Euroland: Deflation Alert - Take II
Eric Chaney (from Paris)
Morgan Stanley
May 12, 2003
Back in March, the second leg of the euro rally seemed over, at 1.08 vs. the dollar. In three months’ time, the euro had gone up 10% against the greenback. “Deflation alert” was the title of a piece I wrote at that time, thinking that a combination of weak economies, fiscal tightening, and excessive unit labour costs was a perfect recipe to initiate a deflationist spiral in the euro area (Weekly International Briefing, 3 March 2003). Eventually, higher commodity prices and a global precautionary inventory build-up ahead of the allied intervention in Iraq moved the deflation theme to the sidelines. The third leg of the rally we are currently witnessing brings it back in the spotlight, I believe. This time, the euro is up against all currencies. In economic terms, this makes a huge difference and enhances the deflationist shock a still very weak Euroland economy is experiencing. The apparent indifference of the European Central Bank to currency misalignment can only make things worse, in my view.
In this note, we show that the third rally is starting from a level that was already painful for the euro zone and even deflationist for Germany. We analyze how the European economy would react in the short term if the euro rose to USD 1.20 and stayed there. We explain why even a bold reaction from the ECB by means of interest rate cuts might not be sufficient to prevent deflation to penetrate in the continent. Our conclusion is that, in order to prevent the Stability Pact from being the first casualty of the super-strong euro, not only monetary policy but also market intervention and an acceleration of reforms are necessary.
On a unit labour cost basis, 0.90 would be a neutral €/$ rate My colleagues Stephen Jen and Fatih Yilmaz consider that the fair value of the €/$ rate is 1.06, this number being the median output of a large set of econometric models, from interest-rate-differential-based equations to more PPP-oriented ones. We take a different approach here, based on relative total labour costs corrected from relative productivity levels, in other words on relative unit labour costs. International institutions do not carry out this sort of statistical analysis, because of significant discrepancies in national statistical concepts. This reservation aside, we think that unit labour costs comparisons are nevertheless useful to assess the over- or undervaluation of a currency on a pure competitiveness ground.
Once numbers are fully crunched, we find that, since 1997, the euro-to-be, then the actual one, should have traded at around 90 cents to equalize unit labour costs on both sides of the Atlantic. I am the first to concede that the concept of “neutral rate” is neither a predictor for the actual exchange rate nor even a policy recommendation. As econometric models, which “fair values” are in a more or less, depending on their degree of sophistication, long term averages, suggest, the euro area-to-be managed to live with an exchange rate significantly above 90 cents over the last ten years. An important reason for that is often labelled as “non price competitiveness.” If customers are willing to pay a premium for German cars, Italian design, or French fashion, companies enjoying this kind of brand premium can be profitable with an overvalued exchange rate. However, there must be a limit to the average customer’s appetite for, say, German quality. At 1.17, the IPO rate markets want to revisit, the euro would stand 30% above its neutral rate. There is little doubt in my view that the euro is now overvalued whichever way you look at it.
Regional differences matter, i.e., Germany matters
Although in general high-wage countries are also high productivity ones and vice versa, yet there are still large differences in unit labour costs. On our estimates, only two countries could afford living with a strong euro, i.e., France and Portugal, for which neutral exchange rates are respectively 1.03 and 1.16. On the other hand, the German neutral rate is 0.80, indicating that, at 1.15, the euro is already 45% above neutrality. Since labour mobility is very low in Europe, unemployment is likely to rise further in Germany, despite Berlin’s commendable efforts to raise incentives to work. The vicious circle in which Germany is caught is indeed a vicious circle for Europe as a whole: Germany still accounts for 30% of Euroland’s output and it is by far the main export market for most of its European trade partners.
A super-strong euro would derail the recovery and initiate deflation
Our recovery script is forecasting a rebound of euro area economies, with GDP growth accelerating to 0.5% per quarter in the second half of the year, after a dip in the current quarter. Note that very poor economic indicators, such as the fall of German orders in March, are consistent with a contraction in the second quarter, without invalidating the prospect for a rebound. However, a super-strong euro could well derail the recovery. Our forecast was based on EUR/USD averaging 1.08 for the remainder of the year. Let us assume for a moment that, because markets often over-react, the correct hypothesis is 1.20, other currencies being stable vis-à-vis the UDS. According to econometric models, a 10% rise of the single currency would cut output by 0.7%. If models converge about the multiplier, they give very different views about the time lags. A fresh view on lags was given by a quarterly model developed by INSEE experts for the euro area (see “Note de Conjoncture,” March 2003). According to INSEE’s model, the maximum impact on GDP, 1% of GDP, is reached after only two quarters. Not surprisingly, the main casualty is fixed investment as companies both anticipate a smaller share of the global pie but also thinner profits. On the other hand, consumers benefit from stronger purchasing power but the rise of unemployment nullifies this positive effect.
Well, if the euro rallies to 1.20 and stays there for a while, the loss of output at the end of this year would be of the magnitude of the recovery we have in our GDP forecast. Practically, GDP would be down 0.2% in Q3 and up only 0.4% in Q4, instead of growing by 0.4% and then 0.5%. On annual average growth numbers, the main loss would be for 2004 (1.8% instead of 2.3%) although 2003 results would also be trimmed (0.6% instead of 0.9%).
Things might turn even worse than the models suggest Models assume that economies react linearly to changes in macro inputs. The real world is different, I think, especially for reactions to exchange rate gyrations. I believe for instance that, when a currency is largely undervalued -- this was the case for the euro in 2000 -- an appreciation can be positive for growth, because positive wealth effects would largely overcome negative profitability effects. In the real world, producers know when a currency is far out of reasonable bounds and, in the case of an undervaluation, they know that super-profits cannot last forever. At the other end of the spectrum, if the initial conditions of a currency rise are already over-stretched, then the negative impact on the real economy is likely to be worse than model multipliers suggest, because of self-reinforcing effects on corporate spending, especially if deflation appears. Models suggest that a 10% rise of the euro would cut inflation by 0.5% to 2%, most of the divergences coming from differently specified wage-price loops. Let’s assume that inflation would be cut by only 1.2%, after four quarters. Since our inflation forecast for 2004 is 1.5%, only 0.3% would be left. Whatever the uncertainties regarding the inflation measurement bias, often assumed to be around 0.8%, there is no doubt in my view that, without monetary reaction, Euroland would enter into deflation territory. Note that if inflation drops below 1% next year, it is most likely to print in red ink in Germany. Also, for highly indebted companies, the vicious circle of real debt increased by deflation would start, with straightforward consequences for banks: non-performing loans would rise faster than write-offs. I guess readers already have a feeling of déja vu and may stop here. Will the rise of the euro force reforms or kill the Stability Pact?
On more structural ground, is it possible that a sustained over-valuation of the euro would prove a “blessing in disguise” and act as a catalyst for structural reforms? According to my colleague Robert Feldman, the Japanese experience suggests that the answer might be yes, but also that a very large misalignment is needed to get there. Being “close to death,” to borrow from PM Koizumi’s own words is a necessary condition to gain the support of public opinion for reforms, in Feldman’s view. According to my colleagues Joachim Fels and Elga Bartsch, reforms in Germany would be accelerated by another year of zero or even negative growth and, in that case, other euro area countries will have to follow. Another school of thought, more popular in Washington than in Frankfurt or Brussels, considers that a super-strong euro would force European politicians to embrace the reflationary policies they have been reluctant to implement so far. No doubt in my view this will be debated at the next G-7 summit in Deauville. Practically, the reflationist camp thinks that the Stability Pact should be scrapped, so that the job of kick-starting the global economy by fiscal means would be more evenly shared between the United States and Europe. This camp also thinks that a super-strong euro will force the ECB to team up with the Fed in a kind of global fight against deflation.
I am not convinced by either camp and I am afraid that on these highly political grounds, there are more opinions than evidence. However, I would concede that the Stability Pact could be one of the collateral damages of a super-strong euro. If, instead of a mild recovery, Euroland is to fall into recession, I do not see how the governments of the three largest countries could cut their deficits below 3% of GDP next year. The rule of thumb is simple: a 1% loss in real GDP growth implies a rise of 0.5% of GDP for budget deficits. In the cases of France and Germany, which are likely to start from 3.5% of GDP this year, abiding with the Pact would imply a discretionary tightening amounting to 1% of GDP, which, I think, is not politically feasible. The same would hold for Italy, which, so far, has managed to conceal the deterioration of its public finances, in my opinion, by selling state assets.
Ditching the Stability Pact would not be in the interest of Europe, where big governments have accumulated huge debts and where baby boomers will start to retire in the coming years. Writing off the Pact would be an incentive for governments to endorse “muddling through” strategies and postpone reforms even more, because it would open the door to higher public spending. Halting the rise of the euro has become an urgent necessity, I believe. The European Central Bank missed an opportunity to send a first signal last Thursday. But even a bold action on interest rates might not be sufficient to stabilise currency markets. Interventions, if they are backed by a strong commitment to respect the Stability Pact and meaningful decisions on structural reforms (pensions and/or labour markets), should be considered as well.
morganstanley.com |
