In my last post, I tried to define the “Europroblem” more accurately. This was necessary because the topic is usually discussed in generalized terms that prevent us to see the root cause of the crisis. I argued that a common monetary unit for the Euro zone does not constitute a problem per se. Indeed common monetary units were the norm in much of the past and are still used in much of the world without having the same negative effects as in the Euro zone. The causes for the crisis lie elsewhere. I claimed that what makes the Euro zone situation unique in history is the fact that when the Euro was created, many countries that had “EM-type” banking systems joined a fractional reserve (DM-Mark) currency block virtually overnight thereby triggering a boom-bust cycle.
In this post I will analyse the problems that arise from this process, first by looking at it deductively and then looking at a natural experiment case study for empirical evidence.
The difference between dollarization and joining the Euro
First, let me clarify what I mean by dollarization and why it is fundamentally different from what happened in Greece and the PIIGS. Dollarization occurs when a country – typically and Emerging Market experiencing high inflation – adopts the USD (or the EUR in which case it is euroization) as legal tender within its borders. The fact that the country changes its legal tender laws doesn’t mean that USD (or EUR) have to be procured, since in most cases the currency in question will already have circulated for quite some time in the country and will be readily available. The government merely stops pretending it can continue to issue widely accepted medium of exchange on its own and acknowledges the facts on the ground. This has happened many times in history, more recently in Ecuador and in Montenegro. Currency boards – a slightly watered down version of the same phenomenon – are even more common.
Note that dollarization is unilateral, i.e. it doesn’t require the approval of the central bank that issues the reserve currency, i.e. the Fed or ECB. No Maastricht treaty or any other document of cooperation has to be signed and there is no discussion about fake budget numbers and “shared values”.
Answer: no treaty is required, since there is no privilege conferred by abandoning the local currency. Life goes on as before for the citizens or even improves as the transaction costs of switching between the (undesired) local currency and the de-facto medium of exchange disappear. Not much changes for the local bank manager either: he has to keep a high enough liquidity reserve against FX deposits just as before and credit growth is strongly tied to savings growth. The bank looses FX conversion fees though a substantial source of revenue for EM banks, which is why currency boards are more common, I suppose.
Joining the Euro(system)
When a country joins the Eurosystem, however, the situation is fundamentally different. There are long negotiations, there are criteria and transition phases etc. Whereas the focus is usually on the free movement of goods and people, the biggest change – as we will see – actually happens in the banking system. Why, because the government debt of the country in question will be eligible for tender at the ECB virtually overnight – a huge privilege and a game changer. Access to the reserve currency via a central bank, in turn, incentivize the banker to run down his “uneconomic” liquidity reserve which before, as a proportion of deposits, had to be much higher than for his peers in the Eurosystem. He now can afford to be less cautious courtesy of the new (marginal) lender of last resort (the ECB). Consequently, underwriting standards are lowered, liquidity cost goes down and a boom bust cycle is set in motion.
For illustration, below is a chart with the evolution of the equity ratio in the US banking system. I use this ratio as a proxy for balance sheet expansion, i.e. credit/money creation. As you can see, the equity ratio experienced a sharp fall twice in the last century. First, after the founding of the Federal Reserve in 1913 which enabled bankers to run down their prudent reserves as there was a “lender of last resort” and another time in the 1930ies with the introduction of the FDIC which reduced the probability of bank runs. Both events set in motion huge multi-decade credit bubbles.
My hypothesis: when a country with a non-reserve currency joins a reserve currency block it is akin to the introduction of a central bank with all the consequences. This is because the local central bank’s role is now enhanced: whereas previously its capacity as “lender of last resort” was limited by its FX reserves, it becomes virtually unlimited once it is part of a reserve currency block such as the Eurosystem.
In the next step I want to test my hypothesis with a case study.
A natural experiment: the Euro introduction in Slovenia
Slovenia became a full member of the EU in 2004 and a member of the Eurosystem in 2007. Its immediate neighbour Croatia (the capital cities are 120km apart), however, has become a member only recently and to this day has not joined the Euro. The two countries are as close as you can get for an experiment in the social sciences.
- comparable degree of development
- similar culture, shared history, close economic ties
- both banking systems historically DM/Euro-based
- local currency after independence pegged to Euro
- 80 percent of the deposits Euro based – characteristics of EM-type banking system
- Comparable legal system
So if my theory is correct, we would see a relatively higher degree of credit creation in Slovenia after 2004 when it became clear that the country would join the Euro in 2007. As markets tend to act in anticipation, it is to be assumed that a large part of the credit creation predates the formal Euro entry.
Indeed, this is the main reason why I prefer this case study to looking at actual data for the PIIGS. Most of the PIIGS joined a long time ago and at different dates, with long transition phases as the Euro project was work in progress during the 90ies. Further, I prefer to use bottom-up single bank data, rather than macroeconomic aggregates and it can be quite hard to get reliable data from more than a decade ago.
I will use the loan to deposit (LTD) ratio as a proxy for credit-money creation, since it does not suffer from some shortcomings the equity ratio has over shorter time spans (loss recognition, different equity definitions etc.). The analysis will focus on the two largest banks respectively: Nova Ljubjlanska Banka (NLB) in Slovenia and Zagrebacka Banka (ZABA) in Croatia. The latter is majority owned by Austrian Bank Austria Group. Here is the chart that depicts the evolution of the LTD ratio for each bank,
As one can see, the credit growth for NLB (blue) really took off with Slovenia’s EU membership (shaded area) with its LTD ratio rapidly rising from under 90 percent in at the beginning of 2004 to more than 130 percent until the financial crisis. No doubt, international credit conditions were generally easy, a trend which allowed Croatian ZABA (dotted red line) under foreign ownership to expand its LTD over the same timeframe as well. Now, the interesting thing is if we compare this to Postanska Banka (green) the biggest domestically owned Croatian bank whose LTD ratio rises slightly over the period: cut off from direct access to the Eurosystem or indirect access via foreign ownership, it had to adhere to much more conservative liquidity policies, just as my hypothesis would predict.
Thus, both Slovenia and Croatia have experienced their own version of ABCT. The specifics were different (they always are) but the results are the same. Credit growth in Slovenia was driven by local banks and their direct link to the Eurosystem, whereas in Croatia it was predominantly foreign banks that transmitted loose international credit conditions to the local economy via group loans, probably in anticipation of a future EU membership and adoption of the Euro. You will not be surprised to hear that both countries experienced serious real estate bubbles that have now busted.
It is clear that similar dynamics must have played out all over Europe when the Euro was finally introduced in 2002. Indeed, looking at the following chart one gets the impression that the introduction of the Euro acted like an aphrodisiac on the continent’s bankers,
Of course, the average masks the all-important regional differences,
As one could expect, credit growth was much higher in the PIIGS (here the example of Spain and Ireland) than in the Eurozone overall. This however is entirely explained by the fact that the introduction of the Euro was a game changer for the bank managers in the PIIGS whereas nothing really changed for the Germans, the Dutch, etc. Most discussions focus on differnces in culture and mentality, but this misses the point: there is no need to resort to cultural differences when looking at phenomena caused by abnormal credit growth.
The Europroblem is the result of faulty design, as opposed to irreconcilable cultural differences between member states. This is not to say that cultural differences do not matter – far from it – it is just that they are irrelevant when discussing questions of currency.
The faulty design lies in the fractional reserve nature of our banking system which is known to generate boom-bust cycles on a regular basis, irrespective of whether we are talking about American, German, Greek or Spanish bank mangers or whether banks are privately or state owned. The stark heterogeneity in performance between the core and the periphery is thus not due to some cultural factor, but rooted in the fact that EM-type banking systems existed in the periphery from whom access to the reserve currency via regular tenders was a much bigger shock than for banks operating in “the core” – the starting point matters.
Thus, the discussion of whether Greece should retain the Euro or not is futile and unnecessary and is repeating in a way the mistake made before: ignoring path dependency and the starting point. Consequently, any viable solution has to start with the fact that Greece currently “has the Euro” and this specific banking system.
In the next and last post of this series I will look at how a sensible solution for Greece could look like (and why it is unlikely to happen)….