Banking exam quantitative aptitude is a speed test in disguise
Most candidates can solve the quantitative section given time. The banking written test refuses to give it. The exam does not test whether you can, but how fast.
Ask a candidate why they failed the quantitative section of a bank written test and they will usually tell you they ran out of time. Ask whether they could have solved the questions given more time, and almost all of them say yes. That admission is the whole story. The banking quantitative section is not primarily testing whether you can do the mathematics. It is testing how fast you can, and those are different exams with different preparation.
A typical recruitment paper gives you a quantitative section of, say, twenty-five questions inside a slice of time that works out to under a minute each once you account for the reasoning and English competing for the same clock. Under a minute includes reading the question, choosing a method, doing the arithmetic and marking the sheet. Any question that takes you two minutes has silently borrowed the time budget of a second question you will now never reach. The section punishes the slow correct answer almost as hard as the wrong one.
Where the quantitative minutes actually go
- Simplification / approximation~25s
- Number series~35s
- Arithmetic word problem~50s
- Data interpretation~75s
sec/question
An estimate of the average time a prepared candidate spends per question type in a bank quantitative section: a simplification or approximation runs around 25 seconds, an arithmetic word problem around 50 seconds, and a data-interpretation question around 75 seconds because it carries a chart or table to read. Figures are illustrative estimates for planning, not measured constants.
The chart is an estimate, not a measured law, but the ranking is stable: data interpretation is where the minutes go, and it is therefore where the strategy has to concentrate. It is also, helpfully, where a single good habit pays several times over, because one dataset usually feeds four or five questions. Read the chart once, carefully, extract everything you might need, and those questions become fast. Skim it and re-read it for each question, and you have paid the reading cost five times over for no reason.
The arithmetic layer nobody wants to build
There is an unglamorous foundation under all of this: raw arithmetic fluency. A candidate who has to compute seventeen times twenty-three the long way, mid-question, has already lost the race to a candidate for whom that product surfaces in a second. Memorised multiplication tables to twenty, the squares to at least thirty, the cubes to fifteen, and the decimal values of the common fractions are not glamorous, and they are the single cheapest speed you can buy. Every one of them shaves seconds off nearly every question, and seconds are the entire currency of this section.
Above the arithmetic sits method fluency: recognising, on sight, that a question is a percentage problem or a ratio problem or a simple-interest problem, and reaching for the shortcut without deriving it from scratch. This recognition is built by volume, not by understanding alone. You can understand percentages perfectly and still be slow at them because you have not done enough of them to make the recognition automatic. The exam rewards the automatic, not the merely correct.
A rough time budget you can actually keep
The table below is a planning tool, not a rule. Its point is to give you a target time per question type so that you can feel, during the paper, when a question has gone over budget and needs to be abandoned. Without a target you have no way to know you are running late until it is too late to recover.
| Question type | Target time | Strategy |
|---|---|---|
| Simplification / approximation | ~25 sec | Approximate aggressively; the answer options are usually far apart |
| Number series | ~35 sec | If the pattern is not visible in 30 seconds, mark and move |
| Arithmetic word problem | ~50 sec | Recognise the type, apply the shortcut, do not derive |
| Data interpretation | ~75 sec | Read the dataset once; answer its cluster together |
| Any stuck question | Fixed box | Abandon at the box and return only if time remains |
How to practise the speed, not the maths
Because the constraint is speed, your practice has to be timed from the first day, not just in the final mocks. Untimed practice trains accuracy, which most candidates already have, and does nothing for the thing that is actually failing them. Practise sets against a clock, record how long each type takes you, and watch those times fall week by week. If your accuracy is high but your times are not falling, you are drilling the wrong thing — probably solving more questions when you should be doing fewer, faster, with your fingers on the specific arithmetic that keeps slowing you down.
I could solve every question I attempted. I just attempted eleven of the twenty-five, and got a respectable score on a section I had lost before I sat down.
That candidate did not have a mathematics problem. They had a throughput problem, and throughput is trainable. It is built from memorised arithmetic, recognised question types, read-once data interpretation and the discipline to abandon a fight you are losing. None of it is difficult. All of it is different from the way most people prepare, which is to solve more problems slowly and hope the speed arrives on its own. It does not.
Frequently asked questions
- How much time do I get per question in a bank quantitative section?
- It varies by paper, but a common shape is a section of around twenty-five questions inside a slice of the paper that works out to under a minute each, once the other sections are competing for the same total time. That budget includes reading, choosing a method, calculating and marking, which is why speed rather than knowledge is usually the binding constraint.
- What is the fastest way to improve my quantitative score late?
- Build the arithmetic layer and drill data interpretation. Memorised tables, squares and common fractions shave seconds off nearly every question, and a read-once habit on data-interpretation sets turns the biggest time sink into a cluster of fast marks. Both are trainable in weeks and both attack speed directly.
- Should I skip data interpretation because it takes too long?
- Usually no — skip it and you have written off one of the largest blocks of marks on the paper. The fix is technique, not avoidance: read each dataset once and answer its whole cluster of questions together, so the reading cost is paid once rather than repeated for every question.
- Is it better to answer fewer questions accurately or more questions quickly?
- Neither extreme. The section rewards accurate throughput: as many questions as you can answer well inside the budget. The way to raise throughput without sacrificing accuracy is to be faster on the questions you already get right, by drilling arithmetic and question-type recognition until they are automatic.
Stop preparing for the maths exam you imagine and start preparing for the speed exam you will sit. The candidates who clear this section are rarely the ones who know the most mathematics. They are the ones who convert what they know into marked answers fastest, and that is a skill you can build on purpose.
Practise this
Banking
Free practice set and a free diagnostic.
Open itFree tool
Try three questions
Timed against the real paper. No sign-up.
Try it- Reading time
- 7 min
- Length
- 1,465 words
- Topic
- Banking