UBS Culture Match (Korn Ferry SJT)
18 scenarios · Untimed (about 20-25 minutes)
What it tests. Alignment with the Three Keys: capital strength, simplification and efficiency, risk management; and the behaviours of Accountability with Integrity, Collaboration and Innovation.
Worked example. Spotting an incorrect historical revenue assumption late the night before a client review, the most effective action is to correct it, rebuild the affected slides and email your associate; the least effective is leaving the error to meet the deadline and raising it only if the client notices.
Common traps. The 'hero' persona (solving high-risk situations alone instead of consulting seniors) and the 'pure profit' perspective (prioritising fast execution over compliance and risk).
How to handle it. Read each scenario through a long-term risk-stewardship lens: pick the option that protects the firm's reputation, streamlines the process and keeps stakeholders informed.
Scales Numerical (Aon cut-e nx)
37 tasks · 12 minutes (about 19.5 seconds per task)
What it tests. Rapid data localisation and high-speed calculation accuracy (percentages, margins, ratios, currency conversion) under pressure.
Worked example. If UK revenue grows 500,000 to 550,000 (10%) and UK operating costs grow 200,000 to 240,000 (20%), the statement that cost growth exceeded revenue growth is True; a statement about regional equity returns the data never breaks out is Cannot Say.
Common traps. The sequential trap (reading every tab in detail before answering) and the extrapolation trap (answering on industry assumptions not in the data).
How to handle it. Spend the first 20 seconds mapping the six tab titles, then isolate the key term, switch straight to that tab and use a calculator; do not guess blindly, since wrong answers can be penalised more than blanks.
Scales Inductive (Aon cut-e cls)
12 tasks · 6 minutes (about 30 seconds per task)
What it tests. Fluid intelligence, abstract spatial logic and hypothesis testing under tight limits.
Worked example. If the given grids both have more shaded circles than unshaded squares, the two choice grids that preserve that ratio are correct, regardless of shape position.
Common traps. The fixation bottleneck (over a minute on one pattern) and overcomplicating the rule when it is built on simple properties like colour, count or symmetry.
How to handle it. Run a quick checklist in the first 5 seconds (count, colour ratio, position, symmetry); if you cannot identify the rule within 25 seconds, make an educated guess and move on.
Scales Verbal (Aon cut-e vx)
18 tasks · 6 minutes (about 20 seconds per task)
What it tests. Reading comprehension, semantic deduction and evaluating claims objectively without external bias.
Worked example. If a risk-strategy text caps emerging-market corporate debt at 5%, the statement that up to 5% is allowed is True; a claim that global targets are met primarily via the London and Zurich offices, when the text only references 'all European complexes', is Cannot Say.
Common traps. Bringing in outside knowledge, and skimming over qualifiers like always, never, at least or exclusively that change a statement's validity.
How to handle it. Read the statement first to pick keywords, open the matching tab, scan for those keywords and answer strictly on the text; if plausible but unconfirmed, choose Cannot Say.
Technical coding (Technology / Quant tracks)
Typically 2 problems · About 60 minutes
What it tests. Data-structure efficiency, algorithm design and runtime complexity.
Worked example. Solve an array-manipulation problem with an efficient single pass, handling edge cases such as empty inputs cleanly.
Common traps. Missing edge cases and over-engineering a solution where a clean array or hash-map approach passes.
How to handle it. Read the constraints, plan the approach, handle edge cases and verify the runtime before submitting.