Check that the member schedules are present
For a non-listed company with share capital, check that member information and relevant schedules are complete, then read them with the capital table. Cover or capital pages alone do not establish individual holdings. Form and schedule requirements can differ by company type.
Companies Registry: specified forms and completion notes ↗Companies Registry: annual return FAQ ↗
Member-schedule exercise: reconcile the denominator
All entities, holders, dates and pages below are fictional. This is a reading worksheet, not a reproduction of the official form or member pages from the ARCH MARKETING sample. Use the headings and pagination in the version you actually obtain.
Record each holder’s source name, share class, holding, relevant date and page. Follow any continuation pages. Retain separately listed entries for a holder, check for duplicate scans of the same page, then aggregate only entries for the same date and class.
Assume the same-date capital table records 1,000 ordinary shares. Reconcile the two member entries, 600 + 400 = 1,000, before calculating B’s direct 60% holding. If page 4 is missing, 600 is not the full denominator and does not identify who holds the remainder. Investigate classes, dates, missing pages and duplicate rows when totals differ.
Scroll sideways to see all columns; focus the table and use arrow keys on a keyboard.
| Holder (fictional) | Share class | Shares held | Record date | Exercise page |
|---|---|---|---|---|
| Company B | Ordinary | 600 | 2024-08-01 | Page 3 |
| Company D | Ordinary | 400 | 2024-08-01 | Page 4 (continuation) |
Companies Registry: specified forms and completion notes ↗Companies Registry: annual return FAQ ↗
What is needed to trace the next layer?
Continuing the fictional exercise, the next question is who holds B. Suppose a separate same-date record shows C holds all shares of that class in B. Record C → B and B → A as separate relationships, each with its own date, source and page. If B’s records are unavailable, leave that layer unresolved.
The arithmetic does not determine voting rights, nominee arrangements or ultimate control. Keep differing share classes, dates, missing pages and unknown intermediate entities separate instead of reporting a definitive ultimate percentage.
Record each ownership edge separately
Use one row per holder-to-company relationship. Separate rows when classes, dates or sources differ. Check that numerator and denominator cover the same scope before calculating a direct percentage. Unknown values are not zero, and multiplying percentages does not establish verified control.
Scroll sideways to see all columns; focus the table and use arrow keys on a keyboard.
| Relationship (all fictional) | What can be recorded | What remains open |
|---|---|---|
| Company B → Company A | Same date and class: 600 / 1,000, direct holding 60% | Source, page and whether other share classes exist |
| C → Company B | A separate same-date record shows all shares of that class | Identity and control need separate checks; do not auto-label as UBO |
| B’s filing lacks a date or member schedule | Retain known document references; leave percentage unresolved | Stop at this layer until the evidence is complete |
Exercise 1: investigate 600 + 400 + 400 before changing the denominator
All numbers and pages here are fictional. The capital table says 1,000 ordinary shares, but extraction yields B 600, D 400 and D 400. Trace each row back: if the last two are the same scanned page twice, the issue is duplication, not an extra 400 issued shares.
Exclude the repeated scan from the working aggregate while keeping the received original and exclusion reason. The corrected sum is 1,000. Matching names alone are not duplicates: different pages, classes or source entries must be retained until reconciled.
Record the cause, excluded source and before/after totals. If the pages cannot be matched confidently, leave it as suspected duplication requiring the original. A convenient total is not evidence that an arbitrary row should be deleted.
Scroll sideways to see all columns; focus the table and use arrow keys on a keyboard.
| Source (fictional) | Extracted entry | Treatment |
|---|---|---|
| Page 3 | B: 600 ordinary shares | Retain |
| Page 4 | D: 400 ordinary shares | Retain |
| Repeated scan of page 4 | D: 400 ordinary shares | Mark duplicate source; exclude from aggregate |
Exercise 2: retain a denominator for each share class
Extend the fictional case: A has 1,000 ordinary and 200 preference shares. B holds 600 ordinary, D 400 ordinary and E 200 preference. Their within-class holdings are 60%, 40% and 100% respectively. The percentage label must identify the class.
Dividing B’s 600 by a combined 1,200 mixes classes. Even where arithmetic is possible, it does not establish voting or control rights. Share counts, capital amounts and paid-up amounts are also different measures; an amount cannot substitute for a missing share-count denominator.
Group by company, date and share class, then reconcile each group to the corresponding issued count. Preserve an unmatched class label and raise an issue instead of renaming it Ordinary for convenience. Ownership and control interpretation remains a separate review.
Scroll sideways to see all columns; focus the table and use arrow keys on a keyboard.
| Holder (all fictional) | Class / holding | Class denominator | Recordable percentage |
|---|---|---|---|
| B | Ordinary / 600 | 1,000 | 60% of ordinary shares |
| D | Ordinary / 400 | 1,000 | 40% of ordinary shares |
| E | Preference / 200 | 200 | 100% of preference shares |
Exercise 3: different dates cannot form a confirmed same-date structure
Another fictional case: A’s August filing names B, and B’s December filing names C. Both dated observations can be recorded. They do not establish that C held B in August or that B’s position in A was unchanged in December. Show the dates separately.
A text diagram could read: “A ← B [D-01, August]; B ← C [D-02, December; August position unknown].” Label missing branches explicitly. Every edge should correspond to a register row; drawing a diagram does not supply missing evidence.
Stop tracing when circular links, ambiguous identities or versions prevent a reliable match. Do not multiply observations from different dates into an ultimate percentage. Deliver the timeline, sourced edges and open branches so the reviewer can decide which further dated evidence is needed.
Frequently asked questions
Are NAR1 shareholder records live?
No. Record the filing’s relevant date and check whether additional or later evidence is needed. A download date is not the date of the underlying information.
References
Use the checklists and examples to prepare your documents. When checking an official form, note which version the guide refers to.