The Promissory Note “savings” revisited

As outlined in the previous post it was projected in the Budget last December that the 2014 deficit would be €8.9 billion.  Just four months later the forecast of the 2014 deficit in April’s Stability Programme Update was €7.7 billion, a decrease of €1.2 billion.

The improvement was due to two things.  The first was around €200 million of revenue buoyancy from taxes that performed better than expected in the final 2012 outturn with carryover effects into 2013. 

The other €1 billion of deficit reduction in the projections was due to the “saving” from the Promissory Note restructuring. This arrangement did have the effect of improving the general government deficit for 2014 by around €1 billion (and subsequent years by declining amounts of around €100 million lower) but the change would not necessarily satisfy the dictionary definition of a saving.

The Anglo/INBS disaster resulted in a capital shortfall of around €35 billion in order to cover their massive loan losses and satisfy their depositor/bondholder liabilities that were repaid. A €4 billion cash injection was given to Anglo in 2009 with the other €31 billion being provided through Promissory Notes (a promise to pay in the future) in 2010. Two payments of €3 billion were made on the notes, one in 2011 and 2012, though the 2012 payment was made with a 2025 government bond rather than cash, meaning that there was around €25 billion of the Notes remaining when the restructure was put in place back in February.

Under the old arrangement, the Exchequer paid interest to the IBRC (the merged entity of Anglo/INBS) as the holder of the Promissory Notes. The IBRC was able to use this money, and the capital repayment portion of the €3 billion payment, to reduce its Exceptional Liquidity Assistance (ELA) liabilities with the Central Bank.  The IBRC used the ELA money from the Central Bank to repay its depositors/bondholders and as the loans it held were not going to be repaid sufficiently to, in turn, repay the Central Bank the money from the Promissory Notes was set to make up the shortfall.

Although the IBRC was outside the general government sector for Eurostat deficit measurement purposes, it was a 100% state-owned entity so the interest payment on the Promissory Notes was not going to an external third party.

The IBRC was itself paying interest to the Central Bank of Ireland for access to the ELA and some of that interest was returned to the Exchequer via the Central Bank surplus on its operations. The net external cost of the arrangement was the interest the Central Bank had to pay into the ECB for the facility to create money under the ELA scheme. The cost of this was the ECB’s Main Refinancing Operations (MRO) rate. This is currently 0.5%.

The interest on the Promissory Notes was high (around 8% from 2013) and fixed but was paid to the IBRC which in turn used it to pay down its liabilities with the Central Bank who then used it to reduce the amount of ELA it had forwarded. The interest rate on the Notes was largely meaningless and as Prof Karl Whelan usefully explained the €3 billion repayment would have resulted in the ELA liabilities to the Central Bank being fully repaid by 2022 even though the Promissory Note repayment schedule extended to 2030.

The IBRC was making a massive interest profit on the Promissory Notes and could use this to repay its only remaining liabilities – to the Central Bank of Ireland. As the IBRC was 100% state-owned the extra interest because of the high rate on the Promissory Notes was merely allowing the repayment of the money used to bailout Anglo/INBS depositors to be repaid quicker. 

The high interest rate on the Promissory Notes did not result in any additional direct costs, so a lower interest rate could not result in any direct savings.

Upon the liquidation of the IBRC, the Promissory Notes were cancelled, and the €25 billion owing to the Central Bank under the old arrangement was replaced with €25 billion of long-term bonds. As with the Promissory Notes the Exchequer is liable for the interest due on these.

Under the Promissory Notes the high interest rate on the €25 billion meant around €1.8 billion was due to be charged to the general government sector in 2014. Under the new arrangement the Central Bank holds €25 billion of long-term government bonds and the floating interest rate on these (currently around 3%) means that an interest bill of around €0.8 billion will be due to the Central Bank next year.

The difference between the figures is the €1 billion “saving” from the Promissory Notes restructuring.

However, just like the ELA arrangement the Central Bank will return any profit it makes from holding the bonds to the Exchequer through the Central Bank surplus. The Central Bank is a state-institution but as the monetary authority it is outside the definition of the general government sector used by Eurostat.

The net external cost to the broader or overall government sector is the cost to the Central Bank of creating the €25 billion to ‘buy’ these bonds from the government. The net external cost is just as it was under the old Promissory Notes/ELA arrangement; it is the ECB’s MRO. There is no actual interest rate saving from the restructuring, so the €1 billion “saving” is illusory though it does appear in the general government accounts. The  cost of the debt to the State disaster remains the same.

There is, though, a saving because the repayment schedule of the €25 billion has changed. Under Promissory Note arrangement the relatively cheap Central Bank funding (at the ECB’s MRO) would have been repaid by 2022. Under the new schedule which gives the minimum rate at which the Central Bank must sell the new bonds access to the funding at the MRO will be extended to 2032. See this earlier post.

There are benefits from the restructuring but they come about because of the slower rate at which the Central Bank sells the new bonds (meaning the interest is paid to a third-party and not recycled back to the Exchequer) compared to the repayment schedule on the former Promissory Notes (requiring borrowing to fund the €3 billion annual payments and thus interest to be paid to a third-party).

The new arrangement delays the rate at which a low-cost Central Bank liability (ELA/money to hold bonds in trading portfolio) is transformed into ‘normal’ government debt with the associated interest costs. The actual savings will be at their greatest roughly between 2016 and 2023 (assuming the minimum resale schedule is adhered to).

Of course, for the Excessive Deficit Procedure it is the general government deficit as defined by Eurostat that matters when it comes to the budget. But some consideration can easily show that the government moving from paying €1.8 billion of interest to the former IBRC to €0.8 billion of interest to the Central Bank doesn’t matter in the broader scheme of things as both have the government as their sole shareholder.

There is no “extra” €1 billion because of the Promissory Note arrangement.  The capital hole in Anglo/INBS was not reduced because of the change.  Nobody said we had to pay €1 billion a year less to pick up the bill.  The interest on the Promissory Notes was not a cost so a reduction in that interest is irrelevant.  It is the ECB’s MRO and how long funding at that rate can be accessed that matters.  There will be a small saving in 2014 on that front because of the Promissory Note restructuring but it will probably be around €100 million.

If there is to be a reduction in the fiscal consolidation in next month’s budget the argument has to be framed about why we should run a larger primary deficit now so that future generations can carry our “legacy of debt”.  And it is to primary balances that we turn next.

Deficit targets and nominal GDP

The upcoming budget is set to be framed around a general government deficit limit of 5.1% of GDP for 2014.  This was set as part of the EU’s Excessive Deficit Procedure (EDP) in an Council Recommendation to Ireland from December 2010.

There was much comment last week on the performance of real GDP in Ireland (up 0.4% in the quarter apparently) but debt contracts are written in nominal terms.  And last week’s National Accounts release from the CSO was not very positive on that front.

In this year’s budget from December 2012 the following GDP/GNP projections were used.

The current estimate is that nominal GDP in 2012 was €163.9 (thus exceeding the figure above) but the increases necessary to reach the subsequent levels seem unlikely.

In the Stability Programme Update published in April the projected general government deficit for 2013 was put at €12.575 billion.  If the actual outturn is close to this, nominal GDP will have to be around the €167.7 billion as projected in order for the deficit to satisfy the 7.5% limit set under the EDP.

The two quarters of 2013 data we have give the following estimates of nominal GDP:

• Q1 2013: €39,106 million
• Q2 2013: €41,679 million

These sum to give a half-year figure of €80.8 billion.  In 2012 (when nominal GDP was €163.9 billion) the figure for the first half of the year was €82.0 billion.

To meet the 2013 deficit target, nominal GDP needs to be nearly €4 billion higher than last year, but it is already running over €1 billion behind the 2012 level.  With revisions, nominal GDP of €87 billion is required in the latter half of the year for the €167.7 billion level to be reached.  The equivalent figure for H2 2012 was €81.9 billion, so year-on-year growth of just over 6% in H2 is required.

There are some indications that economic activity will pick up in the second half of the year.  Employment seems to be edging higher, retail sales were up in July and the change in car registrations all point to some improvement in Q3 and Q4.  Nominal figures are a combination of real changes and price changes and, on the consumption side at least, inflation is heading ever lower and is approaching zero.

Of course, Irish national accounts data is inherently difficult to predict.  A couple of intra-company changes in the MNC sector can have a huge impact on the figures (MNCs account for 90% of the export figure in the national accounts).  Nominal GDP of €167.7 billion could be reached but if it came up short at, say, €165 billion then a €12.6 billion general government deficit (as projected in the SPU) would give a deficit of 7.6% of GDP – in breach of the EDP limit set for Ireland, albeit by a very small amount (0.1 pp). This overshoot is actually the same as the 2013 deficit-increasing impact of the Promissory Note restructuring which took place in February.

It will be March of next year before a first estimate of the 2013 deficit/GDP ratio is known and no extra fiscal efforts will be introduced if it is felt the EDP limit will be breached this year.  It is also possible that developments since April will result in a smaller deficit than the €12.6 billion projected in April.  Whatever about 2013, it is definitely the case that the nominal GDP figures will have an impact on the fiscal effort required for 2014.

Last year’s budget included a NGDP projection of €174.1 billion for 2014.  The indications from the data released since then are that this projection will need to be revised down.  The 1.5% real GDP growth forecast for 2013 is unlikely to be met and the implicit GDP deflator resulting from the CSO’s chain-linking method may be less than 1.3%.  For 2014, real growth of 2.5% is projected with a GDP deflator of 1.3% again expected.

NGDP for 2013 is likely to be behind the budget projection and the 3.8% nominal growth rate for 2014 may also be underachieved.  A €165 billion outturn for 2013 and a 3% nominal growth rate for 2014 gives a 2014 figure of €170 billion.

If this is the case satisfying the EDP limit of 5.1% of GDP requires a general government deficit of no more than €8.7 billion.

Absent the Promissory Note arrangement, this lower nominal GDP would have meant achieving the 2014 EDP limit with a budgetary adjustment of €3.1 billion would have been difficult.  In last year’s budget projections an underlying deficit (i.e. excluding direct banking measures) of €8.9 billion for 2014 was projected.  This was sufficient to met the EDP limit when NGDP of €174 billion is used but would be in excess of the limit if NGDP was actually around €170 billion.  Using last December’s figures it is possible to see that an adjustment of more than €3.1 billion would be necessary to stay under the EDP limit.

There have been changes since then, of course, most notable the Promissory Note restructuring.  By April, the projection for the 2014 deficit (again with an assumed €3.1 billion adjustment) was down to €7.7 billion.  Even with lower NGDP this will be inside the EDP limit of 5.1% of GDP.  At €170 billion such a deficit in 2014 would be 4.5% of GDP.

It may be the case that next month’s budget will have a smaller package of tax increases and expenditure cuts then envisaged at the time of last year’s budget.  However, this not because of any unexpected improvement in underlying economic conditions.  In fact, the numbers last week suggest that more should be done if the fiscal consolidation plan is to remain “on track” not less.

The reason why less can be done is all down to the Promissory Note restructuring and some comments on these “savings” are presented in the next post.

Mortgages in the Covered Banks

The level of arrears in the quarterly statistics produced by the Financial Regulator continues to climb.  The latest figures are for Q2 2013 (30 June) and this is the bleak picture they paint for the €140 billion Irish residential mortgage market.

Only two-thirds of mortgages are being repaid under the terms of the original mortgage contract.  Immediate concern can be directed at the €27.3 billion (19.5%) of mortgages that are at least 90 days in arrears (of which around two-thirds are actually more than 360 days in arrears).

Solving the problem at the level of the borrower should be the priority but a related issue at the level of the bank is whether this level of mortgage delinquency will lead to a further recapitalisation requirement for the covered* banks – and more specifically whether it will lead to one that has to be filled by the State.

[* The group of banks was called ‘covered’ because they were the banks covered by the bank guarantee introduced in September 2008.  The original guarantee expired as planned in September 2010 and the Eligible Liabilities Guarantee (ELG) introduced in January 2010 was phased out this year.  Although the guarantees are no more the covered grouping is still used here to represent the domestic banks the State has an ownership interest in.]

The remaining covered banks are Allied Irish Bank (now merged with the Educational Building Society, EBS), Bank of Ireland and Permanent TSB.  Irish Nationwide Building Society and Anglo Irish Bank were originally part of the covered group but they merged to form the Irish Bank Resolution Corporation which was subsequently liquidated in February of this year.  Irish Nationwide had around €1.8 billion of mortgages while Anglo had no mortgage book of note. 

AIB and PTSB are almost fully state-owned while there is a 15% public ownership in BOI.  Here are their mortgage books in this country (with the percentages representing the proportion of the overall market the banks have).

The first point to note is that €90 billion (64.5%) of the €140 billion across all mortgages are in the covered banks.  The remaining €50 billion are banks in which the state has no ownership interest such as Ulster Bank, KBC Bank, Danske Bank while there are also the remnants of the loan books of lenders who have left the Irish market such as Bank of Scotland (Ireland), Start Mortgages and INBS as mentioned.

If not all of the mortgages are in the covered banks it stands to reason that not all of the mortgage arrears are in the covered banks.  Here are the Irish mortgages in arrears of more than 90 days in the covered banks (with the percentages representing the proportion of non-performing loans the banks have). 

[The figures come from the banks’ mid-year reports and also include loans which are impaired. In general, most of these will already by 90 days in arrears but a loan may be judged by the bank as impaired and not be 90 days in arrears because, for example, of concerns about loan-to-value levels.]

The covered banks had nearly €18 billion of impaired Irish mortgages at the 30 June 2013, with AIB’s buy-to-let loan book looking particularly bad with 48% in 90 day arrears and/or impaired.

In the case of AIB it should be noted that the current figures are for the combined AIB + EBS loan book.  In December 2011, AIB home mortgage loans had 6.6% in arrears of 90 days or more, while for EBS the equivalent figure was 16.6%.

As a side note the last figures on the INBS mortgage book (June 2012) showed that €1.1 billion of the loans were past due and/or impaired – giving a non-performing rate of almost 60%.

What capacity do the remaining covered banks have to absorb losses on their mortgages?  Again we can look to the banks’ financial reports and look at the stock of provisions they have on their balance sheets for losses on these loans (the percentage represents the amount of non-performing loans (NPLs) the banks have as a proportion of the totals in the previous table).

The banks have made a provision of almost 40% against their non-performing mortgages.  The banks can cover over €7 billion of loan losses on their mortgages without having to make a charge on their income statement, reduce the carrying value of the loans and thus reduce their existing capital.

The rate of provisioning varies across the banks with AIB having set aside 33% of NPLs while the cover for PTSB is almost 50%. 

The ability for the banks to be able to have such large provisions on their balance sheet is largely a result of the round of recapitalisation that took place following the PCAR results published in March 2011.  The level of loss provisions the banks now have is fairly close to the Central Bank loss projections that the banks were recapitalised against at the time.

In the PCAR exercise, BlackRock Consultants came up with estimated lifetime losses that the banks might make on their loan books under both a ‘base’ and ‘stress’ scenario.  The banks were recapitalised using the stress scenario with the Central Bank projecting “three-year losses” from the BlackRock figures for the recapitalisation requirement. 

Here are the figures for Irish residential mortgages based on a December 2010 loan book of €98 billion meaning there has been a reduction in the outstanding amount of €8 billion in the past two and a half years.  Most of that is because of capital repayments exceeding new drawdowns.  Anyway here are the loss projections from the PCAR on the banks Irish residential mortgages.

The €16.3 billion of lifetime loan losses estimated by BlackRock under the ‘stress’ scenario was scaled to €9.0 billion of projected losses in the three year recapitalisation horizon used at the time by the Central Bank.

As we have seen the covered banks have €18 billion of mortgages in 90 day arrears and/or impaired.  It would take a very large increase in this and a very low recovery rate (on repossessed properties) for €16.3 billion of losses to materialise.

It can be seen that the provisions for mortgage losses remaining on PTSB’s balance sheet is almost the same as the three-year loss projection used in the March 2011 PCAR (€2.55 billion provision now versus €2.59 billion projection then).  The figures for BOI are reasonably close (€1.69 billion provision now versus €2.02 billion projection then) but their is a big game in AIB (€2.95 billion provision versus €4.39 billion projection).  This likely reflects AIB’s less conservative provisioning highlighted above.

Should this divergence in the AIB figures be a concern? Possibly, but it is still a bit of a guess what the actual level of losses on Irish residential mortgages will be.  The figures don’t have to be the same.  The banks may have crystallised losses in the interim using some of their provisions.  At a recent Oireachtas Committee hearing AIB confirmed that it had written down €93 million of mortgage debt over the past 18 months.

We know the banks have €7.2 billion of provisions set aside.  The process that will result in these provisions being used for specific loans is, five years since the lending bubble burst, still uncertain.

With €9.8 billion of Core Tier 1 Capital, AIB has a CT1 rate of 15% so there is some buffer against losses in excess of those provided for.  A return to operating profitability in the near future would also help.

The tables from each banks’ mid-year reports used in the above analysis are posted below the fold (click to enlarge).

Allied Irish Bank

Bank of Ireland

Permanent TSB

Irish Examiner – 06/09/2013 and 13/09/2013

The Irish Examiner have carried two pieces from me over the past week.  They are available at the following links.

• Facing up to the arrears problem, 6th September
• Getting banks back on track, 13th September

Two minor points on the external trade figures

The CSO have releases the goods trade data for July.  Two minor points spring to mind.

1. The balance of trade in food is lower than last year.
2. Wide-bodied passenger aircraft have a big impact on the figures.

More details on these below.

ONE: Some emphasis is put on the positive performance of food exports with a total of €4,911 million of export in the year to July compared to €4,513 million in the equivalent period last year.  This is an 8.8% increase but it is a nominal rather than real figure – price effects are included. 

It is also worth noting that food imports have similarly increased: from €3,058 million in 2012 to €3,477 million this year.

Thus, the balance of trade in food and live animals has actually decreased slightly, falling from €1,455 million in the first seven months of 2012 to €1,434 million this year.

TWO: Although food imports (and also chemical imports) are up overall imports are down.  There was €29.0 billion of goods imports in the period to July 2012 compared to €28.5 billion this year.  The source of the drop is primarily category 79 in the SITC (Standard International Trade Classification).  SITC 79 is ‘Other Transport Equipment’. 

By this time last year imports for this category were €2,025 million.  This year they have fallen to €778 million.  What other transport equipment have we imported €1.2 billion less of this year? Ans: Wide-bodied passenger aircraft.

Here is an extract from the more-detailed Trade Statistics (to June) for category 792.40

In the first six months of 2012 there were 40 aircraft larger than 15,000kg imported worth €1,823 million compared to just 18 such aircraft this year worth €638 million, with such imports from Brazil and the US substantially down.

This reduction will provide a boost to the balance of trade this year but in reality it is little more than the timing of purchase decisions of some of the aircraft leasing companies who are based here.  The improvement in the balance of trade will be offset by a reduction in Investment as measured in the national accounts (GNP/GDP).  There will be no discernible impact to the economy on the ground.

Inflation approaches zero

Yesterday’s CPI release for August from the CSO shows that general inflation over the past 12 months has been very low.  The following chart has the 12-month changes in the overall CPI and in a ‘core’ measure of inflation (the 85% of the index excluding mortgage interest and energy products).

Core inflation is running slightly higher than the overall inflation rate as both mortgage interest (-5.9%) and energy products (-0.7%) showed price declines over the year.

The core deflation that was seen for two years up to the start of 2011 and the relatively low level of inflation since then has resulted in a substantial narrowing in price differences between Ireland and other countries. 

Here is the Harmonised Index of Consumer Prices (HICP) for Ireland the EA17 since 2002. [The chart says nothing about relative price levels as both indices are set to 100 in January 2002; it merely allows differences in relative price changes to be seen.]

Here is a similar HICP chart for Ireland and the UK showing that the relative gap over the last four years or so has been even greater.

Eurostat does provide some comparative price level data.  The following chart shows comparative prices for HFCE (household final consumption expenditure) for the UK and the EZ17 relative to Ireland (=100) for the past decade.

In all the years shown Ireland has had a higher price level than the other two areas.  The gap to prices in the Euro Area was greatest in 2008 and has narrowed since then.  The gap to UK prices was widest in 2009 and the rate of convergence has been even greater.  As the inflation data above has shown this has continued into 2013 and, although Ireland still has higher prices, we can expect the lines to get closer still in subsequent releases.

Retail sales heat up

Among several data releases from the CSO today is the July Retail Sales Index.  The overall index is heavily distorted by the motor trades with a weight of 21.6% in the index for July.  As a result of the introduction of the ‘132’ number plate Motor Trades in the index show a 25% annual increase over July 2012.

As per usual the focus here is on the RSI excluding the motor trades.  This shows a much smaller increase than the overall index but is not subject to the artificial boost brought about by the push of some new car sales into July (rather than an increase in new car sales).

Here is the index excluding motor trades since the start of 2010.

As can be seen the performance after the local maximum in October 2012 was weak.  The downtrend ended in April of this year and July provided a jump back close to those levels seen in October last year (which were in part helped by the digital switchover).  One explanation for the July increase is potentially the spell of exceptionally good weather at the time.

The largest monthly increases in volume were seen in:

• Bars +2.6%
• Non-Specialised Stores +3.3%
• Food, Beverages & Tobacco +4.2%
• Clothing, Footwear & Textiles +5.3%
• Hardware, Paints & Glass +5.1%
• Books, Newspapers & Stationery +16.2%

some of which could be influenced by the weather.  The largest monthly decreases in volume were seen in Electrical Goods (-4.9%) and Fuel (-4.2%).

The annual changes show that the July increase resulted in a modest annual increase.

While the monthly changes highlight the inherent volatility in the series, and with less exceptional weather since, it is possible that subsequent releases will see a return to the sequence of monthly decreases seen around the start of the year.