|
We recently published an article estimating that around £750m of tenancy deposit money in England and Wales may be sitting dormant or unclaimed.
|
|
It has started a debate, which is exactly what we hoped it would do.
|
|
Eddie Hooker at mydeposits responded with a very different figure. The schemes themselves put unclaimed custodial deposits at no more than £20–25m, ring-fenced against future claims and reported to MHCLG.
|
|
That looks like a very large discrepancy on face value, but it is possible for both figures to be right, because they are measuring different things.
|
|
We Are Measuring Two Different Things
|
|
The £20–25m figure and our £750m estimate are not answering the same question.
|
|
A scheme can only classify a deposit as unclaimed once it knows there is something to claim: the tenancy has ended, repayment has been initiated, but the money has not been released. That population is real, and £20–25m may well be an accurate measure of it.
|
|
We are looking at something different.
|
|
What happens where the tenancy has ended but nobody tells the scheme? No repayment is requested. No release takes place. As far as the scheme’s records are concerned, the deposit is still attached to a live tenancy.
|
|
This is particularly problematic in insured deposits, where schemes have no independent way of knowing whether a tenancy has ended or whether the deposit has already been returned unless somebody updates the record. But the principle applies custodially too. If neither party starts the repayment process, the deposit simply continues to show as live.
|
|
It never reaches the point at which it can be classified as “unclaimed”, because there has been no event to trigger that classification. That is the population our estimate is trying to identify: registrations which remain on the books, but for which there may no longer be a tenancy behind them.
|
|
That distinction matters, because it explains how the schemes can report £20–25m of known unclaimed money while a much larger population of potentially dormant registrations sits outside that definition.
|
|
What We Counted
|
|
Government dwelling stock estimates put the private rented sector at just over 5.2 million households across England and Wales. The English Housing Survey finds that 78% of private renting households paid a deposit when they started their current tenancy.
|
|
On that basis, you would expect an estimated 4.1 million protected deposits to exist. The schemes report about 4.7 million. Here are the specifics:
|
| Private rented households, England and Wales |
5,214,900 |
| Proportion who paid a deposit |
78% |
| Protected deposits we would expect |
4,067,600 |
| Protected deposits the schemes report |
4,706,500 |
| Registrations above the level implied by the household data |
638,800 |
| Indicative value at reported average deposit |
~£750m |
|
|
|
That is not necessarily a hidden pot of cash. It is the value associated with a gap between two counts that ought, within reasonable tolerances, to reconcile.
|
|
Some of that gap will have perfectly ordinary explanations. But until those registrations are reconciled against live tenancies and actual deposit balances, nobody can say how much is administrative residue and how much represents money still sitting in the system.
|
|
In client accounting, that is an unreconciled balance. And an unreconciled balance of this size deserves an explanation. Nobody can say exactly how much is stale administration, how much is genuinely unclaimed, or how much money is still sitting in the system. That uncertainty is not a weakness in the argument, it is the argument.
|
|
A Sense-Check from Scotland
|
|
Scotland is fully custodial, so the unclaimed pile can actually be measured. The Scottish Government reports around £7m unclaimed against a market worth roughly £194m: about 3.6% of the book.
|
|
England and Wales holds around £2.5bn custodially, so £20–25m represents about 1%.
|
|
That is a striking difference. Broadly the same tenants, tenancies and, in several cases, the same operators. I do not think Scottish tenants are four times more forgetful, so what explains this discrepancy?
|
|
Scotland has also now defined the dormancy term in law. Under the Housing (Scotland) Act 2025, a deposit can be treated as unclaimed where no repayment application has been made within five years. England and Wales has no equivalent definition.
|
|
Why This Matters Now
|
|
The government is weighing the future shape of deposit protection, including a possible move to custodial-only.
|
|
That is exactly why this matters now. If insured deposits are going to be scrapped, we need to know which protected deposits still relate to live tenancies and which do not. You cannot sensibly redesign or migrate a system until you know what is actually sitting inside it.
|
|
If the plan is simply to let insured deposits run off over time, that does not solve the underlying problem. It postpones it. Stale registrations, unidentified liabilities, and unreconciled balances do not disappear just because no new insured deposits are being added.
|
|
The risk is that we carry an increasingly opaque legacy book forward for years, without ever properly establishing what is still live, what has already been settled and what money may still be owed.
|
|
Three things would help, and none requires anyone to accept our number:
|
| • |
Publish what already exists. The schemes report unclaimed volumes to MHCLG, but those reports are not public. |
| • |
Define dormancy. England and Wales has no legal definition and no clear rule for what happens to genuinely unclaimed money. |
| • |
Build a reconciliation mechanism. Nothing currently reconciles scheme records against live tenancies and, in the case of insured, deposit funds sitting in agents’ accounts. |
|
|
That last point is where agents come in. National reconciliation is a policy question. Agency-level reconciliation is not. Take your deposit register, match it against current tenancies and against what the schemes hold in your name, and see what falls out.
|
|
It rarely produces anything alarming. It routinely finds registrations for tenancies that ended years ago. We can help any agent with that, and it’s completely free.
|
|
If reconciliation proves our estimate wrong, good. We hope the problem isn’t as alarming as the numbers appear to be today. But what matters is knowing the answer before reforming the system, not afterwards.
|
|
Sources: MHCLG dwelling stock estimates (England, 31 March 2025) and Welsh Government dwelling stock estimates; English Housing Survey 2023–24, rented sectors, Annex Table 2.25; TDS Statistical Briefing; Scottish Government, Tenancy deposit schemes; Housing (Scotland) Act 2025. Full workings and references available on request.
|
|
