The Insolvency Service is now looking at live companies. Creditors should be too.

Whenever I speak to credit and risk teams about phoenixing, I hear the same complaint. A director runs up debts, lets the company go under, and a few weeks later is trading again under a new name. Suppliers, HMRC and staff are left out of pocket. Most teams know exactly how it works. They just tend to find out when it’s too late, once the liquidation notice has landed and the debt has already been written off.

This week’s update from the Insolvency Service suggests government has reached the same conclusion about its own approach.

It is funded by an additional £5m a year for five years, announced at the 2025 Autumn Budget, to build a 50-strong team of investigators focused on this one problem.

Two details in that report matter more than the headline numbers. Investigators are using AI to piece together evidence held across government bodies, including Companies House. And they are targeting live businesses where they suspect wrongdoing, instead of waiting for an insolvency to trigger the case.

Here’s the thing that essentially pushed me to write this piece. There is a common assumption in credit teams that phoenixing is an enforcement problem: something the state deals with after the fact, and something a creditor can only absorb. The Insolvency Service’s own shift says otherwise. The evidence of phoenix behaviour exists on the public record well before the collapse. The question is whether anyone is reading it in time.

Why phoenixing is hard to spot, even when the data is public

We know that the data needed to spot phoenix risk is not scarce. Director appointments, resignations, incorporation dates, registered addresses, SIC codes and filing histories all sit at Companies House, and the Economic Crime and Corporate Transparency Act has strengthened the registrar’s powers to query and correct what is filed there.

What most organisations lack is the ability to connect those records across companies and across time. A single company file rarely looks suspicious. The pattern only appears when you link a director’s history across every entity they have touched.

That is exactly what the Insolvency Service is now doing with AI across government datasets. Private sector credit teams face the same technical problem at a smaller scale.

A four-stage view of phoenix risk

When I think about how a risk team should read these signals, I break it into four stages. Each one is visible earlier than the next.

1. Director history.

Has this director been associated with previous companies that entered insolvency, dissolved with debts, or were struck off? One failure is ordinary business risk. A sequence is a pattern.

2. Rapid succession.

Has a new company been incorporated with the same director, a similar name, the same trading address or the same SIC code, shortly before or after a related company began to fail? This is the core phoenix signature, and it is visible at incorporation.

3. Behavioural change in the live company.

Late or abnormal filings, a move to a virtual office, director resignations, or accounts that sit outside normal ranges for a company of that size. These are the signals that suggest the current entity is being prepared for exit.

4. Formal distress.

CCJs, winding-up petitions and Gazette notices. By this stage the exposure is usually already locked in.

Most credit teams only react at stage four, once the CCJs and petitions start coming in. The Insolvency Service has moved earlier, to stages two and three, and my view is that creditors must do the same.

How to build a custom credit scoring model for UK risk teams

Find out how you can build a custom credit scoring model that works for your team with Company Watch.

Read the article

Where this leaves credit teams: why a clean check on day one isn’t enough

For credit and procurement teams, this means onboarding checks on their own won’t catch risk. A new customer can look fine on day one and start showing warning signs six months later. That’s why phoenix risk has to be tracked all the time, across the director and every company they’re linked to, not just checked once for the company in front of you.

This is the problem we built Vigilance™ to solve. It monitors the whole UK company population at Companies House against 25 fraud indicators across five risk zones, including a rapid succession zone that flags directors forming new companies around failed ones. To date it has flagged more than 58,000 rapid successions. The value is less in any single flag than in seeing the pattern while the company is still trading and you still have choices about credit terms.

Industry recognition for Vigilance™: CICM British Credit Awards 2026

Find out how Vigilance™ helped us win Supplier of the Year and Risk Management Team of the Year in the same awards cycle.

Read more

If I were leading a credit function now

I would do three things. First, extend monitoring from the counterparty to its directors and their wider company network. Second, set alerts on stage two and three signals, not only on CCJs and petitions. Third, treat a phoenix flag as a prompt for a conversation about terms and security, not an automatic decline, because many directors with a failed company behind them are running legitimate businesses.

The Insolvency Service has made its move and is now looking at companies while they’re still trading. If you want to see phoenix risk in your own customer and supplier base before it turns into a write-off, that’s exactly what Vigilance™ is built to do.