Repayment capacity is not a property of one number. Average income can look sufficient while arriving on different dates from instalments. A profitable business can still be waiting for customers to pay. A household may have stable earnings but necessary costs that a simple form misses. In preparing analysis, I would connect the source and timing of money with actual obligations. My purpose would not be to replace the credit specialist’s decision, but to improve the evidence and identify where a headline figure is insufficient. An accurate subtraction can still depend on incomplete assumptions.
A documented case: HSBC in the United Kingdom
On 23 May 2024 the FCA announced a £6,280,100 penalty against three HSBC entities for treatment of customers in arrears or financial difficulty during 2017–2018. Its official announcement describes shortcomings in affordability assessments for repayment arrangements. This is not a finding about all original loan approvals or a Czech legal rule. I use it as a documented reason to examine the customer’s actual circumstances when assessing sustainable payments.
My interpretation is that a customer agreeing to a payment proposal does not prove that it is affordable. Consent does not create the money needed for the next instalment. I would ask how income and expenditure were established, whether assumptions were unduly optimistic and how sustainability was checked. This is my proposed examination, not an extension of the UK regulator’s factual findings. A Czech case would require the applicable requirements and full individual evidence. The transferable lesson concerns the quality of reasoning, not automatic transfer of a foreign enforcement conclusion into another jurisdiction.
Averages can hide timing problems
Comparing regular income and obligations is a useful starting point. I would then examine timing and stability. A one-off bonus is different from recurring salary; sales are different from cash receipts; rental income can include vacancy periods. The same total therefore needs different interpretation depending on its source. I would distinguish verified information from estimates and establish when each input was obtained. Without those distinctions, a calculation can imply more certainty than the evidence supports. The question is not only whether a number is present but whether it describes the expected repayment period.
Obligations also need a completeness check. Two records may use different definitions of recurring expenditure or refer to different dates. Where information moves from a form into a decision tool, I would verify that supplementary facts are retained and that missing values are not silently converted into zero. For an automated result, I would trace the formula and rounding rules. A technical discrepancy does not prove inability to repay, but it can justify reconsidering a decision made using a materially different input. Its actual significance should be checked rather than presumed.
A hypothetical household example
Imagine verified monthly net income of CZK 55,000, essential expenditure of CZK 30,000 and existing instalments of CZK 8,000. Before a new loan, the arithmetic remainder is CZK 17,000. A proposed new instalment of CZK 12,000 leaves a positive CZK 5,000. The assessment does not end there. I would check irregular annual expenses, income stability and whether all evidence describes the same household situation. These invented figures are an illustration, not a recommendation to approve a loan or a suggested universal affordability threshold for lenders.
If necessary annual costs total another CZK 36,000, spreading them over the year adds CZK 3,000 a month. The remainder falls to CZK 2,000. A ten per cent income reduction would change it by another CZK 5,500. This demonstrates sensitivity to assumptions, not the probability of a specific event. A real decision requires all relevant circumstances and the approved specialist process. My contribution would be a calculation whose inputs and assumptions can be checked, including an explanation of what the simplified exercise excludes rather than a misleadingly definitive approval outcome.
A hypothetical company: profit and cash flow
For a business I would apply the same discipline to cash movements. A contract may be profitable while the customer pays ninety days after delivery and materials must be paid for beforehand. Annual profit does not show the deepest monthly funding gap. I would prepare an incoming-and-outgoing cash timeline and identify assumptions about collection dates. A specialist can then assess whether the financing structure fits the actual need or whether regular debt service depends on uncertain receipts. That separates a timing problem from an assertion about the company’s overall profitability.
I would also examine alternative explanations in both examples. An unusually high expense might be non-recurring; lower income might be temporary; an accounting difference may be justified. The objective is not to worsen every assessment whenever something unusual appears. It is to obtain an explanation, evidence and ownership of the judgement. Where information is absent, I would record the specific uncertainty and the next step needed to resolve it. Analytical work should make a decision better informed, not turn every exception mechanically into a negative conclusion without understanding its circumstances.
How I would assess the process
I would select decisions involving different income sources and reconstruct the information available at approval. I would independently recalculate the result and compare it with the application. For amended terms, I would check whether evidence must be refreshed and who confirms that requirement has been met. Exception reasons should match the actual circumstances. I would also examine how changes after origination become visible. Initial assessment and subsequent monitoring are related but distinct controls: a sound original decision does not remove the need to identify later changes in the customer’s ability to make agreed payments.
Recommendations should distinguish methodology, execution and record-keeping. If the policy does not explain irregular income, training staff to complete the same form will not solve the gap. If the process is appropriate but evidence is missing from one file, the remedy may be narrower. If information transfers incorrectly, a system correction and regression test are also needed. Matching the measure to the cause creates something concrete to retest after implementation, rather than a general promise of greater care. It also makes responsibilities clearer because different problems require different owners and verification evidence.
Summary and recommendations
Behind a repayment figure I would examine source, freshness, stability and timing. I recommend making assumptions visible, distinguishing recurring and one-off items, and reconciling evidence to the terms actually approved. A hypothetical calculation helps formulate better questions but cannot replace a full individual assessment. HSBC illustrates the importance of this discipline in another jurisdiction and at a different stage of the credit relationship. Its role here is to motivate careful analysis, not establish an identical defect in the Czech market or imply that one simplified numerical rule can determine affordability.