Credit Appraisal, Underwriting and Decision-Making

Introduction and Context

Credit appraisal is the disciplined process through which a lender determines whether a borrower should receive credit, how much credit is appropriate, on what terms, and with what controls. It is more than an assessment of repayment capacity. It is an integrated judgement involving business quality, financial strength, management conduct, industry conditions, transaction structure, collateral, legal enforceability, risk appetite and expected return.

Underwriting converts evidence into a risk decision. A credit proposal should therefore be persuasive not because it contains extensive information, but because it demonstrates a logical connection between the borrower’s purpose, projected cash flows, requested facility, identified risks, mitigants and proposed monitoring arrangements. The central question is not simply whether the borrower has assets, but whether the proposed exposure will be repaid from reliable sources within the agreed tenor and under reasonably adverse conditions.

For banks, weak appraisal can produce adverse selection, fraud losses, concentration risk and regulatory criticism. Excessively conservative appraisal, however, can reject viable businesses and reduce risk-adjusted growth. Professional credit decision-making therefore balances commercial opportunity with disciplined scepticism.

Learning Objectives

By the end of this chapter, the learner should be able to:

1. Explain the purpose, stages and evidence requirements of credit appraisal and underwriting.
2. Integrate qualitative, quantitative and external information into a credit proposal.
3. Analyse repayment capacity, leverage, liquidity, profitability and debt-service metrics.
4. Conduct scenario and sensitivity analysis and interpret its limitations.
5. Design appropriate risk mitigants, covenants and monitoring conditions.
6. Manage policy exceptions and distinguish sanction, decline and escalation decisions.
7. Prepare a defensible recommendation consistent with governance, ethics and risk appetite.

1. The Architecture of a Credit Decision

A sound credit decision has six linked components:

The traditional “five Cs” remain useful: character, capacity, capital, collateral and conditions. At postgraduate level, these should not be treated as a checklist. They are dimensions of a causal analysis. For example, collateral may reduce loss severity but does not normally cure weak repayment capacity. Similarly, a strong financial position at the application date may not protect a borrower exposed to a severe commodity-price shock.

The lender should distinguish three sources of repayment:

• Primary source: operating cash flow or the borrower’s regular income.
• Secondary source: refinancing, sale of assets or sponsor support.
• Tertiary source: enforcement and liquidation of collateral or guarantees.

A credit proposal is strongest when repayment depends primarily on the first source. Reliance on secondary or tertiary sources should be explicit, tested and supported by evidence.

2. From Application to Decision

2.1 Origination and initial screening

Initial screening establishes whether the request fits the bank’s strategy, target market, risk appetite and legal permissions. Basic questions include:

• Is the customer identifiable and acceptable under customer due-diligence standards?
• Is the purpose permissible and economically credible?
• Is the requested product suitable for the cash-flow pattern?
• Does the exposure exceed sector, geography, group or single-obligor limits?
• Are there immediate red flags such as fraud indicators, sanctions concerns or unexplained repayment stress?

Screening is not a substitute for underwriting. It is a filter that prevents scarce analytical resources from being applied to clearly unacceptable cases.

2.2 Information collection and verification

Information should be relevant, current, independently verifiable and internally consistent. Typical evidence includes audited financial statements, management accounts, bank statements, tax filings, ageing schedules, borrowing statements, ownership records, contracts, insurance documents, valuations and credit-bureau information.

The underwriter should reconcile evidence rather than merely collect it. For example, reported sales should be compared with bank credits, tax returns, receivables movements and inventory purchases. A rapid rise in reported revenue accompanied by falling cash collections may indicate aggressive revenue recognition, customer concentration or working-capital stress.

2.3 Analysis and risk grading

Analysis converts data into an assessment of probability of default, loss given default, exposure at default and expected loss. A simplified expected-loss relationship is:

Here, EL is expected loss, PD is probability of default, LGD is loss given default as a proportion of exposure, and EAD is exposure at default. The relationship assumes that the three quantities are defined on a compatible time horizon and that the product is an appropriate approximation. In advanced models, exposure timing, recoveries, discounting and dependence between variables may require more complex estimation.

Risk grades should be evidence-based and distinguish borrower risk from facility risk. A secured facility may have lower expected loss than an unsecured facility to the same borrower, but collateral should not be used to conceal a high probability of default.

2.4 Structuring and recommendation

The recommended structure should reflect the purpose and cash-flow cycle. A seasonal working-capital requirement may justify a revolving facility, whereas an asset purchase generally requires a term loan with an amortisation profile aligned to asset life and cash generation. Maturity mismatch is a common source of refinancing risk.

The recommendation should state the requested amount, proposed amount, tenor, repayment schedule, pricing, security, covenants, conditions precedent, monitoring frequency and rationale. It should also identify matters that remain uncertain.

2.5 Independent challenge and approval

Credit authority should be proportionate to risk and independent of sales incentives. The approving authority should challenge assumptions, not simply verify document completeness. Material deviations from policy should be escalated to the appropriate committee or higher authority.

3. Qualitative Appraisal

Quantitative statements are historical and often incomplete. Qualitative appraisal provides the context needed to interpret them.

Management and governance

The assessment should consider competence, succession, internal controls, financial discipline, related-party dealings and willingness to disclose adverse information. A technically capable owner may still present high credit risk if decision-making is concentrated, records are unreliable or personal and business finances are mixed.

Business model and competitive position

The underwriter should identify how the borrower creates value, who pays, how margins are generated and what could disrupt the model. Important considerations include customer concentration, supplier dependence, switching costs, pricing power, barriers to entry, technology risk and vulnerability to regulation.

Industry and macroeconomic conditions

Sector analysis should address demand, input costs, cyclicality, regulation, foreign-exchange exposure and competitive intensity. Industry growth alone is not evidence of credit quality. A rapidly expanding sector may contain excess capacity and declining margins.

Conduct and integrity

Repayment history, returned payments, tax arrears, litigation, adverse media and unexplained changes in ownership can be more informative than a single financial ratio. Conduct information must be handled fairly, lawfully and with an opportunity for clarification. Allegations should not automatically be treated as established facts.

Environmental, social and governance considerations

ESG factors become credit factors when they affect cash flow, asset value, legal liability, reputation, access to markets or continuity of operations. Examples include environmental remediation obligations, unsafe labour practices and governance failures. Assessment must be relevant and evidence-based rather than a symbolic scoring exercise.

4. Quantitative Appraisal

4.1 Quality of earnings and cash flow

Profit is not equivalent to cash available for debt service. The analyst should examine revenue recognition, gross margins, operating expenses, non-recurring items, working-capital movements, capital expenditure and taxes.

A simplified cash-flow bridge is:

This is an analytical approximation, not a substitute for a full cash-flow statement. EBITDA means earnings before interest, tax, depreciation and amortisation. The approximation assumes that other operating cash items are either immaterial or separately adjusted. Capital expenditure and financing flows must be considered before assessing total cash available for debt repayment.

4.2 Key ratios

The ratios require consistent definitions. For example, DSCR may be calculated using EBITDA less taxes, maintenance capital expenditure and working-capital requirements, rather than EBITDA alone. The policy definition must be stated in the proposal.

For a simple annual example, if cash available for debt service is 12 million currency units and scheduled principal plus interest is 8 million, then:

This indicates 1.50 units of defined cash flow for each unit of scheduled debt service. It does not prove that future cash flow will remain stable.

4.3 Debt capacity and repayment structure

Debt capacity depends on sustainable cash generation, not merely the maximum amount a lender is willing to advance. A facility should be sized after considering base-case and stressed cash flow, existing obligations, dividend leakage, contingent liabilities and plausible refinancing assumptions.

A level-payment loan can be analysed using the present-value relationship:

A is the periodic payment, P is the principal, r is the interest rate per payment period expressed as a decimal, and n is the number of payments. The formula assumes a fully amortising loan, equal payments, a constant periodic interest rate and payments at the end of each period. Units must match: a monthly rate requires a monthly payment frequency and a number of months.

For a 10 million loan at 10% per year with annual payments over three years, P = 10, r = 0.10 and n = 3. The annual payment is approximately 4.02 million. The figure should be recalculated using the bank’s approved method, especially where rates are floating, payments are monthly or fees are capitalised.

4.4 Collateral and recovery

Collateral analysis asks four questions: Is the asset owned by the correct party? Is the security legally perfected? What is the realistic net recovery value? How long and how costly will enforcement be?

A prudent collateral value is not automatically the market valuation. Haircuts may be required for forced-sale conditions, volatility, illiquidity, depreciation, taxes, enforcement costs and ranking of claims. Guarantees should be assessed for the guarantor’s independent capacity and legal enforceability, not merely for the guarantor’s net worth on paper.

5. Scenario and Sensitivity Analysis

Base-case projections are hypotheses, not facts. Scenario analysis tests coherent combinations of adverse conditions, while sensitivity analysis changes one or more assumptions to observe the effect on outcomes.

Sensitivity analysis

A borrower may be sensitive to revenue, margin, interest rate, collection period, foreign-exchange rate or input prices. If revenue falls by 10% while costs remain unchanged, the effect on cash flow may be much larger than 10% because fixed costs and debt service remain payable.

Scenario analysis

A useful scenario set may include:

• Base case: management’s plan adjusted for independent evidence.
• Moderate downside: lower sales, slower collections and modest margin compression.
• Severe but plausible downside: simultaneous demand decline, cost inflation, higher interest rates and restricted refinancing.

The analysis should identify breach points, not merely produce attractive charts. Relevant outputs include minimum cash balance, DSCR, leverage, covenant headroom, liquidity gap and repayment timing.

For example, suppose base cash available for debt service is 12 million and debt service is 8 million. Under a downside case, cash flow falls to 7 million while debt service rises to 8.8 million:

A ratio below 1.00× indicates that defined cash flow is insufficient to meet scheduled debt service in that scenario. The decision is not automatically decline; the analyst must determine whether the scenario is plausible, whether the structure can be changed, and whether credible mitigants exist.

Important limitations include model risk, uncertainty in management forecasts, correlations between assumptions and the possibility that historical relationships fail during stress. Sensitivity analysis should therefore support, not replace, judgement.

6. Risk Mitigation and Facility Design

Mitigants should address specific risks and have an identifiable mechanism. “Strong relationship” is not a mitigant.

Covenants may be affirmative, negative or financial. Examples include delivery of audited statements, restrictions on additional borrowing, maintenance of a minimum DSCR, limits on dividends and requirements to maintain insurance. Covenants must be measurable, linked to risk and capable of being monitored.

Conditions precedent should be completed before drawdown when they protect the bank from fundamental risks. Examples include executed security documents, evidence of equity contribution, insurance, permits and satisfactory legal opinions. Post-disbursement conditions should have deadlines, responsible parties and consequences for non-compliance.

7. Exceptions, Escalation and Decision Discipline

An exception is a departure from approved policy, product parameters, delegated authority or standard documentation. Exceptions are not necessarily unacceptable, but they increase the burden of explanation and control.

A defensible exception memorandum should state:

1. The exact policy requirement and proposed deviation.
2. The commercial and credit rationale.
3. Incremental risk created by the deviation.
4. Compensating mitigants and their enforceability.
5. Approval authority and expiry or review date.
6. Monitoring requirements and consequences if conditions deteriorate.

The “compensating factors” principle should not become a licence for unsupported optimism. Multiple weaknesses may be correlated. For example, weak financial reporting, owner concentration and customer concentration may all worsen simultaneously under stress.

Sanction

Sanction is appropriate when repayment capacity is acceptable, the structure fits the purpose, risks are within appetite, conditions are clear and the evidence is sufficient. Approval should be conditional where unresolved matters are documentary or operational rather than fundamental.

Decline

Decline is appropriate when primary repayment is implausible, information is materially unreliable, the purpose is unacceptable, legal or conduct concerns cannot be resolved, risk exceeds appetite, or proposed mitigants are ineffective. A decline should be documented respectfully and based on legitimate risk criteria.

Escalation

Escalation is appropriate when the case is strategically important but outside delegated authority, contains material uncertainty, involves a significant exception, or requires specialist legal, technical, compliance or restructuring input. Escalation is not indecision; it is a controlled response to complexity.

8. The Credit Proposal as an Argument

A high-quality proposal has an executive summary that allows the decision-maker to understand the case quickly. It should then provide:

• borrower and ownership profile;
• facility purpose and proposed structure;
• financial analysis and repayment sources;
• qualitative risk assessment;
• downside scenarios and covenant headroom;
• collateral and legal position;
• risks, mitigants and residual risk;
• exceptions and approval requirements;
• monitoring plan; and
• clear recommendation with conditions.

The recommendation should distinguish facts, assumptions, judgements and unknowns. Language such as “management expects” should not be presented as independently verified fact. Contradictory evidence should be disclosed rather than removed from the narrative.

Illustrative Case: Meridian Components Ltd

Background

Meridian Components Ltd, a fictional manufacturer of industrial parts, requests a 30 million term loan to expand production and a 10 million revolving working-capital line. It has operated for eight years and supplies three large customers. The owners have contributed 12 million of new equity.

Problem

The borrower reports revenue growth of 22% and EBITDA of 18 million. However, receivables days have increased from 52 to 79, one customer represents 46% of sales, and the expansion depends on imported machinery priced in foreign currency. Existing debt is 48 million. Management requests a five-year term loan with a two-year principal moratorium.

Analysis

The reported growth is plausible because signed customer contracts support additional volume, but the largest contract permits termination on 60 days’ notice. The increase in receivables days creates a working-capital requirement and may indicate weaker collection quality. The machinery supplier requires a 30% deposit, and the borrower has no documented foreign-exchange policy.

Base-case cash available for debt service is estimated at 21 million, against projected debt service of 15 million, giving:

Under a moderate downside involving a 12% revenue reduction, a 2 percentage-point margin decline and slower collections, cash available for debt service falls to 13 million. Debt service increases to 16 million because of higher interest rates:

The downside exposes a repayment shortfall. Collateral consists of machinery and land. The machinery’s forced-sale value is uncertain because it is specialised; the land is more reliable but subject to a prior mortgage.

Approach and solution

The recommended structure is to reduce the term loan to 24 million, require the owners’ equity to be injected before the first draw, and stage disbursements against verified machinery invoices. The revolving line should be subject to monthly borrowing-base reporting, receivables ageing and exclusion of overdue or disputed invoices.

Additional conditions include assignment of key customer contracts where legally possible, a documented foreign-exchange policy, a minimum quarterly DSCR of 1.15× tested on an agreed definition, restrictions on dividends and no additional secured debt without consent. The bank should obtain an independent valuation and confirm first-ranking security or explicitly price and approve the ranking risk.

Because customer concentration and contract cancellation remain material, the case should be escalated to the credit committee rather than approved solely under normal delegated authority. The proposal should state that the bank is relying on operating cash flow, not liquidation value, as the primary repayment source.

Outcome

The illustrative committee sanctions the reduced term loan and a smaller initial working-capital line, subject to the conditions. It does not treat the collateral as a substitute for cash-flow analysis. Monthly monitoring is required for the first year, with automatic review if the largest customer reduces orders or if receivables days exceed 85.

Learning points

The case demonstrates that growth can increase rather than reduce credit risk; working-capital analysis is essential; scenario analysis should reveal breach points; specialised collateral may have limited recovery value; and escalation is appropriate when uncertainty is material but potentially manageable.

9. Ethics, Governance and Technology

Credit decisions affect borrowers, employees, depositors and shareholders. Ethical underwriting requires consistency, confidentiality, avoidance of conflicts of interest and fair treatment of customers. Analysts should not manipulate assumptions to secure approval, suppress adverse information, or impose unnecessary conditions for commercial leverage.

Governance requires segregation of origination, analysis, approval and monitoring where proportionate. Decisions should be documented sufficiently for later review. Post-approval monitoring is part of underwriting because risk changes after disbursement.

Artificial intelligence and automated decision tools can improve document extraction, early-warning detection and consistency. They can also reproduce historical bias, rely on poor-quality data, obscure reasoning and create privacy risks. Banks should maintain data lineage, model validation, explainability appropriate to the decision, human override controls, access restrictions and ongoing performance monitoring. An automated score is evidence, not an unchallengeable decision.

Contemporary underwriting increasingly incorporates open-banking data, real-time transaction analysis, climate-risk information, digital identity, alternative data and continuous monitoring. These developments improve timeliness but raise questions about consent, data quality, cyber risk, interoperability and whether short-term behavioural signals are being overinterpreted.

Best-Practice Checklist

Before recommending a facility, the analyst should be able to answer yes to the following:

• Is the borrower, ownership and purpose verified?
• Is the primary repayment source specific and credible?
• Are historical figures reconciled with independent evidence?
• Are assumptions transparent and internally consistent?
• Has the facility been matched to the cash-flow cycle?
• Have downside scenarios and covenant breach points been tested?
• Are security, guarantees and legal protections enforceable?
• Does every mitigant address a named risk?
• Are exceptions clearly documented and authorised?
• Is the recommendation proportionate to risk and supported by a monitoring plan?

Challenges and Limitations

Credit appraisal is constrained by incomplete information, uncertain forecasts, changing economic conditions and behavioural responses to lending terms. Ratios can be distorted by accounting policies; collateral values can fall during systemic stress; guarantees may fail when the guarantor is correlated with the borrower; and models can produce false precision. Professional judgement is therefore unavoidable, but it must be structured, evidence-based, reviewable and open to challenge.

Chapter Summary

Credit appraisal and underwriting integrate borrower character, business economics, financial capacity, facility structure, collateral, legal enforceability and external conditions. The primary repayment source should ordinarily be sustainable operating cash flow. Quantitative analysis uses carefully defined measures such as DSCR, leverage, liquidity and expected loss, while qualitative analysis interprets management quality, governance, industry structure and conduct.

Scenario and sensitivity analysis test whether a proposal remains viable when assumptions deteriorate. Mitigants must be specific, enforceable and linked to identified risks. Exceptions require explicit justification, compensating controls and appropriate authority. Sanction, decline and escalation are distinct professional decisions, each requiring a documented rationale. Ethical governance, independent challenge, accurate data and responsible use of technology are essential to defensible lending.

Key Terms

Credit appraisal: Systematic assessment of a borrower, purpose, repayment capacity and risks.

Underwriting: Conversion of credit evidence into a structured risk decision.

Primary repayment source: The main expected source of debt repayment, usually operating cash flow.

Secondary repayment source: Alternative repayment such as refinancing or asset sale.

Tertiary repayment source: Recovery through enforcement and liquidation.

Probability of default (PD): Likelihood that an obligor defaults over a specified horizon.

Loss given default (LGD): Proportion of exposure expected to be lost after recoveries.

Exposure at default (EAD): Amount exposed when default occurs.

Expected loss: Estimated credit loss expressed as PD × LGD × EAD under compatible assumptions.

Debt-service coverage ratio: Cash available for debt service divided by scheduled debt service.

Leverage: Degree to which a borrower uses debt relative to equity or earnings.

Covenant: Contractual undertaking or restriction designed to control credit risk.

Condition precedent: Requirement that must be satisfied before drawdown.

Credit exception: Approved departure from policy, authority or standard terms.

Sensitivity analysis: Testing the effect of changing selected assumptions.

Scenario analysis: Testing a coherent combination of assumptions representing a possible future state.

Collateral: Asset or right pledged to support repayment and recovery.

Residual risk: Risk remaining after mitigants are applied.

Credit escalation: Referral of a case to a higher or specialist authority because of risk, uncertainty or exception.

Review Questions

Short Answer Questions

1. Distinguish primary, secondary and tertiary sources of repayment.
2. Why should collateral not substitute for repayment-capacity analysis?
3. Define a credit exception and identify two matters it should document.
4. What is the difference between sensitivity analysis and scenario analysis?
5. State two limitations of automated credit scoring.

Long Answer Questions

1. Explain the stages of a professional credit appraisal from initial screening to post-approval monitoring.
2. Discuss how qualitative evidence should be integrated with financial-ratio analysis.
3. Explain the assumptions, uses and limitations of DSCR in lending decisions.
4. Analyse the role of covenants, conditions precedent and staged disbursement as risk controls.
5. Compare sanction, decline and escalation decisions and explain when each is appropriate.

Analytical / Application Questions

1. A borrower has cash available for debt service of 18 million and scheduled debt service of 15 million. Under a downside case, cash flow falls to 13 million and debt service rises to 16 million. Calculate both DSCRs and recommend whether the case should be sanctioned, declined or escalated, explaining the additional evidence required.
2. A manufacturing borrower shows strong profit growth but receivables days increase from 45 to 90, while one customer accounts for half of sales. Identify the principal risks, propose mitigants and design a monitoring plan.
3. A proposal meets financial policy but requests a waiver of a mandatory security requirement because the borrower is strategically important. Prepare the main elements of an exception memorandum and explain the governance route you would recommend.

Multiple Choice Questions with Answers

Question 1

Which source should normally be the primary basis for repayment of a business loan?

A. Liquidation of pledged assets

B. Operating cash flow

C. A new loan from another bank

D. A personal guarantee

Correct Answer: B. Operating cash flow

Question 2

If cash available for debt service is 12 million and scheduled debt service is 8 million, what is the DSCR?

A. 0.67×

B. 1.20×

C. 1.50×

D. 2.00×

Correct Answer: C. 1.50×

Question 3

Which statement best distinguishes scenario analysis from sensitivity analysis?

A. Scenario analysis changes only interest rates

B. Sensitivity analysis always uses audited data

C. Scenario analysis tests a coherent combination of assumptions

D. Sensitivity analysis is used only after default

Correct Answer: C. Scenario analysis tests a coherent combination of assumptions

Question 4

In the relationship EL = PD × LGD × EAD, LGD represents:

A. The probability of default

B. The exposure at the time of default

C. The proportion expected to be lost after recoveries

D. The annual lending rate

Correct Answer: C. The proportion expected to be lost after recoveries

Question 5

Which is the strongest example of a condition precedent?

A. A borrower’s general promise to grow sales

B. Execution and perfection of required security before drawdown

C. A forecast prepared without supporting evidence

D. A relationship manager’s informal assurance

Correct Answer: B. Execution and perfection of required security before drawdown

Question 6

A facility should be escalated when:

A. It contains material uncertainty or requires authority beyond delegated limits

B. All evidence is complete and risk is clearly within appetite

C. The borrower submits documents on time

D. The facility has no exceptions

Correct Answer: A. It contains material uncertainty or requires authority beyond delegated limits

Question 7

Why can rapid revenue growth increase credit risk?

A. Growth always reduces liquidity

B. Growth may consume working capital and increase collection risk

C. Growth automatically lowers margins to zero

D. Growth prevents the use of collateral

Correct Answer: B. Growth may consume working capital and increase collection risk

Question 8

Which covenant is a financial covenant?

A. Submission of audited statements

B. Maintenance of a minimum DSCR

C. Registration of a company name

D. Provision of an identity document

Correct Answer: B. Maintenance of a minimum DSCR

Question 9

A five-year loan with equal annual payments and a constant rate is most directly analysed using which relationship?

A. The level-payment amortisation formula

B. The current-ratio formula

C. The inventory-days formula

D. The expected-loss formula only

Correct Answer: A. The level-payment amortisation formula

Question 10

What is a key governance risk of an opaque AI credit model?

A. It cannot process any numerical data

B. It may produce decisions that cannot be adequately explained or challenged

C. It always eliminates human bias

D. It makes monitoring unnecessary

Correct Answer: B. It may produce decisions that cannot be adequately explained or challenged