Given that BoG released its April balance sheet yesterday, it is a nice opportunity to take a fresh look at developments in its major components:

BoG Balance Sheet 2019

The things that stand out are the following:

  1. Regular monetary operations + ELA have been minimized to very low levels of less than €10bn. ELA more or less ended in 2018 since its figures are now basically zero.
  2. Securities purchased in the context of the ECB QE program are now stable around €63 bn.
  3. Total collateral posted recorded an impressive fall from €80bn a year ago to only €16.7bn now, a function mainly of the drop in ELA operations (from €10bn to zero).
  4. Target2 liabilities are down to only €24bn while extra banknotes are now recorded as a (small) net asset instead of a liability suggesting a significant return of cash outside of the banking system.
  5. Government balances have reached almost 2€9bn providing a large enough buffer for debt refinancing needs in the coming future.

The Greek situation has officially been stabilized with any creditor risk substantially minimized. Net Securities (Securities held for monetary purchases minus Target2 liabilities and extra banknotes) are now close to €40bn which means that BoG can easily cover any liabilities towards the Eurosystem even in the case of a Grexit while still maintaining a positive balance of European government securities.


The return of economic growth in the Greek economy appears to have a positive impact on fixed investment. Given the opportunity it is interesting to take a closer look on fixed investment components since before the Greek Great Depression (GGD).

Greek Fixed Investment Components

I ‘ve grouped Fixed investment into two major categories: (Private) Dwelings + Transport Equipment (Ships etc) and Other Investment. Greece experienced a housing boom in the early Euro years while the ship balance exploded after 2004 from roughly balanced to €5.4bn in 2007. While both series increased steadily during the 1995-2005 decade, dwelings + transport equipment surged almost €13bn in a very short time period.

Moreover, while Other Investment stabilized at around €15bn annually after 2012 and has since touched €20bn, the former category dropped to only €5bn per years, a loss of €33bn in a matter of years.

New building permits suggest that this drop is permanent with 2018 permits being 15% of 2007 figures. Decreasing population, stressed bank balance sheets and low household incomes are all factors pointing towards the same direction with only demand for AirBnB apartments being suggestive of positive prospects.

Assuming an accelerator model for fixed (other) investment we can examine Other Investment as a percentage of consumption and exports (total demand) to determine if there is scope for improvement:

Greece Other Investment % Total Demand

Other Investment hovered between 8-10% of total demand in the period before the GGD with a relentless fall to 6.5% until 2012Q2 probably due to animal spirits and lack of access to credit. It now moves between 8-9% suggesting small room for further improvement.

In my view, the above indicate that Greek fixed investment will not return to pre-crisis levels any time soon and projections of 20% GDP investment are quite optimistic and probably overstated. The share of GDP will most likely stabilize around 15% of GDP with dwelings construction being permanently lower.

Given that the external balance is the difference between domestic savings and investment and that the Greek government is destined for semi-permanent primary surpluses of 2-3.5% GDP, my view is that the (Greek) «structural» current account deficit is now quite low. Greek households have already dis-saved almost €40bn while the Greek banking system will not be in a position to provide ample amounts of credit any time soon.

On the other hand, anemic dwelings investment, stable other investment and large Greek government primary surpluses indicate the absence of factors for a large boom in the Greek economy. Demand drivers will include the steady increase in employment (and consumption), external demand and «autonomous» investment mostly in the form of FDI. I am not sure how Greek growth can become impressive any time soon.

I recently finished Thomas Piketty main book «Capital in the 21st Century» which contains a series of very interesting observations on the dynamics of wealth and (labor) income. The one that struck me the most is the fact that in cases where income and population growth is low (something that was basically true for the most part of human history until the Industrial Revolution), inherited wealth becomes very important for wealth holdings. Exactly the opposite happens when population growth is large (hence most children inherit almost nothing) and when income growth is significant and thus wealth accumulated through saved income becomes important in determining total wealth.

The above dynamic is quite evident in the case of Greece. First of all, population is actually falling with fertility rates at 1.3 children/woman. This means that most children grow up without any siblings and thus end up inheriting all of the family wealth (75% of all households have no more than 3 members):

ScreenHunter_1189 Nov. 17 11.08

Moreover, median income has fallen back to 2004 levels. The very large debt burden, high NPL ratios and the requirement for significant primary surpluses for decades to come indicates that Greece faces a significant risk of walking an Italian path in the following decades: Income growth will be stagnant while debt loads will stay high with a large part of income devoted to servicing old debts instead of financing investment and growth.

ScreenHunter_1190 Nov. 17 11.12

In such a setting, past wealth and assets become the determining factor for a person’s total wealth with current income not being able to create new saving and wealth flows. Actually, as I outlined in a previous post, Greek households are consuming past wealth on a grand scale in order to be able to finance current consumption, a reasonable outcome of the 26% fall in median income since 2010. According to recent figures, gross saving was a whooping -7% in 2017.

According to BoG Financial Accounts, more than 60% of household financial assets is in the form of currency and deposits and another 23% in unlisted shares. The rest of household wealth is probably in the form of housing. Listed shares (which typically are the assets that have the largest growth potential) are only 3% of total financial assets (they were 40% in the height of the 1999 stock bubble and close to 10% during the golden Euro years).

Roughly ¾ of Greek households own a primary residence with median total net wealth of 90,000€. The main residence has a median value of 70,000€ making it the most important asset while median financial assets of all households are only a couple of thousand €.

For a risk-averse person (such as the Depression hit Greek population) the main priority is wealth preservation instead of a growth potential (especially if the latter comes with a probability of a destructive tail event). Thus most people would probably prefer a low-yielding Euro asset instead of a re-denomination of wealth in drachmas. An early Grexit might have avoided the 25% drop in output and the current low-growth potential yet, as long as the Greek Depression is now considered a «sunk cost», wealth preservation takes priority and staying in the Euro becomes an attractive option, even if it coincides with low income levels and growth. The «attractiveness» of the Euro option most probably increases with wealth levels which means that it becomes a «hard-line choice» for the ruling class which has most wealth to lose/redistributed in a Grexit scenario.

The Grexit option might as well have come and gone.

This will be a very short post on the way the US used to not allow extreme income inequality during the Golden Age of Capitalism.

The idea was quite simple: Going into WWII, income tax rates on the highest bracket  reached levels around 90% which meant that wage earners had little incentives to increase their annual income above the highest bracket. At the same time, that bracket was set at 100 times the lowest bracket:


The latter was at least 1.5 times personal income per capita although it reached even close to being 3 times larger. The very high level of the top income tax rate remained roughly constant for two decades (40’s and 50’s).


That meant that the highest income allowed was somewhere between 150 and 280 times that of average personal income depending on the specific year. As a result, extreme inequality was reduced through tax incentives. This is in contrast to today when CEO pay hovers around 300 times that of the typical worker:

CEO to typical worker pay

It is quite evident that fiscal (and more specifically tax) policy can be used as a tool for a society to place specific limits on the level of extreme inequality allowed within it.

I ‘ve been reading Robert Gordon’s interesting book concerning economic growth and its drivers during the last 150 years. Gordon’s main thesis is that not all inventions and technological advances are created equal or have the same impact. The major technological breakthroughs which had a significant effect on productivity (and thus on economic growth) took place mostly more than a century ago and involved creating the networks we still use. Networks of water, sewer, electricity, roads, telecommunication, transport. Networks which were introduced only once, had a strong effect on growth for decades and are now present without having any more major impact on productivity.

One can agree or disagree with his (pessimistic) view of the future, yet what I think everyone will agree on is his powerful and extensive description of how life was just 150 years and the way in which these inventions changed everyday life and how everyone used their time.

In my view, there is a clear and strong interaction between social norms, property/voting rights and technological advancement. As long as most peasants (who accounted for the bulk of the population) lived at the subsistence level on mostly self-grown food and self-made clothing, property rights were limited and even forms of slavery were present for centuries (mostly in the form of serfdom in Europe). Once technological advancement changed and enlarged the consumption basket as well as created a demand for factory workers (and an «exit» option for peasants in the country fields), labor services were remunerated in monetary terms and property started to become widespread. The introduction of large massive armies (as opposed to forms of professional army service) also accelerated the abolition of serfdom since nations needed free men to fight their battles.

As a result, while only a few centuries ago most people in the countryside had no property, faced serious restrictions in their movement (even in their most basic personal freedoms in the case of serfs), did not have voting rights and lived in the subsistence level, the 19th (and especially the 20th) century found them enjoying an ever expanding consumption basket, holding property and claiming an expanded array of rights. The Industrial Revolution and technological advancement was the main driving force for these changes.

In particular, not too long ago, even the simplest things that today we take for granted were extremely cumbersome tasks. Most families lived in rural farm houses, with no electricity, no central water and sewer systems meaning that all water (and waste) had to be carried by hand. Light was provided by candles, half the family budget was spent on food while most of the clothing (especially women’s clothing) was made at home by mothers and daughters.

In 1925, wives reported working an average of 6 hours per day for seven days on household tasks, which translates to more than a full-time 5 day/week modern job. Moreover, infant mortality was close to 20% while roughly one in four infants died before reaching the age of 5.

The high children mortality rate, coupled with the needs for household and manual farm work meant that the birth rate was exceptionally high. White women laid birth to an average of 4.6 children while black women to 7.7. Obviously that meant that the wife was usually busy raising a new born child, apart from taking care of the household on a full-time basis.

This all changed in a couple of decades. Clean running water and a sewer system led to a dramatic reduction of deaths caused by infectious diseases (37% of all deaths in 1900) and a drop of infant mortality rates from 20% to 5% in half a century:

ScreenHunter_1180 Oct. 23 23.16

By 1950, the introduction and diffusion of modern conveniences and appliances had reached a critical point. At the same time, clothing was readily available in stores and did not need to be home-made. All this, led to a significant drop of the amount of work required to maintain a household and a large fall in fertility rates. In 1940, while the US was entering WWII only 65% of consumer spending was on categories that existed in 1869, with the share of services in the consumption basket doubling from 24% to 43% (while food decreased by a comparable amount from 44% to 22%).

ScreenHunter_1182 Oct. 23 23.22ScreenHunter_1181 Oct. 23 23.22

WWII full employment along with the millions of men occupied on the front line fighting the enemy, led to the women massively entering into the labor force for the first time. During the following decades, changing social norms meant that women would pursue academic education and a career in all occupations, increasing their participation rate from 35% in 1950 to 50% by 1980 (and 60% in 2000).

Yet what is quite interesting is the fact that, until roughly the end of the 19th century, necessity was the main driving factor for women non-participation in the labor force. Keeping the household was literally a full-time job, while high fertility rates meant that there was almost always an infant present in the family requiring constant attention. One had to choose between working manually in the agricultural fields or taking care of the household. There was no central heating, no running water, no refrigerator, no sewer at the house and every duty had to be done manually, taking hours to complete tasks that take only minutes today.

Obviously this division of labor was also the result of long-standing social norms as well as a function of the hardship associated with manual agricultural work. No work by women with a monetary remuneration meant limited property rights and subordination with regard to men.

Yet necessity and (the absence of) technology played a decisive role in this division. Once advancements in technology dropped the amount of time required for household maintenance coupled with lower infant mortality and a lower requirement of a large number of working hands for agricultural manual work, women’s time was freed on a extraordinary scale. Women started entering the labor force, first as manual factory workers (in WWII) and afterwards in all sorts professions once college education was made available to them (a task that required extensive fighting from their part).

Technology did not only change the consumption basket, create new professions and free human time for more leisure but it also had profound implications on long-established social norms, the division of labor within households and personal property rights, things that had remained quite stable for centuries. Today’s equality between men and women not only rests on decades-long women’s struggles but also on inventions and advances that completely changed the landscape of modern households.

In this post I will do a simple enough exercise: Look into the evolution of the Greek government primary current balance. This is just the balance of current income minus current expenditure excluding interest. Thus it excludes both the investment budget as well as interest payments on government debt.

Although public investment does impact national income (and is added in GDP in the first place), the current primary balance can act as a crude estimate of the «weight» of government on the economic process. Especially its change from one year to the other indicates the expansionary or contractionary stance of fiscal policy.

Greek Government Primary Current Balance 1995 - 2019 %GDP

The general pattern is quite clear. The primary current balance was in surplus in the late 1990s and reached a peak of 7.5% GDP during 2000. Afterwards, the fiscal stance relaxed substantially with a drop of 6.5% GDP in the 2001 – 2007 period to reach only 1% in the final year. The Great Recession took its toll with the balance dropping almost 6% of GDP during 2008-9 to a negative close to 5% GDP.

Then came austerity: The balance grew 8.6% of GDP in the 2010-13 period while it relaxed substantially during 2014 with a drop of 1.1% compared to an increase of 3% in 2013. This can most clearly explain the return to growth in 2014 since the economy experienced a change in the fiscal impulse of 4% GDP.

The effects of the 3rd economic adjustment program are also quite visible with the balance increasing 4.2% in the 2015-18 period which explains why these years had a sense of strong austerity despite a return to economic growth.

The 2019 current primary balance is expected to reach 7% of GDP, roughly similar to the 1999 balance. Yet 1999 registered 6% GDP private credit flow with a further increase to 10% during 2000 (which increased and persisted until 2008) while 2017 closed with a 1% fall in that flow. It is thus clear that the underlying dynamics are quite different and such a large surplus will definitely act as a serious drag on growth.

It is also interesting to take a closer look on the evolution of current revenue and primary expenditure during this time:

Greek Government Primary Revenue & Expenditure excluding interest 1995 - 2019 %GDP

Current revenue starts with a steady increase during 1996 – 2000 (close to 5% GDP total increase), remains stable for a couple of years and then makes a step drop of 1.5% GDP in 2003 to 37.5% around which it hovers until 2009. Since then it starts an uphill march to around 45% in 2013, a change of 7.5% GDP. Although it grew more than 3% GDP in 2014 to 48.4%, this level appears quite unstable since current revenue is projected to return to 45.5% during 2019. Assuming that 45% is the «new equilibrium», future current primary balance will depend heavily on the primary current expenditure trajectory.

The latter appears to grow steadily from 30% in 1998 to 39% a decade later. Even the large austerity package from 2010 onward was not able to break that limit, mostly because of the large fall in GDP which made the denominator fall substantially along with nominal expenditures. The 2013-17 average was 41.7% which dropped significantly during 2018 to 40% GDP and is projected to reach 38.5% in 2019, a cumulative fall of 3% GDP in two years. Whether such a fall can be sustained in the long-run is a question that will be difficult to answer.

If primary current expenditure returns to the 41.5% average, the primary current balance will drop to half its 2019 projection to 3.5% GDP. Since that is the current primary balance target until 2022 this implies that the investment budget will need to be balanced. Given that is close to impossible, the stability of the 3.5% primary surplus target will depend on the Greek government achieving a primary current expenditure level of close to 38.5% GDP.

Actually (Gross Fixed Capital Formation – Capital Transfers Received) has a mean value of -2.4% GDP and a standard deviation of 1.42% for the period 1995 – 2017. If the values were drawn from a normal distribution the probability of a non-negative balance would be around 4%.

Since my last post was mainly focused on the issue of inequality I would like to continue on this road and introduce a concept which I will call «Economic Reproduction Frontier» (ERF).

The main idea is rather simple: Take the threshold of the major brackets of income distribution as a share of average per capita income and examine their course across time. I will be using data from the World Inequality Database on a per adult (equal split) unit for this exercise.

One can think of average per capita income as a crude pointer for both the consumption basket (or frontier) in a given country at a given time as well as the cost of human capital development for an individual. An income say 20% of average per capita income will correspond to a significantly different (and highly constrained) consumption basket and place large obstacles on the opportunities and capabilities of an individual increasing and expanding his human capital. Since most economic output is mass consumed and production is highly connected to human capital, the income available to a large proportion of the population will ultimately act as a constraint on economic growth. Especially if individual income as a percentage of average per capita income falls on a permanent basis.

To do this I calculate the share of threshold income for various income brackets in the US during the period 1966 – 2014 (due to data availability). Below are two graphs, one for P20/P30 percentiles and one for P40/P50. Linear trendlines along with the exact equation and R² for each percentile are also presented.

P20 P30 Percentile threshold % GDP 1966 - 2014P40 P50 Percentile threshold % GDP 1966 - 2014

It is striking how all threshold shares are on a permanent downward trajectory, as well as the very strong R² for all lines (over 0.93 in all cases with stronger results in the P40 and P50 cases). The relative stability of the threshold shares up until the turning point of 1980 is also evident. The regression coefficient points to a fall of roughly 50bp annually for the P40 and P50 brackets which means that each bracket losses almost 5% share of per capita GDP every decade.

P20 and P30 brackets start below 50% and fall to 20% and 30% by 2014. What is striking is that the P40 brackets falls to roughly 40% by 2014 starting from 63% (the P50 bracket falls to 56% of average income starting from 77%). This suggests that at least 40% of the population  cannot maintain a middle-class consumption basket and human capital without going into debt since it is severely income constrained. Even P60 and P70 brackets show a clear drop in the given period (from 90% to 73% for P60 and from 106% to 94% for P70) suggesting an extremely strong middle class hollowing out in recent decades, at least for the p30p70 bracket. Only the top-10% threshold share shows an increase in the given period from 170% to 184% while even the P80 threshold share lost 2.5%:

Change in threshold share of average income 1966 - 2014

Although I am sure this simple exercise will have methodological issues it still suggests a strong loss of income resources for a major part of the general population with serious consequences on long-term growth. If 40% of the population are not able to finance their human capital development through their income nor afford a basic middle-class consumption basket without going into debt, this suggests that long-term growth will be affected in one way or another.

Modern capitalist economies are based on large scale production of mass consumption goods and on using a highly educated workforce in a large part of the production process and sectors.  Linear extrapolation suggests that the P50 bracket will fall to 50% of per capita income by 2026 (while P40 will be 35% and P30 24%) making the above process highly constrained. We might be nearing an inflection point for economic growth due to growing inequality.

Jordan Peterson is a clinical psychologist who’s earned his fair of publicity in recent years, especially through videos of his lectures and speeches on Youtube. Personally, I have found the views expressed on his videos quite helpful and inspiring while his lectures contain a ton of useful information and methodology on how to address life (I especially recommend the Biblical Series).

On the political front, it is clear the Dr. Peterson leans to the right and has been highly critical of far-left and Marxist views. Nevertheless, he makes a solid exposition for the necessity for both broad political parties with the right focusing on personal effort and values, order (which allows people to cooperate and plan for the future) and hierarchy of competence (implying equality of opportunity but not of outcomes). The left on the other hand tends to focus on classes instead of individuals, inequality (which is ultimately a destabilizing force) and change which attempts to balance out the fact that the economic games tends to work like «Monopoly» with most people stacking up at zero and a significant few acquiring almost all the wealth.

In a recent video he analyzes a film by «Future Majority», a team of Democrats trying to bring about change in the political landscape in favor of the working class. Dr. Peterson is especially critical of the main part of  the video which describes the surge in inequality during the last decades. In his own words, he considers this to be the weakest part of the film and that the inclusion was an error.

In my view, this is actually the stronger and most important part of the film and what distinguishes the creators politically. The part in question describes how incomes (for the lower and middle class) have stagnated for decades, how CEO compensation has skyrocketed when compared to worker wages and how almost all of the recovery since the Great Recession went to the top-0.1%.

What I especially like about this part is that it mainly analyzes the inequality problem in relative (rather than in absolute) terms and across time. CEO compensation is compared to their own workers wages and working class income to where it was decades ago. The important part is that the data are absolutely real.

The ratio of CEO to worker compensation has actually increased ten-fold in recent decades:

CEO to worker compensation ratio 1965 - 2014

while the bottom-90% has not seen any actual increase in real income since 1980!

US Bottom-90% real average income (base 1950)

Dr. Peterson basically states (and assumes) that income is mostly earned honestly through hard work with only a small minority being the result of exploitation or counter-productive endeavors. So it seems that CEOs earning 300 times average compensation (instead 30 times during 1980) is the result of exceptionally high productivity and work while the bottom-90% is just not being any more productive for almost 40 years now.

What is missing from the analysis is the fact that work/income outcomes are not only the result of personal choices and work but also of class negotiation power. Worker power will depend on the state of the business cycle (and the level of unemployment), union density and power, outsourcing threats, import penetration (which embody foreign labor acting competitively to domestic labor), the level of sectoral competition (which determines to what point a company will actually act as a monopsony) and a host of other factors.

Since 1980 unemployment has mostly been above the «natural level of unemployment» with only the second half of the 90s accounting for an actual high-pressure economy:


union density fell from 27% in 1973 to 19% in 1986 and 11% in 2011 while the real minimum wage (which mostly determines the floor for the bottom-20% income) has never increased above 85% of its 1980 level:


Even increasing educational attainment has not really helped workers since only post-bachelor degree college education has posted steady increases in weekly earnings. Apart from an increase during the second half of the 90’s (when the economy was allowed to run hot) the rest of the college level education earnings have been flat all the while student loans outstanding have climbed to a total of 1.5tr $.

Real Weekly Earnings by Educational Attainment, 1963 = 100, 1963–2012

In my personal view, Dr. Peterson has allowed his own political views to cloud his analysis of income inequality. It seems that he clings to a stereotype where any actual exposition using hard data of the problem is a manifestation of resentment towards the hard-working wealthy instead of the first step of trying to determine and solve the problem. In so doing and especially by insisting that the film should not have included a section on inequality in the first place he appears to make the case that thinking and analyzing inequality has no value and wealth and income are only the result of personal choices, competence and hard work. Given that the Democrats voter base is within the bottom-90% working class it seems quite strange to insist on not really focusing on income inequality at a time when the working class has not been able to see its real income rise above 1980 levels.

I recently wrote a post on the ECB waiver (for Greek government collateral) and what its expiration on 20 August would entail. Since BoG released its August balance sheet it seems that this question has been answered.

BoG August 2018 balance sheet

As is evident there was no significant change in the relevant amounts of either regular refinancing operations or ELA financing during August apart from a very small fall of €350mn. What actually changed was the total amount of collateral posted by Greek banks which fell by €5.33bn. Since regular financing operations total amount remained constant it appears that collateral quality was enhanced given that the relevant haircut was reduced from 30% to 15%.

It is obviously interesting to know how Greek banks managed such a task. Unfortunately, bank balance sheets data from BoG is not available for August so we will probably need to wait a bit more to find out.

One last interesting fact was the large increase in government deposits of close to €14bn which coincided with an equal drop in Target2 liabilities. This was due to the disbursement of the last round of funds towards the Greek government (by the ESM) which was used to increase its cash buffer. As a result, BoG now only «owes» €26.50bn to the Eurosystem while simultaneously holding €64bn in securities (mostly as a result of the QE program).

According to recent Mario Draghi comments, the waiver allowing Greek government securities to be accepted as collateral in regular Eurosystem refinancing operations will expire along with the end of the Greek adjustment program on August 20 2018.

Based on the above I would like to take a look at what such a move will mean for Greek banks access to ECB (and ELA) lending. I will be using data available in monthly Bank of Greece balance sheet statements as well as Greek bank consolidated balance sheets (available from BoG).

Overall, Greek banks have significantly lowered their refinancing needs with a total balance of €9bn in MRO/LTRO and €7.3bn in Other Claims (ELA). Compared to the end of 2017 regular refinancing operations are down €3bn while Other Claims dropped a more impressive €14.3bn amount. If one compares the figures to a couple of years ago, the amounts are much more remarkable. MRO/LTROs are down almost €24bn while Other Claims decreased a staggering €47bn.

This drop was driven both by large decreases in liabilities towards the Eurosystem (Target2 and extra banknotes) as well as the ECB QE program. The first item is down €23bn compared to 2017 and €57bn during the last two years while ‘Securities held for monetary purposes’ increased by €31bn since June 2016.

Unfortunately it seems that Greek banks also lowered Debt Securities of Other Euro countries (EFSF notes?) by a similar amount of €33,8bn during the last two years. As a result, they now hold only €5.8bn in securities of that category while they also carry €10.6bn in Greek government securities on their balance sheet.

Compared to the total of €9bn in regular refinancing operations outstanding, Greek banks do not seem to hold enough non-Greek government securities to post as collateral. Moreover, they hold €186.7bn in credit claims (before provisions). According to BoG NPL statistics, almost 50% of credit claims are non-performing which means that much less than €100bn credit claims can be used as collateral in some form or another (with significant haircuts given current Greek bank loans quality). Actually, BoG states that Greek banks have already posted €54bn in assets as collateral on ELA operations (and another €12.7bn in regular operations) which suggests that not much is left unusable.

BoG Balance Sheet 2018H1

Consequently, it seems quite probable that at least some part of regular refinancing operations will have to be moved to ELA after the program expiration due to limited availability of high-quality collateral. The amount of financing allowed for ELA (as set by the Eurosystem and announced regularly from BoG) will be an early hint on that. Other developments such as the QE program or a return of deposits to the Greek banking system will act at the opposite direction. Unfortunately, the June 2018 BoG balance sheet statement states that less than €1bn in extra banknotes is outstanding which suggests that most of the ‘cash under the mattress’ has already returned and no major positive developments can be further expected on that front.