Over the weekend I came across a December CEPR paper about the Latvian economy. Authors Mark Weisbrot and Rebecca Ray highlight the Latvian experience with internal devaluation, which may prove to be a case study for the current Eurozone model of internal devaluation by the program countries (Ireland, Portugal, and Greece). Weisbrot and Ray find the following (emphasis mine):

The paper also finds that Latvia’s net exports contributed little or nothing to the economic recovery over the past year and a half. This means that “internal devaluation” cannot have succeeded in bringing about the recovery. Rather, it appears that the recovery resulted from the government not adopting the fiscal tightening for 2010 that was prescribed by the IMF, and also from an expansionary monetary policy caused by rising inflation. The data contradict the notion that Latvia’s experience provides an example of successful internal devaluation.

Note: Internal devaluation is Europe’s favored prescription for any country seeking liquidity assistance; it refers to the process by which an EZ country that cannot devalue its currency reduces relative costs (wages) and prices by raising the unemployment rate in order to shift export income in favor of the deflating economy.

Internal devaluation didn’t work for Latvia, as evidenced by export demand. I wondered, though, how has real export demand performed for the current program countries, Ireland, Portugal, and Greece? Have these countries followed Latvia’s path?

To date, as regards to export income, internal devaluation appears to be working in Ireland and Portugal but not in Greece. For comparison to Latvia, I illustrate the path of real exports and imports spanning 2005Q1 through 2011Q3, as demonstrated for Latvia on page 14 of the CEPR paper.

Similar to Latvia’s experience, real import demand deteriorated in the face of fiscal austerity, especially in the case of Ireland and Greece. In contrast to Latvia’s experience, though, real exports in Ireland and Portugal are roughly 6% and 8%, respectively, above levels in 2007Q1. The Greek experience looks more like that of Latvia, as real imports and exports plummeted below pre-crisis levels, having yet to recover.

As I highlighted in a December post, relative unit labor costs have been cut in the case of all program countries, so internal devaluation is evident. (Note: the chart in the post illustrates the 4-q average Y/Y growth in nominal unit labor costs compared to the EA overall – it’s not intended to be a thorough examination of real exchange rates.) But has that been the driving force behind the export growth?

The resurgence in global trade has been an influential factor in the case of Ireland and Portugal. The chart above illustrates the Dutch estimate of world trade. Notably, there was a resurgence of trade spanning mid 2009 to Q1 2011, which support European exports broadly. It’s difficult, in this respect, to attribute all of the export success to internal devaluation.

Going forward, Ireland and Portugal are very likely slaves to global demand for exports. Prospects there are not looking too bright, given the slowdown in Asia.

Ultimately, I do think that Latvia serves as a warning for EA program countries. Internal devaluation is impossible for those countries with high private sector leverage without a burst of external demand. Unfortunately, the burst of external demand seems to have passed.

(Insert here a discussion of the 3-sector financial balances model, and why Ireland depends exclusively on generous export income to facilitate economic growth.)

The chart illustrates portfolio flows as the 3-month sum total flows in and out of Euro area bond, equity, and money markets, or portfolio investment. The red line plots the dynamics of investment flows by foreign residents in (positive) and out of (negative) the Euro area. The blue line plots the dynamics of investment flows by Euro area residents in (positive) and out of (negative) the Euro area.

When the blue line is positive, Euro area residents are bringing assets home, or repatriating capital. This is rare. When the red line is negative, foreign residents are moving capital out of the Euro area by selling euro assets. This is also rare. Both occurred in 2011.

Foreign investors reduced exposure to euro-denominated assets, -€78.9 billion in Q4, while Euro area investors brought assets back home, +€55.8 in Q4. Spanning 2011, foreign inflows into equity, bond, and money markets turned 180 degrees from +€183 billion net accumulation in the 3 months ending in June to -€78.9 billion in Q4.

For 2011 as a whole, foreigners accumulated similar amounts of euro-denominated assets as in 2008 (the last time the foreign flows turned negative), €233 billion versus €266 billion, respectively. In contrast, the repatriation flows did make a big 2011 dent compared to the 2008 comparable year of repatriation, +€60.2 billion of inflows for the year as a whole versus -€4.9 billion in 2008.

We’ll see if 2012 is a year of normalization – EA resident outflows and positive foreign inflows. Certainly recent actions by the ECB have helped to assuage foreign investors, as evidenced by the +€5.3 billion portfolio inflow in December 2011. The January release will be an interesting one!

Happy Presidents’ Day, Rebecca

The goal of any pay cuts would be to help make Greek workers, who are generally less productive than workers elsewhere in Europe, able to compete more effectively inside the euro zone, where countries share a common currency that does not allow devaluations to help even out differences in labor costs.

Huh? See below. The going line seems to be that the Greeks are lazy. They earn minimum government-negotiated wages without actually doing a whole lot because they’re uncompetitive. This is wrong; the data do not support this view.

First, the Greek people aren’t lazy at all. In fact, Greek workers spent more hours working 2010 (in annual hours actually worked per worker) than those in Chile, Hungary, Czech Republic, Poland, Estonia, Turkey, Mexico, Slovak Republic, Italy, the US, New Zealand, Japan, Portugal, Canada, Finland, Iceland, Australia, Ireland, Slovenia, Spain, the UK, Sweden, Luxembourg, Austria, Belgium, Germany, Norway, and the Netherlands – and in that order. (You can download and view the data from the OECD 2011 Employment Outlook.) Marc Chandler also highlighted this fact back in January.

Sure, one could argue that the Greek workers work a lot of hours, but it’s for less output. Furthermore, labor costs have risen substantially relative to other Euro area countries, so the country’s worse off. That’s the uncompetitiveness route. If you care about productivity and relative wage gains, why not look at the drop in Greece’s relative unit labor costs?

The chart below illustrates the average accumulated gain/loss in nominal labor costs (labor costs per hour) across the EA 12 in the run-up to the crisis, 2005-2008, and then since the recession, 2009-Q32011 (Finland data unavailable). By this measure, Greece is certainly doing what the Troika want of it: relative devaluation in nominal labor costs. Since 2009, Greek labor costs have fallen 5.3%.

(Note: the data are constructed as the percentage gain/loss of the average 2008 quarterly labor costs over the average of 2005 labor costs versus the average of Q4 2010 to Q3 2011 labor costs over the average 2009 quarterly labor costs, all working-day adjusted.)

French and Austrian labor costs appreciated 12% and 10.7%, respectively, spanning 2005-2008, and another 5.7% and 4.0%, respectively, since 2009. In Ireland, the 1.8% average reduction in labor costs since 2009 pales in comparison to the 2005-2008 14.7% surge. Greece saw a lower accumulated gain in labor costs spanning 2005-2008 than most countries and cut labor costs since 2009. The ‘wage’ cost anger towards Greece seems to be misdirected.

Now I’m really wondering what is this obsession with the Greek minimum wage? True, the Greek minimum wage did rise 0.8% spanning 2010-2011 (you can see Eurostat data here). However, as a proportion of average monthly earnings, the 2010 minimum wage in Greece is roughly in line with other program countries, Ireland and Portugal, and lower than that in France, Luxembourg, and the Netherlands.

Only in 2011 do Greece’s policies stick out when monthly minimum wage as a proportion of average monthly earnings surged to 50.1%. However, simple calculations demonstrate that for Greece the higher 2011 ratio of minimum wage to average monthly earnings was largely a function of falling average monthly earnings, -18.7%, rather than the rise in the minimum wage, +0.8%.

Perhaps I am not understanding things clearly here – I am sure that you all will correct me if I am not – but what’s this obsession with minimum wages? It looks to me like the fiscal austerity driven recession is indeed resulting in a reduction in Greece’s relative labor costs irrespective of minimum wage policy. Isn’t that the point?

As highlighted by Clive Crook via Bentolila, Dolado, and Jimeno (or the Vox version), the cyclical aspect of Spain’s two-tier labor market – the two tier system consists of a large share of temporary workers ‘outside’ the permanent employment positions – can explain in part the boom in employment during the bubble and its crash during the recession. But that doesn’t explain the experience of the US. In 2005 (the last measured date for the US), temporary workers accounted for just over 4% of employment compared to 33% in Spain; however, like that in Spain, the US unemployment doubled during the crisis. What’s one country’s problem is not necessarily another country’s structural issue.

In contrast to other key players in Europe, the Spanish rate of temporary employment fell 8.4 ppt since 2005, where most of the drop occurred in 2009 and 2010. The crisis eradicated some of the inefficiencies of Spanish temporary employment by consolidating those industries that tended to favor temporary employment during the bubble, construction for one. More worrisome, though, is the trend in temporary employment in other European markets. The rate of temporary employment increased 3.5 ppt, 3 ppt, and 5.7 ppt in Portugal, the Netherlands, and Ireland, respectively, spanning 2005-2010.

The problem with Spain’s two-tier labor market is well known, while that in Ireland and Portugal, for example, is emerging. The 2002 OECD Employment Outlook Chapter 3 outlines some adverse impacts of temporary employment. Temporary employment is associated with higher wage gaps among the temporary and permanent employees, fewer health benefits, negative effects on well-being for individuals and families, and minimal job security. I haven’t read recent research to this point; however,  those countries with outsized surges in temporary employment – Ireland, Portugal, Netherlands, Italy, Greece, and the UK are among the highest in the sample above  – are likely to experience a drop in welfare relative to other countries, all else equal.

Another interesting aspect of the Spanish labor market on a relative basis is that while Spain has a large presence of temporary workers, these workers do have a relative advantage relating to claim of unfair dismissal than do other global workers. The implication could be, that as the share of temporary workers rise in Ireland, Portugal, or the Netherlands, for example, the employment becomes more volatile and reduces expected lifetime income, hence spending or growth. (Data from the OECD.)

I’m worried and just conjecturing here. But, not only have countries not reformed labor issues already in the pipeline – the Spanish two-tier labor issues date back to 1984 – we’re creating new ones. The Great Recession and forced austerity is likely a very big factor in the surge in Irish temporary employment unemployment. This will bring with it unintended consequences that may require further reform (or alternative scenarios like exit from the EMU) at some point in the future.

It’s pretty easy to see through relative labor performance which economy is pursuing procyclical fiscal policy, namely deficits rise when the economy is booming and fall when the economy is contracting: the EA.

Across the Euro area, 43% of total home loans are made on a variable rate basis – this means that mortgage rates are highly elastic to ECB rate setting policy.

Average mortgage rates started rising well before the ECB actually hiked rates. The bottom in mortgage rates was seen in June 2010 and hit a local peak in August 2011 when rate cut expectations started to pass through. But the high correlation between ECB policy and average mortgage rates was to be expected and very harmful to those economies with a rising household desire to save.

It would be one thing if the variable mortgages were concentrated in the core countries; but they’re not. The Periphery economies drive up the average share of variable rate mortgages. In the most extreme case, Portugal, 99% of all home loans are made at a variable interest rate. It doesn’t take a PhD to figure out the speed at which tighter monetary policy will pass through to the real economy when 99% of all loans are made at a variable rate.

Like I said, I still don’t get why the ECB hiked rates.

Source data: ECB for current rate data, and an ECB structural issues report on EA housing for the variable rate shares.

Put more simply: nominal GDP is diverging across program and non-program countries. If this economic duress leads to early exit, I would posit that the balanced growth clause has been breached.

The charts above illustrate the dynamics of nominal GDP (NGDP) across the largest non-program and program countries (I explicitly refer to a program country as falling under an explicit EFSF program). These charts demonstrate that unbalanced growth may already be in the works. In Q3 2011, Ireland, Greece, and Portugal are producing an average 2.2% above their minimum level of NGDP during the crisis (Greece’s last data point was in March 2011, so this number is clearly biased upward). In contrast, the largest non-program countries are producing at 6.1% above their minimum levels of NGDP during the crisis – a 3.9% differential in recovery patterns. Germany alone is producing 110.3%10.3% above its trough during the crisis. I suspect that the program country average will fall below 100 in coming quarters, as the debt deflationary cycle grabs hold. This view of the Euro area is anything but “balanced”.

Balanced, according to Merriam-Webster online, takes several definitions, but essentially it’s some measure of equality in weight on two sides of a vertical axis. Let’s call the vertical axis the Euro area average NGDP recovery. It’s a pretty close call because France is running just below the EA average – but compared to the minimum level of NGDP attained during the recovery through Q3 2011, 56% of the EA has recovered by a % less than the EA average of 6.3%, while 44% have recovered by more. I’m sure that there are many ways to define balanced growth – but in NGDP terms, this looks unbalanced.

Now, the Treaty defines no explicit time frame for ’balanced growth’ – if it’s a long-term objective, lets say 5-10 years, then one could argue that the forced structural reform in Ireland, Portugal, and Greece (even Spain, Italy, and France) will increase long-term potential growth, thereby not breaching the treaty.

But what if the countries are forced to exit before the structural reform starts producing positive growth in average real GDP? Chapter III of the 2004 World Economic Report highlights two important points that should be considered: (1) it’s rare for countries to tackle multiple levels of structural reform at once; and (2) it takes a long time, as in the case of New Zealand, for aggressive structural reform to pass through to the real potential growth rate. The EA is attempting many levels of reform, including financial, labor, product, and tax. This is rare and history shows that this can take up to a decade to show results (as in New Zealand’s case).

I can only deduce that Greece, Ireland, and Portugal probably don’t have enough time and are likely going to be, if they haven’t started already, weighing the pros and cons of exit. If these countries do choose exit, it’d likely be under economic duress; hence, the EU would have failed to target ’balanced growth’, as outlined in Article 3.

I like the way that Megan Greene (@economistmeg) put it in her response to the Irish Times query “Is austerity the best policy?”: “There is a fighting chance that Ireland can eventually grow its way out of it – but I think the time is too short for Ireland to turn it around.”

Perhaps this aggressive proposal by Germany is one of the unintended consequences of the ECB’s three year long term refinancing operation (LTRO). If eurozone banks have as much access to cheap, three-year ECB funding as their collateral allows, perhaps Germany and the troika have decided that eurozone banks can survive a Greek default.

Perhaps naively so, I look at the BIS statistics and see quite clearly that the Germans and the troika would be wrong as regards the Eurozone banks being able to sustain contagion effects of a Greek default and/or exit. Exposure levels remain too high as a share of equity.

Better put: if I look at key core bank exposure, German, French, Dutch, and Belgian, to the Periphery economies as a share of total equity, these banks could go bankrupt if a Greek exit/default spreads to broad Periphery exposure.

Belgian banks especially are in the the red, with total periphery exposure being valued at just over 2X equity. This means that if the Belgian banks were jointly forced to write down all periphery holdings to 49% of what they are holding on their books – according to the IMF 39% of European banks hold their sovereign exposure to maturity, where sovereign exposure is a subset of the BIS data above – their equity would be completely wiped out and technically insolvent. French and German banks are in slightly better shape, but still exposed at holdings of 1.4X and 1.1X equity, respectively.

Banks have improved solvency ratios, as Q3 2010 periphery holdings as a % of equity were higher:
Germany 166%
France 202%
Netherlands 120%
Belgium 291%

Unless the German, French, Dutch, and Belgian governments are planning to inject equity into their banking systems – they haven’t announced such a grand plan to date – banks remain overly exposed to periphery asset valuations. Thus, I can only conclude the following: the Germans and the troika either (1) have a plan to quickly recapitalize the banks across Europe, or (2) could be making a huge mistake.

Source data: BIS, table 9D; ECB for total equity; and ECB for FX conversion.

There’s an evolution in marketable debt that is telling a very interesting story regarding bank funding through December 2011. As each private funding market shuts down, the ECB compensates by relaxing its lending facilities and collateral rules, effectively shoring up bank liquidity.

Look at the chart below: it maps out the dynamics of the components of marketable instruments in the EA, M3-M2, in levels of seasonally adjusted billion €. See Table 1 of the release, or download the data here. Since September 2011, the level of repo lending dropped 21%, or – €107 billion. Not coincidentally, the ECB started to introduce longer-term refinancing operations starting with the 1-yr in LTRO October. Holdings of debt instruments <2 years increased €40 billion, as banks use the securities for collateral under the ECB’s lending operations.

The ECB is offsetting, at least partially, the crunch in private repo funding markets.

This policy behavior is evident throughout 2010. Spanning the period January 2010 to August 2011, money market securities fell -€ 108 billion while private repo lending rose € 179 billion. The ECB offset fully the dropoff in funding from mutual fund shares by flushing private repo markets with liquidity.

The Table below describes the dynamics of funding through marketable assets more succinctly.

It’s pretty clear what the ECB is doing: plugging up the bank funding holes left exposed by private capital markets. What’s next?

Domestic demand is a large contributor to GDP growth in both Italy and France. Therefore, inferring patterns of aggregate consumption from higher frequency leading indicators, such as confidence, is important. Confidence measures lead real retail sales numbers, and real retail sales lead the quarterly real consumption patterns. Annual real retail sales growth has a reasonably high correlation with aggregate consumption (the ‘C’ of Y=C+I+G+NX) in both Italy and France, 69%; so gauging real retail sales from consumer confidence could potentially be useful.

Consumer confidence could be a useful tool for predicting consumption, hence GDP, in France and Italy…

…but it’s not in France. See, with a correlation of just 38%, household confidence is a terrible coincident indicator of real retail sales and adds practically no predictive value for aggregate consumption or GDP forecasting. French consumers could be just miserable and still post relatively healthy retail sales and aggregate consumption numbers.

…and it is in Italy. When Italians are depressed (not confident), they spend less. And boy are Italians depressed. The same series, consumer confidence and annual real retail sales growth, has a very high correlation in Italy, 76%. The implication is, that with confidence running 12.5 points below its 2000-2012 average I do not expect real retail sales to rise above the current 4.9% annual decline in November (CPI-adjusted).

Given the recent downtrend in Italian consumer confidence, the likelihood of a December decline in real retail sales is high. But even if it did stabilize at current levels, Q4 real retail sales are running 2.5% below Q3 sales. Therefore, household consumption is likely to decline in Q4, and quicker than its 0.2% drop in Q3.

So the moral of today’s European Daily Catch is when it comes to confidence indicators, know your consumers. Unhappy Italian consumers make poor spenders, while unhappy French households may very well hit the shops. Domestic demand in Italy is shaping up poorly.