The CB research behind these proposals

Brendan Burgess · 11 Dec 2014

I think it's well worth reading the Central Bank's Economic Letter no 10

[broken link removed]

published on 1 October.

It answers a lot of questions I had, and raises quite a few as well.

It says in the introduction that "there is a sharp increase in the losses on defaulted loans for loans issued above 85 per cent LTV." although this is not obvious to me from the data.

Controlling for time effects, all loans issued in 2007 are examined, and the sample is split into first time buyers (FTBs) and movers (non FTBs). Figure 4 shows that first time buyer default rates are found to be sensitive to LTV at origination but less so to LTI. The default rate for loans with a LTV greater than 92 per cent (which comprised 40 per cent of 2007 FTB loans) is between 15 and 20 per cent, with the maximum rate of default occurring at LTI ratios of 3.8 to 4.3 times.

For movers, there is a direct relationship between both LTV and LTI
and loan default. For individuals in the top 20 per cent of both originating LTV and LTI, the default rate is greater than one in four loans.

Given this relationship between originating LTV and LTI ratios and later defaults, it is clear that limiting high LTV and LTI lending will improve bank resilience and help protect borrowers from movements in property prices.

There is a very strong relationship between LTV and LTI, so limiting one, will probably substantially limit the other.

Figure 6 shows there is a strong positive relationship for low to moderate LTV and LTI levels for the whole sample. This relationship breaks down for originating LTVs between 80 and 90 per cent, followed by a strong relationship for the very highest LTV loans issued.

Given the large portion of lending in the 80 to 90 per cent LTV region, a combination of both tools is required to reduce risk on the repayment and
collateral channels.

Brendan Burgess · 11 Dec 2014

OK, I have had another look.

What is really interesting from Figure 2 is that the level of defaults rises in a straight line to 70%-80% LTVs and then levels off.

Loans of 90%to100% LTV are no more likely to default than loans of 70% to 80%.

But the lenders' losses rise rapidly from around 55%.

44brendan · 11 Dec 2014

It's somewhat understandable that those with some level of equity in their properties are less likely to default than those with little or negative equity. Perhaps that is simplistic as I haven't the time to read the data and LTV may just reflect historic position at approval rather than at time of default!

44brendan · 11 Dec 2014

OK I see its LTV at loan origination rather than default date! It reflects that those with reasonable "skin in the game" are less likely to default. Not surprising in itself!

The CB research behind these proposals

Brendan Burgess

Founder

Brendan Burgess

Founder

44brendan

Registered User

44brendan

Registered User

The CB research behind these proposals

Brendan Burgess

Founder

Brendan Burgess

Founder

44brendan

Registered User

44brendan

Registered User

Privacy & Transparency

Privacy & Transparency