New research papers from the Central Bank relating to mortgage limits

[Brendan Burgess]

In “Have First Time Buyers continued to default less?”, Raffaele Giuliana highlights that there continues to be a lower likelihood of default for First Time Buyers (FTBs) compared with second and subsequent buyers (SSB). This is a key part of the rationale for the higher LTV requirement for 90 per cent of the FTB group under the measures. The re-testing of the original finding from 2013, which underpinned the calibration of the measures in 2014, forms part of the Central Bank’s commitment to ensure its macroprudential regime is supported by verifiable evidence.

In “Mortgage servicing burdens and LTI caps”, Jane Kelly and Elena Mazza explore the relationship between monthly mortgage repayments and borrowers’ disposable income in the context of the mortgage measures. The paper finds that the average household drawing down a mortgage in 2019 is spending almost one quarter of their net monthly income on mortgage repayments. Today, borrowers face lower debt servicing burdens than in the past, where figures were up to 40% in 2008 for some market segments. The mortgage measures, by regulating Loan to Income (LTI), place an effective maximum on the mortgage-service-to-income ratio that most households will face, thereby improving their resilience to loss of income for example.

Two further papers released today cover “A Measure of Bindingness in the Irish Mortgage Market”, by Robert Kelly and Elena Mazza and “Mortgage borrowers at the loan to income limit” by Edward Gaffney. The research finds that increasing shares of new mortgage holders, almost one in five, are borrowing at the limits. The research points that the mortgage measures have become increasingly binding in recent years, as would be expected in a market where house prices have been rising faster than incomes in the context of housing supply constraints. The share of borrowers close to their maximum drawdown is highest for FTBs, in Dublin and the Greater Dublin Area, those drawing down allowances, and those on household incomes between €70,000 and €80,000. The majority of these borrowers would have taken on higher levels of debt relative to their income in the absence of the measures, suggesting the measures are serving their function of improving household resilience and mitigating pro-cyclicality in the housing market.

 

[Brendan Burgess]

These four papers published today form part of the evidence of this year’s review. The outcome of the review will be announced on 4 December with the Financial Stability Review, following the decision by the Central Bank Commission.

 

[Brendan Burgess]

I have absolutely no idea what "bindingness" means.

The research finds that increasing shares of new mortgage holders, almost one in five, are borrowing at the limits. The research points that the mortgage measures have become increasingly binding in recent years, as would be expected in a market where house prices have been rising faster than incomes in the context of housing supply constraints.

I need to borrow €200k to buy a house, but the LTV or LTI rules limit me to €170k

So I am "bound" by the limits?

If the limits were looser, the bindingness would be lower?
If the limits were tighter, the bindingness would be higher?

Surely there must be a better word and better title for the paper?

“A Measure of Bindingness in the Irish Mortgage Market”,

They [The Central Bank rules] transmit through the direct lending channel,
whereby the level of bindingness alters the size and number of loans relative to a
counterfactual with no measures. This Note provides a measure of bindingness by
combining estimates of credit available and take-up for individual Irish borrowers.

I put this into Google Simplify and it came up with:

If the measures were not in place, people would borrow more. This Note measures the extent to which borrowers were limited in the amount they borrowed by the Central Banks limits (c) Google Simplify .


We derive a threshold
measure of bindingness, defined as the proportion of borrowers with take-up greater
than 90 per cent. This allows for a consistent measure of bindingness across subgroups and
time, which equates to the group of borrowers who likely took on less debt than they would
have had in the absence of the Measures
Results show the proportion of bound borrowers rose from 29 to 46
per cent since the introduction of the Measures. Recently, there has been a stabilisation of
bindingness for borrowers within the limits, reflective of the recent supply increase after a
prolonged period of supply shortage in the housing market. For the Measures to address
these cyclical pressures, bindingness will evolve over the cycle to mitigate the potential
negative impact on borrower affordability and resilience. The rich set of borrower characteristics
also allows for the estimation of some of the distributional impacts of the Measures.
In the first half of 2019, first time and Dublin buyers have significantly higher levels
of bindingness reflecting the high level of Dublin house prices relative to incomes compared
to other parts of the country.


I put this into Google Simplify and came up with

46% of borrowers now borrow more than 90% of what they could borrow.
Before the limits came in, only 29% did so.
More Dubliners and First Time Buyers borrowed more than 90% of what was available to them. (c) Google Simplify .

 

[House2020]

If you were the central bank policy maker would you make any change on Dec 4th?

 

[Brendan Burgess]

Just administrative changes. The exemptions are based on the calendar year. I would change them so that unused exemptions could be carried over or if they exceeded their exemptions in any year, the following year's allocation would be reduced.

I would also signal that 90% LTV is too high and that it will be reduced to 80% in the near future but not yet. However, if the government allowed repossessions the 90% for FTBs could be restored.

Brendan