Banks can now lend as much as they like at high income multiples, so long as the market overall does not. The regulator's own rules say any correction will not show up in the data for eighteen months.
Estimated reading time: 6 minutes
In short. PS19/26, the Prudential Regulation Authority and Financial Conduct Authority's final policy statement published 14 October 2026, permanently removes the 15% cap stopping any single mortgage lender writing more than that share of its new lending at high loan-to-income ratios, in force from 28 October [1]. The policy's own framing is that a market-wide 15% limit, set by the Financial Policy Committee, remains the safeguard [1]. Its companion supervisory statement says that if a correction is ever triggered, it "is not expected to be accurately reflected in data until six subsequent quarters have passed" and applies only to future lending, not mortgages already approved [2]. Neither document defines what counts as a breach serious enough to act on, and the regulators' own consultation response confirms they can choose to do nothing even above the limit. Two documented mistakes: four stars.
Since 2014, no UK mortgage lender has been allowed to write more than 15% of its new residential loans at a loan-to-income ratio of 4.5 or higher, a rule built after the 2008 crash to stop any single bank loading up on the riskiest borrowers. From 28 October 2026, that firm-level rule is gone. In its place sits a market-wide version of the same number, policed by a mechanism that, by the regulators' own account, takes up to a year and a half to tell anyone whether it is working.
A cap removed, a number retained
PS19/26 is blunt about what changed. "The regulators' policy removes the firm-specific 15% cap and provides lenders with greater flexibility to determine their individual high LTI lending strategies, in line with their risk appetite and business models," it states, adding that "while the aggregate high LTI flow is within a level consistent with the FPC's recommended 15% limit, lenders are able to lend above that level" [1]. Two respondents to the consultation asked the regulators to keep some form of individual cap, even a raised one. The regulators rejected that: "simply increasing the current firm-level cap, with or without tiering, would undermine their ability to maintain the aggregate high LTI lending flow at a level consistent with the 15% limit" [1].
This had already been running informally for over a year. A PRA statement from 9 July 2025 recorded that, following the same Financial Policy Committee recommendation, the regulator was "offering a modification by consent that will allow lenders to disapply the 15% limit with immediate effect" while the formal review was under way [3]. PS19/26 makes that interim arrangement permanent and gives it a rulebook.
The backstop that cannot see for eighteen months
The document's own account of what replaces the firm cap is where the claim thins out. The supervisory statement, SS2/26, sets out how a correction is meant to work if the market-wide aggregate breaches 15%: the PRA and FCA "may set expectations for any firm with an individual high LTI flow greater than 15%... to take steps to reduce its future flow," but only "if, at any point, the PRA and FCA judge that the aggregate high LTI flow is at a level not consistent with the FPC's 15% limit" [2]. Even then, correction is not automatic: it applies only "if the aggregate measure becomes inconsistent with the FPC's 15% limit on a persistent basis," a word the document never defines [2].
What happens once that discretionary trigger is pulled is slower still. "This adjustment should apply only to future lending decisions; firms are not expected to withdraw existing approved mortgage offers," the statement says, before adding the line that matters most: "Expected reductions in individual flow measures are not expected to be accurately reflected in data until six subsequent quarters have passed following the announcement" [2]. Six quarters is eighteen months. A policy sold on the strength of a retained aggregate limit cannot confirm, for a year and a half after flagging a problem, whether anything has actually been fixed, and the mortgages written while that clock runs are untouched regardless.
The discretion runs the other way too. PS19/26's feedback chapter records a respondent welcoming "the regulators' proposed approach of exercising judgement when setting adjustment expectations," noting "there may be circumstances in which the aggregate high LTI flow exceeds 15%, but the regulators judge that adjustments are not necessary to return the aggregate flow to that level" [1]. The regulators did not dispute that reading. A breach of the headline number, in other words, is compatible with no action at all.
Industry's own submissions flagged a version of this problem before the policy was finalised. Seven respondents warned that a quarterly measure would make it hard for lenders to "reliably predict single-quarter high LTI flow positions," risking "abrupt changes in lending criteria, more volatility in supply, and poorer outcomes for customers" [1]. The regulators answered by splitting the difference, a single-quarter measure to watch the aggregate, a four-quarter rolling measure to judge individual firms, without resolving the underlying tension either side had raised: fast enough to intervene on, or slow enough to be fair, not both.
The claims, tested
| The document's own words | What the record actually shows | Verdict |
|---|---|---|
| PS19/26 1.3: removing the firm cap is safe because "the aggregate high LTI flow" stays "consistent with the FPC's recommended 15% limit" [1] | SS2/26 3.5: a correction, once triggered, "is not expected to be accurately reflected in data until six subsequent quarters have passed," and applies only to future lending, not approved offers [2] | The backstop exists in name from day one; confirming it works, for lending written from that point on only, takes up to eighteen months |
| SS2/26 3.3 and 3.5: the regulators "may set expectations" to correct a breach "on a persistent basis" [2] | "Persistent" is undefined in either document; PS19/26 3.31 records the regulators accepting they may judge "adjustments are not necessary" even when the aggregate exceeds 15% [1] | The 15% limit's own enforcement trigger is discretionary and explicitly compatible with doing nothing in breach |
Credit where due
The regulators ran a real consultation and published the feedback in enough detail to let this piece test the policy against the objections raised to its own face, including from lenders who support it. UK Finance, representing the industry, called the change one its members "have long called for," crediting it with helping "support more creditworthy people into homeownership while maintaining important overall safeguards" [4]. The shift to a four-quarter rolling measure for individual firms, rather than the more volatile single-quarter basis first proposed, is a genuine concession to the seven respondents who warned the original plan would cause abrupt, unpredictable swings in lending criteria. And the policy does not hide its own mechanics: the six-quarter data lag and the discretionary trigger are both stated in SS2/26's own text, in public, not buried in an annexe only a specialist would find.
Verdict
Four stars, from two documented mistakes. The firm-level cap that limited any one bank's exposure to the riskiest mortgages is gone from 28 October 2026. The market-wide number that is supposed to replace it as the safeguard cannot confirm a correction has worked for eighteen months, and its own trigger for acting at all is left to regulatory judgement the documents do not pin down.
The star score counts two documented mistakes. First, PS19/26 frames the market-wide 15% limit as the safeguard replacing the firm-level cap, but SS2/26's own text says any correction "is not expected to be accurately reflected in data until six subsequent quarters have passed following the announcement" and applies only to future lending, not mortgages already approved. Second, neither document defines "persistent" as used in SS2/26's trigger for action, and PS19/26's own feedback chapter records the regulators accepting they may judge no adjustment necessary even when the aggregate exceeds 15%. Industry's own warning that a quarterly measure could cause abrupt, unpredictable lending swings is reported above as context, not counted separately, since the regulators partly addressed it with the four-quarter rolling measure for individual firms. Two falls in the 1 to 3 band: four stars; the bands are on the ratings page. This piece makes no finding about the personal conduct of any named individual; the claims tested are the regulators' own published words. All sources below were fetched directly on 8 November 2026.
Sources
- PS19/26 - High loan to income lending, Bank of England (Prudential Regulation Authority) and Financial Conduct Authority, published 14 October 2026
- SS2/26 - High loan to income lending, Bank of England (Prudential Regulation Authority), effective 28 October 2026
- PRA review of the LTI flow limit rule and offers interim modification by consent, Bank of England (Prudential Regulation Authority), published 9 July 2025
- FCA and PRA confirm permanent removal of high-LTI mortgage cap, Mortgage Solutions, 14 October 2026