The latest official statistics for England and Wales show 1,946 registered company insolvencies in August 2026, but that single number bundles together very different events. This article explains what the figure counts, how the main procedures differ, and how to read the trend without over-reading one month. It describes the rules and statistics of England and Wales only; Scotland, Northern Ireland and other countries use their own procedures and publish their own figures.
What the August 2026 figure actually contains
According to the Insolvency Service’s monthly release, published on 18 September 2026, there were 1,946 registered company insolvencies in England and Wales in August 2026. The release breaks that total into four types, plus a fifth category that was empty that month:
| Procedure | August 2026 count | Share of total (illustrative arithmetic) |
|---|---|---|
| Creditors’ voluntary liquidations (CVLs) | 1,431 | about 73.5% |
| Compulsory liquidations | 314 | about 16.1% |
| Administrations | 182 | about 9.4% |
| Company voluntary arrangements (CVAs) | 19 | about 1.0% |
| Receivership appointments | 0 | 0% |
The four counts add up to 1,946, matching the published total, and the shares above are simple division of each count by that total. The first thing to notice is that nearly three in four of these cases are voluntary liquidations. Insolvency, in the official sense, is mostly an orderly closing of a failed company rather than a dramatic collapse.
How do the four main procedures differ?
The Insolvency Service describes each procedure in its commentary, and the differences matter because they signal different outcomes.
- Creditors’ voluntary liquidation: the company’s shareholders pass a resolution to wind it up, because it cannot pay its debts. The directors start the process, and the company ends.
- Compulsory liquidation: a court makes a winding-up order, obtained by a creditor, a shareholder or a director. Here the process is usually forced on the company, often by an unpaid creditor.
- Administration: a licensed insolvency practitioner is appointed to try to rescue the company as a going concern or, failing that, to get a better result for creditors than an immediate liquidation would.
- Company voluntary arrangement: a formal agreement to repay creditors some or all of what they are owed. It binds all creditors once 75% or more approve it, and the company can keep trading.
Only administrations and CVAs are, by design, attempts to keep a business alive. Together they made up roughly 10% of August’s total on the arithmetic above. The other procedures end the company’s life.
Is the trend getting better or worse?
The answer depends on the comparison. The release says August’s total was 3% lower than August 2025, when the count was 2,007. It also says the average monthly total for January to August 2026 was lower than the average of the previous three years. On those measures, the overall level is not elevated by recent standards.
The 12-month insolvency rate points the same way. For the 12 months to August 2026 the rate was 50.1 per 10,000 companies on the effective register, about one in 200, down from 52.5 for the 12 months to August 2025. For perspective, the Insolvency Service gives the 2008–09 peak as 113.1 per 10,000 and notes that the register has more than doubled in size since then.
So the headline is not one of surging failure. But the composition tells a more mixed story, which is the next section.
Why the mix of procedures matters more than the total
A flat or falling total can hide movement underneath. The August release compares each procedure with August 2025:
| Procedure | Change vs August 2025 |
|---|---|
| Total | down 3% |
| Compulsory liquidations | up 5% |
| CVLs | down 9% |
| Administrations | up 60% |
| CVAs | up 19% |
Two points follow. Voluntary liquidations, the largest group, fell, which pulled the total down. Meanwhile court-driven liquidations and administrations rose. The release also notes that in 2025 as a whole, compulsory liquidations rose 15% to their highest level since 2012, while CVLs fell 2% and administrations fell 8%.
Why might that matter? One reasonable reading, offered here as analysis rather than something the statistics prove, is that creditor-driven action carries more weight in the failure picture than it did a few years ago. The statistics record what happened, not why, so any explanation involving tax authorities, lenders or suppliers would need separate evidence.
Small counts also swing in percentage terms. There were only 19 CVAs in August, so a change of a handful of cases moves the percentage sharply. A 19% rise in CVAs is a movement of a few cases, not a trend.
Can a single month be trusted?
Not entirely. The Insolvency Service states that all figures are provisional and subject to review. Data for the latest month are extracted about five working days after month end, so some compulsory liquidations may not yet be recorded, and the latest month may be revised. Revisions usually come from data entered into administrative systems after the extraction cut-off.
Seasonally adjusted figures are also revised in later publications as the statistical model is updated. Different pages and releases can therefore show slightly different numbers for the same month. A better habit is to look at several months together and to treat a single month’s percentage change as a prompt for questions rather than an answer. The next release in the series is scheduled for 20 October 2026.
What the figures do not capture
The insolvency count measures formal procedures, so it leaves out several kinds of business difficulty:
- Dormant or dissolved companies. A company can simply stop trading and be struck off the register without a formal insolvency, and it would not appear in these counts.
- Businesses struggling but surviving. Companies cutting costs, stretching supplier payments or borrowing against assets are under pressure but not in the data. The earlier discussion of why cash flow discipline matters more than revenue growth explains how a company can be in difficulty long before any formal step.
- Sole traders and partnerships. The series covers companies. Individuals have their own insolvency routes and their own statistics.
- Other countries. These procedures are specific to England and Wales. The United States, for example, uses a different system built around bankruptcy chapters, and the two should not be compared by name alone.
The count is also a measure of a point in the process, not of cause. Business owners dealing with a sustained higher-cost operating environment may face pressure for many months before a procedure begins, and the statistics record only the end of that sequence.
Why a rate is more useful than a count
The raw monthly count rises mechanically as the number of companies grows. The register in England and Wales is far larger than it was in 2008–09, so a count that looks similar to an earlier period can represent a smaller share of companies. That is why the Insolvency Service also reports the rate per 10,000 companies on the effective register.
A worked example shows the idea. Suppose a register held 100,000 companies and recorded 500 insolvencies over 12 months: that is a rate of 50 per 10,000. If the register doubled to 200,000 and recorded 1,000 insolvencies, the count would double, but the rate would stay at 50 per 10,000. The illustrative figures are not real data. They show why comparing counts across long periods can mislead, and why the 50.1 figure for the 12 months to August 2026 is a more comparable gauge than one month’s total.
Questions to ask when the next release arrives
Rather than reacting to the headline, a reader can work through a short list when the October release appears:
- Has the 12-month rate moved meaningfully from 50.1, or only by a fraction?
- Did the mix shift between voluntary and compulsory procedures again?
- Were the previous month’s numbers revised, and by how much?
- Are the percentage changes based on large counts or small ones?
These questions apply to any official statistical series, but they are especially useful here because the headline total is dominated by one procedure and the smaller categories are volatile.
A steadier way to read business failure statistics
The August 2026 release shows 1,946 insolvencies, a total below the same month a year earlier, a 12-month rate of 50.1 per 10,000 companies and a mix that leans voluntary but with compulsory liquidations and administrations rising. Read together, those facts describe a moderate level of formal failure in England and Wales rather than a spike. The composition, the provisional status of the data and the difference between a count and a rate all deserve as much attention as the headline.
This article is general information, not personalised financial advice; rules, statistics and rates change.
Sources: Insolvency Service; GOV.UK
