Financial DT Test common mistakes
Financial DT Test · Understand the traps and how to avoid them.
Treating 'settlement date' and 'trade date' as interchangeable when computing cash flows, accrued interest, or position exposure in a question about a bond or equity trade.
Candidates read the scenario quickly and anchor on the date first mentioned, forgetting that bond accrued interest and ownership transfer are conventionally measured to settlement (T+1 or T+2 depending on market), while P&L attribution starts at trade date.
What to do instead
Before calculating, write both dates explicitly on scratch paper and label which convention the question requires. For bonds, compute accrued interest to settlement date using the stated day-count convention; for exposure and P&L, use trade date. Practice with one worked example per convention (e.g., 30/360 corporate, actual/actual gilt) until the distinction is automatic.
Confusing real-time gross settlement (RTGS) systems with net deferred settlement systems when asked about intraday liquidity risk, finality of payment, or counterparty exposure.
Both are 'settlement systems', so candidates pattern-match on the word 'settlement' and miss that RTGS settles each payment individually with immediate finality (higher intraday liquidity need, no netting benefit), while net systems batch and net obligations (lower liquidity need, but exposure until end-of-cycle).
What to do instead
Build a two-column comparison and memorise one concrete system per type (e.g., a large-value RTGS versus a retail net batch system). For any question, first ask: does the scenario hinge on intraday liquidity, finality, or netting? That determines the answer. Then check whether the question mentions 'final and irrevocable' (RTGS) or 'multilateral netting' (net system).
Applying the wrong day-count convention or compounding frequency to a money-market or fixed-income calculation, producing an answer that is close but wrong.
Candidates default to the convention they practised most (often actual/365 or annual compounding) without checking the instrument: sterling money markets often use actual/365, euro money markets actual/360, US money markets actual/360, and many bonds use 30/360 or actual/actual. Distractors are built around exactly these near-miss values.
What to do instead
At the start of every calculation question, circle the currency and instrument type, then state the convention out loud before touching the numbers. Keep a one-page convention table (currency, instrument, day count, compounding) and drill ten mixed questions until selection is reflexive rather than habitual.
Mixing up the roles of the clearing house, central securities depository, custodian, and settlement agent when tracing the post-trade lifecycle of a transaction.
The functions overlap in narrative descriptions, and candidates conflate 'clearing' (novation, netting, margin) with 'settlement' (exchange of cash and securities) and with 'custody' (safekeeping and asset servicing). Questions that ask 'who bears counterparty risk after novation' or 'who holds the asset' then get answered with the wrong entity.
What to do instead
Draw the post-trade chain once: execution, clearing (CCP novates and becomes counterparty), settlement (CSD moves securities against cash), custody (safekeeping, corporate actions, income collection). Attach one responsibility sentence to each box and test yourself by asking which entity fails in a given scenario. Revisit the diagram before any lifecycle question.
Selecting the 'most suitable' investment for a client based on the highest expected return or the product the client said they liked, rather than on documented suitability, risk tolerance, and capacity for loss.
Candidates project their own return orientation onto the question and skip the advisory discipline: suitability requires matching risk tolerance, time horizon, objectives, and capacity for loss, and 'attitude to risk' is not the same as 'capacity for loss'. Distractors are designed to reward return-chasing.
What to do instead
For every advisory scenario, extract and write down four facts before looking at options: objective, time horizon, risk tolerance, capacity for loss. Eliminate any option that breaches one of these even if it offers better returns. If the question mentions vulnerability, concentration, or a change in circumstances, check whether a suitability review or rebalancing is the correct action rather than a new product.
Treating 'best execution' as simply achieving the lowest explicit commission, ignoring execution factors such as price, speed, likelihood of execution, size, and settlement.
Commission is the visible number, so candidates anchor on it. Best execution is about the best overall result for the client across execution factors and venues, and total consideration includes implicit costs like spread and market impact, not just the headline fee.
What to do instead
Memorise the execution factors and total-consideration concept, then in each question ask which factor the scenario actually stresses (e.g., a large illiquid order stresses market impact and likelihood of execution, not commission). Reject any answer that reduces best execution to the lowest fee. Link to the firm's execution policy and order-handling obligations.
Confusing market risk, credit risk, liquidity risk, and operational risk when classifying a scenario, particularly funding liquidity versus market liquidity and counterparty credit versus settlement risk.
Scenarios often contain several risk types at once, and candidates label by the first risk word they see rather than by the loss mechanism. Funding liquidity (ability to meet cash obligations) and market liquidity (ability to trade without moving price) are routinely swapped, as are counterparty credit risk and settlement (Herstatt) risk.
What to do instead
Classify by asking 'what causes the loss and when?': inability to meet cash outflows is funding liquidity; inability to sell without price concession is market liquidity; counterparty fails before you pay is pre-settlement credit risk; you pay and counterparty fails before delivery is settlement risk. Drill scenario classification with a decision tree and justify each label in one sentence.
Assuming a regulation applies globally or to all client types, without checking the jurisdictional scope, entity type, and client categorisation (e.g., retail versus professional versus eligible counterparty).
Candidates learn a rule in one context and over-generalise it. Many obligations differ by jurisdiction and by client category, and exemptions or thresholds (e.g., dealing on own account, small firm exemptions) change the answer. The question's facts usually signal the category, but candidates skip them.
What to do instead
Before answering any compliance question, identify three things: jurisdiction, entity type, and client category. Then ask whether the rule in question is modified or disapplied for that combination. Keep a short matrix of the main regimes you are examined on and their scope, and practise questions that deliberately vary one variable at a time.
Confusing data lineage, data quality, and data governance when answering questions about why a risk report is wrong or how to fix a reporting control.
The three terms sound similar and candidates treat them as synonyms. Lineage is the traceable path of data from source to report; quality covers accuracy, completeness, timeliness, and consistency; governance is the ownership, policy, and control framework. A question about 'where did this number come from' is lineage, not quality.
What to do instead
Define each term in one sentence and map typical symptoms to each: unexplained reconciliations point to lineage; stale or missing fields point to quality; unclear accountability points to governance. For scenario questions, identify the symptom first, then choose the control that addresses it, and avoid options that propose a technology fix where a governance or process fix is needed.
Assuming a distributed ledger or cloud deployment removes the need for reconciliation, controls, or regulatory obligations, and choosing 'the blockchain records it so no further control is needed' style answers.
Candidates over-attribute properties to new technology and under-weight the enduring requirements: immutable records still need off-chain reconciliation to fiat cash and legacy systems, smart contracts need legal enforceability analysis, cloud use still requires outsourcing and operational resilience oversight, and data protection obligations persist.
What to do instead
For any technology question, apply a 'what does not change' test: identify which existing controls, reconciliations, and regulatory obligations survive the new architecture. Treat claims of automatic trust or elimination of controls as red flags. Learn the specific supervisory expectations for outsourcing, operational resilience, and DLT settlement so you can name the residual obligation rather than assume it away.