Italy

Research guide to European Monetary Union

Political background

Institutional reforms will remain on the political agenda for some years to come. Although the main parties have agreed a common platform based on the direct election of the head of state, opinions still vary about the distribution of power among the institutions and whether further electoral changes are needed - the 1993 reform created a composite mechanism whereby three quarters of the seats are elected by a first-past-the-post system while the remainder are allocated on a proportional basis.

Implementation of these changes will be arduous, particularly since the constitution requires qualified majorities for any such changes. Since the centre-left ruling coalition holds an absolute majority only once the radical left is included (167 seats out of 315 in the Senate and 329 seats out of 630 in the Chamber of Deputies), the constructive co-operation of opposition parties will be essential.

A parliamentary commission representing MPs of all parties from both the Senate and the Chamber of Deputies has been established with the mandate to outline some reform proposals by the end of June which will then be voted on by parliament. Thus, complex political negotiations will remain a fact of life for Prime Minister Prodi's government or any other.

Monetary policy

Bank of Italy governor Fazio has repeatedly proven his tough anti-inflationary stance over the last year, making clear that he is only willing to ease policy once inflation expectations have safely subsided. This is why in spite of a 4.0% fall in inflation since end-1995, the discount rate has only been cut by 2.25% to 6.75%, and the decline in the repo rate has been managed even more gradually.

The CPI targets, in terms of average inflation were 4% and 3% in 1996 and 1997 respectively (although the government aims at 2.5% this year). Current money market rates still imply a tight monetary policy, in real terms, tighter than in Spain or Sweden. The repo rate will gradually converge towards the discount rate before the next cut in key rates. However, with inflation at below 2.0%, the macroeconomic conditions for further easing are in place. Thus another 50bp-75bp cut may be expected soon. This need not to be the last: Fazio's policy is geared towards lowering structural inflation in Italy and should eventually warrant independent cuts in the future.

The main risk to this scenario comes from wage inflation and indirect taxes which still need to be fully harmonised with the EU. The current wage negotiation framework still implies a virtual CPI-wage indexation with a two-year delay, unless the inflation targets are met. Therefore it is crucial that, as opposed to the last two years, the 1996-7 inflation targets are met, otherwise wages will exhibit downward rigidity in the following round just at the time when Italy should join Emu.

The budget policy

In spite of the corrective measures enacted through 1996 (a total of L48.5 trillion) the state sector deficit far exceeded the original L109 trillion target, coming in at L126 trillion or 6.8% of GDP. However, budget prospects have improved considerably after the government decided to dramatically accelerate convergence to the 3% Maastricht target through a draconian package of measures for 1997 totalling L62.5 trillion. The 1997 budget includes some financial transactions and a so-called "Emu tax" which were submitted to the EU accounting officers. Although most of the measures were approved by EUROSTAT, the EU statistical office, they have no structural impact, so new measures must be found over the coming years in order to keep the primary budget at the level of this year. The L15.55 trillion mini-budget approved on March 27 should virtually ensure the numerical fulfilment of the budget criteria in 1997: our forecast is now 3.1%. However, the cuts mainly come from financial transactions such as the advanced payment of corporate severance tax (L6 trillion) which do not alter the structural imbalances of the Italian public finance position. Although interest rates convergence will contribute to keep the overall deficit under control, it is common knowledge that Italian budget consolidation would achieve much greater credibility if the social security system were further reformed. It seems that the Italian qualification for Emu hinges on two prerequisites: a structural reduction of the "welfare" outlays and the achievement of a permanent non-interest budget surplus in the region of 6% of GDP. This means that the financial law for 1998 should contain structural cuts of at least 1% of GDP if Italy wants to qualify in 1999, a rather ambitious challenge. The cabinet will review its medium-term plan (DPEF) by end-May and submit it to the EU.

There are several reasons why Italy should aim at a medium-term deficit which is lower than the 3% Maastricht target or stability pact ceiling. Some hinge on the poor track record Italy has in meeting its own budget targets. Also, with an outstanding stock of public debt of 125% of GDP and an average duration of less than three years, the debt servicing burden, now more than twice the deficit itself, is outside of the government's control. In light of this the cabinet must include a reasonable cushion against unexpected interest rate developments when formulating the budgetary objectives.

The main argument, however, follows on directly from the debt/GDP ratio criteria in the Maastricht Treaty. According to Article 104c (2), countries with a larger debt may still qualify for Emu provided that the ratio is "sufficiently diminishing and approaching the 60% reference value at a satisfactory pace".

There has been no formal definition to date of what constitutes a satisfactory pace, but it seems obvious that countries with larger debt levels should show a larger reduction in their debt/GDP ratio. This implies running lower deficits than countries with lower debt for some time. Under reasonable mid-cycle growth assumptions we argue that Italy ought to constrain itself to a ceiling of 2% rather than 3%. This would accelerate convergence of the debt/GDP ratio to the 60% threshold, to around 15 years instead of the 20 years required with a 3% annual deficit.

Italy's Emu scenarios and market implications

Can Italy qualify for Emu in 1999? On the one hand Italy should now virtually satisfy four out of the five criteria in 1997. Whether Italy is deemed to satisfy the debt criterion is ultimately a political decision.

Should early membership result from a soft interpretation of the criteria (the debt/GDP ratio has been stabilised since 1995), this is likely to be accompanied by even stricter constraints on Italy's subsequent deficits than the maximum 3% of GDP specified by the stability pact, perhaps a 2% norm, in order to ensure a credible recovery path for the debt/GDP ratio (see above).

This solution would be desirable as it could favour fiscal convergence through a further reduction of the Italian interest rate differential with the core countries (the virtuous circle of the risk premium). However, not only may such a compromise be unavailable, but Italy's membership in 1999 could face political resistance regardless of its economic performance in 1997. In fact, although the exclusion of the Mediterranean countries ("Club Med") as Emu founding members in 1999 would be traumatic for their political leadership, their inclusion could be a political problem in the eyes of Germany which is sceptical about their stability culture. This political trade-off could become crucial for the fate of the whole Emu project since the selection of the founding members is scheduled for April 1998, a few months before the German general elections.

So two main alternative scenarios may be considered: delayed entry for a few countries and a delay to the Emu process.

1) Delayed entry for a few countries

In light of the potential problem a "weak Euro" could present Kohl when trying to sell Emu to the German public, a short, one to two-year delay for the "Club Med" countries could smooth Emu's inception. At the same time it would not remove the pressure for medium term economic policy convergence in these countries. Therefore, the delayed entry option should not destabilise the high yielding markets, and should ensure continued interest rate convergence.

The problem with this scenario is that both Italy and Spain (not to mention Portugal) have made joining Emu with the first wave a matter of national pride. They are hostile to suggestions that their participation could be delayed even if their convergence performance is no worse than some of the core countries. As such Mediterranean countries could reject any overt political discrimination or technical delay and block Emu during the voting process in spring 1998 (Italy, Spain, Portugal and Greece have a total of 28 votes, just above the blocking minority of 26). Thus a "core" Emu may not be politically viable.

One way to overcome the "all-in or delay" dilemma would be an exclusion of one of a few countries so long as the sum of their votes falls short of the blocking minority. In this case Italy (10 votes) is likely to be the first candidate for later membership, after Greece (5 votes) which is already certain to be out in 1999. Since Prodi has tied his fortune to early Emu membership, if the above scenario materialises, the Italian cabinet could be forced to resign causing a temporary sell-off of lira and BTPs.

2) Delay of Emu

If the Club Med/Italy countries continue to converge at a rapid pace and a Club Med delay is unpalatable, then a general delay of Emu could become politically attractive. This would have the added convenience that it would provide the Club Med countries with time to further improve their credibility and so be included in the first round when this finally gets under way. Although this scenario certainly has its appeal there is the question of how much of a delay there would be. There are two main possibilities, each with a significantly different impact on the markets:

a) Unconditional delay

This is unlikely but not impossible. Germany has already stretched its margin for political compromises in sealing the guidelines on the Stability Pact last December. The details of the pact still need to be finalised by June and are likely to bring the discussion on the French proposal for a "Stability Committee" to the fore to which Germany is strongly opposed. So, a disruption of the German-French pro-Emu axis cannot be ruled out. This would be the worst scenario for the high-yielders: BTP and Bono 10-year spreads could blow out to 250bp and 200bp respectively.

b) Stopping the clock

This is what appears to be under discussion at the top political levels in Germany but many legal aspects would have to be examined. Such a decision, if it were taken, is unlikely to be announced before the EU meeting in June in Amsterdam. The market reaction to this scenario would be more contained: perhaps out to 200bp for BTPs and 150bp for Bonos.

Outlook for the Italian lira

The lira's new ERM central parity of L990/Dm was slightly stronger than initially expected, but substantially closer to PPP notions of "fair value" considering that some margin has to be allowed for in anticipation of a persistent positive inflation differential with the core countries in the next few years. Based on most inflation indices, PPP levels are currently within a L900-L950/Dm range (or L1,700-L1,800/Ecu). Since the lira left the ERM in September 1992, investors have consistently required a large risk premium for holding lira assets, reflected partly in an undervaluation of the lira versus its "fair value" level. Following, however, the stabilisation of the debt/GDP ratio in 1995, it is unlikely that the risk premium will rise to such an extent again.

The notion of currency stability required by the Maastricht Treaty is that a country respect the "normal ERM fluctuation margins" without tensions for at least two years prior to the examination for Emu. Aside from the fact that it has not yet been clarified whether the two-year period will be applied in its strictest sense to both Italy and Finland (if so, both countries clearly joined the ERM too late to be eligible for Emu), no formal decision has yet been taken on whether the current 15% bands may be considered abnormal.

It is, however, generally supposed that the currency criterion will be interpreted as a de facto narrow band fluctuation range over the past two years prior to the assessment date. From this it is apparent that the exchange rate once again constitutes a policy objective for the Bank of Italy: no specific target has been announced but the BoI is expected to gear its action to maintaining the lira within notional 2-1/4% bands against the Deutschmark. These stand at L967.7-L1012.3/Dm as compared to the official 15% fluctuation limits of L841.5-L1,138.5/Dm.

The BoI has resumed relatively aggressive foreign exchange intervention and was seen buying Deutschmark/lira around the lower band in January, and selling Deutschmark/lira in February when Emu uncertainty pushed the lira towards L1,000/Dm. The BoI's currency reserves rose a handsome $11.2 billion or 34% last year and at $45.9 billion these provide the BoI with ample ammunition to fend off any temporary crisis of confidence in the lira (even if the experience of 1992 tells us that this strategy is not sustainable in the face of a permanent loss of confidence).

The lira will benefit from good fundamental support in the medium term. In particular, the current account has swung from a structural deficit which peaked in 1992 at L35 trillion (2.3% of GDP) to a substantial surplus in 1996 (L65 trillion, 3.2% of GDP). Although the surplus is expected to narrow following the appreciation in the lira, it should remain above 1.5% of GDP until the end of the decade. In a couple of years' time, the resultant annual accumulation of foreign assets will have virtually reversed Italy's net asset position vis-à-vis the rest of the world, from today's net debtor position (around L68 trillion, including the BoI's gold reserves, 3.6% of GDP) to a net creditor status.

What Emu probability is the market attaching to the lira?

We propose to apply to the exchange rate the same type of analysis which is now frequently applied to fixed income instruments and by this method identify the manner in which the lira is effected by the ebb and flow of Emu sentiment.

The basic concept is that the future value of the lira at a certain date may be expected to depend on whether or not Italy participates in Emu. The uncertainty about the two possible scenarios at any given time must then be reflected in today's value of the lira.

If the lira is perceived to be in Emu by a certain date, today's value should be consistent with a forward exchange rate equal to the (expected) conversion rate, at today's level of interest rates. Alternatively, if the lira was not deemed to be eligible for Emu or Emu did not exist, the lira would be at its non-Emu level, to be defined.

The expected return under both scenarios are then combined in one average scenario weighted according to their associated probability. Assuming that the market is rational, the lira will trade at that level which, for given probabilities, will make the average expected return equal to zero: the break-even level. In formula:

0=P_Emut*(Dm/L_t - Dm/L_Emu-IN)+(1-P_Emut)*(Dm/L _t - Dm/L_Emu-OUTt),

where at any time (_t):

Dm/L_Emu-INt = Dm/L_CONVERSION*[(1 + r_Dmt)/(1 + r_Lt)]^1999-t

The main problems here are the rather strong assumptions one needs to make about both the conversion rates and the non-Emu level for the lira. For the conversion rate we take the ERM central rate.

This is not expected to be the official conversion "rule" (which ought to be announced in spring 1998), which will need to reflect, in some way, actual market valuations, probably some historic average, prior to the irrevocable fixing of the parity on January 1 1999.

However, in the absence of destabilizing shocks, the market may well trade in the proximity of the central parity, especially if, as at present, the central bank intervenes to sustain this level and if there is little likelihood of a last minute, competitive devaluation. For the non-Emu rate one could consider some notion of fundamental or equilibrium value, either a simple PPP concept or a more sophisticated notion such as NATREX (NATural REal equilibrium eXchange rate) which is the favoured approach of Deutsche Morgan Grenfell. The latter, unlike PPP, assumes that the real, or inflation-adjusted exchange rate can vary over time in response to long-term trend developments in an economy's productivity growth and savings rates.

Chart 1 plots the calculated, NATREX level of Dm/L and the actual exchange rate since 1980. It nevertheless needs to be recognised that a non-Emu world need not be a freely-floating or non-ERM world, indeed it is highly unlikely to be.

Nevertheless we used our NATREX estimate of a L1,040/Dm level, on the basis that, if Emu was not there, the lira could well drift at weaker level than the ERM band negotiated with the EU partners: both industrialists and the government often mentioned L1,050/Dm in the past. Chart 2 plots the break-even probabilities on a daily basis since the lira rejoined the ERM.

Given the largely arbitrary exchange rate levels constructed under the alternative scenarios, the absolute probability levels have no great significance.

However, the actual changes to the implied probabilities of Italy's Emu participation in 1999 does gives us an immediate indication of how the market's confidence in Italy's Emu membership was scaled back dramatically in February.

Since during this period the broad economic picture can be considered to have been stable, it is plausible that this deterioration in Emu expectations was the main source of exchange rates movement over this period. Included in the same chart are the probabilities derived from the more familiar exercise based on forward swap rates.

The derived probabilities are not of course the same, but are nevertheless in roughly the same ball park and do demonstrate roughly contemporaneous shifts. One point of interest is that Emu probabilities derived from the FX market are less volatile than those derived from the swaps market. Intuitively this is because the BoI is increasingly smoothing out short term fluctuation in the lira exchange rate but not necessarily in the forward interest rates.

Lira: market structure and developments

The collapse in the lira in early 1995 had a significant effect on FX market liquidity, especially amongst the local banks which scaled back their market profile. The recovery in the lira, and general trend decline in volatility (albeit somewhat reversed since November) has benefited liquidity conditions, and domestic banks continue to recoup their market share.

Domestic trading probably now accounts for half of the total, up from one third last year, with the rest being heavily concentrated in London. The BoI continues to enhance its market awareness through deeper contact with domestic players and as a result its intervention and trading tactics have become increasingly sophisticated.

The latest bout of Emu uncertainties has had a relatively limited effect on Dm/L volatility, and part of the reason for this has been the more pro-active and aggressive manner in which the BoI has participated in the market.

The Bank for International Settlements (BIS) does not produce aggregate figures for lira trading in its triennial survey of the FX market. Average daily turnover in the domestic market alone, however, amounted to some $19 billion in April 1995, or some 82% of the total turnover of the local foreign exchange markets.

The bulk of interbank transactions is naturally against the Deutschmark, but in customer transactions up to 40% can be against the dollar.

The French franc, Ecu and Swiss franc are the next most popular currency pairs, but only in comparatively small volumes. Spot transactions continue to dominate the local market; foreign exchange swaps, however, have doubled in importance since the last BIS survey in 1992 and in April 1995 enjoyed a total daily turnover of $8.5 billion compared to the spot turnover of $14.8 billion (a growth rate of only some 20% over 3 years).

Liquidity in the spot market is deepest during European hours. The market can occasionally suffer a dearth of liquidity in late US or Asian trading, resulting in wider spreads and potentially sharper price movements in response to moderate flows. In Europe, the average spread is 30 ticks for a standard Dm5-Dm10 million lot. Transactions of up to Dm100 million can usually be accommodated without too much difficulty, but for amounts much in excess of Dm 100 million the bid/offer spreads do tend to widen. Electronic brokerage is gaining in popularity and may now handle up to 20% of spot transactions.

Liquidity in the forward and swap markets has suffered something of a setback in recent months, with a commensurate decrease in transparency and a stabilisation in bid-offer spreads which had formerly been contracting. Players wishing to conduct large transactions of say greater than $75 million in tenors of between three to 12 months are now being forced to deal direct with banking counterparties rather than through the brokers. This decline in liquidity is partly a function of a general trend in forward FX markets, but may also have something to do with Emu and all its attendant uncertainties.

The growth of the FX options market in Italy has continued from 1996 into the first quarter of 1997. Corporates have continued to take advantage of the exotic hedging products available to them and Outperformance Forwards, Conditional Forwards and Bonus Forwards are still in vogue. The recent increase in Dm/L implied volatilities from around 4% to a peak of around 7% has prompted increased interest in convergence trades from both corporates and investors looking at stability within spot Dm/L ranges. This increase in volatility as well as the attractive carry on short Dm/L positions has made these trades more attractive. While uncertainty over this currency pair continues, it is expected that these types of trades will continue to be in demand. If convergence were to occur, a decline in implied volatility would be expected and further increase returns on these structures.

Chart 1: Dm/L: Actual versus Natrex estimate

Source: Deutsche Morgan Grenfell

Chart 2: Measure of Italy's Emu 1999 entry probability

Source: Deutsche Morgan Grenfell