Key Post How much more will mortgage defaults cost the taxpayer?

[Brendan Burgess]

Summary

• The potential defaults on home loans and buy to lets have been fully provided for and it is unlikely that the taxpayer will be required to put more money into the state-owned banks.
• This assumes that the current economic structures continue. If, for example, Ireland exits the eurozone, all bets are off.
• It also assumes that new insolvency legislation provides for the write off of some negative equity for home loans in arrears, but that it does not provide for an across the board write-down of negative equity for those who are not in arrears.
• It also assumes that the state does not subsidise debt write downs for the mortgage lenders it does not own.
• The main factors which determine the levels of the banks' losses are
  1. the level of unemployment
  2. the interest rate
  3. future falls or rises in property prices
  4. The borrowers' efforts to repay their mortgage
• The last bank recapitalisation over capitalised the banks. It's quite possible that the next one will do the same and require further provisions.
• The outstanding arrears on home loans due to the state banks at the moment is around €650 million.
• If the state-owned banks transfer their trackers to the IBRC at fair value, the capital requirements of the state-owned banks would increase by around €4 billion. The actual amount which the state would have to put into the banks would depend on the overall capital position of the bank.

 

[Brendan Burgess]

There is a very serious arrears problem. This is most serious for the 60,000 or so familes who are in arrears. The taxpayer has recapitalised the banks. But is it enough?

The government owns AIB, PTSB and the EBS. So it will bear the costs of any default on these. I have excluded Irish Natiowide because its loan book is very small, and the losses will be dwarfed by the losses on development loans. The shareholders will fund any losses incurred by Bank of Ireland - the government owns only 15% of the shares. The losses in Ulster Bank, KBC, NIB etc. will be borne by their shareholders and will not affect the Irish taxpayer.

The state owned banks published annual reports which give huge detail on their mortgages, arrears, and loan to values. The following is a summary for the three state owned banks for both owner occupied and buy to let mortgages.

Table 1 Mortgages held by AIB, EBS & PTSB

|Total|Owner occupied|Buy to let
Positive equity and not in arrears|€25 billion|€21 billion|€4 billion
Positive equity but in arrears| €3 billion|€2 billion|€1 billion
Negative equity but not in arrears|€31 billion |€23 billion|€8 billion
Negative equity and in arrears|€8 billion|€4 billion|€4 billion
Total mortgages outstanding |€67 billion|€51 billion|€16 billion

.
Table 2 Estimated negative equity AIB, EBS & PTSB

|Total|owner occupied| buy to let
Total mortgages in negative equity|€39 billion |€27 billion|€12 billion
Estimated negative equity|€13 billion |€9 billion|€4 billion

.

 

[Brendan Burgess]

How much will mortgage defaults cost the taxpayer?

So how much are the banks likely to lose on home loans?

On a home loan, a bank has two sources of security

• the income of the borrower
• the underlying property value

If the borrower can't keep up the repayments, but the property is worth more than the mortgage, the bank won't lose.

If the borrower is in severe negative equity, but can keep up the repayments, the bank's loan will be repaid.

The lender will lose money when the income of the borrower is insufficient to keep up the mortgage payments and the property is sold for less than the outstanding mortgage.

The factors affecting the levels of losses include

• the level of unemployment
• the interest rate
• the rate of income and other taxes
• government subsidies e.g. Mortgage Interst Supplement
• future falls or rises in property prices
• The borrowers' effort to repay their mortgage

Table 3 Arrears outstanding on owner occupied mortgages at AIB, EBS &PTSB at 31 March 2012

Total mortgages outstanding| €51 billion
Mortgages in arrears|€6 billion
Estimated arrears |€650 million

.
The total arrears on all home loans for all lenders at 31 March 2012 was €1.3 billion. I have estimated that the state owned banks' share of this is €650 million.

Table 4 Estimated losses on owner occupied loans

|||Estimated losses
Positive equity and not in arrears|€21 billion|2.5%|€500 million
Positive equity but in arrears| €2 billion|5%|€100 million
Negative equity but not in arrears|€23 billion |5%|€1 billion
Negative equity and in arrears|€4 billion|30%|€ 1.2 billion
Total mortgages outstanding |€51 billion||€3 billion
Provisions already made |||€4.5 billion
Additional provisions needed|||nil

Under the Central Bank recapitalization, the banks have already provided for €4.5 billion of losses due to defaults.

The mortgages which are most likely to cause a problem for the banks are the ones who are in both negative equity and in arrears. They owe around €4 billion and the property behind it is worth around € 2.8 billion. Most of these will recover in time and pay off their mortgages in full. Some are unsustainable. The property should be sold and the negative equity written off. To be conservative, I have assumed that the lenders will lose the entire negative equity on all loans in arrears and negative equity.

There must be a significant risk from the €23 billion who are in negative equity but not in arrears. Some will go into arrears and some will default. I have estimated a loss of 4% on these.

Two other factors could reduce the losses further, although I have made no allowance for these factors in the above calculations.

• The owners of the securitised mortgages have to bear some of the losses.
• In the case of EBS, Mortgage Indemnity Guarantee insurance should also reduce some of the losses.

 

[Brendan Burgess]

Table 5 Alternative method of estimating cost of defaults on home loans

Homes in arrears over 90 days at present|60,000
Assume it doubles |120,000
Total defaults 25%|30,000
State owned banks' defaults|15,000
Average mortgage balance| €250,000
Average shortfall on sale (70%)| €175,000
Cost of defaults | €2.6 billion

This is less than the €3 billion estimate above and 60% of the provisions already made.

The current level of provisions could cope with up to 50,000 defaults or repossessions.

Note. At 31 March, there were 77,630 accounts in arrears with a total balance of €15 billion. This represensts about 60,000 homes with an average mortgage balance of €250,000.

 

[Brendan Burgess]

A third way of estimating the cost of defaults on home loans

Table 6 Analysis by LTV ratio of of owner occupied lending that is more than 90 days past due

|total|AIB & EBS|Permanent tsb
|€m|€m|€m
less than 50%|439|223|216
50% - 100%|1,503|853|650
101% -150%|2,468|1,439|1,029
>150%|1,772|957|815
Total|6,182|3,472|2710

Assume all properties in arrears over 90 days are repossessed and all negative equity is just written off

Table 7

|mortgage balance |estimated negative equity
|€m|€m
101% -150%|2,468|822
>150%|1,772|886
Total|6,182|1,708

It would cost the banks €1,708 m to write off all the negative equity on mortgages in excess of 90 days arrears.

 

[Brendan Burgess]

So how much are the banks likely to lose on buy to let mortgages?

The factors affecting the losses include the factors mentioned above plus the level of rental income.


Table 8 Arrears outstanding on buy to let mortgages at AIB, EBS &PTSB at 31 March 2012

Total mortgages outstanding| €16 billion
Mortgages in arrears|€5 billion
Estimated arrears |€1.25 billion

.

The arrears level on buy to let mortgages is around 25%

Table 9 Estimated losses on buy to let loans for AIB, EBS and PTSB

||Estimated losses
Positive equity and not in arrears|€4 billion|€80 million
Positive equity but in arrears| €1 billion|€50 million
Negative equity but not in arrears|€8 billion |€400 million
Negative equity and in arrears|€4 billion|€ 1.2 billion
Total mortgages outstanding |€16 billion|€1.7 billion
Provisions already made ||€2.4 billion
Additional provisions needed||nil

The 25% arrears figure overstates the buy to let problem.

An investor has two sources of income to pay the mortgage - the rent and their other income.

Around 60% of buy to let mortgages are on cheap trackers. The rental market is very good at the moment and the state pays generous rent supplement. So the majority of landlords are well able to pay the interest on their loans from the rental income.

The problem facing landlords is that they are unable to make the capital repayments on their loans. So they are paying the interest proportion and often a bit more, but they are going into arrears for the capital element. This is not very serious. The mortgage balance is not increasing. I have been astonished by the attitude of some property investors. They tell the lenders that they can't pay any more than the rent coming in. They seem to have made some mental partitioning of their rental income and their mortgage repayments. But they should be using their other income to meet their repayments.

Although the arrears are very high, this won't translate into correspondingly huge losses for a few reasons. Investors don't have the same protection which borrowers with home mortgages have. In many cases, the investment property is cross secured on another investment property or the borrower's home, so the negative equity figures are overstating the problem.

 

[Brendan Burgess]

left blank

 

[Brendan Burgess]

The tracker problem is potentially bigger and much more difficult to quantify

The three banks have around €30 billion in trackers between them. In normal circumstances, they would probably be losing around 2% on each tracker or €600m a year. But these are financed by the ECB at 1%, so they are not actually losing money on them at the moment.
If the banks retain their trackers, their future profitability will be reduced, but I understand that they don't need to make any provisions against the future losses.

If the banks shift €30 billion in trackers to the IBRC, presumably they will also shift €30 billion in cheap loans from the ECB.

If the trackers are sold to the IBRC at fair value, presumably the IBRC should get a discount on their nominal value. That discount should be around 20% which would be €6 billion. The banks would need to be recapitalised by a further €6 billion to provide for these losses.

There is some overlap between the losses due to defaults and the losses due to the trackers. So the €6 billion would be reduced probably to €4 billion.

 

[hastalavista]

Brendan,
Interesting data here.

Are there any public numbers available for the maturity profile of the debt and then has anyone overlaid a yield curve on the maturity profile to show what might happen say in 2017/8 0r 2022/2023 if ECB rates go back up to the 4% or so they hit before the bust

To put it in context, at the peak, my interest only exceeded what I am now on with a 33%/67% mix of i only and p + i over 15 years.

My quarterly interest bill is the same as a monthly payment so 3 steps forward/one step back on the treadmill.
Hasta

 

[Brendan Burgess]

I don't think that there is although one could estimate it.

To put it in context, at the peak, my interest only exceeded what I am now on with a 33%/67% mix of i only and p + i over 15 years.

My quarterly interest bill is the same as a monthly payment so 3 steps forward/one step back on the treadmill.

I don't really follow this. What I think you are saying is

"My monthly repayments are artificially low now. They are less now than the interest element of the repayment was when the interest rates were 4%"

The primary determinant of arrears is the level of unemployment. It seems to be more significant than the actual interest rates. If you are unemployed, you can't pay your mortgage, even if the interest rate is very low. If you are employed, you can usually pay your mortgage or repayment, even if the rate is high.

Brendan

 

[Shivvers]

Summary

• The potential defaults on home loans and buy to lets have been fully provided for and it is unlikely that the taxpayer will be required to put more money into the state-owned banks.

Not necessarily, according to the Financial Regulator today. I am not allowed to post links but it is the top story on RTE Business news that Elderfield says the banks need another 3 billion or 4 billion over next few years. This is just to meet new Basel regulations. Elderfield says he hopes they will be able to raise private capital. BOI maybe... but AIB and PTSB haven't a hope.....

When the wall of mortgage defaults hits ( and it is coming, no doubt about it), they will need a lot more than 3 or 4 billion.....

 

[hastalavista]

Shivvers
While BB is big enough and bold enough to defend himself, and I am sure he will do so in due course, 2 points occur:

BB was talking in the context of the potential defaults.... and the need for more capital to underwrite these.

Elderfield say that new capital will be required to meet Basle III requirements
as described below, which is not what BB was referring to.

[The package, known as Basel III, sets a new key capital ratio of 4.5 per cent, more than double the current 2 per cent level, plus a new buffer of a further 2.5 per cent. Banks whose capital falls within the buffer zone will face restrictions on paying dividends and discretionary bonuses, so the rule sets an effective floor of 7 per cent. The new rules will be phased in from January 2013 through to January 2019.
Banks will be required to triple core tier one capital ratios from 2 per cent to 7 per cent by 2019. This ratio measures the buffer of highest quality capital that banks hold against future losses.]

 

[Shivvers]

Shivvers
While BB is big enough and bold enough to defend himself, and I am sure he will do so in due course, 2 points occur:

BB was talking in the context of the potential defaults.... and the need for more capital to underwrite these.

Elderfield say that new capital will be required to meet Basle III requirements
as described below, which is not what BB was referring to.

Why would he need to defend himself? I wasn't criticising anybody. I was just pointing out that the banks would need more capital, and I believe I said it was for Basel in my post.

But I also believe mortgage defaults alone, Basel or no Basel, will also mean the Irish banks need more capital, and lots of it.......

Have you ever seen No Country for Old Men?
"You can't stop what's coming........"

 

[Brendan Burgess]

When the wall of mortgage defaults hits ( and it is coming, no doubt about it), they will need a lot more than 3 or 4 billion.....

Hi Shivvers

The reason I put in so much time into doing this Key Post was to counteract these sort of broad brush statements. I have taken the numbers in some detail. I have allowed for a doubling of the current level of arrears and I have shown that the current level of provisions for mortgage default is adequate.

I am not sure what "a wall of mortgage defaults" is? Can you quantify it? Then you can work out what the cost will be to the taxpayer.

 

[Shivvers]

Brendan,

I find your figures somewhat hard to follow and also very hard to accept. I can't work out the methodology/assumptions you have used to estimate the level of losses you believe the banks will accrue. I also believe you are woefully underestimating the real levels of mortgages in difficulty.

Take table 4, for example. All estimates no matter who does them are essentially guesswork, but I think your guesses are overly optimistic. Coming down from the top in table 4, you seem to have estimated losses of, respectively, about 2.5%, 5%, 4% and 30%. Is this right? Where did you get those figures from? What economic assumptions re house prices/unemployment etc underpin those estimates?

They look to me like sky blue thinking. I'll take one category as an example. Losses of just 5% on homes that are not in NE but are in arrears? That appears fanciful. A further 10% drop in house prices (and that's conservative), knocking thousands of those homeowners into NE, would decimate that provision.... And when the banks start selling all the homes that they would have to repossess, house prices would fall like a stone, making the losses worse.

You say you have assumed that the number of 90 day arrears will eventually double to 120,000. I say that your doubled figure is probably closer to where the number of overall distressed mortgages is today.

30 days, 90 days, it doesn't matter.... they are in distress and unless employment picks up (no sign of that), most of the current 30 dayers (who you ignore altogether in table 5) will become 90 dayers soon enough. The number of mortgages in arrears of 30 days or more, added to the number that have been restructured in 2009-2011, comes to almost 110,000. That is the real number of mortgages in distress today, not 60,000 you have used. Those who were restructured in 09-11 were all put on interest only. When those interest only periods end, they will be in distress again, because economically and in terms of house prices, nothing will have changed.

There is no point in me doing a back of the envelope calculation to estimate what I think the losses will be. I will only get it wrong, like everyone else has so far. To assess whether the banks will need more capital, we need only measure what we have seen in the data against the worst-case scenario of the PCAR last year. If we are surpassing the worst case (against which the banks were capitalised) the banks will need more cash.

According to the banks provisioning, by the end of this year, they will already have recognised losses approaching 60% of the worst case envisaged under PCAR for the period 2011-2013. But the arrears problem is accelerating, as the latest central bank figures show, not slowing down. Throw into the mix the effect of the upcoming personal insolvency legislation, together with this breakneck acceleration we are seeing in the figures as people exhaust their savings and redundancy cheques from 2009 2010 (this is the 'wall' I was referring to), then by the end of 2013, the PCAR worst case scenario will be easily breached. It is clear from the trajectory of the figures.

Thus, the banks will definitely need more capital. They have not even begun to get to grips with the mortgages problem, because they have only just finished dealing with the developers' problem. Nobody on earth know how much more cash they will need, which is precisely why the banks are still unable to fund themselves and remain hooked on ECB cash.

You say the current provisions could cope with 50,000 defaults or repossessions. Morgan Kelly (who has been scarily accurate with all his estimates and figures, right throughout the crisis - go back and read all his stuff, it is unbelievably accurate) has estimated that there will be about 200,000. Unless (for the first time) Kelly is way way off, the current provisions are a sick joke and it is not inconceivable that the banks are in danger of completely collapsing all over again. I will put my trust in the figures of the guy who has gotten it right all the way through. Even if he is half wrong, the provisions are still too small.

You have also discounted Bank of Ireland entirely because it is majority private owned. That doesn't mean the state won't have to rescue it again. Remember, when the state stepped in to rescue the banks last time, they were all 100% privately owned. The state had to step in because the private shareholders were wise enough not to cough up because the problem was so big. Don't assume BOI's 85% owners will cough up this time round when the mortgage 'wall' hits. They may walk away, just like last time. If they do, the state will have to step in once again. Although I concede that scenario is probably more unlikely than likely.

So, to conclude, when you add up the €3-4 billion Elderfield is estimating they will need just for Basle 3; the acceleration of the mortgage crisis and the mass repossessions that will ensue after the insolvency legislation is brought in this year; the drop in house prices that will kick in when the banks flood the market with all those homes they repossessed (they have only repossessed 1,000 homes so far - what a joke); and the fact that it is now certain that the banks will breach the PCAR worst case scenario..... in my opinion, it is absolutely certain that the banks will need more state cash. And we don't have it to give to them......

 

[ballaboy]

One interesting fact from your figures is that 17% of the total mortgages are in arrears,you can say one in five,not very promising.

That figure is only for State backed Banks,add in sub prime and other Foreign Banks,would seem the figure could rise to 25% in arrears and rising

 

[Brendan Burgess]

Hi Shivvers

I am glad to see my figures challenged in such depth.


Firstly , let me stress a point which most people seem to overlook

The lender will lose money when the income of the borrower is insufficient to keep up the mortgage payments and the property is sold for less than the outstanding mortgage.

Most commentators seem to miss this point.

90% of borrowers are not in arrears. They are keeping up their mortgage repayments. So although over 50% of them are in negative equity, it doesn't really matter.

Negative equity only matters when the borrower is no longer able to make their mortgage repayments.

So what will determine the level of defaults? The biggest factor will be the level of unemployment. It is currently at 14.4%. It is not forecast to rise dramatically. In fact, the OECD expects it to fall marginally.

What is odd about the Irish experience is that arrears continue to rise, although unemployment peaked about 12 months ago (?). Normally there is a 6 month time lag as people use up their savings.

 

[Brendan Burgess]

Secondly, a general point about forecasting.

There is no point in me doing a back of the envelope calculation to estimate what I think the losses will be. I will only get it wrong, like everyone else has so far.
...
the acceleration of the mortgage crisis and the mass repossessions that will ensue after the insolvency legislation is brought in this year; the drop in house prices that will kick in when the banks flood the market with all those homes they repossessed

Economists, and people generally, are woeful at forecasting. I wouldn't pay much attention to what they say.

But bank capitalization is a critical subject, so we have to make forecasts about what might go wrong in a worst case scenario. And that is what I have done.

I have exaggerated all the problems.

• I have shown a doubling of the arrears levels, although I think that they are unlikely to double
• I have assumed that 25% of them will default - although I believe that most of those in arrears will recover in time
• I have assumed that the banks will lose 70% of the mortgage value on those defaults, although the average loss is likely to be around 44%, and not 70%
• I have made no allowance for the fact that the owners of the securitised loans will take some of the cost, especially in the worst case scenario.
• I have made no allowance that the providers of Mortgage Indemnity Guarantees will take some of the hit.

Then I approach the estimate from two different angles to see how the figures match. In both cases, the losses come out well below the existing provisions.

 

[Brendan Burgess]

Now, some detailed responses

Losses of just 5% on homes that are not in NE but are in arrears? That appears fanciful. A further 10% drop in house prices (and that's conservative), knocking thousands of those homeowners into NE, would decimate that provision.

These people are not in arrears. The underlying value of the house is irrelvant if they can continue to pay their mortgage.

You say you have assumed that the number of 90 day arrears will eventually double to 120,000. I say that your doubled figure is probably closer to where the number of overall distressed mortgages is today.

I am using the Central Bank's figures for 90 days in arrears. These will be the source of the defaults. Many more are struggling, but they are paying their mortgage and will continue to do so, unless they lose their jobs.

Those who were restructured in 09-11 were all put on interest only. When those interest only periods end, they will be in distress again,

If you check out the accounts, you will see that this is incorrect. Around 50% of PTSB's rescheduling were term extensions and reduced payments but more than interest only. For AIB, it was around 30%

But a more fundamental point is that if someone can pay the interest on their mortgage, they are ok. Their home is not going to be repossessed while they are able to do so. Take a Bank of Scotland full term interest only customer. If they can pay the intersest, they will never be in arrears. Someone with the same profile but with a 20 year repayment mortgage will appear in the arrears statistics. People will lose their home, when they are unable to afford to pay the interest and so the balance outstanding keeps rising.

But the arrears problem is accelerating, as the latest central bank figures show, not slowing down.
...
this breakneck acceleration

Here are the actual arrears figures

|31 March 2010|31 March 2011|31 March 2012
Number of mortgages in arrears over 90 days| 21,187|49,602|77,630
Increase||28,415|28,028
% in arrears over 90 days|2.7%|6.3%|10.2%

So, while it has increased, it has not been "accelerating", much less "breakneck acceleration". At this rate of increase, it would take over two more years to double.

Morgan Kelly (who has been scarily accurate with all his estimates and figures, right throughout the crisis - go back and read all his stuff, it is unbelievably accurate)

He has made 4 separate forecasts on the costs of mortgage defaults/debt forgiveness ranging from €6 billion to 200,000 defaults. One of these might be right. 3 will definitely be wrong. Check out Why Morgan Kelly is spectacularly wrong on the cost of mortgage defaults By the way, I respect his opinion which is why I sent him a link to this thread last week.

You have also discounted Bank of Ireland entirely because it is majority private owned. ... If they do, the state will have to step in once again. Although I concede that scenario is probably more unlikely than likely.

The Bank of Ireland has around half the average level of mortgage arrears of the other banks. They will probably rise, but their current levels of capital would comfortably withstand a doubling in arrears and defaults.

the mass repossessions that will ensue after the insolvency legislation is brought in this year; the drop in house prices that will kick in when the banks flood the market with all those homes they repossessed

I have been the loudest voice calling for the level of unsustainable mortgages to be recognised. I have estimated to be 10,000 homes, but it may be as high as 20,000 homes. They will take place over a period of time and won't all hit the market on 1 January 2013.

The insolvency legislation is the big uncertainty. If it allows people in negative equity who can otherwise repay their loans, to hand back the keys and wipe out their negative equity, then the losses would be much higher. I don't expect that to happen.

 

[Brendan Burgess]

I think that it's worth looking at Table 5 again

Table 5 Alternative method of estimating cost of defaults on home loans

Homes in arrears over 90 days at present|60,000
Assume it double |120,000
Total defaults 25%|30,000
State owned banks' defaults|15,000
Average mortgage balance| €250,000
Average shortfall on sale (70%)| €175,000
Cost of defaults | €2.6 billion

This is less than the €3 billion estimate above and 60% of the provisions already made.

The current level of provisions could cope with up to 50,000 defaults or repossessions.

We all know people who

• bought their house with a 100% mortgage
• at the top of the market in 2006
• which has since fallen by 70%
• who now can't make their repayments

But the danger is that we assume that all borrowers in arrears meet this profile. Of course they don't.



Most borrowers in arrears

• Did not buy at the top of the market
• so their property has not fallen by 70%
• And they did not have a 100% mortgage

A more typical or average case would be as follows

House price|100
90% mortgage|90
house value today after 50% fall|50
bank's loss on repossession|40
Percentage loss 40/90|44%

But to be conservative in my estimates, I have assumed a 70% loss on all defaults.