United States
Hong KongHKSI study manual vs question bank: which one is failing you
The manual is the authoritative text and the only place the full syllabus is properly explained. A question bank is where you find out what you actually retained. You need both, weighted roughly one part reading to two parts testing once your first pass through the manual is done.
Candidates ask which one to buy. It is the wrong question, and the right one is what proportion of your remaining hours should go to each, because almost everyone who fails Paper 1 got that ratio wrong in the same direction.
They read too much and tested too little. Reading feels like progress. Recognition is not recall, and the exam only pays for recall.
| HKSI study manual | Question bank | |
|---|---|---|
| What it is | The official text, written by the body that sets the paper, against the current syllabus | A set of multiple-choice items with answers, and sometimes explanations |
| What it is best at | Coverage and authority. Nothing else explains all nine topics with the same standing | Exposing what you do not know, and building speed at 1.5 minutes a question |
| Failure mode | Passive reading. You finish it, feel prepared, and score badly | Learning items instead of rules. You ace questions you have seen and miss their cousins |
| Best used | First, once, properly. Then as a lookup | Continuously, from earlier than feels comfortable |
| Handles syllabus changes | Yes, HKSI revises it | Depends entirely on the provider. Check the update date |
| Gives you a mark out of 60 | No | Yes, which is the whole point |
| Enough on its own? | For a minority of disciplined self-testers | No |
Why the manual alone leaves people short
Because Paper 1 does not ask you to describe the licensing regime. It gives you a scenario and four plausible options, and asks which regime applies. Those are different mental operations, and only one of them is trained by reading.
The manual also cannot tell you where you are weak. You finish a chapter and your sense of how well you know it is generated by the same brain that just read it, which is the least reliable possible source. A question bank replaces that feeling with a number.
The pass mark makes this expensive. You need 42 of 60, and the published mean pass rate over the twelve months to June 2026 was 51.9%. A lot of the people inside that missing 48.1% had read the manual.
Why a question bank alone leaves people short
Different failure, same result. Work only from questions and you build a library of specific items rather than a model of the rules, so you handle anything you have seen before and fall apart on a rephrasing.
Paper 1 samples across nine topics and 66 second-level syllabus headings. No bank covers every heading with enough items to make pattern-matching a safe strategy, so the gaps in the bank become gaps in you.
There is also a currency risk. The syllabus moved to v3.5. The manual gets revised by HKSI; a third-party bank gets revised when its owner gets round to it. Ask when it was last updated before you rely on it.
The split we would actually recommend
- One full pass through the manual, reading at pace. Do not highlight everything. The goal is a map of what lives where, not retention.
- Switch to questions much earlier than feels right. Around a third of the way into your available time, not at the end.
- From then on, roughly two hours of testing for every hour of reading, and let your scores decide which chapters you go back to.
- Reserve the last stretch for full timed mocks in the real 60-question, 90-minute shape, with the manual closed.
- Reread only what you got wrong. Rereading what you got right is the most comfortable and least useful thing you can do.
If you have read a topic twice and never scored it, you are not preparing, you are revising your own sense of competence. Score it. If the number is worse than you expected, that is the method working.
What a good explanation looks like
This is where question banks differ from each other most, and it is the only feature worth paying a premium for. An answer key says B. That teaches you one item.
A useful explanation says why B, names the Part of the Ordinance or the code the rule comes from, and ideally says why the closest wrong option is wrong. That teaches you the rule, and the rule answers the next twenty questions you have not seen yet.
Our own bank is 131 questions with a cited explanation on each one, and we should say plainly that they are written against the Ordinance and the SFC codes, and cite where each point comes from. For anything that matters professionally, go to the source text.
If you can only afford one
Buy the manual. That is not a hedge, it is a genuine recommendation against our own interest.
You can generate crude self-testing for free by covering a page and writing down what you remember, or by turning syllabus headings into questions you answer aloud. It is worse than a real bank and it is not nothing. What you cannot do for free is reconstruct the authoritative explanation of a rule you have never read, which is why the manual is the base layer and everything else, ours included, sits on top of it.
Once your reading is done, the useful next step is a timed run rather than another chapter. The exam format page sets out what that run should look like, and the study plan puts the reading-to-testing ratio on a calendar.
Common questions
Is the HKSI study manual enough to pass Paper 1?
For a disciplined self-tester, sometimes. For most candidates, no. The manual builds recognition and the exam pays for recall, and the gap between those two is where a large share of the 48.1% who fail actually lose their marks.
How many practice questions should I do before Paper 1?
Enough that your score stops moving, which is a better test than any fixed number. Practically, that means several timed runs at the full 60-question, 90-minute format plus repeated work on whichever topics keep scoring badly.
Should I buy a question bank before or after reading the manual?
Buy it before, use it during. Waiting until you have finished reading wastes the diagnostic value, because you want your weak topics identified while there is still time to fix them rather than in the final week.
Do question banks reflect the current syllabus?
Some do and some do not. The syllabus moved to v3.5, so ask any provider when their material was last updated against it. If they cannot answer, use their questions for practice and not as a source of rules.