The Ongoing Trickle of Repossessions

The latest mortgage arrears statistics from the Central Bank provide an update for the end of September (Q3) 2023. They show there continues to be a trickle of repossessions.

During the third quarter of 2023, there were ten court-ordered repossessions and ten abandonments/voluntary surrenders of primary dwelling houses.  During 2023, court-ordered repossessions have been occurring at a rate of around one per week.

We are now also provided with a more complete breakdown of repossession activity.  The number of repossessions shown above can be broken down into those carried out by banks and non-banks. Non-banks include regulated lenders such as Pepper and Start and also unregulated loan owners.  The chart below shows a split out of the maroon area in the opening chart.

In the last 15 years, there has been a total of 3,400 court-ordered PDH repossessions. Of these around 70 per cent have been undertaken for banks. However, the banks stalled their execution of repossession orders in early 2020 (likely pandemic related) and there has been no increase since.  In the year to the end of September, there were 57 court-ordered repossessions and 42 (nearly 75 per cent) were carried out for non-banks.

Figures on legal proceedings suggest that the banks have pretty much ceased using legal proceedings to obtain repossessions orders.  Here is the number of accounts with legal proceedings in progress.

In early 2016, the banks had almost 12,000 PDH mortgage accounts with legal proceedings in progress. The last figures show that this had fallen to just 1,100 by the end of September. This is 0.2 per cent (1 in 500) of the 594,000 PDH mortgage accounts held by the banks.

Some of the reduction, of course, came about as the non-performing loans were sold to non-bank entities but the fall in the overall total is very clear. And it can be seen that the level of non-banks has been stable in recent years. In Q1 2019, non-banks had 4,400 PDH accounts with ongoing legal proceedings. The latest figure for these entities is 4,300.

That is not to say there has been an improvement in the mortgage accounts held by non-banks.  The arrears on these accounts is still very high.

Of the 710,000 or so PDH mortgage accounts in the latest update, almost 115,000 (16 per cent) were held  by non-bank entities.  And of those almost a quarter (27,300) were in some form of arrears.

The arrears ranged from relatively small (less than 90 days arrears) to incredibly large (over 10 years arrears).  It is bizarre that we have to produce figures for such significant arrears.  Non-bank entities had 21,000 PDH mortgage accounts that were more than 90 days in arrears.

Included in that are 4,500 accounts that are more than ten years in arrears.  The total balance outstanding on these loans is €1.3 billion, giving an average balance of €280,000.  There have been €833 million of missed payments on these giving an average level of arrears of just under €185,000.   That is an incredible amount of missing payments.

It is worth noting again that arrears is not a good measure of current loan distress. It does not tell us when the payments were missed.  An account that had two years of missed payments a decade ago but has had every payment made since will be counted as being two years in arrears.

Also, the measurement of arrears in terms of days past due is impacted by the repayment required.  If an account has had €6,000 of historical missed payments and the required monthly payment is €1,000 then that account will be 180 days (six months) in arrears.  If the required monthly payment rises to €1,200 – due to. say, interest rate increases – and there is no new arrears, then the €6,000 of historical arrears becomes the equivalent of 150 days (five months) of arrears. 

This measurement issue won’t impact a count of the total number of accounts in arrears.  There has been no rise in the total number of accounts in arrears.

As shown above, there remains around 29,000 PDH mortgage accounts which are more than 90 days in arrears.  Of these 21,000 are held by non-bank entities.  The banks have 8,000 such accounts but this represents just 1.3 per cent of the total PDH mortgages they hold (594,000).  The banks do not have a problem with a long-term mortgage arrears.

In some cases this will have occurred with the borrower getting back in track either with or without a cure or modification. In other cases it is because the banks have simply sold the loans.  There were plenty of claims that these sales would lead to a surge in repossessions. All we have seen so far is a trickle.

No Surprise in November CT Receipts

When the June figures were published the €4.2 billion of Corporation Tax collected in the month pointed to receipts of around €6 billion in November.  The year-on-year declines for August, September and October were attributed to largely idiosyncratic volatility in such a concentrated source of tax revenue. 

And so it proved, with the monthly receipts for November coming in broadly in line with expectations at €6.3 billion or five percent above the crude projection based on the June figure.  With November being the most important month of the year for Corporation Tax, November 2023 was a record month, surpassing the €5.0 billion collected in the same month last year.

In year-to-date terms, November was enough to push 2023 back up above the equivalent total for 2022.  The difference looks small on the chart but the €22 billion collected so far in 2023 is almost €1 billion higher than what was collected in the same period last year.  The difference to 2014 is staggering.

The recent volatility in the receipts can be seen with the 12-month sum.  If December 2023 just matches what was collected last December then the total for the year will be €23.5 billion, up on the €22.6 billion from last year.

The volatility is highlighted in the annual changes from the above chart.  This shows some extraordinary changes. 

Towards the end of 2022, CT was growing at an annual rate of over 60 per cent – exhilarating but unsustainable. This growth plummeted during 2023 and approached zero in October. The November numbers mean that the annual change in the 12-month sum rebounded a little and rose to 2.8 per cent – and such is the scale of these receipts now that even such a small relative increase is a pretty significant sum.

It is hard to say if the December returns will garner much attention, and as a standalone month there is no discernible pattern.  2022 was a boom year for CT but December 2022 was actually lower than December 2021 – that volatility thing again.

Corporation Tax is a highly volatile source of revenue. While there can be some information in the monthly changes there is also an awful lot of noise.  Here the year-on-year comparisons of monthly CT receipts since January 2016 – and note that the chart excludes some of the volatility as the vertical axis is cut off at +100.  The final point shown is November 2023 which was was 27 per higher than November 2022.  Missing values are used when the calculation of the annual change involved a negative number (three instances since 2016: January 2019, April 2019 and April 2021).

Monthly figures that are lower than the same month of the previous year are not uncommon, though some occur in months that are not important from a CT perspective. Around one-third of months in the chart above have negative values – and this was a period when aggregate CT revenues grew at an extraordinary rate.

Four of the six months from December 2022 to May 2023 were negative in year-on-year terms.  The run of three negative figures in a row that was seen from August to October this year was a bit unusual – we have to go back to early 2021 to see such a run, when the pandemic would have had an impact, and the only other instance shown of three negative values in a row was in early 2017. 

The proximity of the recent three-month negative run to the key month of November also likely contributed to the focus. Those negative outturns will also have been seen as an opportunity to try and dampen expectations. Who knows what December will bring. Probably more volatility but structural shifts remain absent for now.

Volatility to the fore in Corporation Tax revenues

For almost a decade now, the year-to-date revenues from Corporation Tax have exceeded the equivalent amount from the previous year (with early 2021 being the only minor exception).  We can see below that the 2023 year-to-date figure for October has dipped below the 2022 line.  The €15.7 billion collected so far in 2023 is €0.4 billion (3%) below the amount collected to October 2022.

We can also see that this is a reversal of the position seen only a few months ago. By July, 2023 was running nearly €2 billion ahead of 2022 (€10.9 billion versus €9.0 billion).  The reversal since means that this year’s Corporation Tax receipts for August, September and October have been around €2.3 billion lower than the equivalent months from last year.

These three months last year saw €7.2 billion of Corporation Tax collected, this year they brought in €4.8 billion - a fall of one-third, which does appear to be dramatic. 

As seen above, this year was still the second-highest on record for these three months but is the fall just volatility that can be expected in such a concentrated revenue source or a harbinger of a more systematic decline?  For the time being volatility seems the more likely landing spot.

To put the recent declines in context we can look at the 12-month rolling sum of receipts.

The change in direction in the past three months is pretty clear.  November is the most important individual month for Corporation Tax and a one-third drop there relative to November 2022 certainly would be alarming.

The recent falls have put the 12-month sum back close to where it was this time last year, with the annual change in the 12-month sum having plummeted towards zero in recent months. Annual growth rates of 40 per cent plus while exhilarating are not sustainable.

In the 12-months to October 2022, €21.9 billion of Corporation Tax was collected. The 12 months to October 2023 have seen revenues of €22.2 billion – an increase of 1.25 per cent.

To assess the concern that could be raised after the falls in recent months we go back to the Revenue Commissioners guidance on when large companies need to pay their Corporation Tax.

Large companies

Large companies can pay their preliminary CT in two instalments when their accounting period is longer than seven months.

The first instalment is due on the 23rd of the sixth month of the accounting period. The amount due is either:

• 50% of the CT liability for the previous accounting period
• 45% of the CT liability for the current accounting period.

The second instalment is due on the 23rd of the eleventh month. This will bring the preliminary tax up to 90% of the final tax due for the current accounting period.

The key months are month six and month 11 of the company’s financial year, by which time 90 per cent of the estimated tax due for the current year must be paid.  Any remaining tax due will be paid when the company files it tax return.

A company must file its return and pay any tax due nine months after the end of the accounting period. The company must make this payment on or before the 23rd of the ninth month.

Clearly any individual month can be any of these deadlines: month six, month 11 or month plus nine.  The likely candidates for the months we are looking at are:

• August: month 11 for companies with a September year-end
• September: month +9 for companies with a December year-end
• October: month 11 for companies with a November year-end

August

Here are the monthly receipts for August since 2009.  There was an exceptional increase in August 2022 which was up 167 per cent on August 2021.  Some of this was reversed in 2023 but August 2023 was still significantly higher than all years prior to 2022.

The impact of payment timings can be seen if we combine the receipts for March and August – month six and month 11 for companies with a September year end.

Across the two months of March and August there was no fall at all in 2023.  This points to a company having a large increase in its tax liability in 2022 and paying most of this in August 2022 (when it needed to have 90 per cent of its preliminary tax paid).  For 2023, the payments were more balanced as the March payment would have been based on 50 per cent of the 2022 liability with a smaller amount due in August this year to bring the payments up to the required 90 per cent.

The fall in August 2023 seems mainly volatility due to concentration and the timing of payments.

September

The fall in September was small which also points to volatility.

September 2022 was pretty much double September 2021 (€2 billion versus €1 billion). Some of this increase was reversed in 2023 but it is hard to point to anything systematic that might be going on. 

It could just be that companies with December year-ends that filed their 2021 tax returns in September 2022 had more of their final tax payment outstanding than when they filed their 2022 tax returns in September 2023. With lots of companies having December year-ends we can’t point to any company-specific factors but it could be just a general timing issue.

October

The monthly figures for October seem particularly volatile and it is hard to discern anything that might be going on. 

There was a local peak in October 2018 of €1.6 billion.  This fell to €1.0 billion in 2019 and pretty much collapsed to just €200 million in 2020.  This put October 2020 as one of the lowest Octobers in the last 15 years. 

There was a strong bounceback in 2021 and a further increase to €2.3 billion in 2022.  Relative to that there was a large fall in 2023, back to €1.3 billion (which was also lower than both 2018 and 2021). 

It is because of this fall in the October receipts that the year-to-date total for 2023 is running behind the 2022 level.  As we have seen, the August fall was offset by a rise in March, and the September fall was pretty modest.  So, the October fall could be significant – if we could link it to something.

We could try to link the October payments to those made the previous May, which are month 11 and month six for companies with November year ends, but that doesn’t reveal a lot.  May is also also month 11 for companies with a June year end.

The monthly receipts for May have actually been pretty stable over the past four years.  This could mean rising payments from companies with June year ends and falling payments from companies with November year ends but there is no way of telling that. 

For companies with November year ends we can see the October (month 11) payments but as noted above, outside of being especially volatile, there is no discernible pattern.  It is possible that a company with a November year end (for the purposes of paying Irish Corporation Tax) has had volatile profits over the last few years and this has likely been driving the volatility in the October Corporation Tax receipts.

This again points to the concentrated and volatile nature of these Corporation Tax receipts with a company-specific factor driving changes in the aggregate figures.  It does not point to a structural shift in the overall pattern of receipts.

November

November is a month that could be a bell-weather for such structural changes.  It is the most important month for Corporation Tax.  November is month 11 for the most common year-end: December. Last November saw €5 billion of Corporation Tax collected – the highest monthly total ever, and higher than the annual total for 2014. 

The data back to 2009, show a strong relationship between the June and November receipts.

June 2023 was the highest June ever with receipts of €4.3 billion.  If the above relationship holds, this points to record receipts again in November, this time to the tune of €6 billion.

One aggravating factor could be the rise shown above for receipts in August.  August is month 11 for companies with September year ends.  These companies’ tax returns and their final payments will be due nine months later – the following June.

Up to recently, very little Corporation Tax was collected in August (and also in March).  This would translate into very little final tax due the following June.  Now however, the combined receipts for March and August exceed €4 billion.  This can be expected to lead to additional tax payments the following June when companies with September year ends file their tax return and make their final payments. 

The above relationship between June and November was almost exclusively based on the June payments been the first preliminary payments for companies with December year ends so they were a strong predictor of the second preliminary payment in November.  This relationship may no longer be as clear cut as more of the June receipts could be due to companies with September year ends. 

Even with this we would still expect this November’s receipts to be strong. Talk of structural decline in Ireland’s booming Corporation Tax seems premature – for now. But we can certainly expect them to be volatile.

Numbers of FTBs continue to edge up

A couple of indicators during the week show that the number of first-time buyers (FTBs) continues to rise. First, the volume of market transactions tagged as FTBs in the CSO’s residential property price index.  This is updated monthly and the volume of transactions is shown here on a 12-month basis.

There were 17,500 stamp duty filings tagged as FTBs made in the 12 months to the end of August.  Of these, 12,500 were for existing properties and 5,000 were for new properties.  The increase in the last year has been exclusively in existing properties. 

The volume of FTB transactions has been rising for more than ten years and the total is now pretty much back on the increasing trend evident up to the onset of COVID. 

The second source is the mortgage drawdown data from the Banking and Payments Federation.  This is available quarterly with this week’s update giving figures for Q3 2023.  Again, the figures are presented on an annual basis.  One plus of the mortgage data is that it is available back to 2004 (though the new/existing split is only available from 2006).

The pattern in the mortgage series from the BPF matches that in the volume series from the CSO.  In the year to the end of Q3, there were just over 25,000 mortgage loan drawdowns by FTBs in the BPF data.  This is around 60 per cent of the levels seen during the height of the credit bubble.

There are a number of reasons why the level is higher in the BPF data compared to the CSO (25,000 versus 17,500).  Most of it is likely because they are counting different things: mortgage loan drawdowns versus market transactions.

People who self-build can draw down an FTB mortgage, but obviously there is no market transaction to be included in the CSO data.  FTB mortgages can be drawn down for non-market transactions, such as family transfers or someone may take out a mortgage to renovate a property they received as a bequest.  It may also be that a mortgage drawn down in stages is represented by multiple drawdowns in the BPF data.

With these in mind it could be taken that the CSO figure gives the lower limit for the number of FTBs with the BPF figure giving an upper limit.  The true number of FTBs may be closer to the upper limit.  Whatever about those differences, the trends match, and show both rising.

The Household Sector in the first half of 2023

A few weeks ago, the CSO published the Q2 2023 update of the non-financial institutional sector accounts. We will use those to check in on the status of the household sector for the first six months of this year (H1) and how it compares to last year and to 2008.

First, the current account.  In the prevailing environment, it is important to note that the figures in the table below are in nominal terms, i.e. not inflation adjusted.  We will look at some of the aggregates in constant prices towards the end of the post.

The start point of the current account here is Gross Domestic Product. For the household sector this is the value added produced, a large share of which is due to the imputed rents attributed to owner-occupier households with the remainder arising from the activities of the self-employed.

From there the account proceeds through a series of inflows and outflows until the bottom line is reached, Gross Savings, which is disposable income not used for consumption.

Most of the items in the table show an increase for H1 2023 over H1 2022.  One of the most significant is the continued growth of compensation of employees received.  This is up 7.3 per cent (remember in nominal terms) and stood at €64.6 billion for the first six months of the year.  Non-financial corporations are the most important source of COE and their pay outlays are up 8.0 per cent in year-on-year terms.

The largest relative changes are for the interest flows. Both interest paid and interest received are up almost 60 per cent in 2023. Taxes on income and social contributions, particularly those paid to the government sector, rose in line with the rise in income.  Social benefits received rose slightly for H1 2023 but remain lower than they were in 2020 and 2021 as COVID-19 supports were withdrawn.

All told, the national accounts estimate of household disposable income for H1 2023 was €73.8 billion.  With consumption expenditure of €64.8 billion and a €1.8 billion adjustment for the excess of contributions to private pensions over benefits received from them, the gross saving of the household sector is put at €10.6 billion for the first six months of the year. 

This is lower than each of the previous three years and gives a savings rate of 14.5 per cent for H1 2023, which remains slightly higher than the 12.1 per cent recorded for H1 2019 in pre-COVID times.

To see how that saving might be used we can turn to the capital account.

Again, the figures are in nominal terms.  The key figures in the capital account are the amount Gross Savings taken from the current account and the amount of Gross Capital Formation (investment in new capital assets) undertaken by households.  These are the main items that give rise to the net borrowing/lending of the household sector.  There are some other minor flows for capital taxes paid, investment grants received and other capital transfers.

With gross savings of €10.6 billion and with the household sector “only” undertaking €4.5 billion of gross capital formation, the household sector had €6.6 billion available for net lending. This goes on the financial balance sheet.

The CSO don’t do quarterly updates of their financial institutional sector accounts but the Central Bank do with estimates for Q1 2023 the latest available.  These show the household sector adding a further €2.7 billion to their burgeoning deposit balances. Between currency and deposits, the Central Bank estimate that the household sector had €198 billion on hand in Q1 2023, up from €154 billion in Q1 2020.  Households also used savings to make the contributions to private pensions referenced above.

On the liability side there was very little movement in Q1.  Repayments on existing loans were pretty much equivalent to drawdowns of new loans, with a net increase in loan liabilities of just €36 million in Q1.  Though it should be noted that repayments have exceeded drawdowns for the past 15 years or so.

All told, the Central Bank estimate that the household sector had €507 billion of financial assets at the end of Q1 2023, with an offsetting amount of liabilities of €143 billion.

Some Constant Price Series

The CSO also some of the main aggregates using constant prices. These series are also seasonally adjusted.  Here are the series for Disposable Income and Consumption Expenditure.

It can readily be seen that the income increase shown in the table for the current account in nominal terms is eliminated with the adjustment to constant prices. In real terms, aggregate household disposable income has been declining in recent quarters.  Real consumption has continued to rise and is not far off its pre-COVID trend.  In aggregate terms, both series are well ahead of the previous peaks reached in 2008.

To conclude, we make one additional adjustment to the CSO constant price series – put them in per capita terms.

As the population is growing fairly rapidly this shows that the decline in per capita income is more pronounced than that shown by the aggregate figures, and that per capita real consumption has essentially been flat for the past year or so.

It is also interesting to make comparisons back to the previous local maxima for each series from 2008.  Both the disposable income and consumption series are actually little different to their peaks in early 2008 – the latest readings put both series about 2 per cent above those 2008 peaks.

This might suggest a “lost 15 years”.  However, the environment on which each were achieved are wholly different.  We can see this by looking at the ‘Saving minus Investment’ of the household and government sectors.  This is each sector’s contribution to the current account of the Balance of Payments and is the Net Lending/(Borrowing) of a sector excluding capital taxes, investment grants and capital transfers.  For this we return to nominal figures and annualise them by looking at their four-quarter moving sum.

The difference between 2008 and 2023 is huge. In 2008, the household sector was in deficit to the tune of €20 billion.  Via the circular flow, this borrowing was contributing to household income.  When the crash came the borrowing was taken over by the government as tax revenues collapsed and spending on income supports increased.  Without this borrowing, income (and consumption) would have been much lower.

For 2023, there is a combined surplus of €20 billion.  There is a €40 billion difference between the positions of 2008 and that of 2023.  Now both the household and government sectors have an excess of disposable income over their consumption and capital formation expenditure.  In terms of GNI*, it is equivalent to going from borrowing 12 per cent of national income to lending 8 per cent of national income.  Simply comparing the income and consumption outcomes misses the scale of the turnaround in the borrowing/(lending) position of the economy.

We are in an income position to spend more – Corporation Tax concerns aside! But do we have the resource capacity to produce the things that we would like to buy, such as new housing units?

The aggregate Corporate Tax calculation enters an unsettled spell

In recent years when looking at the annual update from the Revenue Commissioners of the aggregate corporate tax calculation we noted that things were relatively calm, albeit it with elevated levels of receipts.  These were for tax returns filed for financial years ending during the 2018, 2019 and 2020 calendar years.

We now have the update for tax returns filed filed for financial years ending during 2021 (the last of these returns would have been filed in September 2022).  The 2015 upheaval is a well-worn track so we will just focus on the five most recent years.

The Determination of Taxable Income

We’ll start with the determination of taxable income.

Right from the top we can see big changes for 2021.  There an increase of over €55 billion in gross trading profits recorded on tax returns for the year, reaching €250 billion.  This carries right the way down the table with the end showing that taxable income increased by over €40 billion.

However, there are some significant changes along the way. Early on, we see that after being relatively stable from 2018 to 2020 the amount of capital allowances used jumped again in 2021, coming in at just under €100 billion as a result of a €23 billion increase.

Next we see a large increase in foreign income included in the tax returns of Irish-resident companies.  For 2021, this exceeded €20 billion for the first time and was close to €10 billion more than the next highest year.  The next part of the table will show the impact of this on tax payments (answer: very little). 

Other income also saw a jump in the amount of capital gains included to reach €5 billion in 2021. This is actually a regrossed amount.  The applicable CGT rate is 33% but the gains are included in the tax return to be taxed at 12.5 per cent. Hence, the gains are regrossed and multiplied by 2.64 (= 33/12.5) to get the amount to give the necessary amount of tax.

In charges and deduction we see that trade charges, mainly certain royalty expenses, rose back to levels seen up to 2019.  This was the largest change among these items.

All told, the CT returns filed for years ending during 2021 had just over €150 billion of taxable income with €6 billion of that subject to tax at 25 per cent.  We now turn to the calculation of tax due.

The Determination of Tax Due

The starting point of this part of the calculation, Gross Tax Due is simply the amount of taxable income multiplied by the applicable rates (either 12.5 or 25 per cent).

The relative simplicity of the Irish CT regime means there isn’t a whole lot going on here.  The biggest reason for the reduction of gross tax due is because of tax already paid. 

The largest single item in the above table is double taxation relief and this was almost €3 billion in 2021.  The additional foreign tax credit of nearly €450 million can be added to this.  These reflection the foreign corporate taxes paid on the €23.5 billion of foreign income included in the top half of the table.  Little additional Irish tax is due as the amount already paid, albeit abroad, covers the gross tax due in Ireland.

The item for gross withholding tax on fees is somewhat similar but in this case it is Irish tax that has already been paid.  In some circumstances, when the buyer of services is paying them they will pay 80 per cent of the invoiced amount to the supplier and 20 per cent to the Revenue Commissioners.  The 20 per cent represents a withholding tax at the standard rate of Income Tax.  When filing their tax returns, companies will record any service fees that have been withheld from them and reduce their tax due figure accordingly.  The Revenue Commissioners will already have received the amount.

The most significant item in the table that actually reduces companies’ tax bills is the R&D tac credit.  Between the R&D credit used and the excess R&D credit refunded the total cost was just over €750 million in 2021, a slight increase on the outturn for 2021.

All told, the bottom line is a tax due figure of €15.1 billion for tax returns filed for periods ending during 2021.  There are some timing differences but we can that, in recent years, the tax due figure from the aggregate CT calculation, closely matches the CT receipts collected for the Exchequer.

We will conclude with a look at some of the tumult in the aggregate figures, with most of this seeming to concern capital allowances, in particular those for intangible assets.

Snowballing Claims for Capital Allowances

As before we will look at the total amount of capital allowances claimed, the total amount used (that is, actually offset against gross profit) and the consumption of fixed capital from the National Accounts.

Up to 2019, the most significant feature was the growth in the figures. he ratios shown were relatively stable.  Up to 2019, capital allowances used were just over 90 per cent of the amount claimed, and consumption of fixed capital (depreciation) in the National Accounts was just under 90 per cent of the amount claimed.  This stability did not continued into 2020.

In 2020, we can see that capital allowances used fell to just over 50 per cent of the amount claimed.  We also see a break in the link between capital allowances claimed and consumption for fixed capital in the national accounts.

The break of this link is interesting as it suggests that the increase in capital allowances claimed is not fully linked to assets that would be included in the capital stock for National Accounts purposes.  In the last two years, capital allowances claimed in the aggregate CT calculation have risen by over €80 billion (from €86 billion to €173 billion) while consumption of fixed capital for NFCs in the national accounts has increased by “just” €20 billion (from €78 billion to €98 billion).

A second pointer comes from looking at how the unused capital allowances, which came to almost €75 billion in 2021, show up elsewhere in the CT stats.  Typically, we might expect an increase in unused capital allowances to result in an increase in loss carried forward.  In most situations, unused capital allowances from one period are carried forward as a loss to use against income in a subsequent period.

However, changes in losses carried forward were nowhere near large enough to accommodate the scale of unused capital allowances in the above table.  The are lots of losses forward sloshing around the Irish CT system and lots of them are the result of unused capital allowances but the changes in 2021 (+€7 billion in losses forward) are of little help in explaining what is going on with capital allowances.

Of little help except it tells us where to look.  There is one instance where unused capital allowances are not carried forward as a loss but as capital allowances to be claimed again in subsequent periods (until they are eventually used).  And those are capital allowances for intangible assets.  And perhaps, unsurprisingly, this is where there have been the largest changes in recent years.

Unfortunately, the Revenue do not provide figures for capital allowances for intangible assets used but we do have figures for such capital allowances claimed.  Here they are by sector for recent years with very large growth recorded for a number of sectors notably manufacturing, wholesale and retail, and ICT.

We can see from the total (€131 billion in 2021) that, on their own, capital allowances claimed for intangible assets significantly exceed the total amount of all types of capital allowances used in 2021 (€99 billion).  We can safely conclude that capital allowances for intangible assets are the reason those ratios broke down in the previous table.

Is there cause for concern? It is hard to know.  We know that there were significant onshoring of IP assets to Ireland in 2021 and 2022.  This would have increased the amount of capital allowances claimed.  And we know that there are now legislative restrictions on the amount of such capital allowances that can be used.

Since October 2017, new claims for capital allowances for intangible assets have been restricted to offsetting 80 per cent of the profit (before deduction of interest and these capital allowances) by such capital allowances.  Prior to this the cap was 100 per cent.  Any available capital allowances above this amount cannot be used in the current year and are carried forward to be claimed as a capital allowances in subsequent years.

The reason for the cap on capital allowances for intangible assets is to ensure that losses cannot be artificially generated for use elsewhere by a company or group.  Capital allowances for intangible assets are ringfenced for use only against profits generated by the acquired intangible asset.  If unused capital allowances could be carried forward as a loss they could be used against any profits.  All that has changed in recent years is the cut-off for the cap.

The cap will likely result in a “snowball effect”, at least initially, increasing the gap between the capital allowances claimed in any year and the amount used.  In the early years a company will have capital allowances that they can claim each year (following either the accounts or fixed rate approaches).  The cap may mean they cannot use all of these in the current year.

Then they may also have unused capital allowances from previous periods that they will also claim in the current year.  These will also be unused in the current year.  Thus the amount of unused capital allowances to be carried forward grows.  This could continue until they are no new capital allowances to be claimed and the unused capital allowances carried forward will be unwound until they are fully exhausted.

We don’t have precise figures but maybe a bit of guesswork can put us in the ballpark.  Claims for capital allowances for intangible assets have skyrocketed in recent years.  For the sake of exposition, let’s say the cap means that €10 billion of the amount claimed cannot be used.  That would mean that €10 billion would be carried forward to be claimed in the next period.

In the next period let’s assume that the cap again means that €10 billion of the amount claimed for that year cannot be used.  We than also have the €10 billion from the previous period that is brought forward and again claimed.  This means there are now €20 billion of unused capital allowances.

In the next period there’s another €10 billion of that year’s claim that can’t be used and this is added to the €20 billion from previous periods that are brought forward claimed again.  We are now up to €30 billion to be carried forward and so on.

We cannot say that this is the only thing that is going on but it does seem likely to be a large part of the story. And for the IP that was onshored in 2020 and 2021 this will go on for a few years yet. The gap between the amount of capital allowances claimed and the amount used will grow ever larger.  Down the line the full amount of the capital allowances will have been claimed and the amount available in any subsequent year will only be those which have been carried forward and these will eventually be fully exhausted.

The figures are so large that there may be something else going on but it is hard to make out. What we can see are the growing claims for capital allowances for intangible assets.  Due to the snowball effect outlined above this is likely to continue. It could be some time before calm returns to the aggregate CT calculation.

Corporation Tax motors along

Corporation Tax receipts continue to pour in for the Exchequer. 2023 seems set to extend the decade long run of each year exceeding the previous year – though the gap to last year has narrowed.

In July, CT receipts for 2023 were running about €1.5 billion ahead of those for 2022.  August and September weren’t as strong as last year reducing the gap to €600 million. 

August seems to have been affected by some firm-specific idiosyncrasies that will wash out, while there is little to be taken from the September figure.  September is month T+9 for firms with a December year-end and is when they file their tax return and make their final tax payment for the previous year.

On a 12-month basis, CT receipts seem to have plateaued around €24 billion which is an extraordinary amount.

Which means that the growth of the 12-month sum has also eased considerably.

Optimism for the remainder of 2023 is mainly due to the strength seen in June. Companies with December year ends pay the bulk of the CT in June and November (corresponding to month 6 and month 11 of their financial year).

The €4.2 billion collected in June of this year points to November receipts of around €6 billion, which would be €1 billion more than the same month last year.

And with the rate for large companies set to rise to 15 per cent and the exhaustion of capital allowances for onshored intangible assets the risks to forecasts would seem to be on the upside.

Population Projections

Almost all of the work the CSO does is retrospective – collect data on what has already happened.  There is one area where part of the work undertaken is forward looking – demographics.

After every Census the CSO use a set of fertility, mortality and migration scenarios to project out the country’s population by several decades.  These are not forecasts; merely projections of what might happen under different scenarios.  The most set available are those undertaken following Census 2016 and they give projections out to 2051.

The projections were based on three scenarios of net inward migration:

• M1: net inward migration of 30,000 per annum
• M2: net inward migration of 20,000 per annum
• M3: net inward migration of 10,000 per annum

and two fertility scenarios:

• F1: fertility rate remains at 2016  level of 1.8
• F2: fertility rate declines from 1.8 to 1.6 by 2031 and stable thereafter

When combined, the scenarios give six projections with the two fertility scenarios used for each of the three migration scenarios.

We have now passed the subsequent census and can compare the projections to the actual outturn of recent years.  Obviously, there are things the projections, which were published in June 2018, could not be expected to account for such as COVID-19 and the Russian invasion of Ukraine.

Anyway, here are the six projections and the actual outturns seen.

Reality was outstripping the projections even before they were published.  Of course, this reality was not confirmed at the time and only fully revealed with the results of Census 2022.  The latest population figure for 2018 (4.885 million) is 20,000 more than the highest projection for that year (M1F1 4.865 million).

For 2020, (with the figures covering April of each year) the gap between the actual population (5.029 million) and the highest projection (4.988 million) was 42,000.  The gap to the lowest projection for 2020 (M3F2 4.904 million) was 126,000. And that is for a time before COVID or Ukraine could have any impact.

For 2023, just five years after the projections were published and the gaps now range from 117,000 to 274,000.  Now clearly, the impact of those fleeing the war has impacted that but there were already significant, and growing gaps, by 2018.  Under the lowest projection (M3F2), the actual population of 2023 was not due to be reached until 2034.

We can compare the migration scenarios set out in the projections to what has actually happened since 2016.

• 2017: +39,200
• 2018: +44,400
• 2019: +44,000
• 2020: +44,700
• 2021: +21,800
• 2022: +51,700
• 2023: +77,600

Hindsight is 20/20 vision and all that, but right off the bat it can be seen that even the highest scenario of +30,000 per annum was significantly exceeded.  More recent years have been impacted by COVID19 and Ukraine.

Are such population projections important?  They are to the extent that they impact policy.  Our current National Planning Framework was published in 2018.  The baseline projection on which this was based was for the population to reach 5.7 million in 2040, just 400,000 more than the 2023 population. The NPF stresses the need for plans to be flexible and the ability to adapt for higher population outturns.  We are already there.

Wages and Salaries in the Labour Costs of the EU14

Eurostat produce a quarterly labour cost index. A number of sub-components are also produced. Here is the ‘wages and salaries’ component for the Business Economies (NACE B to N) of the EU14 (the previous EU15 now excluding the UK) with Q1 2013 set equal to 100.

All the selected countries have seen nominal hourly wage growth over the period ranging from 43 per cent in Austria to 13 per cent in Italy.  Ireland is second-ranked with an increase of 33 per cent over the period, along with Germany.

Next, we look at the annual growth rates with a four-quarter trailing moving average taken to smooth out some of the volatility in the series.

We seen that annual nominal hourly wage growth in Ireland is fairly stable over the past few years, averaging close to 4 per cent since 2018. This would enough to be Ireland, top of the group in 2019 and again briefly in 2020 but accelerations in many countries has seen Ireland’s relative position drop towards the bottom and is now close to Sweden, Denmark and Finland.

Of course, nominal changes are just changes in numbers. We want to assess real changes, i,.e. changes in the purchasing power of wages. To do that we must adjust for changes in the price level. We will do that with Eurostat’s Harmonised Index of Consumer Prices (HICP).  Here it ins since 2014, with the start of that year set equal to 100.

We can see that the HICP for Ireland was remarkably stably through the most of the period shown. In the HICP for Q4 2020 was little changed from what it has been in Q1 2014, i.e. there was no inflation over the period.  This certainly changed in 2021.  In the last couple of years, prices have risen by around 20 per cent in most of the selected countries. However, the reasons for the inflation have largely been common the relative ranking is largely unchanged.

Compared to Q1 2014, Ireland’s HICP was 18 per cent higher in Q2 of this year. This was the third-lowest increase in the EU14 with Greece and Denmark seeing smaller increases.  The largest increase was in Austria where the HICP was over 30 per cent higher compared to its level in 2014.

Inflation rates are coming down. The use of a four-quarter moving average understates the extent of the recent falls in inflation rates.

Our purpose is to adjust the nominal wage growth rates for inflation. We see that Austria has both the highest nominal wage growth since 2014 (43 per cent) but also the highest price level increases over the same period (HICP +32% since 2014).  Here we combine both Eurostat datasets to get an index of real, i.e. inflation-adjusted, hourly wages.

Combining wage growth and HICP inflation sees the green line move to the top, but the last few years have seen a steep decline. From 2013 to the end of 2020, real wage growth is put at +20 per cent for Ireland, around five percentages points higher than any of the other selected countries. 

During 2021 and 2022, real wages declined as the rate of inflation outpaced the growth of nominal wages.  In recent quarters they have moved closer together and real wages have been stable in Ireland in 2023 though are now only +12 per cent compared to the start of 2013. Much of the gains made up to 2021 have been reversed.  Real hourly wages in Ireland are now back to where they were in 2019.

To conclude here are the latest real annual growth rates for hourly wages:

As shown by the levels, Ireland is back close to zero. Nominal hourly wages are up close to 4.5 per cent and the HICP is around five per cent higher than it was last year.

To better see the trends for individual countries we will use a four-quarter moving average to smooth out some of the volatility.

Ireland has ground to make up to get back to the top spot it held in the pre-COVID years.

The past is a foreign country

We all like looking to the past for evidence and pointers to help understand the present.  A deterioration in the current of the balance of payments has been such a pointer for instances of economic mismanagement in Ireland. 

Following Honohan and Walsh (2002) we can see that in the last fifty years the Irish economy has gone through two loops of:

1. Imbalances building up via that deterioration of the current account,
2. Weaknesses being exposed and resulting in a shooting up of the unemployment rate,
3. A period of painful readjustment before,
4. Recoveries took hold.

These loops from 1975 to 1998 and from 2003 to 2018 are shown below using the unemployment rate (vertical axis) and the current account (horizontal axis).

The thing is though, these are not really useful for assessment the current position of the Irish economy.  2022 saw Ireland with an average unemployment rate of around 4.5 per cent while the surplus on the modified current account of the balance of payments (CA*) was equivalent to around 7.5 per cent of modified gross national income (GNI*). 

The economy had reached a near identical position in 2019 but then COVID19 reared its head.  Relative to the economic history of the State since its foundation, it is an unprecedented position: near full employment and large balance of payments surpluses. From a policy perspective, the unemployment rate could be a bigger constraint to achieving priorities. Sure, we should not forget the mistakes of the past, but they do things differently there.

Addendum: It is somewhat nonsensical but a further way of looking at it is to consider the sum of the unemployment rate and the current account deficit (thus a deficit is represented by a positive number).  This shows:

Again, we can see the imbalances building up after 1975 and 2003.  But the main thing is to highlight just how far 2022 is from even those starting points.  We know things can change quickly but we should also recognise where we are now is pretty much unprecedented.

Still Waiting for US GDP to be Revised Up

We have been tracking the impact in Ireland’s national accounts of the structures of US MNCs for some time.  In recent years, many of those structures have changed leading to changes in several key metrics in Ireland such as the stock of intangible assets and the destination of outbound royalty payments. 

As a result of the changes in the destination of royalty payments we have been expecting revisions to US GDP.  See previous post here with an even earlier one here. 

The previous post goes through the impact of the changes introduced by Google and Facebook in 2020 when they ended their use of “double-irish” type structures.  We won’t repeat that detail here but will highlighted the updated position of the mismatch between the Balance of Payments figures for Ireland and the U.S.

One point we will reiterate is the changed nature of outbound royalty payments from Ireland.

Obviously the first thing to notice is the scale. These outbound payments are enormous. In the 12 months to the end of March they came to almost €140 billion.  What the chart also shows is the change in the destination of these royalties.

Up to 2020, most of the royalty payments from Ireland ended up in offshore financial centres such as Bermuda and the Cayman Islands, now the bulk of them are going to the US.  The non-US portion is still quite significant (€34 billion in the year to March) but due to suppressed data we don’t have direct insight into the destination of these. It looks like most of these payments are going to The Netherlands, Switzerland and Singapore (which would also be among the potential locations for US MNCs to locate their IP). 

Anyway our interest here is in the royalty payments from Ireland to the U.S. and these now exceed €100 billion in annual terms.

To see the mismatch (and avoid any classification issues) here are the balance of payments figures for services trade between Ireland and the US. What is shows are Eurostat figures for services imports by Ireland from the US and BEA figures for service exports from the US to Ireland. In principle these should correspond. Is that what we see? Hmmmm, no.

There is a few chart crimes going on there, most notable that the series are in different currencies, but the conversion from dollars to euro has nothing to do with the gap that has emerged since 2020 (and would actually only increase it).

We can see that Balance of Payments data on this side of the Atlantic is showing that, in 2022,  Ireland had almost $200 billion of services exports from the US.  Over on the other side of the pond, their Balance of Payments data shows that the US $85 billion of service exports to Ireland.  And as noted above, in principle these should be the same.

The chart is for all services so it cannot be a classification issue – unless the payments are accounted for by the BEA as goods exports or primary income, both of which are unlikely.

We cannot immediately assume that there is a GDP effect.  The trade could be in IP assets which, initially at least, would be GDP neutral.  There would be disinvestment (-) and exports (+) on the US side with investment (+) and imports (-) on the Irish side. 

But as might be expected given the contents of the post it is mainly a story of royalties.  Here are the royalties components from Eurostat and the BEA of the above overall services trade figures.

Mind the gap! Ireland reporting €102 billion of royalty imports from the US in Eurostat’s data becomes the US reporting just $16 billion of royalty exports to Ireland in the BEA’s data.  And as we started by looking at total services this cannot be explained by classification issues.  We have a difference of something approaching €90 billion.

For what it is worth, it can be noting that the revision to Ireland’s national accounts for 2015 (the 26% growth rate and all that) was of the order of €40 billion.  Some US commentators should perhaps be wary of skeletons in their own closets.

However, as the previous post goes through the likely GDP impact would not be the full difference shown above as a share of the royalty payments that previously went from Ireland to the like of Bermuda did subsequently flow on to the US as payments for R&D services exports.

Statements from Google and Facebook have confirmed that they are now licensing their technology directly from the US rather than offshore locations.  This should have led to a reduction in US R&D service exports and an increase in US royalty exports. 

The rise in royalty exports would be larger (reflecting the fall in profit reported in Bermuda etc.). It is possible there could be a GDP impact of $40-50 billion for 2022 – equivalent to around 0.2% of US GDP.  Not exactly headline grabbing.

Why the gap?

In part, the answer to this apparent puzzle comes down to scale. In a small economy like Ireland, the activities of companies such as Google and Facebook are relatively enormous. The Large Cases Unit (LCU) in the CSO will be in regular contact with these companies for filing updates and notice of any significant changes.  For an economy of the scale of the US, such companies are simply are not as important in and of themselves when it comes to the BEA’s production of their National and International Accounts.

In the BEA’s case they undertake “benchmark surveys” every few years of a wide number of market participants to get a deep understanding of what is going on.  Between these comprehensive “benchmark surveys” the BEA do quarterly surveys of a much smaller sample and extrapolate the data they publish from those.

The most recent benchmark survey for Transactions in Selected Services and Intellectual Property with Foreign Persons was carried out for 2017.  The data was collected in 2018 with the results published in 2019.  As it so happens, the BEA are in the process of collecting data for the latest such survey.  In this instance the benchmark year is 2022 with the data collected in 2023 and results not due until sometime in 2024.

So it looks like the significant changes implemented by US MNCs operating in Ireland fell in the middle of the gap between the BEA’s benchmark surveys that would have picked them up.  They were too early for the survey for 2017 and while they should be picked up by the ongoing survey for 2022 the results won’t be pulled together and incorporated in the BEA’s International Accounts until 2024.  It seems our wait for US GDP to be revised up will be going on for a while yet.