The UK filing regime is unusually complete compared to most jurisdictions. Most data vendors extract the headline numbers from annual accounts and stop there. What they leave behind is a set of signals that, read together, tell a far more detailed structural story about a company than any single financial measure can.
This is a note on what those additional filing layers contain, how we approach them at Thema, and where the limits of the public record genuinely sit.
What the filing registry actually contains
Companies House holds several distinct filing streams, each carrying a different signal type. The major categories are annual accounts (the financial position statement, with disclosure depth dependent on the company's size threshold); confirmation statements (registered address, SIC code, shareholder information, and since 2016, PSC data); Persons with Significant Control entries (beneficial ownership down to the 25 percent threshold); mortgage charges (registered security over company assets, with nature, date, and satisfaction status); director appointment and resignation events (names, dates, and occupation descriptions); and capital event filings covering share issuances and restructurings.
No commercial database surfaces all of these in consistently queryable form. Most vendors have reasonable annual accounts coverage for medium and large filers, thinner coverage for small and micro filers, and patchy or absent coverage of PSC, charges, and appointment histories. That leaves a substantial portion of the structural picture unread.
Annual accounts: reading across the size thresholds
UK company law creates four filing thresholds. The data yield varies significantly across them, and understanding those differences is a prerequisite for extracting anything useful from accounts at scale.
Micro-entity filings and their limits
Micro-entity companies can file a simplified balance sheet with no profit and loss account. The practical yield for competitive intelligence is: total assets, net assets, and sometimes creditor splits. You can infer approximate asset intensity and solvency, but not revenue or profitability. For sector mapping, micro-entity filers typically sit below acquisition relevance thresholds. They matter primarily as inputs to supply chain analysis and as early-stage tracking of companies that may grow into the target range over a three-to-five year horizon.
Small company filings: the most analytically useful tier
Small companies (broadly: turnover under 10.2 million GBP, fewer than 50 employees) represent the most productive tier for mid-market PE work. Many file abridged accounts that omit a full profit and loss statement, but the balance sheet and note disclosures are often sufficient to triangulate a revenue range using indirect methods. Net trade debtors, staff cost disclosures where present, and the fixed asset base each support different estimation approaches. A company with 1.8 million GBP in trade debtors and a debtor days pattern consistent with its sector is telling you something specific about its revenue run rate, even without a P&L line.
For medium and large companies, disclosure is substantially more complete: full P&L, segmental notes where applicable, director remuneration, and related-party transactions. Data quality at this tier is high, but these companies are already visible through other channels. The discovery opportunity is smaller.
The PSC register as an ownership signal
The Persons with Significant Control register, introduced in 2016, requires UK companies to disclose any person or entity holding more than 25 percent of voting rights or economic interest. For competitive intelligence work, it does two things that no other public source replicates in this form.
First, it identifies whether a company is part of a larger group, and which entity sits above it. Second, it flags when a financial investor holds a meaningful stake. When a fund entity appears as a PSC, you can distinguish founder-owned businesses from portfolio company assets. A PE-backed business may be approaching the end of its typical hold period. A founder-owned business may not have had a structured exit conversation at all. That distinction matters significantly for assessing the timing and nature of a potential approach.
Compliance quality varies. Some companies list no PSC because their ownership sits in trust or overseas holding structures that fall outside UK disclosure scope. Others lag on updates following changes of control. We treat PSC data as indicative rather than definitive, and cross-reference it against charge filings and director appointment patterns where there are apparent contradictions in the record.
Mortgage charges: capex and financing intelligence
A registered charge is filed when a company grants security over its assets to a lender. The filing records the date, the lender (or a nominee), the asset class covered, and satisfaction status when the charge is later discharged. Outstanding charges tell you that the company is carrying secured debt, who the lender is, and when the financing was raised. Charge dates cluster around capital expenditure cycles, so a pattern of multiple charges filed over a short period is often an indicator of an expansion phase.
Satisfied charges, read with their original dates, allow you to reconstruct a rough debt history. The absence of charges on a capital-intensive business is also meaningful: either the company is genuinely unlevered, or its financing sits in a group structure that does not appear in this entity's filing record. Both interpretations are worth following up.
Director appointment patterns as an organisational signal
Director appointment and resignation events are filed on standardised forms and carry dates, nationality, and an occupation description. This information is timestamped and available in the filing history for every company, which makes it usable for event-detection analysis across large numbers of companies simultaneously.
Several patterns are worth monitoring. A cluster of resignations shortly after a PSC change often indicates a management transition following a change of control. Multiple new appointments of directors with financial services backgrounds in a short window can precede a transaction or refinancing. A consistent pattern where the same individual appears as director across many companies in a sector often points to nominee or group holding structures rather than operational management.
Director data works best as a corroborating layer. Its value is highest when combined with account and charge data to date a transition: if the accounts show a balance sheet change, the PSC register shows a new owner, and director filings show new appointments within the same twelve-month window, you have a fairly complete picture of a change of control event that may not have been publicly announced.
What we do not use in isolation
A clarification worth stating directly: none of these filing streams is precise enough to use alone, and none is current. Annual accounts are filed up to nine months after the accounting period end date, so the most recent accounts you can access may reflect a position that is eighteen to twenty-four months old. PSC entries are self-reported and subject to update lag. Charges require active deregistration when satisfied; not all satisfied charges are promptly removed.
We treat Companies House as a structural anchor, not a real-time operational source. Its value is the multi-year structural picture: ownership history, financing behaviour, directorship changes over time, and balance sheet trajectory across multiple filing periods. When combined with employment data from hiring feeds, which operates on a shorter lag and captures forward-looking organisational signals rather than backward-looking financial ones, the filing record becomes considerably more useful than either source in isolation.
This is also the dimension that most commercial vendors do not build. Standard market data products optimise for point-in-time financial extraction. The structural read, across multiple filing types and multiple periods, is what tells a deal team whether a company has been preparing to transact, is mid-cycle, or is under financial pressure. Those three states look meaningfully different in the filing record, and being able to distinguish them from the public record before making the first call is what makes the approach worth the extraction work.