The Bank of England’s Financial Stability Report produced by the Financial Policy Committee (FPC) said interest rates had risen since the requirement to apply the stress test was removed last June.
It went on to say the withdrawal enabled households to borrow larger amounts than before the ATR was in place, as the FPC expected. It also said as the base rate rose, it allowed borrowers who would have otherwise not been able to get a mortgage to be approved.
Its analysis suggested that the reduced buffer allowed households to borrow more on average.
The FPC’s loan data found that the withdrawal had resulted in a two to four per cent increase in loan sizes. It said the ATR removal offset the fall in loan sizes which was a result of higher rates.
Further analysis showed that until the end of 2022, the stress test removed increased total mortgage approvals by one per cent. As the base rate rose, this increased to an estimated one per cent to five per cent by August this year. The FPC said this equated to less than 0.5 per cent of the current mortgage stock, which was small compared to the 44 per cent drop in approvals seen since it was withdrawn.
Cap on reversion rates
The FPC said the withdrawal on the ATR also limited how much lender’s increased the stress buffer applied to reversion rates such as the standard variable rate (SVR).
Between August 2022 and September this year, the average SVR increased from 4.4 per cent to 7.8 per cent as lenders passed the base rate rises onto borrowers.
While this increased lender stress rates, the removal of the ATR meant the increase was smaller, the FPC said.
Insight from lenders suggests that stress buffers above reversion rates fell from around three per cent at the time the ATR was removed to one per cent by 2023. By September this year, the median stress rate applied to new loans was 8.8 per cent. The FPC said if the ATR was still in place, this would have reached nearly 11 per cent.
The FPC said the withdrawal of the ATR had a limited effect on the resilience of borrowers, which was expected.
It estimated that the share of new high loan to income (LTI) lending would have decreased to around 4.5 per cent to 4.9 per cent of the market if the test was in place, but instead by Q3, the share of mortgages at this LTI stood at 5.5 per cent compared with 10 per cent in Q2 last year. This is below the FPC’s limit of 15 per cent.
Additionally, the share of households with high debt-servicing ratios was just 0.1 per cent to 0.3 per cent higher due to the withdrawal and remained low at 0.8 per cent.