Credit Analysis and Lending: The Best Books, in Order
Credit analysis is one question asked at four levels: will this borrower repay, what happens if it does not, what does that mean for a portfolio, and what does it mean for the lender's own solvency. This path follows that escalation. It starts with the financial statements — you cannot analyse credit you cannot read — then corporate credit, then the bank's own risk position, and ends with the quantitative models. It assumes accounting and basic statistics, and it is a professional reading list rather than an investment recommendation.
The Statements Come First
IntermediateRead a set of accounts for the purposes of a lender rather than an equity investor, and know the mechanics of a lending decision end to end.
▸ Study plan for this stage
Pace: 8-10 weeks. Fridson and Alvarez's Financial Statement Analysis (400 pp) over three weeks with a real annual report open beside it, Sathye, Bartle and Boffey (579 pp) over three, and Koch and MacDonald's Bank Management (717 pp) read selectively over three - it is a full banking course text and only
- Cash flow versus reported earnings, and why a lender underwrites the first
- The specific manipulations Fridson catalogues: revenue recognition timing, capitalised costs, reserve releases, related-party transactions
- Off-balance-sheet obligations - leases, guarantees, receivables sales, pension deficits - and where to find them
- The working capital cycle and cash conversion, and how a growing profitable borrower runs out of money
- Coverage and leverage measures a lender actually uses: DSCR, fixed-charge coverage, funded debt to EBITDA
- The lending process end to end - origination, credit assessment, documentation, security and perfection, monitoring, workout and recovery
- Loan pricing as a function of the bank's own funding cost, not just the borrower's risk (Koch and MacDonald)
- Collateral and security: what a lien is worth in a liquidation versus on the appraisal
- How do you reconstruct free cash flow available for debt service from a published set of accounts?
- What does a lender look for in a set of statements that an equity analyst would ignore entirely?
- By what mechanisms can a profitable, growing company default within twelve months?
- How does the bank's funding profile change the price of an otherwise identical loan?
- Which of these books' regulatory and provisioning assumptions have been superseded, and what replaced them?
- Take one of Fridson and Alvarez's worked revenue-recognition cases and redo the same adjustment on the statements of a company you choose, showing the restated cash flow
- Build DSCR and fixed-charge coverage as Fridson defines them from one real 10-K, then recompute them treating all operating leases as debt and note the difference
- Walk one hypothetical borrower through Sathye, Bartle and Boffey's full origination-to-recovery sequence and write the credit memo their template calls for
- Price a loan using Koch and MacDonald's spread-over-funding-cost model, then mark which of its regulatory capital assumptions Basel III has replaced
- Find the off-balance-sheet disclosures in one annual report and quantify each against the on-balance-sheet debt
Next up: With the statements and the mechanics of a lending decision in hand, the next stage is the framework rating agencies use to turn those numbers into a credit opinion on a specific corporate borrower.
Fridson and Alvarez, and the right first book because it is written by credit people: it is about how statements mislead — revenue recognition, off-balance-sheet obligations, the gap between reported earnings and cash. Read it before anything else here.
Sathye, Bartle and Boffey, a course text covering the full lending process — origination, assessment, documentation, security, monitoring and recovery. Read it second for the procedural picture; it is the most straightforwardly instructional book on this path.
Koch and MacDonald's standard text, included for the parts most credit books skip: asset-liability management, funding costs, and how loan pricing has to relate to the bank's own balance sheet. Read it third so that a credit decision stops looking like a standalone judgement.
Corporate Credit
BeginnerAnalyse a corporate borrower the way a rating agency does — business risk, financial risk, structural subordination, covenants and recovery.
▸ Study plan for this stage
Pace: 9-10 weeks. Ganguin and Bilardello's Fundamentals of Corporate Credit Analysis (428 pp) over three weeks, Glantz and Mun's Credit Engineering for Bankers over four with a spreadsheet open, Colquitt's Credit Risk Management (372 pp) over two. Note who is speaking: Ganguin and Bilardello wrote from in
- The business-risk-plus-financial-risk matrix and how the two axes combine into an indicative rating
- Industry risk assessment: cyclicality, barriers to entry, capital intensity, and competitive position within the industry
- Structural subordination - where in a group the debt sits and which cash flows it can reach
- Covenants: maintenance versus incurrence, the standard baskets, and what a covenant actually gives a lender
- Recovery analysis and loss given default by instrument and by security
- Through-the-cycle rating philosophy and why agency ratings move less than market spreads
- Cash-flow modelling and sensitivity: which two or three variables actually move the credit
- The step from single-name analysis to portfolio thinking, which Colquitt introduces
- How do business risk and financial risk combine in the Ganguin framework, and what happens when they conflict?
- Where does structural subordination bite in a holding-company group, and how would you document around it?
- Which covenants would have caught a specific deterioration early, and which would only have triggered after the fact?
- What is the recovery difference between senior secured and senior unsecured in a real default, and what drives it?
- Which of Colquitt's risk-transfer instruments behaved as described in 2008, and which did not?
- Run one real issuer through Ganguin and Bilardello's business-risk and financial-risk grid, land on an indicative rating, then compare it with the agency's published rating and account for the gap
- Build Glantz and Mun's cash-flow model for that same issuer and stress the two variables the book identifies as most sensitive, recording the rating implication of each
- Draft a covenant package using Glantz's constructions, then test it against the issuer's last three years of numbers and see whether it would ever have tripped
- Map the legal entity structure of one group with operating subsidiaries and mark where each tranche of debt sits relative to the cash flows
- Read Colquitt's securitisation chapter and mark every claim about the market that the 2008 crisis falsified
Next up: Corporate analysis assumes the lender is a stable observer; the next stage turns the same tools on the lender itself, which is a genuinely different and harder problem.
Ganguin and Bilardello, written from inside a rating agency and the clearest exposition of the business-risk-plus-financial-risk framework that most corporate credit work still uses. The catalogue files it under the bare title. Start the corporate work here.
Glantz and Mun on cash flow modelling, ratio analysis, industry comparison and covenant construction, with a heavier quantitative slant than Ganguin. Read it second as the modelling counterpart.
A practitioner's treatment covering the underwriting process, credit derivatives, securitisation and portfolio management in one volume. Useful here as the bridge from single-name analysis to the portfolio question that follows.
The Lender's Own Position
BeginnerMove from the borrower to the institution — how a bank measures, prices and provisions for credit risk across a whole book of loans.
▸ Study plan for this stage
Pace: 10-12 weeks. Golin and Delhaise's The Bank Credit Analysis Handbook (800 pp) over five weeks with a bank's annual report to hand, van Greuning's Analyzing and Managing Banking Risk (367 pp) over three, and Caouette, Altman, Narayanan and Nimmo's Managing Credit Risk (528 pp) over four. Van Greuning
- Why a bank is a different borrower: opaque asset quality, leverage that would be fatal in any other industry, and dependence on confidence
- The bank analysis ratio set - NPL ratio and coverage, capital adequacy, liquidity, net interest margin, cost-to-income - and the CAMELS framing
- Funding structure and run risk: deposit stability, wholesale reliance, maturity transformation
- The supervisory perspective: corporate governance, board risk oversight, and the internal control layers van Greuning organises around
- Portfolio credit risk: concentration, correlation, and how much diversification a marginal name actually buys
- Credit risk transfer instruments - CDS, securitisation, sub-participation, loan sales - and their basis risks
- Economic capital, RAROC and risk-adjusted pricing as the link between a single loan and the institution's solvency
- Provisioning and reserve adequacy, and how the accounting standard chosen changes the reported number
- What can you not learn about a bank's asset quality from its published accounts, and what proxies does Golin offer instead?
- Why does the same leverage that would be reckless for a corporate borrower work for a bank, and when does that stop being true?
- How does van Greuning's supervisory framework differ in purpose from Golin's analytical one, and where do they ask the same question?
- At what point does adding names to a loan portfolio stop reducing risk, and what determines that point?
- Which of Caouette's risk-transfer mechanisms move risk and which only move where it is recorded?
- Pull one bank's published accounts and complete Golin's ratio template in full - asset quality, capital, liquidity, earnings - then mark which of his ratios Basel III has redefined (CET1, LCR, NSFR)
- Apply van Greuning's governance and risk-management questionnaire to that same bank's annual report and note every question the disclosures cannot answer
- Build a hypothetical twenty-name loan book and compute concentration using Caouette's measures, then add a twenty-first name and quantify the diversification benefit
- Take one bank that failed after 2008 and run Golin's framework on its last pre-failure accounts, recording which ratios flagged and which did not
- Compare a loan's economics under RAROC pricing against a flat spread and identify the borrower type each approach favours
Next up: The portfolio question raised here has a formal answer, and the last stage works through the models that supply it - along with what each of them assumes.
Golin and Delhaise on analysing banks as borrowers, which is a genuinely different problem: opaque asset quality, leverage that would be fatal anywhere else, and dependence on confidence. Read it first here; it inverts everything the corporate stage taught.
The World Bank framework for bank risk assessment, organised around corporate governance and the supervisory perspective. Later editions were retitled Analyzing Banking Risk and appear as a separate catalogue record; buy one edition, not both.
Caouette, Altman, Narayanan and Nimmo — the standard treatment of credit as a portfolio problem, covering concentration, correlation, transfer and the instruments used to move risk off a balance sheet. Read it last in this stage; it is the direct setup for the quantitative work.
The Quantitative Layer
BeginnerWork with the formal models — default probability, loss given default, correlation and portfolio optimisation — and know what each one assumes.
▸ Study plan for this stage
Pace: 16-20 weeks and the hardest stage by a wide margin. Read Saunders and Allen's Credit Risk Measurement (328 pp) first over three weeks - it is a survey written to orient you before the primary literature and requires only comfort with statistics. Lando's Credit Risk Modeling (324 pp) is the graduate
- Structural models: Merton's option-theoretic default, distance to default, and the KMV EDF implementation
- Reduced-form and intensity-based models, and why the two families disagree about short-horizon spreads
- Rating transition matrices, the Markov assumption behind them, and the evidence against it
- Default correlation, copulas, and the specific role the Gaussian copula played in structured credit before 2008
- Loss given default as a modelled quantity rather than an assumption, and its correlation with default rates
- Duration times spread as an empirically tested risk measure for corporate bond portfolios (Dynkin et al.)
- Economic capital allocation and portfolio optimisation under a credit loss distribution
- Model validation, data quality and the organisational politics of an internal transfer-pricing regime (Bohn and Stein)
- What does each model family assume about the default event itself, and which assumption fails first in practice?
- How is distance to default computed from equity market data, and what does it require you to believe about the equity market?
- Why does a Gaussian copula understate joint defaults in a downturn, and what alternatives does the literature propose?
- What does duration times spread claim over conventional spread duration, and what is Dynkin's evidence?
- Which parts of these models survived 2008, and how would you know from the books themselves?
- Implement the Merton model on one listed firm following Saunders and Allen's exposition and back out distance to default, then compare it with the firm's actual rating
- Work Lando's derivation of the survival probability under a Cox intensity model line by line, and reproduce the numerical example he gives
- Reproduce Dynkin, Hyman, Ben Dor and Phelps's duration-times-spread test on a corporate bond index sample and check whether their result holds on your data
- Run Bohn and Stein's model validation checklist against the Merton implementation you built and record every point where your data fails their standard
- Read Gregoriou's correlation-trading and CDO chapters noting their pre-crisis vintage, and write down which of their assumptions the 2008 experience breaks
Next up: This is the end of the path; from here the live material is regulatory rather than academic - the Basel framework documents, CECL and IFRS 9 guidance, and the agencies' own published criteria, all of which these books cite but none reproduce.
Saunders and Allen's survey of the modelling approaches — structural, reduced-form, actuarial — written to be read before the primary literature. Start the quantitative stage here; it maps the territory without committing you to one school.
The rigorous academic treatment: intensity-based models, rating transitions, dependent defaults, and the pricing of credit derivatives. Requires stochastic calculus. This is the reference the practitioner books cite.
Bohn and Stein on implementing a Moody's KMV-style framework inside a real institution — data problems, model validation, and the organisational politics of a transfer-pricing regime. The most practical book at this level.
Dynkin, Hyman, Ben Dor and Phelps from the Barclays research group, on constructing and hedging corporate bond portfolios with empirically tested measures such as duration times spread. Read it for the buy-side view of the same models.
An edited collection covering securitisation, CDOs, correlation trading and regulatory capital. Uneven, as edited volumes are, and worth having for the chapters on instruments the single-author books treat briefly. Note that much of it predates the 2008 crisis; read the correlation chapters with that in mind.
Discussion
Keep reading
Paths that share books, cover the same subject, or open a related topic.
