Irish Times 29/03/13

The previous post on EU corporation tax revenues reminded me of an article for The Irish Times that was published in the Friday business section about a month ago.  I don’t think it was available online and was based on this post.  Here is a scan of the piece and the text is reproduced below.

Increasing headline rate of corporation tax will do little to unravel ‘Double Irish’

Ireland's headline rate of corporation tax is 12.5% despite some pressure to increase it.  Such pressure seems to have been successful in getting Cyprus to increase its corporate tax rate as part of its EU/IMF programme and now only Bulgaria has a lower corporate tax rate in the EU than Ireland.

However, it is not necessarily the headline rate but the calculation of the tax base that is the most important feature of the Irish corporate tax regime.
Recent data published by the Revenue Commissioners show that, in 2010, companies in Ireland reported an aggregate gross profit of just over €70 billion. On these profits, companies paid €4.2 billion which gives an average effective tax rate of 6%.

Effective tax rates will always differ from headline tax rates. For example, the headline rates of personal income tax in Ireland are 20% and 41% but the average effective income tax rate for the €78 billion of personal income reported in 2010 was less than 13%.

Personal tax credits are the primary reason for lower effective rate for personal income, though some allowances such as pension contributions mean that taxable income is not the same as gross income.

The difference between gross income and taxable income is also true for companies but to a much greater extent. While gross company profits in 2010 were €70 billion deductions and allowable charges meant that taxable profit was €40 billion.

These deductions include a provision for past losses as only cumulative positive profits are subject to tax. Many companies made significant losses in years prior to 2010 and €4 billion of such losses were used to offset gross profits earned in 2010 for tax purposes.

Companies can also spread the charge for capital expenditure against their gross profit over a number of years. Although capital expenditure, for example on new plant or equipment, occurs at the time of purchase, the allowance for such capital expenditure against company income is spread over a number of years to better match the income flows from the capital item purchased. In 2010, companies in Ireland used €12 billion of capital allowances.

The third major deduction from gross profit are trade charges. These are expenses that companies incur but which are not deducted when the gross profit calculation is made. The most significant trade charge incurred by companies in Ireland is in relation to patent licenses and royalties. 

The corporate tax treatment of payments for intellectual property is what makes Ireland an attractive location for computer services companies such as Google, Microsoft and Facebook as well as large pharmaceuticals.  Companies incurred around these €12 billion of allowable trade charges in 2010.
Finally group relief for intra-company transfers meant that a total of €30 billion of deductions were made from gross profit to determine the amount of taxable profit.

Applying the 12.5% rate to €40 billion of taxable profit meant that the tax payable was around €5 billion. A number of tax credits such as incentives for research and development expenditure in Ireland as well relief to prevent the double taxation of income meant that the total tax paid was €4.2 billion.

The data from Revenue Commissioners also show that the effective rate declines with company income. Companies reporting a net trading income of between €50,000 and €100,000 faced an effective tax rate of 10%. This declined with income and companies with a net trading income of €10 million or more had an effective tax rate of 6%.

This suggests it is the larger companies which are using the reliefs and charges to reduce their tax bill.

Google Ireland has around €12 billion of revenue in Ireland but it makes a much smaller annual profit of around €100 million. Changing the Irish tax rate would generate more tax from the €100 million but would do very little to the €12 billion of revenue flows.

Through a process known as the Double-Irish the profits made by Google Ireland on its €12 billion revenue accrue to a company resident in Bermuda.
It is features of the Irish and Dutch systems and, most importantly, a tax treaty between them that allows the money to be transferred without a tax liability to Bermuda. The profits are taxed in Bermuda where the corporation tax rate is zero.

Changing the Irish tax corporate tax rate will not change the effectiveness of this scheme in reducing global corporate tax bills.  If we change our business-friendly environment the income flows, and more important the associated jobs, will simply move to another country.

Corporate Tax Revenues

The release yesterday by Eurostat of some tax revenue statistics rekindles the debate of whether Ireland is/is not a “low-tax economy”.  With receipts of 28.9% of GDP (35.4% of GNP) Ireland is below both the EU27 (38.8% of GDP) and EA17 (39.5% of GDP) weighted averages.  One tax that attracts attention is the Corporate Income Tax. 

At 12.5%, Ireland has the third-lowest headline rate of Corporate Income Tax in the EU.  Only Bulgaria and Cyprus are lower (and as part of the recently negotiated “rescue” programme Cyprus has committed to raising its rate to 12.5% leaving only Bulgaria at 10%).

This clearly supports the “low-tax” hypothesis.  But looking at Corporate Tax revenues as a percent of GDP paints a somewhat different picture.

The presence of countries with large financial sectors relative to the GDP is notable.  But the fourth member of that group is not at the top with them. 

Using this approach Ireland is pretty much an average taxer of corporate income.  So which is it?  Both sides in the “low-tax” debate can provide support for their position.

One determinant of the latter ranking is obviously the amount of corporate income in the country.  A graph from an earlier post gave the Gross Operating Surplus of the Corporate Sector as a proportion of GDP.

Relative to GDP there is more corporate income to tax in Ireland than in all but one EU country.  Even with a low rate this alone could push Ireland into mid-table in the previous chart.

With this in mind we can calculate a crude ‘implicit tax rate’.  This is akin to an average effective tax rate.  In the chart below it is simply Corporate Income Tax Paid divided by the sum of Net Operating Surplus and Net Mixed Income.  The 2011 figures are missing for five countries (Bulgaria, Greece, Malta, Portugal and Romania ) but the remaining data produces the following chart, where the EU27 and EA17 are simply the unweighted arithmetic averages.

The missing countries are unlikely to alter the result.  Here it can be seen that Ireland is a “low-tax” country on corporate income.  The implicit rate in France is almost four times greater, which may go some way to explaining the pressure applied to have the Irish headline rate changed as part of the rescheduling of Ireland’s loans from the EU in the summer of 2011.  The other half of the “Double Irish” – the Dutch Sandwich – shows up with the Netherlands also being a low-tax country for corporate income.

The position of Cyprus seems incongruous.  It has (for the time being) the joint-lowest headline rate of corporation tax in the EU but has the highest implicit rate.  This is confirmed by the implicit rate calculated by Eurostat (see Table 86 on page 257 (pdf page 259) which is reproduced below). 

The above implicit rates are different to Eurostat’s because missing data mean Ireland is excluded from the Eurostat table but the relative ranking is generally the same for the countries included in both.  The difference is mainly due to the treatment of dividends with dividends paid excluded from the calculation of corporate income (dividends are included in the tax base for the corporate income tax but it is done on a withholding basis).  The methodology used by Eurostat and their results for the implicit corporate income tax rate are below the fold.

1. How Eurostat calculate an implicit tax rate for corporations.

2. What Eurostat finds (for some countries).

GDP and International Comparisons

It is frequently stated that Gross Domestic Product (GDP) figures for Ireland are not suitable for international comparisons because of the distorting effect the presence of MNCs here and their use of tax arbitrage activities.  Gross National Product (GNP) is sometimes provided as an alternative (though a hybrid is probably best).

GNP is GDP plus net factor income from abroad (the flows of wages, rents, interest and profits into and out of a country).  The outflow of MNC profits means net factor income is negative for Ireland and GNP is less than GDP.

This can easily be seen using data from Eurostat.  Here is a chart of Gross National Income (GNI) as a proportion of GDP for 2011.  GNI is GNP plus net transfers from abroad (the flows of foreign aid, EU contributions/subsidies and other transfers).  GNI is largely the same as GNP.

For most countries the difference between GNI and GDP is small.  Ireland and Luxembourg are clear outliers.  International comparisons based on GDP (such as tax revenue statistics) do not take this into account.

Here is a similar chart but this time from Eurostat’s Institutional Sector Accounts.  This shows Gross Disposable Income of the household sector as a proportion of GDP. (Data for Malta and Bulgaria is not available).  Gross Disposable Income of households is

1. Self employed profits + net property income + wages earned = Gross Income
2. –  taxes paid – social contributions + social transfers = Gross Disposable Income 

Again, Ireland is towards the bottom of the chart but it is no longer in a small group of outliers.   It is difficult to determine the reasons for the relative rankings in the chart.  That is not the purpose and the reasons are manifold.  There may be something to the ranking but I’m not sure.

The purpose is simply that while some dislike the use of GDP in an Irish context because of the relative gap to GNP, maybe there are some in other countries who dislike GDP because it does not reflect the relative disposable income of their household sectors.  National and Sectoral accounts have something for everyone.

[I should be noted that Household GDI is 66% of Irish GNP which is bang on the EA17 average as a % of GDP.]

Finally here is Gross Operating Surplus of the corporate sector (financial and non-financial corporations) as a proportion of GDP.  Gross Operating Surplus is gross value added less wages paid less production taxes.

10-year yield at 3.49%

Here is a snapshot of Irish government bond yields as calculated by this website (go to Live Quotes ⇒ Bonds ⇒ World Government Bonds and select Ireland).

Really?

Recapitalising the Banks

The issue of further capital for the banks has attracted some attention in recent days.  Prof. Brian Lucey had a piece in Saturday’s Irish Examiner and yesterday’s Sunday Business Post led with the headline ‘IMF warns of new €16bn black hole in Irish banks’.

The issue in the SBP piece is actually about the contingent liabilities of the State rather than the banks and the IMF have actually been making the same point for at least a year.  Here is a quote from the IMF’s fifth review issued this time last year with the same 10% of GDP (€16 billion) contingency.

Recognition of contingent liabilities would constitute a one-off increase in the level of debt. Ireland’s contingent fiscal liabilities relate to the covered banks, the IBRC, and NAMA. There is no expectation of losses from these entities as the covered banks have been recapitalized under PCAR 2011, the IBRC meets capital adequacy requirements, and NAMA received assets at heavy discounts—averaging 58 percent—to protect its viability. Under the standard scenario, the assumption of 10 percent of GDP in contingent liabilities by the Irish government would raise the debt-to-GDP ratio to 124 percent in 2012 and cause it to peak at 129 percent in the following year, but starting from 2014 debt would start to decline steadily, reaching 123 percent by 2016. However, the debt trajectory would be higher if the higher debt level resulted in higher interest rates on new market funding.

Although the level and composition of the contingent liabilities have changed over the year (NAMA Bonds, ELG guarantees, ELA Guarantees), and are subject to further change because of the IBRC liquidation, the IMF have not adjusted the 10% of GDP contingency in their scenario analysis.  It is not clear that they have given this issue much consideration recently.

In fact if we go all the way back to the IMF’s first review (May 2011) we find this graph in the annex on public debt sustainability (page 41).

And even before November 2010, the IMF included a ‘one-time 10% of GDP contingent liabilities shock’ in their debt sustainability analysis.  Check out page 37 of the Article IV Report on Ireland published in June 2009.

So the IMF is not warning of a ‘new €16 billion black hole in the Irish banks’ but the broader question still stands:  will the Irish banks need more capital?  Maybe or maybe not.  When Craig Beaumont, the IMF Mission Chief to Ireland was asked as part of the conference call on the publication of their latest report on Ireland (the ninth review) he was non-committal as can be seen below the fold.

SPEAKER: Great. Do you think the Irish banks will need more capital, leaving aside the Basel standards in terms of the --

MR. BEAUMONT: Well, like I said before, the analysis hasn’t been done and we can’t prejudge it.

The analysis is the latest PCAR (Prudential Capital Assessment Review) due to be completed over the next few months and published in the autumn.  Of course we have been here before.  There was the inadequate 2010 PCAR but this was followed by the much more granular analysis in the 2011 PCAR.

In the 2011 PCAR exercise the banks were recapitalised for expected losses in the adverse or stress scenario which, along with the baseline scenario, is summarised in the following table.  Click to enlarge.

The projected losses used were those in the adverse scenario.  A quick comparison to what has happened subsequently shows that the outturn has been worse than the baseline scenario but in most cases is slightly better than those in the adverse scenario.

The first estimates from the CSO are that, in 2012, GDP grew by 0.9% and GNP by 3.4%, Consumption declined by 0.9%, Investment grew by 1.2% and Government Consumption fell by 3.7%.  All of these are better than the adverse scenario, though in some cases not by much, as were the outturns for 2011.  The Balance of Payments current account recorded a surplus of close to 5% of GDP in 2012.

According to the QNHS, employment grew 0.1% in 2012 and the like-for-like unemployment rate with the above table finished the year at 14.2%.  The CSO’s residential house price index shows that prices fell 16.7% in 2011 (close to the adverse scenario projection of 17.4%) but the 2012 fall of 4.5% was significantly better than even the baseline scenario projection of 14.4%.  We don’t yet have full year data for personal disposable income but the data to Q3 (source) show a 4.1% increase on the first three quarters of 2011, while an annual decline of 1.2% was used in the adverse scenario.

It is notable that the level of mortgage arrears are not included in the projections and it would be interesting to see what level, if any, was included in the analysis.  The projections gave rise to the following capital requirements in AIB (now merged with EBS), BOI and PTSB (now separated from Irish Life).

Some €18.7 billion of capital was required to cover the projected loan losses under the adverse scenario that were expected to arise by the end of 2013.  Additional capital buffers (equity and contingent capital) brought the total requirement to €24 billion. 

The following shows how the €24 billion was reached in aggregate across the banks (click to enlarge).

The expected operating profit over the three years of close to €4 billion is unlikely to be achieved.  In the stress case BlackRock Consultants projected that there would be €40.1 billion of lifetime loan losses in the banks.  The Central Bank estimated that €27.7 billion of these would occur within the three-year timeframe of the exercise and these were deducted from the opening capital stock and balance-sheet provisions from the end of 2010.

A further €13.2 billion of losses were allowed for from the deleveraging of non-core loans.  Although specific details of the deleveraging have not been provided it has been stated on a number of occasions that the “overall cumulative discounts incurred have been within PCAR assumed discounts.”  The lower losses here, such that they are, may cover the shortfall on the €3.9 billion of expected operating profits.

Much of the focus will be on whether the €27.7 billion of three-year losses allowed for up to the end of 2013 will be enough and whether lifetime losses after 2013 will continue to erode the capital base.  Last week’s IMF report has a useful summary of the position of the PCAR banks.  Click to enlarge.

The banks’ core-tier one (CT1) capital has increased to €22.3 billion (from €13.3 billion at the end of 2010).  The stock of loan loss provisions on their balance sheets has increased to €26.8 billion (from €9.9 billion at the end of 2010).  When summed to €49 billion these are large relative to the €223.8 billion of gross loans that the banks have. 

Finally, here is a breakdown of the €27.7 billion of three-year losses from the lifetime total of €40.1 billion used in the exercise.  Again click to enlarge.

The banks have recognised some loan losses since the PCAR but if the banks were to fully recognise the losses shown above and write-off the projections as loan losses it could be almost fully covered by their current stock of provisions.  Having no provisions would not leave the banks in a particularly sound state but given their existing capital they can set aside a further €6 billion for provisions (assuming no operating losses) and still remain above the 10.5% CT1 ratio.

It remains to be seen if this is enough but there is considerably capacity in the covered banks to deal with loan losses. 

Yield Curve

This site gives a snapshot of Irish government bond yields at different maturities.

Net Income changes since 2008

Here is a summary table of changes to net income at different gross annual incomes since 2008.  There is two figures for 2009.  The first corresponds to Budget 2009 introduced in October 2008 and the 2009* figures are the result of the Supplementary Budget introduced in April 2009.

All the figures are derived from this tax calculator which is the source of any errors and omissions.  The figures are based on a single, private-sector, PAYE worker aged under 65 who does not have a medical card.  No allowance is made for pension contributions, medical expenses or the like.  The table is merely to highlight some of the relative changes that have occurred in net pay.  Net pay is the starting gross figure in each column after the deduction of Income Tax, the Universal Social Charge (or the Health and Income Levies as appropriate) and PRSI.

As can be seen a worker on €15,000 has seen their net pay fall by €399 as a result of the introduction of the USC in Budget 2011.  This represents a drop of 2.7%.  The reductions increase through the income distribution.  At an income of €250,000 the reduction is just over €20,000 leading to a 13.3% fall in net pay. 

Around two-thirds of this reduction resulted from the measures introduced in the April 2009 Supplementary Budget.  The changes were a reduction in the thresholds and a doubling of the rates for the Health and Income Levies as well as an increase in the (now abolished) PRSI ceiling.  Some details of these changes are here.  The Income Levy had only being introduced the previous October and the two levies were subsequently merged into the Universal Social Charge in Budget 2011.