Savings Confusion and an Investment Collapse

It has been a confusing day for reports on the savings behaviour of Irish households.  One story on the Irish Times site tells us that:

The savings index fell from 98 to 81 in December, the lowest ever level since the index's inception in April 2010, as increased negative sentiment towards the economic environment discouraged saving.

While another story posted to the same site says:

The institutional sector accounts, which brings together information on the activities of households, businesses and the Government, show that the gross amount of household savings was €11.92 billion for the first three quarters of 2012. This is more than the €9.3 billion of total savings recorded during 2011.

The derived gross savings ratio increased by 14.5 per cent in the second quarter to 16 per cent in the third quarter last year. This ratio expresses household savings as a percentage of gross disposable household income.

Of course, the Savings Index from Nationwide UK (Ireland) and the Institutional Sector Accounts from the Central Statistics Office are dealing with very different meanings of the term ‘saving’.  The focus here is on the measure produced by the CSO which relates to the gap between disposable income and consumption expenditure and is shown in the following chart.  The savings rate is the percentage of gross disposable income (plus an adjustment for pension funds) that is not spent on consumption.

It is often suggested that the current rate is somehow “too high”.  A recent Irish Examiner report states that:

At a press briefing to announce the end-of-year exchequer figures on Thursday, Finance Minister Michael Noonan said that the national savings rate was now 14%, compared with 1.5% during the Celtic Tiger years. If people spent more and brought the savings rate down to 10%, then it would add 1% to growth, the minister said.

Is Ireland’s savings rate “too high”?  The Q2 2012 figure shown in the above chart is 14.5%.  Eurostat figures show that the EZ17 figure for the same quarter was 12.9%.  In fact, after going above the EZ17 rate  for the first time in Q1 2009 the Irish savings rate was below the EZ17 average for the next three years.  It is only in Q2 2012 that the rate has exceeded the EZ17 average.

While the Irish household savings rate has shown some volatility over the past decade, the most dramatic changes in household behaviour have not been to do with savings but to do with capital formation (investment in non-financial assets).  In national accounts, household investment mainly consists of the purchase of new dwellings and the renovation of existing dwellings. 

Household investment has gone from close to 30% of gross household disposable income in 2006 to less than 5% now.  Again the comparison to the EZ17 average is revealing.

Compared to the EZ17 average, the household investment rate in Ireland has gone from being significantly “too high”, to what seems to be a small bit “too low”.  The household investment rate fell dramatically from 2007-2009 and has continued to fall, albeit more slowly, since then.  By Q3 2012, the household investment rate in Ireland had dropped to a low of 3.6% of gross disposable income.  The EZ17 average is around 9%.

The gross investment rate of households is defined as gross fixed capital formation divided by gross disposable income, with the latter being adjusted for the change in the net equity of households in pension funds reserves.

In the first three quarters of 2006, the household sector in Ireland invested €19.5 billion in gross capital formation in non-financial assets (mainly buying new houses but also paying stamp duty on second-hand houses).  For the first three quarters of 2012, the equivalent figure is €3.8 billion, a drop of more than 80% from the peak. 

If the EZ17 average of a 9% household investment rate applied in Ireland then the 2012 figure for the first three quarters would be €6.2 billion.  The quote from Michael Noonan above shows that he thinks that reducing the savings rate to 10% (below the EZ17 average) would add 1% to GDP growth.  However, if the household investment rate rose to 9% (just equal to the EZ average) it would add 2% to GDP growth.

Where is this saving going?  Excluding the adjustment for pension funds the household sector “saved” €31 billion in the two and a half years from the start of 2009.  As it wasn’t used to fund consumption where did this €31 billion go?

The Central Bank produces quarterly financial accounts which show the changes in the financial position of the household sector.  For the household sector, the key measures are currency and deposit assets and loan liabilities.

Since the end of 2008 the currency and deposit asset of the Irish household sector has hardly moved.  It was €120 billion in Q4 2008 and was €124 billion in Q2 2012.  Over the same period the loan liabilities of the household sector declined from €204 billion to €180 billion.

Since 2009, Irish households have not used €31 billion of their disposable income to fund consumption expenditure.  In the main this unspent money has been used to pay down debt rather than accumulate deposits.  The Nationwide UK (Ireland) survey shows that Irish households are not building up “savings”; the CSO data clearly show that Irish households are “saving” but that this is going to pay down debt.

The Irish background data used above in the non-financial charts and some additional comments are below the fold.

First, the components of Gross Household Disposable Income which is the sum of wages, self employed earnings/mixed income, property income and social transfers less taxes on income and wealth.  Click to enlarge.

For the first three quarters of 2012, gross household disposable income was €66.4 billion, which is a 4.1% increase on the €63.8 billion recorded for 2011.  This increase in income is not reflected in household consumption expenditure which continues to flat-line.

The factors giving rise to the increase in gross disposable income are in the following table (all in €millions).

Wages received, non-wage earnings and mixed income, and social benefits all contributed to the increase in gross disposable income over the year.  Offsetting factors were the reduction in net property income (mainly interest) and the increase in taxes on income and wealth.

Compared to the peak in 2008, gross household disposable income has fallen 9.3%.  Over the same period there has been an estimated 2.2% increase in the population (4,485,100 to 4,585,400) so the fall in per capita income is even greater. 

And here are the consumption, savings and investment figures.  The savings and investment rates in the final two columns are four-quarter moving-averages. Click to enlarge.

Although the first table shows some positive signs with the recent rise in gross household disposable income there aren’t even hints of so-called green shoots from the household sector in the data in the second table – consumption expenditure is moribund and household investment continues to fall. 

There were anecdotal claims that retail sales during the Christmas period and January sales were ahead of expectations.  As “the plural of anecdote is not data” we will wait until the final column in the above table is filled in before drawing actual conclusions.

Flows in the Household Sector

A couple of recent posts have looked at the financial stocks of assets and liabilities in the household sector.  The second part of the Institutional Sector Accounts is that which looks at non-financial flows; earnings, income, consumption and savings, investment and depreciation.

First, here are the figures for the household sector current account for the past four years.

For most of the aggregates there was very little change in 2011 on the 2010 figures.  The large drops seen in 2009 and 2010 eased somewhat in 2011, but there is little sight of a comeback in these numbers.

Production (value of output) in the household sector rose slightly in 2011 leading to a small rise in the operating surplus of the sector (+1.6%).  Wages paid to households (including employer’s PRSI) from the government, business and household sectors all declined in 2011, though at 1.3% the drop was the lower than previous years. These and other minor changes led to a decline of 1.3% in the Gross National Income of the household sector.

The aggregate that shows the largest change is "Taxes on Income and Wealth” which increased by 22.1% on the year.  However, much of this is a reclassification change rather than an increased tax burden on households.  In 2010, €2,018 million was collected via the Health Levy which was classified as a social contribution.  For 2011, the Health Levy was abolished (along with the Income Levy) and replaced by the Universal Social Charge and all the revenue collected under the USC is classed as a tax rather than a social contribution.

Reflecting this social contributions paid to the government sector from households fell 10.7%.  In 2010, social contributions paid to the government (all by the household sector) were €11.5 billion.  For 2011, this was €10.3 billion.  The remainder of the social contributions paid by the household sector were made to the financial sector (private pension contributions). 

These were largely unchanged on the year at €4.3 billion.  Social contributions to and from the household sector will be related to non-profit organisations which are included in the household sector.

In 2010, the total of income and wealth taxes and social contributions from the household sector paid to the government sector was €23.9 billion.  For 2011, this increased to €25.4 billion.  Social contributions paid from the government sector to the household sector were €23.8 billion, an increase of 2.5% on 2010.

The sum of these changes meant that the gross disposable income of the household sector declined from €86.2 billion in 2010 to €84.2 billion in 2011, a fall of 2.3%.  From this household consumption was €77.5 billion, a decline of 0.8% on 2011.

Here are the capital accounts of the household sector.

The household sector has moved from being a net borrower to a net lender.  This is largely as a result of the collapse in investment in non-financial assets by the household sector (buying new houses).

In 2008, gross capital formation by the household sector was €17 billion.  This itself was down from €26 billion in 2006 and the fall continued through to 2011 when investment by the household sector was just over €4.7 billion.  With depreciation of existing household capital slightly under €4.7 billion, the net capital formation of the household sector in 2011 was just €44 million.  Repair of existing capital and investment in new capital barely covered the depreciation of existing capital.

The household sector is not using its gross savings to invest in fixed capital.  Nor is the sector using it to build up deposits; it is going it to pay down debt.

The Household Sector Financial Balance Sheet

Here is a summary of the 2011 financial balance sheet of the household sector from last week’s Institutional Sector Accounts.

The numbers are little changed from 2010 when a net financial wealth of €118 billion was reported.  The main reason for the slight increase in net financial wealth is the continued reduction in household loan liabilities.  The total wealth of the household sector would need non-financial assets (property, land and valuables) added to this €117 billion figure.  The value of houses owned by the household sector is probably around €300 billion.

One notable feature of the balance sheet of the household sector is the declining level of loan liabilities.  According to the CSO data, this peaked at €203 billion in 2008.  Since then it has fallen by €24 billion.

The fall in short-term loans may hit a lower bound but the drop and long-term loan liabilities has averaged around €6 billion per year and if current repayment and drawdown patterns remain this is likely to continue.

There is massive debt in the household sector in Ireland but it is clear that there are massive repayments being made.  The reduction to €179 billion is significant but the end-2006 figure was €169 billion so the progress in unwinding the huge borrowing of the boom is still small.  The 2002 figure was €110 billion.

Reductions in household gross disposable income means that the drops in the level of debt are not reflected in drops in the key debt-to-disposable income metric for the household sector.  This is reflected in this chart from the CSO.

Mortgage Repayments

Irish households are making capital repayments of around €5 billion per annum on their stock of mortgage debt.

The first release of the Financial Regulator’s Residential Mortgage Arrears and Repossession Statistics shows that in September 2009 there was €118.6 billion of outstanding balances on all owner-occupied mortgages in Ireland.  The most recent release gives data for March 2011 and shows that there was €116.0 billion on the outstanding balances. 

Over the 18-month period in question the total balance had decreased by nearly €3 billion.  However, this does not give the full capital reduction on the mortgages that were outstanding in September 2009 as new lending issued since them would be included in the March 2011 figure from the Financial Regulator.

Although not perfect, we can use the Irish Banking Federation’s Mortgage Market Profile data to gauge new lending in the mortgage market.  This data is estimated to represent 95% of the mortgage market and provides details of new lending in five categories

• First-time buyer
• Mover-purchaser
• Residential investment
• Re-mortgage
• Top-up

For our purposes here we can ignore residential investment mortgages as the mortgage arrears data only includes owner-occupied mortgages.  There are two categories that are definitely new lending: first-time buyers and mortgage top-ups.

In the 18 months between September 2009 and March 2011, there was €3 billion of mortgages issued to first-time buyers and €0.8 billion of top-ups provided to existing borrowers.  There will have been some capital repayments on this €3.8 billion but for simplicity we will assume it to be small.

The treatment of the lending to the other two categories is less clear-cut.  Re-mortgages are borrowing to repay existing loans (switching product or lender).  We will assume that the €0.7 billion of re-mortgage lending was equal to the value of the existing mortgages.  This would mean that no new lending was issued.

For mover-purchases the outcome is even more uncertain.  The purchaser may or may not have had a mortgage on the original property.  If there was no mortgage then the lending will have been new lending.  If they have an existing mortgage, then the new mortgage may be more or may be less less than the mortgage on the original residence. 

Although there is no indication to believe it is true we will assume that the €2.3 billion issued to mover purchasers matches the loans they had on their existing property.  Again this would mean that no new lending was issued.

It is probable that both of these assumptions are conservative.

Once we account for the €2.8 billion reduction in the overall outstanding amount and the €3.8 billion of new lending for first-time buyers and top-ups then the €118.6 billion owed in September 2009 was actually reduced by around €6.5 billion over the next 18 months. 

Some of this capital reduction may be due to loan write-downs but the vast majority will be due to repayments.  There are substantial mortgage arrears problems in Ireland, but it must also be remembered that substantial mortgage payments are also being made.  The Financial Regulator’s data show that nearly 90% of mortgages are neither in arrears or restructured.

Over the 18-month period in question here, the household sector had a gross savings rate of around 10% and had about €15 billion in gross savings.  As we have said before most of that is going to pay down debt rather than increase deposits.  And as shown here a lot of that debt repayment is going to pay down mortgages.  I would guess that around €5 billion of mortgage capital repayments are being made per annum.

Looking briefly at the number of mortgages.  In September 2009 there were 794,609 mortgages outstanding.  By March 2011 this had fallen by more than 12,000 to 782,429. 

Mortgages to first time buyers are definitely an increase in the number of mortgages.  We will take it that there is no increase in the number of mortgages as a result of mover purchasers. Remortgages and top-ups do not change the number of mortgages.  There were nearly 15,500 mortgages given to first-time buyers between the two dates in questions.

An arrears rate of 6.3% (which is only going to increase in the medium term) is signal enough that there are significant problems in the mortgage market.  Any proposed solution should be targeted at this problem as the majority of people are still meeting their mortgage obligations.

Jumbo Mortgages. Take Two.

These issue of €1 million+ mortgages is back in the news today following a presentation by Prof Morgan Kelly to the Irish Society of New Economists yesterday.  He first made the claim of 10,000 million plus mortgages in his Hubert Butler lecture in Kilkenny two weeks ago and which was examined on this site here and I also have a piece here.

An article in today’s Irish Times provides a fresh defence of the claim.

Yesterday he told a meeting of the Irish Society of New Economists in Dublin that this “anecdote” had “taken on a life of its own”.  He had been called “irresponsible” for using it.  “I read this in a newspaper a year ago, it has to be true,” he joked.

Prof Kelly said he had since used econometric calculations to analyse how many of these large investment mortgages there were, concluding that the anecdote “seems to be correct”.

Yesterday he said the investors probably took out more than one mortgage so it was 10,000 mortgages not 10,000 people who owed €11 billion.

Prof Kelly also calculated that two-thirds of investor loans were interest-only.  “These interest-only loans seem to concentrate among investors, and my guess would be this is large properties.”  This large number of interest-only investor mortgages was “ bad news” for the Irish banking system and the taxpayers, he said yesterday.

These investors “typically bought property that was designed for investors”.  Prof Kelly predicted “very large losses” on these properties.  Demand for property was driven by the flow of lending from banks. “The flow of lending to these investors is only 1 per cent of what it was back at the peak,” he said.

“There is no demand for this stuff, so I think there is going to be very large losses on these things.

This time the focus is on investment or buy-to-let mortgages.  In most cases the most important demand for these properties is in the rental rather than real estate market.  There may be some investors who bought investment properties for resale but the majority would have been bought to rent.

The most recent Daft report indicates that rents have fallen by an average of 25% since 2007.  This would be bad news for investors but it is cancelled by the drop in interest rates, particularly tracker rate mortgages. 

The key ECB rate has fell from 4.25% in July 2008 to 1.00% in May 2009.  It now stands at 1.50%.  This is still lower than at any time during the 2002 to 2007 period.  Recent indications are that further rate increases by the ECB will be put on hold. 

If two-thirds of these loans are interest only as claimed then the repayments on these loans will have fallen by more than the drop in rent.

We can get an insight into the residential investment market in Ireland from an interview  in The Irish Examiner with Hubert Kearns, chairman of the project agency which handles the collection of the Non-Principal Primary Residence charge of €200 on behalf of the country’s local authorities.  This includes all second properties and not just those for investment.

One of the most surprising features of the new tax regime was the level of compliance, about 80%, for the self-declaration tax on non-principal private residence (NPPR).

"There was a very high level of compliance by people before the due date both last year and this year. Another thing that surprised us is that there is a very large number of individuals who own a sizeable number of properties.
"There are 99,000 people with one property, but there are 35,000 people who have between two and 10 properties and who have paid the charge on two to 10 properties. In effect, this means of course that they own between three and 11 properties.

"And the figures go up, 970 people have between 12 and 21 properties; 230 people have between 22 and 31 properties and 100 people have between 32 and 41 properties."

Although there are 35,000 people who have paid the charge on between two and 10 properties, it is likely that most of these are in the range of two to four and are unlikely to have mortgages of more than €1 million.  Still there could be a couple of thousand people with five to 10 properties.

There is 1,300 people who paid the charge on more than 10 properties.  There is no doubt that a sizeable proportion of this group could have one (or more) mortgages in excess of €1 million.

If we look at the banks we see that the covered banks had a buy-to-let loan book of around €24 billion as reported in the stress tests.  The loan balances for AIB, BOI, PTSB and EBS are summarised in this table.  There was also about €600 million of buy-to-let loans in INBS.  Click to enlarge.

Of the total buy-to-let loan book of €24 billion it is hard to imagine that €11 billion would be concentrated in fewer than 10,000 people.  It could be the case though.

The stress tests allow for €6.3 billion of loan losses in buy-to-let loans.  The “three-year loss provision” of the Central Bank means that capital was provided for around €3.5 billion of losses between now and 2014.

Unlike owner-occupied mortgages we do not get data from the Financial Regulator on arrears for buy-to-let loans.  The banks themselves have given us an insight into this. It’s not pretty.  When announcing their half-year results AIB said.

One in five of its 44,000 Irish buy-to-let mortgages was in arrears or had been restructured to help borrowers at the end of June, compared with one in 12 of the bank’s 126,000 home loans.

The stress test loss of €6.3 billion in the adverse scenario would require about €12 billion of defaults in buy-to-let mortgages assuming that, on average, the property repossessed is worth 50% of the loan value.  That is a default on half the loan book.

The buy-to-let loan book is a mess but we still lack evidence that these 10,000 jumbo loans exist.

UPDATE: This morning we have had some useful information on whether the banks have 10,000 mortgages of more than €1 million on their books.  The numbers come from the Askaboutmoney.com website and can be seen here.  A fairly robust defence of the numbers was provided over on Namawinelake in this comment.

The numbers provided by the banks seem more realistic to me.  Although it seems to be “sources within the banks” rather than an official publication, the evidence is that the covered banks have about 2,500 mortgages on their books of more than €1 million.  This includes owner-occupied AND buy-to-let mortgages.

If 20% of these loans defaulted the banks are looking at losses of around €250 million.

Household Borrowings, Repayments and Debt Forgiveness

There is no doubt that the credit market has slowed to a virtual standstill in Ireland.  This was further emphasised by the mortgage data that was released by the Irish Bankers Federation during the week.  Here are the amounts of mortgages issued since 2005.

In the the second quarter of 2011 just €624 million of mortgage lending took place.  This is gown over 90% from the peaks since in 2006.  This pattern is reflected if we look at the monthly mortgage transactions of the banks’ balance sheets from the Money, Credit and Banking Statistics produced by the Central Bank.

The huge monthly increases up to 2007 are evident.  For more recent times it should be noted that the monthly transactions for mortgages on the banks’ balance sheets are negative.  For the second quarter of 2011 these summed to around €600 million.

Given the new mortgage lending that was issued it is probable that the balance of loans that existed at the start of Q2 2011 fell by around €1 billion during the quarter.  It is also likely that most of this is due to repayments rather than write-downs.  Irish households might have huge amounts of debt but in most cases it is being repaid.

Although we only have figures to the end-of March, the mortgage arrears data from the Financial Regulator show that entering Q2 2011, nearly 90% of all mortgages were being paid according to the terms of the original contract.  Some of these may actually be being repaid quicker than the agreed schedule.

This level of repayment is being reflected in the Quarterly Household Sector Accounts also produced by the Central Bank.  Here are the loans of the Household Sector in those accounts.

Again, the rapid rise in indebtedness is evident, but it is clear than since peaking in 2008 household debt liabilities have been decreasing.  This is because repayments on existing loans is greater than the amount of new loans issued. 

Long-term loans make up 95% of the total with about two-thirds of that again being mortgages for owner-occupied houses.  The same mortgage arrears data linked above show that there were €116 billion of owner-occupied mortgages in Ireland.

Household loans peaked at €203.3 billion in the 4th quarter of 2008.  By the first quarter of 2011 this had fallen to €184.9 billion.  The figures above suggest that this fall continued into the second quarter. Even still the level of household debt in Ireland is significantly above international norms, but it is falling.

There are massive debt problems, private and public, in Ireland.  Most of household debt can be repaid.  Some of it will never be repaid, and this could run to many billions – but maybe the problem is “not enormous ”.

A DEBT forgiveness scheme to relieve homeowners in mortgage distress would cost “in the region of €5-€6 billion”, UCD professor of economics Morgan Kelly has said.

While I am not a fan of widespread debt-forgiveness it is useful to note that the stress tests from last March allowed for about €6 billion of losses over the next three years in the covered banks on owner-occupied mortgages.   Although the precise details are not provided the scheme could be funded with the money already provided to the banks.

Maybe we should keep Prof. Kelly in this relatively good mood but I wonder what will happen when he learns that there is €116 billion of owner-occupied mortgages in Ireland rather than “about €55 billion”.  About two-third of this €116 billion are in the covered banks.

UPDATE: Here are more details of the scheme proposed by Prof Kelly.

“The good news is that if you leave investment mortgages out [of total mortgages owed], which are largely the banks’ problem, and look at mortgages people have on their own houses, there are about €55 billion of these out there,” he said.

“A lot of people can’t repay these mortgages and this is causing people terrible agony,” he said.

Prof Kelly made his estimations based on 20 per cent of people having difficulty paying their mortgages. This was the default figure in Florida where there was a similar housing bubble, he said. He estimated that mortgages would need to be halved on average.

“I would reckon that the ultimate cost of this very useful social programme is something in the region of €5 billion to €6 billion.”

I don’t know where the €55 billion figure came from.  The mortgage arrears data from the Financial Regulator show that there was almost €116 billion outstanding across the 782,00 of owner-occupied mortgages in Ireland at the end of March 2011.

If it is expected that 20% of people will get into difficulty, than that is around 150,000 mortgages.  If these people have have an average mortgage of €200,000 then the forecast is there will be €30 billion of mortgages in difficulty.  Reducing these mortgages by half would cost €15 billion.  This is two-and-a-half times greater than the suggested cost of €6 billion.

There is a bit more on mortgage debt forgiveness in this post.

10,000 mortgages

In his Hubert Butler Lecture delivered in St. Canice’s Cathedral, Kilkenny on Saturday 6th August as part of the Kilkenny Arts Festival, Professor Morgan Kelly made reference to 10,000 interest-only mortgages of between €1-2 million. 

Here is the actual quote (which is from about 31:35 in the audio file posted here)

“What worries me increasingly are mortgages, and in particular there is a group of mortgages given out - interest only mortgages - that were given out to professionals; to lawyers, solicitors, estate agents, at the peak of the bubble. And about 10,000 of these were given out, and this seems trivial, there are three quarters of a million mortgages out there, why should we care about 10,000 of them?  However, it turns out that these mortgages were for properties of €1-2 million each. So these guys would put up like 20% of the price, so 10,000 of these loans of €1.1 million each, which means that there’s like €11 billion in loans to these high-rollers from the boom, most of whom could barely buy you a cup of coffee now.”

Kelly forecasts that that banks would be lucky get back half of €11 billion issued on these loans and predicts a total of €30 billion of additional losses in the banks on top of what has already been accounted for already.  Following the huge economic disasters of the last few years this would be the one to push us over the edge, if it comes to pass.

The relatively good news is that there is no evidence that these 10,000 loans of a million plus actually exist.   On the other hand there is also no evidence that these 10,000 loans do not exist.  They might be out there, or they might not. We just don’t know.

Kelly based his statement on a May 2010 press release from the Irish Brokers Association (IBA) which made reference to this number.  I was in contact with the IBA last week and Ciaran Phelan sent an explanatory email. See here.

In the email it is explained that:

As no exact figures are available ( which I believe they should be) we could only estimate the numbers involved which we did using data from Moody's, Central Bank and our own internal conversations with members involved in the mortgage area who have considerable expertise in mortgage lending.

It is very likely that the The Sunday Independent  made the same enquiries and received the same email.  They carried a front-page story and a longer article that seemed to make a good deal more of the email than might be expected.

There were some misgivings that Morgan Kelly’s claims were based on an article from by Jack Fagan in The Irish Times on the 6th May 2010 (ungated version here).  However, when I went searching for a reference to these 10,000 loans, I first came across them in an article by Charlie Weston in The Irish Independent!  This article was published on the 30th April 2010, almost a full week before the same details appeared in The Irish Times.

The Sunday Independent article attempted to use Stamp Duty data from the Revenue Commissioners to counter the claim that these 10,000 €million plus mortgages exist.  This is inconclusive as the register of transactions included for Stamp Duty is not the full list of all mortgage transactions that occur. 

The Stamp Duty numbers show that between 2004 and 2007 around 13,000 stamped transactions took place on properties with prices of more than €635,000.   Using this number it is hard to imagine that 10,000 million plus mortgages are in existence.

The problem with using the Stamp Duty data in this manner is that while there was over 50,000 property transactions with Stamp Duty in 2006, there were over 200,000 mortgage transactions.  Stamp Duty only covers about a quarter of all mortgage transactions.

In my opinion it is very unlikely that there were 10,000 interest-only mortgages of more than €1 million given to professionals at the peak of the boom.  There is no evidence that they exist, but neither is there sufficient evidence to say that they don’t.  For the moment let’s assume that they do exist.  

Even if these mortgages were issued it is certain that only a fraction, rather than all, of them are in arrears.  With the typical tracker-rate mortgage that was available at the time, the annual interest bill on a mortgage of just over €1 million would be around €30,000 or less.  Even with the recent downturn this would be a manageable burden from the incomes of most professionals. 

Although no figures are available, there is no reason to expect mortgage arrears on these specific loans to be substantially higher than the average for all mortgages.  The Financial Regulator reports that nearly 90% of all mortgages are being repaid according to original contract, while 94% have avoided going into arrears of more than 90 days.

If we take it that these 10,000 mortgages are a special case and assume that 25% will default.  The banks will be have to write-down a loan of around €1.1 million but will take ownership of the house, which with a 60% price drop, would be worth around half a million.  In this very adverse scenario there is likely to be a loss of around €1.25 billion.

It is also important to realise that there is no certainty that all of these loans, such that they exist, were issued by the “covered” banks, most of which have now been nationalised.

Although not proven by the Stamp Duty statistics, it is likely that the actual number of million plus mortgages is lower than the 10,000 figure used by Kelly.  Looking at the aggregate mortgage data it could be less than half of this figure.  This is as much of a guess as the original estimate provided by the Irish Brokers Association but guessing seems to be the order of the day.  If we apply a more realistic default rate of 10% to this estimate, then the losses for the banks will be around €300 million. 

This is nearly 20 times lower than the €5.5 billion if there were to 10,000 defaults and such a loss was more than provided for in the bank stress tests published at the end of March.   The banks can handle €300 million of losses on jumbo mortgages.  If it comes to it they can handle €1 billion of losses on jumbo mortgages. 

It is possible that some of these loans do not appear in the mortgage statistics as they could be structured as commercial loans through a professional’s business.  For the present analysis we will stick with mortgages.

The stress tests allowed for €9.5 billion of residential mortgage losses over the next three years. That would probably require something around €16 billion of mortgage defaults.

The four main covered banks (AIB, BOI, EBS and PTSB) had €97 billion of residential mortgages at end-2010 (€74 billion in owner-occupied mortgages and €23 billion buy-to-let mortgages). A default rate of 17% over the next three years is allowed for in the stress tests.

If the losses to the covered banks on these 10,000 loans were €1 billion this would be easily absorbed into the €9.5 billion allowed for in the stress tests and there would still be nearly 90% of the loss provision available for the other 90% of mortgages.  

We cannot be sure if these loans exist or not.  For the wider economy it would probably be a good thing if they did exist.  The Financial Regulator report shows that there was €116 billion outstanding at the end of March 2011 on the 782,000 mortgages covered in its mortgage arrears data.  This gives an average mortgage balance of €148,200

If just 10,000 loans account for €11 billion of the total, then the average balance on the remaining 99% is actually €135,900.  With 10% of the burden carried by so few people, the average burden across everyone else is much lower.  The notion of “average burden” means nothing for someone who has lost their job and is trying to service a €300,000 mortgage but the exercise does give caution when using the aggregate figures.

As for these 10,000 mortgages – it really is hard to know.  It will take actual data from the banks to prove or disprove their existence.  Even if they are on the books of the banks it is hard to imagine that they will necessitate the banks getting additional capital. 

There may be other skeletons remaining on the banks’ balance sheets but this is not one of them.  BlackRock Consultants were paid plenty of money to  analyse the banks’ loans books so it is unlikely they would have missed these mortgages.  This was an interesting anecdote to provide to an audience at an Arts Festival lecture but is by no means another nail in the economy’s coffin.

What are we saving?

There has been a lot of discussion following from Michael Noonan’s comments last week that “what we really need is for people to go into the shops and start buying again”.   Stephen Kinsella has a good post on this.  As a follow up to this it is useful to consider just what it is Irish people are saving.  To start here is a highly stylised and artificial example.

Consider that society is made up of two groups and that the disposable income of each group is 50.  The first group is the Thrifty Group. Of their disposable income they spend 40 and save 10.  The second group is the Profligate Group. They spend their income of 50and also borrow an additional 10 to increase their consumption to 60.

In this first period aggregate disposable income is 100 and total consumption is 100.  The savings rate is zero.

We then move to the second period.  The situation and behaviour of the Thrifty Group does not change.  They continue to have a disposable income of 50, spending 40 and saving 10 of this.  We will also assume that the disposable income of the Profligate Group does not change but they are now unwilling/unable to borrow to fund their consumption and they spend their entire disposable income of 50.

In this second period aggregate disposable income is 100 and total consumption is 90.  The savings rate is 10% but this was driven by a change in borrowing behaviour.  There has been no change in savings behaviour.

Finally, consider a third period.  Again we will assume no change for the Thrifty Group but now the Profligate Group have to pay back the money they borrowed in Period One.  We will assume that this requires a repayment of five, thus reducing their consumption to 45.

In this third period aggregate disposable income is 100 and total consumption is 85.  The savings rate is now 15% but again it is not because of any increase savings behaviour.

In Ireland the savings rate has shot up in recent years. 

The suggestion from Michael Noonan is because this is as a result of an increase in savings behaviour.  It is true that a gap has emerged between household disposable income and household consumption.

Although the data in this graph only go to 2009 all indications are that this continued into 2010 and will likely continue in 2011 and 2012.  In economics, the definition of savings is disposable income that is not spent. 

In 2008, both disposable income and consumption expenditure were around €90 - €92 billion.  In 2009, household disposable income fell slightly but consumption expenditure dropped from €90 billion to €81 billion.  The assumption is that this €9 billion is money that households could have spent, but are now choosing to save.  As we saw with our stylised example above, this could be a factor of borrowing rather than saving.

If we look at household financial accounts we see that this is more than likely the case.  First here are household deposits.

During 2006 to 2008 when the savings rate was low household deposits were rising.  When the savings rate shot up in 2009, the increase in deposits slowed and had even begun to decline slightly during the end of 2010.   This is an aggregate of deposits.  It is likely for some households that precautionary deposits are increasing while for other households deposits are decreasing as they use them to offset significant reductions in income.

The savings rate might be a hefty 12% but this is not been seen in household deposits.  Michael Noonan may want people to go out shopping but they do not have an extra savings to draw on, household deposits are falling.

So where is the 12% savings rate going?  Well, as explained above it could be driven by a fall in new borrowings.  The Credit, Money and Banking Statistics from the Central Bank show that household credit is now contracting.

The drop in growth of consumer credit is evident and since 2008 has moved from growing at 18% per annum to falling at 18% per annum.  Even if household income and savings amounts had remained unchanged this huge turnaround in consumer credit would have seen consumption expenditure fall.

Finally here is the aggregate amount of loans owed by households.

Since peaking in the middle of 2008, the amount owed on household loans has fallen from €203 billion to €185 billion.  Short-term loans have fallen from €14 billion to €9 billion.

The household sector may have a savings rate of close to 12% but there are no swelling deposits that could be used to boost consumption.  The savings rate has shot up for two reasons:

1. Households are no longer borrowing to fund their consumption expenditure.
2. Households are paying back the loans they previously used to fund their consumption expenditure.

One issue that cannot be addressed here is whether people are paying back these loans in an accelerated fashion.  The Financial Regulator produces data on mortgage arrears, those behind on their repayments.  It would be useful if we could get even a snapshot of those who are in mortgage advances, those ahead on their repayments. 

There may be households who are saving by paying down loans quicker than in the original contract.  It is difficult to say if this is true but the total amount of loans outstanding to households is falling quite rapidly.  It must be stated that it is falling from a huge height and is still 120% of GDP, well north of international averages.

So what impact is this huge savings rate having on household deposits? Very little.  What extra money is available for consumption? Not much.  There may be some money that is currently being used to accelerate the repayment of debt. This will continue for the medium term and once households have repaired their balance sheets they will be in a position to increase consumption relatively quickly.  It will take more than words from the Minister for Finance to get the tills rolling again.

Where does our spending power come from?

For those who can’t wait, the answer to this question is that 50% of it comes through the State.

In a previous post we looked at the rapid reduction in liabilities in the household sector which has seen the total amount of household loans fall from  €203 billion in Q4 2008 to €186 billion in Q4 2010.   Household liabilities fell by an average of €3 billion a quarter in 2010.

This post is based on an interest point that was raised in the comments (HT: Dreaded_Estate).  The question was:

What percentage of the economy do you think is being paid by the public and private sector?

There is a brief answer in the comments to the original post and we largely reproduce it here with some minor additions.

Unfortunately we cannot answer this question with 2010 data as the Non-Financial Institutional Sector Accounts won’t be published again by the CSO until October.  Instead we use 2009 data but the changes are unlikely to have been of major significance.  We examined the 2009 household accounts when they were released and the numbers here are from that post.

In 2009, households had a net disposable income of €89.6 billion.  Of this €81 billion was used for consumption.  The represents money that was saved and used to either build up deposits or pay down loans.

The accounts show that in 2009 total household wages were €72.8 billion. Separate figures from the Department of Finance show that the public sector pay bill was around €18 billion so about a quarter of wage income in Ireland comes from the public sector.   It is likely that around one-third of this would have been paid in income taxes and social contributions, thus public sector pay contributed around €12 billion of the €90 billion of household disposable income.

The State also provided €26.4 billion of tax-free social benefits which directly added to household disposable income.  There was also a public sector pensions bill of around €1.8 billion which would be partially taxed.  It is likely that the State provided around €40 billion of disposable income in either wages, pensions or direct payments to households.

It is also true that some self-employed and private wage income actually comes from the State through State contracts and other processes.  It is very difficult to estimate how much this might be.  In 2009 the government spend €9.3 billion on goods and services and both central and local government spent €6.3 billion on capital expenditure.  It is likely that a significant portion of this expenditure further added to household disposable income through one means or another.

All told, it is probably that a low estimate is that somewhere around €45 billion or 50% of 2009 net household disposable income came either directly or indirectly from the State. 

In relation to the original post on the payment of household loans it is impossible to say how much of the €6 billion reduction in 2009 or €12 billion reduction in 2010 was made from disposable income that came from the State but it is sure to be something above zero.

There is of course a circular flow element to the above analysis with a good share of the disposable income provided by the State returning through consumption taxes. Here is a breakdown of the 2009 total government expenditure.

Quarterly Financial Accounts

The Central Bank has released the Q4 2010 data for its Quarterly Financial Accounts series.  The release is here and the data is here.  The numbers give an insight into the improving balance sheet of the household sector in Ireland.

Net financial wealth in the household sector has been increasing since the beginning of 2009 but is largely unchanged from where it was in 2002 when the series begins.  This has been brought about by a combination of increases in financial assets and decreases in financial liabilities.

Although the total amount of financial assets held by households increased from 2002 to 2007, this increase was offset in net financial wealth by a similar increase in liabilities.  The graph below gives the largest of those household financial liabilities: loans.

The total amount of loans owed by households has been falling since end of 2008.  This is the result of a huge slowdown in the drawdown of new loans, particularly mortgages, and repayments of existing loans.  The amount of outstanding loans in the household sectors peaked at €203 billion in Q4 2008.  In the eight quarters since then it has fallen to €186 billion.

At 120% of GDP this is still very high but the data show that it is declining as households undergo significant deleveraging.  The vast majority of this reduction will be the result of loan repayments rather than loan write downs.  Most of the loan write downs applied by the banks to the loans books so far have been as a result of the transfer of their developer loans to NAMA.

Cash transfers

According to the Household Accounts, disposable income in 2009 was supported by record payments of €26.3 billion in direct cash transfers from the State.  This represents 29.3% of the €89.6 billion net disposable income that households had in 2009 and 15.4% of GNP.  Here is a breakdown of the €26.3 billion in cash transfers provided by the State in 2009.

The €26.3 billion paid out by the State to supplement the incomes of those in need (or not in some cases) exceeds the €20.6 billion collected in Income Tax and PRSI contributions from those with earned income. 

Over the past ten years the proportion of Income Tax and PRSI contributions to GNP has remained relatively stable, and at 15.7% in 2009 were actually greater than then the 15.0% figure of 2000.  On the other side, cash transfers from the State have increased from 9.8% of GNP in 2000 to 20.1% in 2009.

We now have apparently “high-cost” services in a so-called “low-tax” economy.  Much as we’d like, we cannot have both.  I would like to hear social and economic reasons to justify why we should try to balance this inequality rather than have the outcome the result of a piece of arithmetic.  Social welfare recipients and tax payers deserve more than that.

Pension rates

In 1997 the contributory old-age pension was £80.30 per week (€101.96).  As of today the contributory old-age pension is €230.30.  This represents an increase of 126%.  The progression of this payment is shown below. 

Here is the same data in a table with some additional CPI inflation numbers.

Prior to 2010 there was only one year since 1998 when the increase in the contributory old-age pension was less than the rate of inflation.  This was in 2000 when the pension was increased by 5.1% while inflation was running at 5.6%.  In each subsequent year to 2009 the increases in the pension outstripped inflation.

We can also see that the total amount paid by the State in contributory old-age pensions has increased from €413 million in 1997 to €3,340 million in 2009 – an increase of 709% which is driven by demographic changes as well as increases in the rates.

Here is a graph comparing the contributory pension rates to the consumer price index can be seen here. 

This comparison is similar to a previous comparison we did with one public sector worker’s salary.   Some of the same caveats to that comparison also hold in this instance.

First, the CPI is a price index and not a cost-of-living index.  Second, we are not making a judgement on what the “right” weekly pension rate is.  The pension could have been “too low” in 1997 just as easily as it might be “too high” in 2010.  These are judgement calls.  Here we just make a comparison of two series.  Make of it what you will.

The interest cost of household debt

We have been considering the household sector in recent weeks.  We look at the aggregates from the household sector accounts, the use of the €11 billion saved in 2009 and the components of the €84 billion of private consumption expenditure.

One further issue to consider is the role of debt and interest in the household account.  Interest form parts of “property income” which was part of the table produced with the first post linked above.  Household debt is a feature of the capital account which we have not yet considered.

Interest expenditure does not form part of private consumption expenditure.  The graph below shows household interest outgoings since 2002 while a comparison of this to interest earned can be seen here. 

After rising to a peak of €8.1 billion in 2008, the household interest bill fell to €4.0 billion in 2009.  This is likely because of the fall in the ECB base rate which was raised to 4.25% in July 2008 but had plunged to 1.00% by May 2009.

There  was also a decrease in the amount of debt on which this interest was accruing.  According the Central Bank data outstanding household loans in December 2007 was €153.0 billion.  By December 2008 this had fallen to €144.6 billion with a further fall to €140.1 billion by December 2009.  This €140 billion in debt was made up of €110 billion of loans for house purchase, €24 billion of consumer credit and €6 billion of other loans.

The household sector capital account for the period 2006-09 is shown below the fold

It can be seen that 2009 was the first year in which the household sector was not a net borrower (and this is actually the case since this series began in 2002).  This is was driven by the surge in savings to €11 billion and the collapse in capital formation (house purchases).

Components of Consumption

The release of the Non-Financial Institutional Sector Accounts prompted us to the consider the Household Sector.  In a subsequent post we looked at what we had done with the close to €12 billion we had saved in 2009.  It is clear that this money was used to pay down debt rather than build up deposits.

We now turn to an examination of the money used for personal consumption expenditure.  In 2009 personal consumption was €84.3 billion down from a peak of €94.8 billion in 2008.  In nominal terms these are increases from the €29.7 billion in personal consumption expenditure in 1995.  Here is the pattern of the nine components of consumption expenditure in current prices as reported by the CSO since 1995.

The graph shows dramatic increases across a number of sectors, particularly food, housing, transport, recreation and education.  The largest increase over the 15 year period is seen in the category labelled miscellaneous goods and services.   The effect of the recession is most visible in the miscellaneous, housing and transport and communications categories, all of which recorded drops of more than 10% in 2009.

Using current prices it is clear that some of these changes in nominal consumption could be due to price effects.  Here is the same graph using constant prices.

Looking at the aggregate figures is reasonably information but it is probably more intuitive to look at the relative proportions of the components of personal consumption expenditure, i.e. the proportion of consumption was devoted to each category.

The stand-out feature of this graph is the drop in the proportion of expenditure that goes on food, beverages and tobacco.  The category comprised 29.6% of expenditure in 1995.  This had dropped to 18.8% by 2008, though increased slightly to 20.2% in 2009.   The breakdown of this category can be seen here.   The sub-category food (excluding meals out) has fallen from 14.0% of consumption in 1995 to 8.9% in 2009.  The other sub-categories have also shown declines in the proportion they comprise of consumption: non-alcoholic beverages (-0.2%), alcohol (-2.8%) and tobacco (-1.3%).  This reduction in the amount of our income we have to devote to food expenditure was the subject of a previous post.

With less of our expenditure devoted to the food, beverages and tobacco category we used a greater proportion of consumption on housing (as a proportion up 4.6% from 11.9% in 1995 to 16.5% of consumption in 2009), miscellaneous goods and services (up 4.4% to 22.3%) and expenditure outside the state (up 2.4% to 6.8%).  There are no sub-divisions provided for the housing and expenditure outside the state categories, and the rather inconclusive sub-categories of the miscellaneous category can be seen here, though there does appear to be an increase in the proportion of consumption on health services.  It is also worth noting that the housing category does not include mortgages as these are treated as investment expenditure.

The only other category for which notable sub-categories are provided is the transport and communication category which up 13.2% of consumption expenditure in both 1995 and 2009 though it had risen to 15.7% in 2006.  The breakdown of this category can be seen here which shows a sharp decline in expenditure on personal transport equipment (down from 6.1% in 2000 to 2.4% in 2009) in the last two years with expenditure on communication showing a steady increase (up from 2.0% in 1995 to 3.3% in 2009).

[ASIDE: One side issue that this analysis raises is the weighting given to the group ‘Motor Trades’ in the CSO’s Retail Sales Index.  Across the year this category makes up about 23% of the Retail Sales Index.  Within this group 82.2% of the weighting is given to the purchase of vehicles.  This would indicate that the purchase of vehicles has a weight of about 19% in the Retail Sales Index. 

The consumption figures explored here suggest in 2005 (the year in which the Retail Sales Index weights are based) that expenditure on personal transport equipment made up 4.6% of total consumption expenditure.  Although the Retails Sales Index only deals with the purchase goods while the consumption expenditure data cover goods and services this would not to sufficient to close the gap.  If expenditure was split evenly between goods and services this would suggest a weighting of 9.2% for the purchase of vehicles.  Why then is the weighting over twice as large at 18.9% in the Retail Sales Index?

Could it have to do with how vehicle purchases are financed, i,e,  cash versus finance?]

UPDATE: A little thought suggests a possible answer.  Figures in the Retail Sales Index are likely to include only the value of the car sold.  This is a measure of turnover which is the total value of sales.  Personal consumption expenditure is likely to be the value of the car purchased less the value of any trade-in used as part of the deal.  The measure of expenditure is likely to be the amount spent above any trade-in value. Q.E.D?

Where did the money go?

As we saw in an earlier post the CSO’s Institutional Sector Accounts revealed that households saved just over €11 billion in 2009.  What was done with this money?

To get a brief insight into this we can turn to the Financial Data on the household sector provided by the Central Bank.  Table A11.1 provides the figures for deposit by Irish residents by sector and category.  Here are the endpoints of 2009 (in €millions). 

Month	Deposits from Irish Households	Change
Jan 2009	97,684	-
Dec 2009	99,148	+1,464

At the start of 2009, Irish households had €97.7 billion on deposit.  During the course of a year in which €11 billion was saved, deposits of Irish households increased by just €1.5 billion.

Although the level of deposits remained relatively unchanged in 2009, there was some change in the type of deposits used by Irish households.


Over the year deposits in accounts with a maturity of up to two years fell from €37.1 billion in January to €31.5 billion.  At the same time deposits in accounts with notice periods of three months or less rose from €7.8 billion to €13.6 billion.  It seems households want to be in a position of having quicker access to their deposits.

So if households didn’t put the money that wasn’t spent on deposit where did it go?  To answer this we can turn to Table A5.1 with gives details of loans to Irish households.  Here again are the endpoints of 2009, in €millions.

Month	Loans to Irish Households	Change
Jan 2009	151,209	-
Dec 2009	140,085	-11,124

The total value of loans extended to Irish households fell by over €11 billion in 2009.  This €11.1 billion reduction was split across loans for house purchases (down €5.3 billion to €110.2 billion), consumer credit (down €5.2 billion to €23.8 billion), and other loans (down €0.7 billion to €6.1 billion).  The largest proportionate decrease was in consumer credit which fell by 17.9%, loans for house purchases fell 4.5% and other loans were down 10.4%.  It is likely that a significant portion of these reductions is due to repayments by households but some may be due to revaluations and write-offs.

Still, the pattern is pretty clear.  Most of the money saved in 2009 as indicated by the Household Sector Accounts was not used to build up deposits but was used to pay down debt.

If we extend this analysis through to the most recent September 2010 data we see that the patterns seen in 2009 have not carried through to 2010.

Month	Deposits from Irish Households	Change		Loans to Irish Households	Change
Dec 2009	99,148	-		140,085
Sep 2010	96,221	-2,927		139,096	-989

It appears, in fact, that the overall saving trend has reversed.  In 2010, household deposits have fallen by €2.9 billion while household loans have fallen by close to €1 billion.  This is indicative of dissaving rather than saving. 

The jump in the savings rate seen in 2009 may prove to be only a transitory effect.  And with retail sales now below last year’s level this suggests that household disposable income has dropped significantly in 2010.