuine collections vs. payables-stretching or one-time items), not just the total.
Early warning signals
•Monitor proactively well before the 90-day NPA mark: DSCR trending down across consecutive quarters, deteriorating receivables aging, delayed financial reporting, proactive covenant-waiver requests, key management departures.
•Distinguish a single weak data point (explainable, temporary) from a genuine multi-quarter TREND across several signals together — waiting for actual default means missing the entire proactive-intervention window.
Industry/Conditions and Collateral
•Industry-level risk (demand trends, regulatory change, competitive dynamics, input cost exposure) is a distinct analytical layer — even a strong individual borrower can face real risk from industry-wide headwinds that historical borrower-specific financials don't yet reflect.
•Collateral book value ≠ realizable value in a genuine default/recovery scenario — inventory and receivables typically realize significantly less under distressed liquidation (and receivables carry correlated risk with the borrower's own distress); real estate/fixed assets are generally more reliable. This is why LTV ratios differ meaningfully by collateral type.
Rating migration and restructuring
•A rating downgrade reflects a genuine reassessment of default probability, not just a label change — monitor external rating TRAJECTORY as an independent cross-check against internal analysis; a borrower with stable internal metrics but a declining external rating warrants specific investigation.
•Restructuring decisions require distinguishing underlying CAUSE, not just responding to the symptom: temporary/addressable stress → extend/modify terms; ongoing risk needing better protection → add covenants/security; fundamentally non-viable business → recovery/enforcement action. Defaulting to simple extension regardless of cause risks "evergreening" a non-viable loan.
Portfolio-level and stress-test thinking
•Individual borrower analysis is necessary but insufficient — portfolio-level concentration risk (industry/geography/single-borrower exposure) and correlation risk (do borrowers' defaults move together under shared stress) require separate analysis, since individually sound underwriting doesn't protect against aggregate, correlated exposure.
•Stress testing (modeling adverse scenarios — rate rises, sector downturns) surfaces vulnerabilities current-conditions analysis alone can't reveal, and is often what exposes hidden portfolio correlation.
Genuine judgment over mechanical rules
•Real credit skill means understanding WHY thresholds exist and recognizing when a generic rule doesn't fit a specific industry's normal financial structure (e.g., capital-intensive infrastructure businesses have different normal DSCR profiles) — mechanically applying universal thresholds risks incorrectly declining genuinely sound credits.
•Governance/qualitative factors (related-party transactions, management transparency) matter as much as quantitative ratios and are systematically missed by ratio-only analysis — the "Character" C deserves genuine analytical attention, not a quick formality check.