Recent Economic Developments and Selected Issues

This paper reviews economic developments in Italy during 1995–96. The performance of the Italian economy in 1996 was characterized by a marked decline in inflation and a strengthening of the external position, accompanied however by lackluster growth. The slowdown in economic activity was more pronounced than in most European countries. Unemployment increased, albeit marginally, for the third consecutive year. The paper shows that during Italy’s economic cycle in 1992–96, key macroeconomic variables did not exhibit their typical cyclical pattern. The paper also briefly reviews developments in the public finances.


This paper reviews economic developments in Italy during 1995–96. The performance of the Italian economy in 1996 was characterized by a marked decline in inflation and a strengthening of the external position, accompanied however by lackluster growth. The slowdown in economic activity was more pronounced than in most European countries. Unemployment increased, albeit marginally, for the third consecutive year. The paper shows that during Italy’s economic cycle in 1992–96, key macroeconomic variables did not exhibit their typical cyclical pattern. The paper also briefly reviews developments in the public finances.

I. Introduction and Overview

1. The performance of the Italian economy in 1996 was characterized by a marked decline in inflation and a strengthening of the external position, accompanied however by lackluster growth. The slowdown in economic activity was more pronounced than in most European countries and unemployment increased, albeit marginally, for the third consecutive year. Regional disparities persisted, with the gap between unemployment rates in the North and the south widening further, and the protracted stagnation in the South putting pressure on the financial Situation of banks in that region. Despite a marked deceleration in export growth, the current account surplus rose further, reflecting a strong decline in imports and an improvement in the terms of trade. The best news was on the inflation front, as the 12-month rate declined consistently during the year to about 2½ percent by year-end. Together with an appreciable strengthening of the exchange rate, this performance laid the ground for a return of the lira to the ERM at end-November 1996. With the fiscal outturn largely exceeding original targets in 1996, the need for fiscal adjustment measures continues to dominate the policy debate, as Italy aims to comply with the Maastricht criteria in time for early EMU participation. This background paper contains four studies that focus on selected policy issues arising from the above recent economic developments.

2. Chapter II shows that during Italy’s latest economic cycle (1992–96), key macroeconomic variables did not exhibit their typical cyclical pattern. It goes on to argue that the atypical behavior of these variables can be explained, at least in part, by a macroeconomic policy stance that differed from those prevailing in previous cycles, and a different institutional environment affecting the wage formation process and labor markets. The chapter also argues that the efforts of recent years to achieve low inflation and redress fiscal imbalances may have increased the degree of synchronization between the Italian cycle and that of core European economies, thus improving the prospects for the Italian economy in an environment of increasingly coordinated policies within the EU.

3. Chapter III deals with fiscal issues. It briefly reviews recent developments in the public finances, the new three-year plan for 1997–98, and its first installment in the form of the 1997 budget and its “special effort for Europe.” The sustainability of this effort will need inter alia to entail action affecting social expenditure in Italy: in this regard, the chapter summarizes the findings and recommendations of the special commission on social spending (the so-called Commissione Onofri), who presented its report at end-February 1997. The report is vocal in pointing out the shortcomings of a system that is fragmented and inequitably targeted, with excess spending on pensions overprotecting retirees (and other support mechanisms benefiting mainly participants in the formal labor market), at the expense of needy and socially excluded segments of the population, particularly the young.

4. Chapter IV examines developments in the external sector. Following a brief review of recent balance of payments developments, it provides an analysis of the sustainability of the lira’s new central rate in the ERM by looking at indicators of competitiveness, market expectations, and model-based estimates of equilibrium exchange rates. The analysis concludes that the chosen central parity of the lira is well within a range of estimated sustainable rates, and—with the support of lasting fiscal consolidation and wage moderation—appears appropriate and not problematic from an economic viewpoint.

5. Chapter V reviews the structure and recent performance of the banking system in Italy. The chapter points to a deterioration in the performance of the banking system in recent years, as financial liberalization and a regionally unbalanced economic cycle brought to the fore some long-standing structural problems. It notes, however, that current problems cannot be characterized as systemic: they have an important cyclical and regional component, and in several cases reflect individual bank mismanagement in the context of public ownership. All in all, Italian banks appear to have weathered the problems arising from the economic cycle and the transition underway since 1990. Nonetheless the system’s transition is incomplete: significant structural problems persist and need to be addressed, especially in light of the increasingly competitive environment facing the banking industry. A reduction in operating costs, particularly labor costs, and privatization, including notably of banks owned by charitable “foundations,” are the primary steps required in this direction.

II. The Atypical 1992–96 Cycle: The Role OF POLICY1

A. Introduction

6. In its characterization of Italian business cycles, the Italian Institute of Conjunctural Studies (ISCO), measures economic cycles from an initial turning point marking the beginning of an upswing to a final end-point signaling the end of the subsequent downturn (Table 1). Since Italy is currently in the third year of a recovery phase, following a recessionary period that began in 1992, the reference economic cycle used in this chapter is measured from 1992 onward and will be referred to as the 1992–96 cycle. The initial downturn began in the second quarter of 1992 and ended in the first quarter of 1994, while the ensuing upswing set in the following quarter and continues today. While formally not recorded as a recession, a marked slowdown in economic activity started since late 1995 and continued throughout 1996 (Chart 1). A casual observation of the features of this most recent cycle seems to suggest that macroeconomic developments have not coincided with typical cyclical patterns. During the recessionary phase, losses in employment were unprecedented and real wages did not display their typical countercyclical pattern. During the recovery, the contribution of private consumption was unusually weak, with consumption of durables in particular failing to exhibit its usually strong procyclical pattern. Net exports—usually a countercyclical variable—were strongly procyclical during the earlier part of the recovery. Unusual cyclical developments in prices were also present.

Table 1.

Italy: Economic Cycles since 1970

article image
Source: ISCO.

Turning points are based on ISCO’s own monthly cyclical indicator.


ITALY: Cyclical Developments Since 1970 1/

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: Bank of Italy.1/ Shaded areas represent recessionary periods.

7. This chapter examines the behavior of a set of macroeconomic variables during the 1992–96 cycle, contrasts it with the typical pattern exhibited by these variables in previous cycles, and sketches out possible explanations for the change in cyclical behavior. In this regard, the chapter shows that a combination of a different policy mix (a gradual tightening of fiscal—and, later, also monetary—policy, and a floating and highly volatile exchange rate) along with the effects of structural policies implemented since the early 1990s (including a fundamental change in the wage formation mechanism and measures aimed at enhancing labor market flexibility) can explain, at least in part, the atypical features of the latest cycle.

8. The chapter is organized as follows. Section B reviews briefly the empirical literature on postwar economic cycles in Italy and identifies typical cyclical patterns; section C examines the behavior of key macroeconomic variables during the most recent cycle; section D looks at the synchronization of the Italian cycle with that of other European countries; and section E concludes with a brief discussion of the policy implications of the chapter’s findings.

B. Main Features of the Cycle

9. Most studies find no striking differences between Italy and other industrial countries regarding the qualitative behavior of macroeconomic variables during cycles or in the “underlying” model governing economic dynamics. In this sense, Italian cycles are considered to be quite conventional and, as in other industrial countries, not to have changed significantly over the last two to three decades.

10. Using data for the period 1970–90, a recent study by Schlitzer2 finds that economic fluctuations in Italy conform qualitatively to the general character of the business cycle in developed economies: (i) output of services and industry are strongly procyclical, as are most demand components; (ii) consumption is less volatile than GDP, but various categories of investment, exports, and imports are all more volatile than output; (iii) due to the strong procyclical nature of imports, the trade balance is countercyclical; (iv) prices are countercyclical;3 (v) employment lags GDP but the correlation is not strong;4 (vi) labor productivity is markedly procyclical; and (vii) real wages are countercyclical, although the link with GDP is rather weak.5

11. Giannini and others (1995) also find that most features of the Italian cycle are similar to those of other industrial countries. Unlike Schlitzer, the authors’ results are based on dynamic simulations produced by a structural VAR, identified through the imposition of standard neo-Keynesian properties. One of the key results of their work is that nominal wages behave as an exogenous variable, and that shocks to prices and wages represent the main source of business cycles in Italy. In particular, they find that a positive price shock, in the absence of accommodating monetary policies, causes a temporary decrease in real GDP, thereby generating a negative correlation between output and prices; nominal wage shocks are found to have an important positive effect on real output in the short run.

12. Work by Backus and Kehoe (1992) on a sample of 10 industrial countries, including Italy, finds that most of them exhibit many of the properties of postwar business cycles in the United States.6 In terms of real variables, the authors find that: (i) consumption has been procyclical and approximately as variable as output, with consumption of durables being more volatile and strongly procyclical; (ii) investment is uniformly procyclical and is typically two to four times as variable as output; (iii) government consumption exhibits no systematic cyclical tendency, and has been more variable than output; and (iv) net exports have generally been countercyclical. The authors also find that patterns of price fluctuations have changed considerably: before WWII prices were predominantly procyclical; since then, they have been consistently countercyclical and have exhibited higher persistence in almost all countries.7 Christodoulakis and others (1995) confirm Backus and Kehoe’s findings and extend them to all EU countries, in a study based on data for the period 1960–90. Their results highlight the fact that variables that do not behave similarly across countries are those under the control of the government—fiscal and monetary, and to a lesser degree labor market variables.

13. Given that the 1992–96 Italian cycle took place under a floating exchange rate, it is also pertinent to discuss the role of the exchange rate regime in the economic cycle. Baxter (1991) and Baxter and Stockman (1989) find that a floating exchange rate regime leads to higher real exchange rate variability, but that this increased variability does not seem to alter the typical cyclical pattern of most variables. In addition, floating exchange rates are found to reduce the synchronization of international business cycles.

C. The Most Recent Cycle (1992-present)

14. The latest recession began in 1992, intensifying toward the end of the year amidst substantial instability in financial markets, and leading to an output decline in 1993 comparable to that in Germany and France, and among the sharpest in large industrial countries. The 1994–95 recovery was characterized by relatively strong growth (Chart 2), but differed markedly from previous upswings: it was export-driven to an unusual extent and was characterized by an atypically slow response of domestic demand (Chart 3). The recovery took place in the context of two temporary shocks which deeply affected its pattern: (i) a substantial real depreciation of the lira, especially in the spring of 1995, that began to revert itself as from mid-1995, and (ii) a temporary tax incentive scheme aimed at promoting investment, enacted in mid-1994 and scheduled to expire at end-1995.8 In addition, the overall macro-environment had changed in various respects: earnings were growing very moderately as a result of the new framework for wage determination resulting from the accords of 1992 and 1993, and fiscal—and later monetary—policy was tightened. Toward the end of 1995, economic activity slowed down significantly, reflecting the fading out of both temporary shocks. A further tightening of monetary policy and the increasing prospects, later confirmed, of a stepped-up fiscal adjustment in the run up to EMU also depressed domestic demand in 1996.


ITALY: International Comparisons

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: IMF, World Economic Outlook.

Italy: Domestic Demand and Net Exports

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Sources: ISTAT.

Output and demand

15. The behavior of private consumption during the 1992–96 cycle was unusual: its fall during the recessionary phase was of an unprecedented magnitude, while the pickup in the post-1993 recovery period was atypically moderate (Chart 4). Moreover, during the upswing consumption of nondurables did not display its traditional procyclical pattern. This behavior reveals an uncommon asymmetry in the relative variability of consumption (from recession to recovery), possibly related to important changes in fiscal policy and exchange rate developments, as well as institutional factors. The sluggish behavior of private consumption during the 1994–95 upswing was largely the result of moderate growth in earnings in the context of the 1993 labor agreement, only modest gains in employment, and relatively large swings in consumer confidence associated with political instability and its related uncertainty. While the renegotiation of wage contracts has since late 1995 led to an appreciable acceleration in wage growth, a combination of other factors has continued to prevent a recovery in household spending, especially on durables. These factors include: (i) the uncertainty of the fiscal outlook and the anticipation that renewed efforts will be needed to achieve the fiscal targets, affecting household confidence; (ii) continued pessimism more generally over future employment prospects; and (iii) high real interest rates, as a result of a tight monetary policy stance and the sluggish response of bank rates in the context of rapidly falling inflation.


ITALY: Cyclical Developments in Consumption and Investment 1/

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: ISTAT.1/ Shade area represent recessionary period.

16. The behavior of investment was, in contrast, more consistent with past cyclical patterns. It was driven by a surge in purchases of machinery and equipment which, although very strong, was not more intense than those observed in previous upswings in the 1970s and early 1980s. However, this investment surge was strongly influenced by the enactment of the above-noted incentive scheme. The Tremonti Law led to a sizable temporary reduction in the cost of capital at a time of high and rising profits—a key determinant of investment in Italy—and a favorable economic outlook both domestically and abroad. As a result, it exerted a one-off, procyclical effect on the timing of investment decisions, strengthening the upturn in purchases of capital equipment that began in early 1994 and sustaining it through the end of 1995, with the anticipation of future investment plans. The extension of the Tremonti Law to April 1996 did not prevent a drop in investment in early 1996 and only slight growth thereafter, especially in the face of a marked decline in capacity utilization, rising stocks, and an increasingly worsening economic outlook.

17. Exports, which had already shown signs of revival during 1992 reached impressive rates of growth following the exit of the lira from the ERM. Starting with the third quarter of 1993, year-on-year growth of real exports of goods and nonfactor services remained near or above 10 percent for almost two years. In the third quarter of 1995, the volume of exports recorded a small decline for the first time in 13 quarters, followed by a strong contraction (over 10 percent in annual terms) in the last quarter. Real exports are estimated to have increased only marginally in 1996. The marked slowdown in export growth as from mid-1995 exceeded the decline in partners’ imports and led, in 1996, to the first loss in export market shares in four years.9 The slowdown also coincided with a relatively strong appreciation of the lira, suggesting that the evolution of the exchange rate may have played a strong role in shaping the dynamics of export growth during the most recent cycle.

18. The relationship between import growth and the real exchange rate appears to have been weaker than in the case of exports, and the procyclical nature of imports was manifest throughout the cycle. Thus, imports declined in 1993, in the midst of a fall in domestic demand and a depreciating exchange rate, but recorded strong growth during the 1994–95 upswing, despite a further real exchange rate depreciation. Moreover, import growth decelerated markedly in 1996, driven by very weak domestic demand and despite a pronounced strengthening of the lira.

19. As a result of the procyclical nature of imports, and the atypical evolution of exports, net exports exhibited an asymmetric cyclical behavior. They displayed their usual countercyclical nature during the recession, but appeared procyclical during the recovery (Chart 5). Moreover, net exports continued to increase during the slowdown of 1996. As a result, the current account balance improved continuously during the entire cycle, with the swing since 1992 reaching over 6 percentage points of GDP. About 5 percentage points are accounted for by an improvement in the trade balance, with the remainder largely reflecting a reduction in net service payments abroad as Italy’s net foreign asset position improved.


ITALY: Exports, Imports, the Trade Balance and the Current Account

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Sources: Bank of Italy; and ISTAT.1/ Shade area represents recessionary period.

Employment, wages, and prices

20. After falling substantially since mid-1992, employment began to pick up again in 1995, lagging as usual the recovery in economic activity.10 Given the depth of the downturn, the drop in employment was unprecedented (Chart 6). A study by Christofides (1996) interprets such increased sensitivity of employment to output as possibly resulting from the enactment in the early 1990s of various measures aimed at increasing labor market flexibility11 and to a policy change, whereby the government ceased acting as “employer of last resort.” In this light, econometric simulations carried out by the author predict only a small reduction in employment until mid-1994 and a recovery thereafter. In the event, the absence of proportional gains in employment following the pickup in economic activity suggests an asymmetric behavior of employment during the 1992–96 cycle.

21. In a novel development, the number of service sector jobs declined together with those of the industrial sector during the recessionary phase of the cycle. For years, employment in the service sector had been growing steadily, showing almost no correlation with the cycle; however, in the 1992–94 recession, employment in the service sector fell. During the recovery, employment grew in both the industrial and service setors but since late 1995 gains in employment have been confined to the service sector, reflecting mainly the relatively good performance of tourism-related activities


ITALY: Employment

(In millions)

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Sources: Bank of Italy; and ISTAT.1/ Shaded areas represent recessionary periods.

22. The decline in employment during the 1992–94 recession is remarkable also in that it was accompanied by falling real wages, in contrast with the typically countercyclical behavior of wages described in Section B (Chart 7). In 1993, for the first time, nominal wages had been negotiated for the following two years on the basis of a targeted (low) inflation rate. Thus, with nominal wages largely predetermined, neither movements in the lira nor increases in indirect taxes fed into wage formation. Moreover, as actual inflation exceeded targeted inflation, real wages declined. The pace of contractual increases picked up in late 1995 and 1996 as wages were renegotiated, with new contracts incorporating some catchup for past inflation. Thus, largely as a result of structural changes, the behavior of wages in the recent cycle was atypical. It should nevertheless be stressed that, as in the past, movements in real wages still appear to be largely determined by institutional factors.


ITALY: Real Wages

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: ISTAT.l/ Shaded areas represent recessionary periods.

23. The behavior of inflation in recent years also appears to have differed from its typical countercyclical pattern. In the latter years of the expansionary boom ending in 1992, consumer price inflation accelerated from about 4 percent in early 1987 to over 6 percent in mid-1991. Inflation began to decline since then, and with the onset of the recession it fell below 4 percent in late 1994. During the early part of the recovery, inflation again appeared to behave procyclically, only to fall steadily since then, and with most of the decline taking place during 1996: the year-on-year rate fell from 5.6 percent at end 1995 to 2.6 percent a year later.12 It should be pointed out that, although the effects of the depreciation of the lira on inflation were contained they may help explain the apparently procyclical behavior of inflation: the increase in inflation in 1995, as well as the decline in 1996, were undoubtedly affected by exchange rate movements.13

D. International Comparisons of Recent Cyclical Developments

24. While macroeconomic developments in Italy during the latest cycle do not correspond to the country’s standard cyclical pattern, many of these atypical cyclical developments can also be found in other European countries. Since policies of fiscal restraint and anti-inflationary monetary rigor were being pursued in other EU countries as well, and large devaluations were experienced by some of them, the presence of similar atypical cyclical developments in other EU countries is consistent with the notion that changes in policies exerted an important influence on the dynamics of key economic variables during the most recent cycle, especially in the recovery phase.

25. In its review of economic developments in 1994–95, the BIS Annual Report identifies several features shared by European countries: relatively weak growth in private consumption; buoyant investment in machinery and equipment, a particularly striking development given the slow pace of recovery; a sharp rundown of inventories during 1995, with negative consequences for growth; and an unusually strong, albeit declining, contribution to growth of net exports in countries that experienced large depreciations in 1992–93.

26. The report indicates that the sluggish behavior of consumption reflects a combination of factors: (i) stagnant or declining real disposable income as a result of rising unemployment and the contractionary effect of tightening fiscal policies; (ii) a reluctance by households to reduce their saving rates, given the low levels already reached during the recession and prospects of further fiscal tightening; and (iii) low consumer confidence given bleak prospects for employment creation. All of these factors appear to have been critical in Italy, where the effects of the recession on employment were particularly pronounced, and where the intended fiscal correction is the largest among EU countries.14

27. The remaining common features identified in the BIS report also constituted salient aspects of the Italian recovery. As explained earlier, Italy experienced a strong (but not unprecedented) cycle of purchases of machinery and equipment, induced, at least in part, by a procyclical investment incentive law. As for inventories, survey-based data suggest that, throughout 1995 and the first quarter of 1996, entrepreneurs may have revised downward the level of inventories that they would consider appropriate. This may have been prompted by increasing signs of weakness in domestic demand, especially consumption (private and public), and may have led to a sharp contraction in stocks. Not surprisingly, the contribution of inventories to GDP growth declined markedly in 1995 and became negative in 1996.

28. The effect of the exchange rate on net exports in countries that experienced devaluations in 1992–93 is shown in Chart 8. The top panel compares the evolution of the ULC-based real exchange rate in four countries: two countries whose currencies appreciated or remained broadly stable (Germany and France) and two that experienced a large depreciation (Italy and Sweden).15 The remaining four panels show the domestic and foreign contributions to growth. The differences are quite clear: while Germany and France saw minimal or no contribution from the external sector, Italy and Sweden recorded sharp positive contributions. Moreover, Italy and Sweden experienced slower domestic demand growth, particularly in consumer spending. These developments suggest that a combination of fiscal adjustment and a depreciated exchange rate may be responsible for an atypical behavior of demand components in the most recent cyclical upswing.


Selected Countries Cyclical Developments

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Sources: WEO.

E. Final Remarks

29. The preceding sections have shown that (i) many features of the 1992–96 cycle were unusual; and (ii) that other European countries where similar policies were being implemented also experienced some of these atypical cyclical developments. This chapter suggests that changes in exchange rate, monetary and fiscal policy may explain at least part of these developments. The large depreciation of the lira may have led to a relative price effect strong enough to offset the standard income effect behind the countercyclical pattern of the trade balance. It may also have led to an apparently procyclical behavior of inflation in the first part of the recovery phase. Increasingly tight fiscal policy—and to some extent tighter monetary policy—and the prospects of further tightening in the run up to EMU, also seem to have affected the qualitative cyclical behavior of aggregate demand components. Structural policies, especially those affecting the labor market, also appear to have led to a change in the behavior of employment and wages.

30. The findings reported above have important policy implications. The study by Formby and others (1992) finds Italy to be an outlier in terms of cyclical developments at the EU level, with the lack of synchronization explained by the absence of coordination of policies between Italy and other EC countries. However, in a later study, Christodoulakis and others (1995) find Italy’s cycle to be positively and significantly correlated with the average EU cycle, with deviations in economic policies explaining the lack of a higher correlation. It is interesting to point out that the Formby study was based on data running through 1986, while Christodoulakis’s was based on data running through 1990. It is likely that developments in the late 1980s led to an increase in statistical measures of synchronization, since it was precisely during the 1986–90 period that Italy strengthened its commitment to the exchange rate, maintained a relatively tight monetary policy stance, and achieved an improvement, albeit insufficient, in the public accounts (Visco 1995). Thus, to the extent that the move from an absence of synchronization to some degree of it can be attributed to a change in macroeconomic policies during the late 1980s, it would seem that further tightening of fiscal policy after the exit of the lira from the ERM in 1992, and, more recently, the renewed commitment to an exchange rate target within the ERM, and a planned substantial improvement in the fiscal position in the run up to EMU, could lead to a considerable degree of synchronization of the Italian cycle with that of other EU members, thereby preparing the Italian economy for a smooth transition to monetary union.

31. Further refinements to the analysis in Section C of this paper is warranted. Corroboration of the statistical significance of atypical features of the most recent cycle in Italy, and their relation with macroeconomic and structural policies implemented in Italy in recent years, would be needed to verify further the extent to which Italy has indeed made progress in getting itself ready for economic life in an environment of coordinated economic policies.

III. Public Finances16

A. The 1996 Budget and Outcome

32. A combination of weaker-than-anticipated growth, the ineffectiveness of a number of budgetary measures, inadequate expenditure control, and some extraordinary unanticipated expenditures contributed to a stalling of the fiscal adjustment process in 1996. The fiscal deficit outcome exceeded the revised and relaxed targets formulated in the course of the year. In late September 1996, the authorities decided to strengthen fiscal adjustment, and stepped up their fiscal plans to aim for compliance with the Maastricht criteria a year ahead of original plans, with the intention of joining EMU in the first wave.

33. Based on a projected GDP growth rate of 3 percent, the 1996 budget targeted a reduction in the state sector deficit to Lit 109.4 trillion (5.8 percent of GDP), from a deficit of 7.4 percent of GDP in 1995. The process of primary adjustment was set to continue, with the balance net of interest outlays set to reach a surplus of 4.3 percent of GDP, 0.7 percentage points of GDP above its 1995 level. Interest rates were projected to decline through the year, by more than 200 basis points, thus alleviating the debt service burden by almost 1 percentage point of GDP. To achieve these results, the budget approved in December 1995 relied on both revenue-raising and cost-cutting measures. The former were primarily geared to increasing excise duties and enlarging the direct tax base, while spending cuts focused on budget transfers to firms and to other public entities, primarily social protection agencies, as the effects of a number of savings measures in the public health care system and the newly-enacted pension reform were expected to reduce their financing requirements during the year.

34. In the early months of 1996, however, it became clear that the achievement of fiscal targets was at risk. The effects of an insufficient yield of a set of measures included in the 1996 budget, an unbudgeted capital transfer to Banco di Napoli, and a less favorable interest rate path than originally envisaged, were compounded by revenue shortfalls as the effects of a weakening economy began to make themselves felt. In the April quarterly budget report, projected GDP growth was revised downward to 2.4 percent, and it was estimated that, in the absence of corrective measures, the state sector deficit would reach 6.3 percent of GDP, with the primary surplus not exceeding 4 percent of GDP.

35. Toward the end of the spring, as growth weakened further and the inadequacy of certain budget measures became increasingly manifest, revised estimates of trend developments set the deficit at almost Lit 130 trillion (close to 7 percent of GDP), prompting the new government to submit a supplementary fiscal package, officially valued at Lit 16 trillion (0.9 percent of GDP). The additional budget, approved in June 1996, comprised a cancellation of funds allocated for new expenditure programs to be authorized during the rest of the year, changes to the payment schedule of a number of tax obligations, and provisions to curb tax avoidance. GDP growth was revised further downward to 1.2 percent, and on this basis, the additional measures were considered sufficient to achieve a revised overall deficit target of 6.1 percent of GDP. In the October Forecasting and Planning report, the authorities accepted a further relaxation of the overall deficit target to Lit 123 trillion (6.6 percent of GDP), in view of a new downward revision of growth, to 0.8 percent.

36. The final, albeit still preliminary, outcome of Lit 139.8 trillion (7½ percent of GDP) for the central government borrowing requirement marks a considerable overrun against the above estimates.17 Not only was growth significantly weaker than originally anticipated, but its composition also adversely affected the structure of transactions subject to indirect taxation, with a sizeable fall in the share of durable goods, subject to a relatively higher VAT rate. Overly optimistic projections of the yield of a number of measures in the 1996 budget (notably those aimed at enhancing tax compliance) and in the mid-year corrective package account for another significant part of the shortfall. A sharp rebound in spending by various decentralized authorities, as well as an unanticipated withdrawal of funds from the Treasury by the EU, also played a role in widening the deficit.

37. When measured on the basis of new accounting practices recently endorsed by Eurostat,18 the state sector deficit (Lit 132 trillion, or 7 percent of GDP) is lower. Preliminary data for the general government elaborated according to the same methodology put the overall deficit at Lit 127.5 trillion, or 6.8 percent of GDP, marking a slight improvement on the comparable figure recorded in the previous year (7 percent of GDP). The primary surplus, however, deteriorated by ½ percentage point of GDP, to an estimated 3½ percent of GDP in 1996.

B. The 1997–99 Three Year Plan and the 1997 Budget

38. In June 1996, the government presented a new three-year plan, covering the period 1997–99. Despite a lower growth outlook, it maintained the previous plan’s targets: a state sector deficit of 4½ percent of GDP in 1997 (5½ percent of GDP at general government level), and achievement of the Maastricht criteria in 1998.19 Measured against the official trend projections (a no-measures tendenziale), the plan was front-loaded, envisaging a corrective package for Lit 32.4 trillion (1.7 percent of GDP) in 1996.

39. The three-year plan indicated the possibility of accelerating adjustment in the fall of 1996, if growth and interest rate developments were deemed to allow it. This was indeed the decision taken in finalizing the 1997 budget, and the package submitted to Parliament at end-September 1996 was divided into two parts. The first part included the “core” adjustment whose broad outline had already obtained parliamentary approval with the June plan, slightly augmented to compensate for the adverse effects of the projected 1996 overshoot. The second part set out an additional effort (in an amount of Lit 25 trillion), officially referred to as the “intervention for Europe,” and specifically designed to achieve the Maastricht reference value in 1997, a year earlier than previously planned.

40. Although parliamentary discussion has not modified the broad structure of the 1997 package, the amounts of the “core adjustment” and of the “intervention for Europe” have since October both changed slightly.

The “core” adjustment

41. More than half of the “core” adjustment of Lit 34.8 trillion for the general government accounts is made of cost-saving measures (Table 2), with cutbacks in transfers to public utilities accounting for a large part of the expenditure cuts. In particular, a more restrictive framework contract for the provision of public transportation services by the railways, together with a new condition regulating provision of services by the Postal Administration, is expected to yield some Lit 5 trillion (27 percent of total general government expenditure cuts). Other measures are centered on equalizing the social security contribution regime among different categories of public employees, controlling the demand for health care by users, extending the limits on the filling of vacancies introduced in 1996, and promoting part-time public employment.

Table 2.

Italy: General Government: Breakdown of 1997 Fiscal Package

article image
Source: Data provided by the authorities.

The “intervention for Europe”

42. The “intervention for Europe” (put at Lit 25 trillion) was originally to be split equally between a still-undefined “Tax for Europe” and “selective deficit-reducing measures” (in practice, methodological accounting revisions). In the event, the amount to be collected from revenue measures was reduced to Lit 11.5 trillion (0.6 percent of GDP), and total accounting revisions in 1997 are estimated to amount to Lit 14.8 trillion (0.8 percent of GDP).

43. On the revenue-raising side of the “special intervention for Europe,” Lit 5.5 trillion is to be levied as a temporary personal income tax surcharge (the so-called “Euro-tax”);20 Lit 3.5 trillion by the advance taxation of funds set aside by firms to cover workers’ severance payments (trattamento di fine rapporto);21 and, finally, Lit 2.5 trillion by speeding-up collection of already assessed direct taxes.

44. The remainder of the “special intervention for Europe,” in the form of “selective deficit-reducing measures,” concerns accounting revisions, for a total impact in 1997 of Lit 14.8 trillion (0.8 percent of GDP). These revisions may be summarized as follows:

Accounting of interest on postal savings certificates (0.3 percent of GDP). Accounting of capitalized interest on postal savings certificates is to be changed from an accrual to a cash basis, with payments to be recorded when they fall due, in a manner analogous to that used for zero-coupon bonds. This change was approved by Eurostat in early February 1997.

Redefinition of loans to railway company and other public entities (0.2 percent of GDP). The government has recognized as debt of the State all the loans incurred in the past by the railway company and by a number of other minor public entities. As a consequence, capital amortization, previously included as a “transfer to enterprises,” is to be recorded below the line as a reimbursement of a public liability. At the same time, the interest paid on the loans is reclassified from “transfers to enterprises” to “interest payments.” Eurostat approved this change on February 21, 1997.

Shift of effect of Constitutional Court rulings (0.2 percent of GDP). Under this item, back payments (through bond issuance) of the liabilities arising from two Constitutional Court rulings on pensions are to be imputed to the years 1993–95. As a consequence, the general government deficits recorded in those years will be increased by Lit 7.4 trillion, Lit 7.6 trillion, and Lit 2.5 trillion, respectively, while the payments planned effectively to take place in the years from 1997 to 2000 will not contribute to the deficit in those years. There is, however, a contrast with the treatment of such back payments in the stock of debt. While, as noted, the payments contribute to the deficit in the years to which they are imputed (1993–95), they will be accounted for in the stock of debt in the years in which the bonds to cover the back payments will be actually issued (from 1997 onward).

C. The Welfare System in Italy: Recent Reform Proposals

45. There has, in recent years, been a growing consensus that public expenditure on social protection in Italy suffers from severe shortcomings. The system is recognized to be inadequately targeted, fragmented, and excessively complex. It is, in particular, inequitably skewed toward the elderly and “insiders”: excess expenditure on pensions overprotects retirees, and other support mechanisms are largely limited to present or past participants in the regular labor force. The system thus fails in its primary task of providing an effective safety net for the truly needy, in particular for those who never joined the labor market, notably among the young. The system’s failings are evidenced by trends in poverty and inequality indices—hence the characterization of the Italian social protection system as “the poverty of welfare.”

46. The awareness of these shortcomings (evidenced also by the work of a commission on poverty and social exclusion)22 led the new government to appoint, in early 1997, a special commission on social expenditure. The commission, known as the Commissione Onofri, presented its final report on February 28, 1997.23 This section briefly reviews the report’s main findings and recommendations, which—while still general in their formulation—would, if implemented, constitute an ambitious overhaul of the system.

A European comparison

47. In comparing the Italian system to that in the rest of the EU, the Commission’s report finds it to be marked by “some glaring anomalies.” While the level of social expenditure, at about one-quarter of GDP, is not out of line with the EU average, its internal composition is clearly atypical. The report notes that the comparison with partner countries highlights two types of distortions: first, as regards the risks that are covered and, second, the categories that are protected. With respect to the coverage of risks, protection for “old age and survivors” absorbs 61½ percent of total social spending in Italy, versus an EU average of slightly over 45 percent.24 Expenditure on other forms of assistance (notably unemployment, family, and housing benefits) is markedly lower than in partner countries (about 18½ percent of the total, versus almost 32 percent); public spending on active labor market policies is also comparatively low. Health expenditure, for its part, is not out of line with that elsewhere (amounting to about one-sixth of total social spending, or some 5 percent of GDP, and, as in other countries, on a declining trend in recent years).

48. As regards the distortion affecting the categories that are protected, the Commission found a marked divergence from other EU countries between the benefits accruing to (present or past) participants in the formal labor market, and the treatment of other workers or the nonemployed. Protection is strongly tilted toward the first group, and within it to those employed in large enterprises or in the public sector: payments from the Wage Supplementation Fund, and the so-called “mobility allowance” in case of collective dismissals, both of which exclude small manufacturing firms and most service activities, are significantly more generous than ordinary unemployment benefits (with the “mobility allowance” being more than double the ordinary benefit). Furthermore, access to these schemes is often dependent upon administrative discretion. There has in addition been a tendency to introduce emergency measures targeted to specific situations, thus leading to a host of overlapping ad hoc programs benefitting single groups rather than tailored to protect truly needy individuals. The overprotection of retirees who previously had formal jobs is also evident in a significantly higher ratio between average old age retirement benefits and the social pension than that observed in other partner countries. Finally, the Commission noted that Italy stands out in comparison to other European countries in lacking any form of minimum income support for the needy, as well as an adequate network of support services for families.25

Reform proposals

49. Against the above background, the Commission identified a need to reform the system of social protection in Italy with a view to shifting expenditure away from the pension system and toward: (a) targeted forms of support to cover income and employment risk, designed to promote greater labor mobility and provide an equitable and effective safety net (by, inter alia, reducing benefits for “traditional” categories of workers so as to ensure, or indeed introduce, adequate protection for the socially excluded), and (b) more extensive active labor market policies. In designing the reform, the Commission stressed the importance of avoiding forms of moral hazard and work disincentives that could promote welfare dependency. The Commission was also particularly attentive to the macroeconomic constraints on expenditure, and formulated its suggestions within the twin constraints posed by the government’s intention to participate in EMU from the start, and by the recognition that the level of taxation and social security contributions is already excessively high.26

50. As regards the specifics of the reform proposals, the Commission’s recommendations are briefly summarized below.

Pensions. Recommendations in this area were sympathetic to those (including the staff) who see a need for a faster phasing in of the 1995 reform.27 In particular, the Commission called for a rapid harmonization (and eventual unification) of the various regimes, now characterized by a privileged treatment, among others, for the self-employed and civil servants;28 an increase in the contribution rate of the self-employed; a more rapid increase in the minimum retirement age to European levels; the pro-rated extension of the new, contributions-based formula for the calculation of benefits also to those with at least 18 years of contributions (who were grand fathered by the 1995 reform); a strengthening of rules prohibiting the cumulation of benefits; and the promotion of private pension funds. The Commission also advocated more timely corrections, based on automatic criteria, of the system’s demographic-related parameters, notably the transformation coefficients applied to capitalized lifetime contributions to calculate a retiree’s first-year pension.

Health system. The Commission recognized that significant progress has been made in recent years to contain health expenditures in Italy, to the point that per capita health expenditure is currently among the lowest in Europe. In these circumstances, it saw a twofold need: to maintain control over the evolution of such expenditure, and to restructure the system away from one based prevalently on external controls to one assigning a greater role to appropriate incentives, personal responsibility, and accountability. Among specific measures, it called for an extension of copayments (on hospital stays, emergency treatment, and home visits), a review of current widespread exemptions from such payments, and the possibility of selling over-the-counter drugs in general stores. On the administrative side, it pressed home the importance of ensuring that Regions (responsible for health expenditure) be subject to a hard budget constraint.

Income support for the unemployed and other workers at risk. The Commission, characterizing the current unemployment protection provisions as “a disorganized and virtually ungovernable system of successively overlapping instruments,” recommended its simplification into a two-tiered apparatus. First, an insurance-based scheme of short-term income support to tackle cases of redundancies due to temporary reductions or suspensions of activity, financed through a new version of the current Ordinary Wage Supplementation Fund, so as to avoid termination of the work contract and the ensuing loss of firm-specific human capital in situations viewed as temporary.29 Second, in order to enhance both equity and efficiency, it recommended the introduction of a uniform two-year scheme for all eligible unemployed, to replace the wide array of programs that currently graduate protection according to sectoral and occupational groupings.30 Replacement rates would be gradually raised to European levels to allow more effective job search; at the same time, coverage would gradually decline with duration and be made conditional on the beneficiary’s effort in terms of retraining and job search intensity. The system would be complemented by an upgrading of employment services, where the Commission recommended an end to the public monopoly of job placement, a marked decentralization of services, and a strengthening of the links with local institutions. Much greater emphasis should also be placed on active labor market policies aimed at improving job matching (through orientation programs and work counseling) and investment in human capital. The Commission saw, in particular, a need to broaden the training content of apprenticeship and trainee contracts introduced in the recent past (including in the September 1996 Patto per il Lavoro) to increase work opportunities for new entrants.

Welfare expenditure (assistenza). In the Commission’s view, welfare spending in Italy reflects an “obsolete model, far removed from that followed in other European countries.” It occupies a residual role (absorbing only some 3½ percent of GDP), “squeezed by a cumbersome and inequitable pension system and a scantly efficient health system.” It is in this area that the Commission thus made its more far-reaching and innovative proposals. Besides emphasizing the long-standing need to draw a clear separation between social security (previdenza), to be financed from contributions, and welfare (assistenza), properly financed from general taxation, the Commission’s report takes a strong position on the current incomplete and insufficient coverage of the residual risk of being poor. In response, it recommends the introduction, in a gradual and experimental way over a period of time, of a means-tested minimum income scheme (minimo vitale), administered by local authorities and providing a safety net available to all needy adults (of any age) willing to participate in the production process, unless infirm or too old to do so. To avoid poverty trap dynamics, the report advocates that the benefit should only partially compensate the shortfall from the poverty line; and be contingent on the beneficiary’s effort to (re)join employment. To avoid overprotection, it should rest on radically reformed mechanisms of award, based on an extended definition of own resources31 and centered on households, rather than individuals, as a more reliable unit of income assessment. The new scheme would imply the abolition of the current system of family allowances. The report envisages that, in the initial reform phase, spending on these forms of assistance should not increase from its current level in relation to GDP, but that—once advocated savings are realized elsewhere, notably in the pension system—it could rise gradually from the current 3½ percent of GDP to 4.2 percent of GDP by the year 2001.

51. As noted above, the final report of the Commissione Onofri was submitted at end-February 1997. It is currently in the hands of the government, who will decide when and how to proceed on its wide range of recommendations.

Table 3.

Italy: Incidence of Poverty (1995)

(In percent)

article image
Source: Commissione di indagine sulla povertá (1995).
Table 4.

Italy: Incidence of Poverty on Different Age Groups

(In percent)

article image
Source: Cannari and Franco (1997).

Italy: Incidence of Poverty

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: Committee on poverty and exclusion (1995, 1996).

IV. External Developments and the Lira’s Return to the ERM32

A. The Balance of Payments

52. Italy’s external position has improved steadily during the past five years. Following a string of current account deficits that reached Lit 35.2 trillion (corresponding to 2.3 percent of GDP) in 1992, the current account swung to a surplus of Lit 17.8 trillion (1.1 percent of GDP) in 1993, and rose steadily thereafter to an estimated surplus of nearly Lit 70 trillion (3.7 percent of GDP) in 1996. Improved competitiveness, owing to a depreciated exchange rate compared to the period before September 1992, wage moderation, and actions to reduce the fiscal deficit have helped to bring about a shift in the composition of demand that has made for growing current account surpluses (Chart 10).


ITALY: Current Account, Exports, and Imports

(In percent of GDP)

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Sources: Bank of Italy; and IMF, World Economic Outlook;1/ Staff projections.

53. The current account improvement in 1996 was due to an increased merchandise trade surplus; the deficit on invisibles was little changed. After recording large gains in 1995, the trade surplus in 1996 rose to an estimated 5¼ percent of GDP in 1996. This performance reflects both an improvement in the terms of trade following three years of steady deterioration, and a relatively strong decline in import volume. Import volume growth turned negative during 1996, despite a strong appreciation of the lira, reflecting weak domestic demand, particularly investment (Chart 11). The volume of exports remained essentially flat, lagging growth in demand from trading partners, and leading to the first loss in export market share since 1992. External performance was not homogeneous in terms of direction of trade: export growth to East and Central European countries reached 23 percent and those to the newly industrializing countries of East Asia approached 10 percent, while exports to EU countries stagnated. Indeed, the share of Italy’s merchandise exports directed at non-European Union countries has expanded considerably since the turn of the decade as the share of exports directed to the European Union declined from 63 percent to 55 percent. At the same time, however, the share of imports from EU countries remained broadly unchanged at slightly over 60 percent (Chart 12). After rising for the past three years, the share of Italian exports in world trade declined in 1996 (Chart 13). While the share of imports declined only marginally from 62 percent to 61 percent (Chart 12).


ITALY: Recent Trade Performance

(Year-on-year percentage change)

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: Bank of Italy.

Italy: Direction of Trade 1990 and 1996

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: Direction of Trade Statistics.1Data includes January - November 1996.

Italy: Export Growth

(Year-on-year percentage change)

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: World Economic Outlook.

54. The deficit on invisibles rose slightly (by Lit 0.7 trillion in the first ten months of 1996 over the corresponding period in 1995), as the increase in net transfers—mainly to the EU—was only partly offset by the rise in the surplus on services and the smaller deficit on investment income. In the services category, the surplus on tourism narrowed due to the loss of competitiveness and weaker international demand, which led to the first reduction in spending by foreign tourists since 1989. The deficit on investment income declined owing to the continued contraction of the external debt and of the narrowing of the differential between domestic and foreign interest rates.

55. Capital movements generated outflows of Lit 7.8 trillion in 1996, compared to Lit 6.6 trillion the preceding year. Although the pattern of capital movements was not uniform during the year, it was accompanied by an almost constant appreciation of the exchange rate. Starting in the second quarter, the restoration of confidence by international markets, following the elections and expectations of reductions in interest rates, resulted in a sharp rise in net inflows of foreign portfolio investment. These amounted to Lit 122.3 trillion for the year as a whole, double the 1995 inflows and the largest such inflow in the postwar period. Over three quarters of this stemmed from purchases of government securities. The past year also witnessed the further internationalization of nonbank residents’ portfolios, a process that had been checked after the 1992 exchange rate crisis by the need of italian investors to reduce their net foreign currency liabilities to resident banks. These liabilities were more than halved between 1992 and 1995, allowing Italian investors to resume the diversification of their portfolios while maintaining their foreign currency liabilities essentially unchanged. Portfolio investment abroad rose from Lit 11.9 trillion in 1995 to Lit 55 trillion in 1996, spurred by the reduction in the differential between Italian and foreign securities. Nonbanks’ net external liabilities stood at Lit 121.4 trillion at the end of 1996, an increase of Lit 34.7 trillion during the year.

56. Resident banks generated net capital outflows of Lit 42.5 trillion in 1996, of which Lit 13.4 trillion were in foreign currency, reducing their net external liabilities to Lit 53.4 trillion. Official reserves (including gold), however, rose by Lit 20.5 trillion during 1996 to Lit 106.7 trillion. Thus, Italy’ s net external liabilities declined to Lit 68 trillion at the end of 1996, well below 4 percent of GDP and down considerably from 11 percent at the end of 1992. The current account surplus projected for 1997 should bring the net external position roughly into balance by the end of this year. Indeed, in view of the sizable, negative “errors and omissions” of Lit 31.3 trillion in the first ten months of 1996—a negative cumulative total since 1992 of around Lit 100 trillion—Italy’s’s net external position is most likely already in balance, a situation last observed in 1984.

B. The Lira’s Reentry into the ERM

57. The lira reentered the ERM on November 25, 1996, at a central rate of Lit 1,906.5 per Ecu, equivalent to Lit 990 per deutsche mark, which was less than 1 percent lower than the market rate on the preceding business day. Reentry followed approval of the 1997 budget by the lower house of Parliament, incorporating the government’s stepped-up fiscal plans aiming for early EMU participation, and was made possible by the improvement in underlying fundamentals: the year-on-year rate of inflation had more than halved from a peak of 6 percent in November 1995 to slightly above 2½ percent; the differential between Italian and German long-term interest rates had declined from close to 6 percentage points in early 1995 to less than 2 percentage points in November 1996; as noted above, the current account had recorded an improvement of 6 percentage points of GDP since 1992, while over the corresponding period Italy’s net external liabilities had declined steeply, to the point that Italy was projected to become a net creditor in the proximate future.

58. ERM reentry presents certain advantages in terms of meeting one of the prerequisites for participation in EMU from the start, providing a focal point for exchange rate expectations and acting as a discipline device on domestic macroeconomic policies. Reentry could be costly and counter-productive, however, if it were to prove unsustainable. The success of the return to the ERM will rest mainly on two factors: the appropriateness of the chosen central parity for the lira, and the enactment of supportive macroeconomic policies. The remainder of this chapter examines the first of these two factors, namely, whether the central rate at which the lira reentered the ERM is a sustainable exchange rate from a medium-term perspective.


59. In the year and a half or so prior to rejoining the ERM, the lira appreciated considerably, both in nominal and in real effective terms and against the deutsche mark (Chart 14). In real effective terms it appreciated by about 25 percent between March 1995 and November 1996, more than reversing the steep depreciation observed from mid-1994 to early 1995. Thus, in November 1996, the real effective exchange rate of the lira based on (normalized) unit labor costs was about 10 percent below its level at the start of the EMS in 1979 and its level at the beginning of 1986, when Italy strengthened its commitment to stable nominal exchange rates within the ERM. Based on consumer prices, however, the real effective exchange rate was around 10 percent above its 1979 level and at around the same level as at the start of 1986.33 The behavior of the lira’s real effective exchange rate in late 1996 was not far removed from that of several other currencies that had come under severe pressure during the 1992 and 1993 ERM crises, such as those of Spain, Sweden and the United Kingdom (Chart 15).


Italy: Exchange Rates

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: IMF, International Financial Statistics.

Real Effective Exchange Rates

(Index, 1979 = 100; based on ULC)

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: IMF, World Economic Outlook

60. On a bilateral basis, which permits a more direct comparison with EU trading partners and main ERM participants, the ULC-based real exchange rate of the lira vis-à-vis the French franc in November 1996 was at about the same level as in 1979 and slightly below that of early 1986, while it remained substantially depreciated vis-à-vis the deutsche mark (Chart 16). Against the Spanish peseta, which had also experienced considerable volatility in recent years, the real bilateral rate of the lira was around the levels prevailing at the time of the 1992 ERM crisis.


Italy: Real Bilateral Exchange Rates of the Lira

(Index, 1979 = 100; based on normalized ULC)

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: IMF, International Financial Statistics.Note: A decline in the index indicates a depreciation of the Lira.

61. On the whole, despite a considerable erosion of international competitiveness in the second half of 1995 and in 1996, Italy’s competitive position appears to remain favorable. However, without an entrenchment of price stability it would tend to be eroded in the years ahead. Inflation has declined substantially from the 6 percent year-on-year peak in November 1995 to around 2 ½ percent in early 1997, but it is still around 1 percentage point above that of Germany and somewhat above the EU average (Chart 17). The reduction in inflation must be sustained to avoid the experiences that marked Italy’s previous ERM membership: that is, the frequent nominal devaluations that prevailed in 1979–86 (an accommodative policy clearly inconsistent with early entry into EMU), or a repetition of the overvaluation of the lira and the experience of 1987–92. As this experience showed, monetary policy alone cannot sustain exchange rate stability in the absence of wage and fiscal discipline.


Inflation Rates

(In percent change from a year earlier)

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: IMF, International Financial Statistics.

Market expectations and the lira’s ERM reentry rate

62. Several considerations suggest that market exchange rates in late-November 1996 provided a reasonable basis for setting a new central parity for the lira. As noted above, the real effective exchange rate of the lira was close to its level in 1979, when Italy first entered the ERM—this was the case not only on a broad effective basis but also more narrowly with respect to ERM currencies. Between the late 1980s and early 1995 the lira had ranged from a rate prior to the ERM crisis (Lit 800 to the deutsche mark) that was clearly overvalued and which resulted in a declining export market share and widening current account deficit, to an undervalued rate in the course of 1994–95 (reaching Lit 1,200 to the deutsche mark) and a rapidly growing current account surplus. By late-November 1996, the lira stood at around the middle of this range, but the range is too wide for competitiveness indicators alone to provide a helpful indication of the appropriate reentry rate.

63. Market expectations of the future value of the lira, however, also suggested that the lira was close to its sustainable level. Indeed, the appreciation during the year had been associated with a reversal of exchange rate expectations. In October 1995, the consensus forecast was for a nominal appreciation of the lira with respect to the deutsche mark at horizons of 3 months, 1 year, and 2 years ahead. But as the lira strengthened (by more than forecast), market participants began projecting a depreciation. In November 1996, the consensus forecast anticipated a nominal depreciation of 2.3 percent against the deutsche mark two years ahead (Charts 18 and 19), but only a slight real depreciation when forecast inflation differentials are taken into account. Furthermore, implied forward exchange rates indicated that over the next two years the lira was expected to depreciate by at most 6 percent against the deutsche mark in nominal terms. When inflation differentials and possible risk premia are taken into consideration, the implied real exchange rate depreciation is slight.


Consensus Forecasts

(Italian Lira per Deutschemark; percentage change from current rates)

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: Consensus Forecasts.

Consensus Forecasts

(Italian Lira per Deutschemark forecast rates)

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: Consensus Forecasts.

The lira’s medium-term sustainable exchange rate

64. Model-based estimates of equilibrium real exchange rates over the medium term provide additional evidence in support of the view that market exchange rates prevailing in November 1996 constituted a reasonable basis for setting the ERM central parity for the lira. These estimates are based on a macroeconomic balance framework that focuses on the requirements for simultaneously achieving internal and external balance. They involve estimating and comparing the current account that is likely to emerge in the medium term under prevailing exchange rates (the underlying current account) to the “equilibrium” or desired current account in the medium term. To the extent that these differ, the real exchange rate will have to adjust in order to move from the underlying current account position to the equilibrium current account. In the calculations reported below, it was assumed that the medium-term positions are reached over a 5-year horizon (ending in 2001) during which output gaps are closed and structural fiscal balances are adjusted to desirable levels.

65. The underlying current account is the cyclically-adjusted current account that would result at current exchange rates if these rates were sustained. It is calculated by adjusting the actual current account for existing output gaps at home and abroad and for the effects of changes in lagged real exchange rates:34




where CA and CAU denote the actual and the underlying current accounts as a ratio to GDP; X and M denote nominal exports and imports as a ratio to GDP; GAP and GAP denote the domestic and the foreign output gaps; αx and αm denote the elasticities of export and import volumes with respect to foreign and domestic income, respectively; βx and βm denote the long-run exchange rate elasticity of exports and imports; R denotes (the logarithm of) the real effective exchange rate; and RL is a weighted average of current and past exchange rates.

66. Estimates of the underlying current account were based on updated data for the October 1996 World Economic Outlook; specifically, it was assumed that Italy would register a current account surplus of 3.6 percent of GDP in 1996 and an output gap slightly above 3 percent. On these assumptions, and based on the exchange rates prevailing during October 1996, it was estimated that Italy had an underlying current account surplus of 1.4 percent of GDP.

67. The “equilibrium,” or desired, current account is derived from a model of the equilibrium saving-investment balance that links saving and investment flows to their structural determinants in a multilateral framework.35 The sustainable exchange rate is then defined as the level of the real exchange rate that would equate the cyclically-adjusted current account balance to the desired current account balance (CAd) in the medium term:




Rd - R, the deviation of the current real exchange rate from its desired level, is the estimated degree of “misalignment.” This approach involves some strong assumptions, in particular, that potential output can be taken as exogenous to saving, investment, and the exchange rate, and that output gaps are closed in the medium term.

68. The saving-investment balances implied by the fiscal adjustments necessary to meet the Maastricht Stability and Growth Pact conditions were taken as objective ways of generating desired current account balances in the medium term. Two cases were considered. In one case (Case I), the medium-term desired current accounts were based on the structural fiscal balances assumed in the October 1996 WEO, except for Italy, for which a somewhat greater degree of fiscal consolidation was assumed in order to meet the Maastricht criteria. The assumed structural fiscal deficit for Italy in the medium term is equivalent to 2 percent of GDP. In a second “Stability and Growth Pact” case (Case II), the structural fiscal deficits were assumed to decline to 1 percent of GDP in the medium term in all EU countries likely to participate in EMU, including Italy.

69. In case I, the medium-term equilibrium current account balance for Italy is estimated to be a surplus of 1.1 percent of GDP. The estimated change in the multilateral real exchange rate necessary to move from an underlying current account surplus of 1.4 percent of GDP to a surplus of 1.1 percent is a multilateral real appreciation of 0.8 percent. In case II, the equilibrium current account surplus widens to 1.6 percent of GDP, necessitating a multilateral real depreciation of 1.6 percent. Taking into consideration the margins of uncertainty involved, these estimates suggest that, prior to rejoining the ERM, the lira could not be said to be “misaligned” in a multilateral sense.36

70. The above exercise was carried out in a broad multilateral framework, while the lira’s reentry rate into the ERM was, of course, set in relation to other ERM currencies. What is relevant in this context, therefore, is an assessment of the lira exchange rate against a weighted average of other currencies participating in the ERM. Against this narrower group of currencies, the lira would have to appreciate by 3 percent in case I and by 3.4 percent in case II. Again, given the margins of uncertainty, the lira could not be said to have been “misaligned” vis-à-vis the ERM currencies, and market rates in November 1996 provided a reasonable basis for ERM reentry.

71. The discussion thus far has been couched in terms of real exchange rates. Participation in the ERM, however, involves central parities for nominal exchange rates. With the lira rejoining the ERM at a rate close to then-prevailing market levels against other ERM currencies it might be viewed as having done so at a somewhat depreciated rate with respect to the point estimates reported above.37 The level of Italian unit labor costs over the next five years is, however, according to the October 1996 WEO projected to rise by about 3½ percent (and the price level by about 3 percent) relative to that in other EU countries, eroding any competitive advantage. The inflation outcomes are, of course, not independent of the ERM reentry rate and the fiscal consolidation path. In particular, the more ambitious fiscal consolidation path corresponding to the Stability and Growth Pact scenario would result in a lower inflation rate path than that assumed in the projections.

72. In sum, competitiveness indicators, market expectations, and model-based estimates of equilibrium exchange rates, all suggest that the lira could not be said to have been “misaligned” against other ERM currencies as a whole, and that an ERM reentry rate of between Lit 950 and Lit 1,000 to the deutsche mark could be considered broadly appropriate.

Financial indicators and sustainability of a narrow exchange rate band

73. In late 1996, movements in financial indicators pointed to a clear improvement in the prospects for a sustainable reentry of the lira in the ERM. They nevertheless underscored the lingering effects of past political uncertainties and the magnitude of the task ahead. In particular, interest rate differentials continued to reflect premia on lira-denominated assets in excess of expected inflation differentials. As for the challenges ahead, the Bank of Italy’s task of maintaining the lira within a narrow band will not be facilitated by the large share of short-term and floating rate public debt, or by the still relatively low holdings of international reserves, and will hinge on effective progress on the fiscal front.

74. Reflecting the decline in inflation, as well as the appreciation of the exchange rate, differentials between Italian and German short- and long-term interest rates narrowed by between 3 and 4 percentage points between April 1995 and November 1996 (Chart 20). Nevertheless, differentials remained sizable, especially at the short end of the maturity spectrum, and the implied forward exchange rates corresponded to a depreciation of the lira vis-à-vis the deutsche mark of 6 percent over the next two years; over a longer period, forward differentials (Chart 21) suggested that inflation in Italy will continue to converge to “core” EU-average levels. Although forward rate differentials remained above 150 basis points for all horizons, when interpreted as a measure of expected values, they should be seen as reflecting not only the likely path of inflation in Italy and partner countries, but also the probability markets attach to the event of a crisis and the resulting depreciation of the lira.


ITALY: Selected Interest and Exchange Rate Indicators

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: Bloomberg; IMF, Treasurer’s department; and staff calculations.

ITALY: Forward Interest Rates

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: Bloomberg.

75. Since late 1992, daily volatility of the lira/DM exchange rate has been around 0.5 percent, with abrupt increases in crisis periods (Chart 20).38 Volatility of this magnitude would seem excessive in the context of narrow exchange rate intervention bands. However, it would likely prove difficult to reduce the lira’s volatility without the support of a clear strengthening of the fiscal stance, as the ability to intervene in exchange markets will be constrained by Italy’s still relatively low level of international reserves. Although the ratio of reserves to imports recently improved (in part owing to the recent fall in import flows), it remains low both from a historical and comparative perspective (Chart 22). International reserve holdings are also relatively small as a proportion of portfolio investments flows. Gross monthly portfolio flows (Chart 23) increased substantially over the past year, to ten times the total stock of international reserves, reflecting greater nonresident interest in Italian securities. Although these flows reflect in part a higher turnover and the short maturity of many positions taken by market participants, they nevertheless suggest that, to be effective in support of the lira, any intervention would likely need to be substantial. Interventions in exchange markets are thus likely to be effective only in smoothing the exchange rate of the lira when there are no major sources of turbulence. At the same time, hikes in interest rates in defense of the lira can have a substantial impact on the government budget because of the large stock of short-term or floating rate debt, corresponding to about 55 percent of GDP (Chart 24)—although this share compares favorably to the 65 percent ratio in 1992.


Total International Reserves

(In ratio to total imports)

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: IMF, International Financial Statistics.

ITALY: Portfolio Investment

(Gross Monthly Flows, Trillions of Lire)

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: Bank of Italy.

Italy: Structure of Marketable Public Debt

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: Ministry of Finance

76. The above considerations suggest that despite a clear increase in confidence during 1996, financial markets may have still harbored a substantial degree of uncertainty about Italy’s financial outlook. Market developments also appeared to reflect “convergence plays” which, as past experience has shown, can unravel rapidly and disruptively when confidence is adversely affected. In these circumstances, and in view of the costs and constraints of an excessive reliance on monetary policy, fiscal adjustment that is sustained and permanent in its effects, and continued wage restraint, are essential conditions for a successful participation of the lira in the ERM.

C. Conclusion

77. In several respects, in the fall of 1996 the prospects for a successful reentry of the lira into the ERM were the best they had been since the lira’s exit four years earlier. In particular, the market exchange rate was close to a rate that may be considered as sustainable in the longer run, consistent with a fiscal consolidation in line with the targets being considered in the Stability and Growth Pact and an inflation rate converging to the “core” EU average. Also, financial indicators had improved substantially in the previous twelve months, even though they still pointed to some uncertainties with regard to future economic policies. As to the lira’s reentry rate, various considerations, including international competitiveness, estimates of longer run sustainable exchange rates, and market expectations, suggested that the market rate that had prevailed in late 1996 was appropriate from a medium-term perspective and provided a reasonable basis for setting the chosen central parity.


A. Introduction and Summary

78. During the past few years the performance of the banking sector in Italy has deteriorated further, from an already lackluster situation. This performance has highlighted the existence of longstanding structural problems, brought to the fore also by the liberalization of the Italian financial markets since 1990 and by the unbalanced nature of the economic upswing initiated in late 1993. Banks’ profitability decreased sharply in the period, with the rate of return on equity for the sector as a whole standing close to zero (0.4 percent) in 1995. By the end of 1996, over 10 percent of total bank loans (equivalent to 7 percent of GDP) were non-performing, with about 40 percent of the value of these loans expected not to be recovered.

79. These problems, although reflecting trends observed also in other developed economies, such as the narrowing of interest margins and increasing disintermediation, were exacerbated by the lack of adjustment in banks’ operating costs, most notably labor costs. The predominance of public ownership (some 60 percent of the banking sector is controlled either by the Treasury or by charitable “foundations” whose boards are largely appointed by local authorities) in many cases compounded the difficulties.

80. Macroeconomic conditions had an early, but limited, impact on northern banks, but have been one of the key elements accounting for the poor performance of southern banks. The compression of firms’ profits driven by the overvaluation of the lira in the early 1990s was more than fully reverted by the lira’s sharp depreciation following ERM exit and its further decline in 1994–95. The beneficial effects of the export-driven economic upswing that ensued were felt by most northern banks (whose performance nevertheless continued to be burdened by structural factors). In the South, however, the economy remained in the doldrums, suffering also from the reduction in public investment and employment, cuts in government transfers to households, and the nonutilization of structural funds from the European Union, due to deficiencies in local authorities’ planning and monitoring procedures. The weakness of the regional economy showed up in a sharp deterioration in the quality of loans to small enterprises and the self-employed. As of late 1996, close to 25 percent of all loans in the region were classified as nonperforming.

81. Prudential supervision has been effective in enforcing adequate provisioning for nonperforming loans and imparting a substantial degree of transparency to the sector, as illustrated inter alia by the frequent publication of detailed data. The supervisory authorities have also steered the sector through a fundamental overhaul of its institutional set up, reflected in the incorporation of most banks that were previously ruled by special provisions in the public law. Public ownership, however, continues to predominate and—as in other countries where this is the case—may have delayed bank managers’ responses to problems in certain cases, despite supervisors’ injunctions, especially where traditional and large banks were involved.

82. The direct fiscal cost of banking problems has been limited to 0.2 percent of GDP in capital injections in a few southern banks, most notably Banco di Napoli, which was ultimately sold in late 1996. Indirect costs, illustrated by the transfer of government-owned assets to some banks, have been of a similar magnitude. Market participants generally believe the scope for further interventions on behalf of public banks is limited. Other macroeconomic effects of the banking sector’s difficulties appear minor. In particular, there is no evidence of a “credit crunch” at the national or regional level.

83. Turning to the future, a reduction in labor costs and overstaffing, and privatization are the primary steps required to address the sector’s problems. The government and the supervisory authorities have been vocal in advocating early and profound changes in the sector. In particular, discussions are underway to seek a solution to the labor problem, in particular the required cut in personnel—estimated at around 10 percent of the sector’s current workforce of 340,000. The Italian Banking Association has recognized that it will be necessary for the sector to bear most of the costs related to the reduction in overstaffing.

84. The government has also taken steps to provide new impetus to privatization through a set of measures approved by the Council of Ministers in January 1997. These comprise mainly tax incentives to induce “foundations” to sell their controlling stakes in banks. Although the new rules do not set a specific deadline for the “foundations” to relinquish their control of banks, the recent improvement in the performance of several northern banks and of the Italian stock exchange appear to have induced some major “foundations” to accelerate their privatization plans.

85. Italian banks are also increasingly considering alliances with foreign partners in response to the anticipated intensification of competition in the run up to EMU. Such alliances can indeed play an important role in allowing Italian banks to catch up with foreign banks in terms of technology and the development of new products. Competition is also likely to accelerate the consolidation of the Italian banking sector, which remains among the most fragmented in Europe.

86. The complete transition of the Italian banking sector toward an efficient market-based system will also largely depend on the contemporaneous development of other segments of the Italian capital market. In this connection, the authorities have been taking several steps to strengthen the stock exchange and enhance the transparency of financial markets in general. Avoiding widespread minority cross-holdings between domestic banks would also appear to be an important element favoring greater transparency and improved governance in the sector. With privatization, the Bank of Italy is also likely to move toward even greater reliance on adequate capitalization as the main prudential tool. In this context, consideration could usefully be given to the adoption of a set of structured early intervention and resolution rules, with a view to enhancing market discipline.

87. In conclusion, current banking problems cannot be characterized as systemic: they have an important cyclical and regional component, and in several cases have stemmed from individual banks’ mismanagement. The supervisory setup has been able to address and contain the problems adequately, within the constraints posed by public ownership. All in all, Italian banks appear to have weathered the problems arising from the latest economic cycle and the transition underway since 1990. Nonetheless, this transition is incomplete: significant structural problems persist and need to be addressed, especially in view of the increasingly competitive environment facing the industry. Important challenges thus remain for the industry and its supervisors.

88. This study reviews the developments in the banking sector in Italy in recent years, the main features of which were the comprehensive change in the institutional organization of the sector and the deterioration in banks’ performance, in a broader setting of an uneven economic recovery that exacerbated longstanding regional imbalances. The paper is organized as follows. Section B outlines key features of the banking system and the regulatory framework. Section C examines the macroeconomic dimension of current problems. Section D reviews the steps required to address the sector’s difficulties, and discusses the prospects for banking in Italy. Section E concludes the study with an overall assessment.

B. Overall Structure

89. The institutional organization of the Italian banking system has changed significantly in the 1990s, but the system remains one of the most fragmented in Europe (Tables 5 and 6; and Appendix I) and public ownership predominates. Public banks, owned by the Treasury or controlled by charitable foundations (fondazioni) whose boards are largely appointed by local authorities (see Box 1), account for around 60 percent of outstanding loans.

Table 5.

Italy: Structure of the Banking System

article image
Source: Bank of Italy.
Table 6.

Italy: Selected Indicators of Size and Concentration

article image
Sources: BIS, 1996 Annual Report, and OECD, Bank Profitability, 1996.

Commercial banks only.

90. The structure of the banking sector in Italy has undergone notable changes as a result of the passage of the so-called Amato Law in 1990 and of a new Banking Law in 1993. The Amato Law provided a framework and incentives for the incorporation and consolidation of numerous banks that were formerly organized as semi-public institutions regulated by special legal provisions.40 The new banking law, which revised and consolidated the large number of laws and regulations enacted in the previous six decades, bringing them together into a unified text (Testo Unico), provides the framework under which the sector operates (Appendix II).

The Development of Italian Savings Banks and the Creation of the “Foundations”

Modern savings banks originated in Germany at the end of the eighteenth century, and spread throughout Europe in the following Century. By 1850, there were 60 such banks in Italy (several reflecting the reorganization of banks founded in previous centuries, some of which still exist). This number had tripled by the turn of the century, before the sector was overhauled in 1938, with only 91 institutions left operating, among them some of the largest savings banks in the world (e.g., the Lombard bank Cariplo). Savings banks had a strong regional character, and because of their large and highly dispersed deposit base, they represented a prime avenue for channeling liquid funds to the banking system. They were also an important source of funding for local government through their branches specialized in long-term credit, which also offered home mortgages.

Formally, savings banks were organized as either associations or foundations. All were non-profit or charitable organizations based on a trust fund, but the former were controlled by a general assembly and the latter by a more restricted council. These distinctions were broadly maintained when the banks were incorporated in 1990–92, with these different organizations becoming owners of the banks, which became joint stock corporations (società per azioni)—the generic name of “foundation,” by which the controlling institutions (enti conferenti) are known, thus in some cases being a misnomer. Currently, most “foundations” are controlled by a board of trustees largely appointed by local and central government, the accountability of which has been questioned by several commentators.

Incorporation increased the scope of former savings banks’ activities, but narrowed the objectives of the “foundations” to include only some of their traditional fields of activity, such as charities and support of scientific research and the arts. It is envisaged that “foundations” will continue to earn a large part of their income from interests in financial and nonfinancial enterprises. But they will only control business concerns connected with narrowly defined activities in one of their traditional fields. Such a restriction is expected to force a gradual divestment from banking. The sale of controlling stakes in the former savings banks would amount to the placement of about Lit 35–40 trillion on capital markets. Such privatization has not yet taken place, although promising signals of the willingness of some major “foundations” to divest are emerging.

91. Before 1990, the organization of the market reflected long-held traditions, such as the special chartering of savings banks (started in the early 1800s), 41 and the segmentation of financial markets enshrined in the banking law of 1936 (a separation which was introduced, as in several other countries, in the aftermath of the collapse of universal banks in the early 1930s). Under this law, banks were organized along the type of operation they could engage in (e.g., the acceptance of either short- or long-term funds), as well as their legal status. Banks accepting long-term funds were called Special Credit institutions. Among them, the Istituto Mobiliare Italiano (IMI), Mediocredito, and Mediobanca lent mainly to industry and for investment in energy and telecommunications, while Crediop specialized in public works financing; there were also a number of regional development institutions such as ISVEIMER and IRFIS in the South. Banks accepting short-term funds could be classified in four groups. The first group included six old-established public chartered banks (banche di diritto pubblico) and three banks of national interest (banche di interesse nazionale), nationalized in the 1930s; the second group comprised the savings banks (casse di risparmio); the third group comprised incorporated commercial banks (banche ordinarie); the fourth group included hundreds of small cooperative and mutual banks.42

92. After 1990, most of the banks in the first and second groups, as well as those accepting long-term funds, were incorporated (i.e., became limited liability ventures organized as joint-stock companies) and became controlled by a “foundation.” In 1993, the new banking law eliminated most barriers across banking operations and reduced the categories of banks to only three: incorporated banks, mutual banks, and cooperative banks.43 Incorporated banks now operate most branches and hold the majority of deposits, despite still being outnumbered by the mutual and cooperative banks. Most barriers across banking operations were eliminated by the 1993 banking law, and banks are now often classified for statistical purposes into five broad categories based on size and geographical coverage instead of function and legal status 44.

93. The number of branches, which was tightly regulated until 1990 (also with regard to their geographical distribution), has since increased by more than 50 percent, totaling 23,927 at the end of June 1996. The number of branches per inhabitant is still lower in Italy than in Germany or Spain, but about the same as in France and higher than in the United Kingdom, the United States, or Japan. Most of the new branches are small.

Regulation and supervision

94. The 1993 banking law provides the legal basis for banking activity, regulation, and supervision, and was drawn up in accordance with European directives. The law covers all banks and any nonbanking intermediary which engages in deposit taking, loan granting, payment services, and foreign exchange trading, thus creating a level playing field for market integration. The law also includes special provisions concerning the licensing and supervision of conglomerates(gruppi bancari). On the other hand, it does not regulate mutual fund and security firms, which are nevertheless regulated within a broader framework which comprises both prudential and consumer protection aspects. Supervisory authorities consist of the Interministerial Committee for Credit and Savings, the Minister of the Treasury, and the Bank of Italy (see Appendix II); the responsibilities and powers of each authority are clearly established in the law. The Bank of Italy is independent, the Governor is appointed for an unlimited term, and cannot be removed.

95. The supervisory authorities dispose of a broad set of instruments designed to address problems in a timely manner. The Bank of Italy, which is responsible for implementing banking supervision, has powers to summon bank officials and request corrective actions when it detects banking practices that are inconsistent with the regulations or that suggest the existence of anomalies. It can also propose administrative sanctions, which are made effective through decrees by the Minister of the Treasury. Resolution of problem institutions has usually been through mergers with stronger banks (or through an orderly transfer of assets and liabilities), rather than through banking failures. Competitive pressures, privatization, the liberalization of branching, and the new legal framework are likely to impinge on the ability of supervisors to arrange these operations in the future. The Bank of Italy still retains a role in overseeing the reorganization and merger of banks, with the objective of guaranteeing competition in the sector (as assigned by Law 287 of 1990) and the sound and prudent management of banks (article 57 of the banking law). Banks’ participation in deposit insurance schemes was made mandatory in 1996 in accordance with European directives; the insurance ceiling of Lit 200 million adopted for banks is higher than the minimum set by the European Union.

Size and concentration

96. Measured in terms of the amount of deposits or of capitalization, the banking sector in Italy is among the largest in Europe, but the largest Italian banks are much smaller than the equivalent institutions in partner countries. Concentration is also lower than in France, Spain, or the United Kingdom. There are seven “large” banks in Italy, but the largest Italian bank in asset holdings ranks only forty-seventh in the world and eighteenth in Europe (Table 7). The absence of large banks by international standards is a result of the previous institutional segmentation of the market. This feature has not yet been reversed by the provisions in the Amato Law favoring mergers and acquisitions in the sector, and banking consolidation in Italy still lags that observed in other industrial countries. To date, consolidation has involved mostly the merger or acquisition of small banks, with the purchase of a few medium-sized banks by larger institutions.

Table 7.

International Ranking of the 10 Largest Italian Banks

article image
Sources: The Banker, July 1996; and ICBA.

Codes for major (controlling) shareholders are the following: Pub.= a Foundation, the Treasury, or other public sector banks; Priv= Private sector.

97. Banks’ assets grew rapidly during the economic upswing of 1988–92 (outstanding loans grew by 60 percent in constant lira), and contracted in the aftermath of the 1993 recession (credit fell by 30 percent in real terms in 1993–95). Financial disintermediation, as well as the recent fall in households’ saving rates driven in part by the decline in inflation, also contributed to this contraction.

Asset quality and portfolio risks

98. The ratio of nonperforming loans (sofferenze) to total loans doubled in 1991–96, reaching 10.1 percent in 1996 (Table 8 and Chart 25).45 In addition to Lit 120 trillion in bad loans, loans with minor problems (partite incagliate) amount to about Lit 45 trillion. According to the estimates banks provide the Bank of Italy, about 40 percent of the face value of bad loans is not recoverable. The bad loan problem appears, however, to have peaked, with the ratio of bad loans to all loans likely to stabilize or fall slightly in 1997. The proportion of loans to firms considered at risk or facing problems, which peaked in 1992, has fallen to levels close to those prevailing in the late 1980s.

Table 8.

Italy: Stock of Nonperforming Loans

article image
Source: Bank of Italy



Italy: Non-performing Loans

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: Bank of Italy

99. The incidence of bad loans differs substantially across economic sectors and geographic regions. The ratio of bad loans is significantly higher in construction than in industry (Chart 26), a feature common to many countries but accentuated in Italy by the uneven nature of the last economic upswing, based mainly on the expansion of the tradeable sector. The geographic distribution of bad loans is also very uneven. Their share in total loans is three times lower in the northern regions than in the South, where the tradeable sector is small and economic weakness has been pronounced and prolonged (Table 9).


Italy: Proportion of Bad Loans by Sector of the Economy and Region

(In percent of total loans)

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: Bank of Italy
Table 9.

Italy: Regional Breakdown of Nonperforming Loans 1/

(In trillion of lira)

article image
Source: Statistical Bulletin of the Bank of Italy.

As of end of June 1996.

100. The large share of bad loans also results from past inadequacies in the tax law (recently corrected) and in the bankruptcy process. On the one hand, charge-offs could be deducted from taxable income only after all legal action had been exhausted. On the other hand, bankruptcy procedures take on average six years (although they can often extend over a period of more than eight years), lengthening the period during which bad loans tend to remain on the books.46

101. The increase in the share of bad loans has prompted the Bank of Italy to intensify the scrutiny of banks’ accounts, particularly in the South. In 1992–95, the Bank of Italy conducted on-site inspections in 93 percent of southern banks and reviewed on average about 70 percent of these banks’ loan portfolios. On-site inspections covered 70 percent of northern banks; the Bank of Italy’s objective would be to perform an on-site inspection in each of the 970 banks at least once every four years, but the cycle has in practice been longer, especially as regards large and foreign banks.47

102. The proportion of banks’ large credit exposures has fallen in recent years, but is still high by international standards, in part reflecting the relatively small size of Italian banks vis-à- vis their customers.

Liquidity and capital adequacy

103. Liquidity has seldom been a problem in the Italian banking system. “Free” capital,48 on the other hand, has in some cases been scarce (Table 10), despite banks’ relatively comfortable position in terms of supervisory capital (which is enhanced by large holdings of government securities). The overall (risk-adjusted) solvency ratio of banks fell slightly, to 12.1 percent in 1995 (12.4 percent in the North and 9.5 percent in the South).

Table 10.

Italy: Capitalization Indices

(In percent of total assets)

article image
Sources: Prometeia, and The Banker.

Risk-Adjusted Capital to Asset Ratio (1995), according to the Basle Committee guidelines.

Banco Rolo merged with Credit in 1996.

Intermediation margins and operating costs

104. Interest margins in Italy have been historically large when compared with those prevailing on average in other G-7 countries. They nevertheless appear to be converging toward that average in recent years (Chart 27, and Box 2). Commissions and fees have traditionally represented a comparatively small source of income for Italian banks (except in 1993, when there was a windfall in trading commissions). Their share in banks’ gross income has been around 25 percent (well below the 35–40 percent share recorded in the United States and the United Kingdom, but close to that prevailing in continental Europe), reflecting the narrow range of services offered by banks and the small contribution of fees from “managed savings.”


Italy: Relative Performance Indicators for the Banking Sector 1/

(In percent of total assets)

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: OECD1/ Selected partner countries: France, Germany, Japan, Spain, United Kingdom, and United States.

Why Interest Margins Have Narrowed

The recent narrowing of interest margins appears to be driven by several factors. First, the spread between lending and deposit rates narrowed after the lira left the ERM in 1992, as the stance of monetary policy was relaxed and deposit rates proved to be stickier than lending rates (Chart 28). Deposit rates have been sticky in part because of competition from government securities (individuals can easily buy these at banks or post office branches), with banks fearing a swift customer response to any fall in the remuneration of deposits. Second, there has been a significant shift in the structure of banks’ liabilities, which has tended to increase banks’ funding costs. Since 1990, CDs have increasingly substituted for current and savings accounts as a source of funds for banks. The volume of securities issued by banks and repurchase agreements has also expanded significantly. These sources of funds are on average more expensive than deposits, even when monetary policy is relaxed (on the other hand, banks have benefited from reductions in the level of required reserves). Finally, weak demand for credit since 1992 has raised the share of loans granted at a rate below the prime rate, from 16 percent to about 45 percent of the total.


Italy: Selected Determinants of Banks’ Interest Margins

Citation: IMF Staff Country Reports 1997, 044; 10.5089/9781451819700.002.A001

Source: Bank of Italy

105. Operating costs, and in particular staff costs, are well above the average in partner countries (Chart 27), and their failure to adjust appropriately to the decline in gross income has been a key factor in the 17 percent fall in banks’ operating income (measured at constant prices) observed in the last three years. The average cost per employee, corrected for fluctuations in exchange rates, has been about 40 percent higher in Italy than in partner countries. Moreover, as intermediation margins have narrowed, labor costs have taken a larger share of gross income (Table 11). Although the component of wages set by national contracts has increased moderately since 1993 (Table 12), wages at firm level have continued to increase at a relatively fast pace, despite the flagging performance of most banks.

Table 11.

Italy: Labor Cost as a Percentage of Gross Income

article image
Sources: OECD, Bank of Italy, Bank of France.

Commercial banks only.

Table 12.

Italy: Evolution of Labor Costs

article image
Source: Assicredito.

106. The current workforce of 340,000 bank employees is also generally judged to be excessive, with several observers estimating the overstaffing to be in the order of 10 percent of the workforce. The Italian Banking Association also estimates that the greater security problems faced by Italian banks, when measured against those in other European countries, entail an additional cost to Italian banks of some Lit 1.5 trillion a year, or 0.7 percent of banks’ equity (ABI, 1997).

Provisions and profitability

107. The deterioration of asset quality since 1993 has entailed an increase in provisions for prudential reasons, which for a large part were not tax-deductible (Table 13).49 Fluctuations in bond prices also had an impact on banks’ profitability in certain years. In 1995, provisions absorbed about half of banks’ operating income, in the South, however, they corresponded to twice the operating income.50

Table 13.

Labor Costs, Provisions and Profits of Italian Banks

(In trillion of lira)

article image
Sources: OECD, Banks Profitability; and Bank of Italy.

108. In recent years, profits have been particularly disappointing for incorporated banks, with mutual banks and foreign banks maintaining a positive rate of return (Table 14). In 1995, returns in the sector were significantly reduced by the losses posted by banks located in the South (Table 15), most notably Banco di Napoli (see Box 3), Banco di Sicilia, and Sicilcassa, that required financial aid from central and local governments. Overall, banks’ performance appears to have improved in 1996, despite the persistence of structural problems. Net income was spurred by a temporary increase in intermediation margins and a rebound in bond prices. There were also lesser charges against nonperforming loans. Profits for the first half of 1996 are estimated to have been twice as large as those recorded in the corresponding period of 1995.

Table 14.

Italy: Profit and Losses Accounts of Banks

(As percentage of total assets)

article image
Source: Bank of Italy.

Includes foreign banks and special credit institutions.

Table 15.

Italy: Regional Indicators of Banking Performance, 1995

(In trillions of lira)

article image
Sources: Bank of Italy; SVIMEZ; and Fund staff estimates.

Measured as the ratio of After Tax Income to Tier One Capital.

The Crisis and Rescue of Banco di Napoli

Banco di Napoli is one of the oldest banks in Italy, its origins dating back to 1539. The eighth largest Italian bank at the time of its incorporation in 1991 (it used to be one of six public chartered banks), Banco di Napoli was thereafter reorganized as a universal bank, in line with the framework provided by the new banking law. As required, these changes were cleared with the supervisory authorities, following the approval of a business plan. Very soon, however, it became clear that the bank’s performance was departing from that envisaged in the plan. While, similarly to other public chartered banks, Banco di Napoli received an allowance from the Treasury when it was incorporated (it also benefited from an exemption on the income taxes connected with the doubling of its equity base arising from the Lit 2 trillion revaluation of its fixed assets), operating costs remained high, and the bank failed to develop new products or improve its technical and organizational resources. Moreover, following a period of rapid credit expansion aimed at increasing market share during the early 1990s, the negative effects of the economic slowdown in the South became evident, with a sharp deterioration in asset quality, which had previously been close to the national average. The deterioration was especially marked in the portfolio of credit to small- and medium- sized firms (40–45 percent of the bank’s total lending), although problems with loans to households (15 percent of the portfolio) were also present.

A detailed analysis by the supervisory authorities of the credit portfolio of Banco di Napoli, based on prudential reports, revealed that the data reported by the bank understated the actual level of credit risk. As a result, throughout 1994 the Bank of Italy held several meetings with the bank’s managers and addressed letters and injunctions to the bank’s governing bodies, formally asking them to take remedial action. Subsequently, following an on-site inspection of the bank’s foreign branches and the bank announcement of losses of Lit 1.14 trillion in 1994, the Bank of Italy initiated a deeper probe into the bank’s accounts. In early 1995 a new management team was appointed, and a full-blown inspection started. The inspection revealed that, in addition to a large amount of bad loans, the bank had a large share of its capital locked-in, requiring a restructuring of its balance sheet and the immediate supply of liquidity from outside sources (losses in 1995 amounted to Lit 3.16 trillion). To this end, a Lit 2.4 trillion loan was extended by the Deposits and Loans Fund and leading Italian banks, and a decree was issued in March 1996, authorizing the government to inject up to Lit 2 trillion of funds before the end of the year. Government funds were to be provided in connection with the restructuring and privatization of the bank.

After a preliminary restructuring (see Appendix I), which included the hiving off of about Lit 12.5 trillion in bad debts to a special structure, 60 percent of the capital of the Banco di Napoli was sold through an auction in December 1996, which drew scant interest (and no private bidders). The bank was jointly acquired by the Treasury controlled Banca Nazionale del Lavoro (BNL) and the publicly-traded insurer INA (in which the government has only a small stake), in an operation that extends the network of BNL in the South and will permit INA to strengthen its leadership in the life-insurance market, by securing its access to a combined network of 1,400 branches. On the other hand, while it will probably reduce BNL’s reliance on interbank funds, the acquisition streched the bank’s capital base. Together, BNL and INA will soon have to recapitalize Banco di Napoli, whose losses in 1996 are estimated at Lit 1.6 trillion. In the short-run, a key factor for the success of the purchase will be a reduction in operating costs in both banks, in particular with the elimination of duplications at the managerial level.

The special structure created to dispose of the bad assets (SGA-Società per la Gestione di Attività) was set up in accordance with the provisions in the so-called Sindona Law passed in the 1970s. The counterpart to the assets transferred to SGA is a loan from Banco di Napoli to SGA (Table 16). The rate at which Banco di Napoli will be remunerated for this loan is computed as an average of the gross yield of three-month Treasury bills, the three- month interbank rate, and the prime rate. Capital losses will be supported by the Bank of Italy, who will compensate Banco di Napoli for the difference between the prices at which assets were transferred to SGA (not all assets were transferred at face value) and those at which they will be disposed by SGA. Such compensation will be effected through the temporary granting to Banco di Napoli of credit lines at subsidized rates. These arrangements were discussed with the European Commission in early 1997.

Table 16.

Italy: SGA Simplified Balance Sheet

(In billions of lira)

article image
Source: Ministry of the Treasury.

109. The deterioration in the profitability of Italian banks in recent years should, however, be put in perspective, with the distortionary effects of inflation on real rates of return taken into account. On an inflation-adjusted basis, although profitability undeniably deteriorated in 1993–95 (especially when southern banks are considered), it was already lackluster at the height of the economic expansion in 1990. That year, banks’ profits were barely enough to offset the monetary depreciation of banks’ capital implied by the inflation rate (which was running at around 6.5 percent, versus a return on equity of 8 percent).

C. Macroeconomic Dimension of Current Problems

110. Macroeconomic factors, such as the overvaluation of the lira before its exit from the ERM in 1992 and the unbalanced nature of the recovery from the 1993 recession, contributed to the current weaknesses of the Italian banking system. The first compressed firms’ profit margins, with some spillover effects on northern banks, while the second prolonged the weakness of economic activity in the South, with severe consequences for southern banks (see Box 4).51

111. The impact of current banking problems on the economy have, in turn, been minimal in part due to their highly localized nature. There is no evidence of a “credit crunch” at the national level. The reduction in loan portfolios appears to have been driven mainly by demand, which has been weak since 1993 largely due to the turnaround in firms’ profitability and to the relatively high level of interest rates. Therefore, the transmission mechanism of monetary policy remained intact, while the negative impact on the overall economy of the efforts made by several banks in recent years to restrain the growth of certain segments of their portfolios was contained. The impact of current banking problems on the economy of the South, although not negligible, has been alleviated by the sale of several banks to stronger institutions in the North. Despite the implementation of stricter rules for the granting of credit, new loans have generally increased after the restructuring of those banks.

The Italian Economy in the 1990s

After a relatively favorable performance in the second half of the 1980s, economic growth in Italy started to weaken in 1990, and faltered in 1992 partly owing to the overvaluation of the lira and ensuing turbulence in exchange markets. Export growth, following the exit of the lira from the ERM in September 1992 and its sharp depreciation, drove the economic revival that began toward end-1993 and peaked in 1995. On the other hand, private consumption remained sluggish for most of the decade, as wage moderation and the contraction in dependent employment restrained the growth in total compensation. A sizeable fiscal consolidation further dampened domestic demand. Together with the landmark accord on wage bargaining of July 1993, these factors contributed to the remarkably small pass-through of the lira depreciation to prices and to the ensuing fall of inflation. They produced, however, an unbalanced recovery.

The recovery was also unevenly distributed geographically. It was pronounced in the North, where the share of tradeable goods in output is large, but almost absent in the South, where this share is small. Although manufacturing industries in the South increased their exports substantially in 1991–95, their contribution to regional growth was outweighed by the sharp contraction in leading sectors such as construction (hit by a 20 percent reduction in public investment in the region).1 Employment in the region fell by over 8 percent, affected also by the restructuring of state-owned enterprises which account for a large share of employment, and the curtailment of tax incentives to firms. Government transfers were also cut back, while some Lit 30 billion in funds from the European Union could not be invested because of deficiencies in the planning and controlling procedures of local governments. Overall, per-capita output fell by 1 percent in the South, while growing by more than 5 percent in the rest of the country. The effects of such a long recession on local enterprises were devastating, with inevitable consequences for the banking sector in the region. The recession in the South also affected households’ income and housing prices, with, however, lesser consequences to banks. Despite a marked increase during the boom years, consumer credit still accounts for a small proportion of banks’ assets; and households usually finance only a small share of the purchase price of their houses (50 percent on average).

1 Public works in general suffered interruptions as Italian courts investigated allegations of widespread corruption in the sector.

112. The direct fiscal cost related to ailing southern banks has been limited to outlays amounting to 0.2 percent of GDP. Indirect costs, comprising mainly the recapitalization of a number of banks using shares of companies partially owned by the government, and the implicit costs of special loans in support of some of these banks, have been of a similar magnitude. The scope of similar action in the near future on the part of the government in its role as shareholder, or in support of any foundation, is viewed as being limited. On the other hand, pressures to provide for a mechanism aimed at reducing overstaffing in the sector as a whole could result in some tax or other expenditure, although banks are expected to bear most of the costs of adjusting their workforce (the cost of reducing the current workforce by 10 percent in one or two years is estimated at close to 0.2 percent of GDP, see below). Banks privatization receipts (estimated at 1½–2 percent of GDP) would also have only a minor direct effect on the budget, since “foundations” are not formally part of the general government, even if controlled by local authorities. Nevertheless, the law envisages the investment of part of these receipts in activities connected with health and research, with the aim of having charitable institutions supplementing government involvement in these fields, potentially entailing a corresponding reduction in government outlays in the medium term.

113. Although the government has also been requested to examine the possibility of changing the tax treatment of the banking sector, any action in this area is more likely to be part of the comprehensive tax reform contemplated by the authorities rather than a specific response to the sector’s problems. In this connection, the Italian Banking Association has requested the exemption of banks from the new local value added tax (IREP), whose introduction is contemplated in the current tax reform, and the crediting of the national VAT incorporated in the goods and services purchased by banks. The sector has also pressed to have the current corporate income tax rate (53 percent) lowered. Recently, the tax law was changed to permit banks to earn tax credits on past provisions and to carry over losses. According to the Italian Banking Association, such credits are projected to increase banks’ income by Lit 4.5 trillion in the course of the next nine years.52

D. Current Issues and Prospects

114. From a broad perspective, the increasing role of market forces in Italian economic life has driven the process of change which is also affecting the banking sector (Box 5). The authorities have undertaken to steer this transition, mainly through the liberalization of financial activities and the privatization process. In this, they have been mindful of the risks associated with financial liberalization. Since early 1993, the Bank of Italy has encouraged banks to concentrate their efforts first in the development of savings and payments services, and in improving the efficiency and quality of the services they already delivered, rather than hastily venturing into a multitude of new areas. The authorities have also been placing increasing emphasis on the need for the banking sector to rise to the challenges posed by increasing competition in the run up to EMU, and have focused on eliminating certain sectoral features that are perceived to place an undue burden on banks.

Developments in the Italian Capital Markets and their Bearing on the Banking Sector

Since the late 1980s, Italy’s participation in the process of European integration has entailed a move toward a more open and competitive financial environment. An important effect of this process has been that of changing the setting in which the government finances its borrowing requirements, reflected inter alia in the creation of a large government bond market. In addition, with all capital controls being abolished in 1990, large-scale foreign investment flows in and out of Italian government securities motivated and rewarded the entry of numerous foreign banks. The development of other capital markets has, however, been sluggish, despite important reforms aimed at improving their efficiency. Stock market capitalization still corresponds to only one-fifth of national GDP (well below the equivalent proportions in France or the United Kingdom), and non-bank corporate bonds account for less than 1.5 percent of the stock of Italian bonds. Investment banking is therefore still accessory, limiting the scope of domestic banks outside the narrow confines of standard loan-granting activities. The loan market, on the other hand, has been constrained by firms’ increased reliance on self-financing, driven by firms’ improved profitability, a slowdown in investment rates, and relatively high real interest rates, especially after cutbacks in subsidized credit. Alternative financing instruments benefiting from tax advantages—such as leases—have not been successfully marketed by banks, their use being restricted mainly to operations among financial branches of industrial firms.

Short-term issues and the authorities’ response

115. The Bank of Italy has followed a two-pronged approach to strengthening banks’ balance sheets, while the government has focused on accelerating the privatization of banks controlled by the “foundations.” On the one hand, the Bank of Italy has been placing greater emphasis on improving banks’ internal control systems. In this connection, the Bank has in the recent past made numerous and explicit requests (through letters or meetings with officials) for improvements in banks’ internal controls of loan selection, review, and monitoring. It has also issued a number of specific provisions in this area,53 and has announced plans to issue new, more comprehensive rules to promote the establishment and the maintenance of effective internal control systems within banks. On the other hand, the Bank has pressed the parties concerned to facilitate and share the costs of reducing the excess workforce in the sector, and to agree to greater flexibility in working arrangements.

116. Improvements in internal controls of loan selection will require the development of new analytical tools by banks, both those specialized in short-term lending, and the leading sources of long-term debt (such as IMI and Mediocredito). The Italian banks have traditionally relied, to a comparatively large extent, on collateral or government guarantees for the concession of credit, partly because of the past prevalence of short-term credit lines and subsidized credit as important sources of resources to Italian firms. The relatively widespread absence of an in-depth analysis of the borrower’s prospects is also associated with the fact that medium- and large-sized firms in Italy maintain relationships with numerous banks, thus hampering any individual bank’s knowledge of the overall financial flows generated by the borrowing firms.54

117. Dismissal costs, which are a chronic problem in most sectors of the Italian economy, have been estimated at around Lit 100–150 million per bank employee, depending on the number of years to retirement. Among the reasons for such a high cost is the virtual lifetime job guarantee enjoyed by bank employees and provisions that make early retirement unattractive unless accompanied by generous severance payments. Excess labor costs and overstaffing are particularly evident in problem banks and in larger institutions (where some 70 percent of estimated overstaffing is reportedly lodged). Proposals advanced by some of the interested parties have drawn on previous sectoral plans (e.g., for the steel industry), which included a role for the government in financing part of the severance payments and compensatory transfers to social security. Despite the sense of urgency shared by social partners and the authorities, a comprehensive plan has yet to emerge from the ongoing discussions.

118. Proposed changes in working arrangements include a greater emphasis on wage differentiation and its linking to individual and firm productivity, as well as more flexibility in working hours, geographical mobility job contents, and types of contract (part-time jobs still account for a very small fraction of total employment in the sector). Several of these changes are clearly required in the Italian labor market as a whole and would also entail tackling obstacles in other markets (e.g., housing, transportation, etc.). Nevertheless, the rigidities are particularly marked in the banking sector, because of national and firm-level contracts, and their alleviation would have an immediate impact on bank profitability (a 10–15 percent reduction in labor costs would add some 2 percentage points to the sector’s rate of return on equity). Moreover, the evidence from problem banks in the South indicates the scope for labor agreements entailing significant savings in labor costs, as illustrated by the successful negotiations between labor and management of the southern savings banks Carical and Caripuglia following their take over by the northern savings bank Cariplo. Such negotiations led to a 21–27 percent cut in labor costs achieved through reductions in personnel, hours worked, paid holidays, and contributions to the firm’s pension scheme, and the introduction of part-time labor and flexible working hours (the latter are increasingly important for the development of new retail activities, such as 24-hour customer services).

119. A reduction in labor costs will also be imperative in view of the large investments that banks will be required to make in the next few years. Investments in connection with the introduction of the single currency alone have been estimated by the Federation of European Bankers to cost on average around 2 percent of banks’ operating costs for a period of three years, or about Lit 1.2–1.5 trillion in the case of Italy.55 In addition, Italian banks will have to invest heavily in technology, including automatic teller machines (ATMs) to catch up with European partners, and telecommunication services (especially in view of increasing competition from cross-border services). Although most of these investments will tend to reduce operating costs and generate new revenues in the future (e.g., ATM deposits cost about a third of the cost of processing a check, and often can be more easily charged to customers), they will entail a large up-front cost at a time when other revenues (e.g., commissions from exchange rate operations, which amounted to Lit 0.5 trillion in 1995) will dwindle. The extreme fragmentation of the sector aggravates the problem, as some of these investments entail fixed expenses that will especially burden small banks.

120. The government has attempted to provide new impetus to privatization in the banking sector through a set of measures approved by the Council of Ministers in January 1997. The bill to be discussed in Parliament establishes a number of incentives to encourage sell-offs by foundations with, for example, foundations being exempt from taxes on the capital gains originating from the sale of stakes in the capital of banks.56 The project also permits banks to recapitalize through the issuance of debentures and warrants. The relevance of such a measure stems from the current low profitability of banks and the significant revaluation of fixed assets that accompanied the incorporation of many banks, which together have sharply reduced the earnings-to-book-value ratio of most institutions and therefore increased the dilutive effects of issuing new equity. By providing an alternative to stock floatings, the bill reduces the risks of administrators in charge of foundations being perceived as squandering the capital of these institutions, while facilitating the attraction of fresh capital.

121. Other incentives for the privatization of banks contemplated in the bill include the creation of a public agency on nonprofit organizations responsible inter alia for determining if the foundations are earning adequate returns on their assets and penalizing those found wanting, and a number of requirements for the organization and appointment of the governing bodies of foundations and savings banks. Some commentators have expressed doubts about the effectiveness of the incentives selected, which appear to leave a large degree of discretion to foundations regarding the timing and scope of the sale of their assets; others have been skeptical about the benefits from the creation of another public agency, in charge of overseeing what will eventually be private institutions (the bill also provides for a change in the legal status of the foundations). Recent moves by major foundations would suggest that even in the absence of a definitive legal framework, privatization will proceed as long at stock market conditions and individual banks’ performances improve. The risk of foundations owning less performing banks choosing to delay the privatization of these institutions is, however, present despite the pressure which will presumably be exerted by the supervising agency contemplated in the draft bill.

Medium-term issues and prospects

122. The authorities are well aware of the importance of the development of other segments of the Italian capital markets for the establishment of a competitive banking sector. In this connection, the law implementing the EU Directive on Investment Services (EU Directive 93/22) provided for the privatization of the stock exchange (and the government securities’ secondary market, MTS). Steps to further strengthening supervisory bodies (such as the Stock Exchange Commission, Consob) have also been taken. These actions will inter alia help to expand the scope of banking, by creating room for the development of investment and merchant banking. Nevertheless, the ultimate success of these reforms will depend on the banks themselves, and on their capacity to develop new products and services, to acquire expertise in new areas, and respond quickly to changes in market conditions.

123. Regarding retail activities, banks are likely to continue to rely on the high degree of confidence shown by households (according to recent surveys, 52 percent of Italians rely on banks to make their savings choices, and a large number of savers expects to use banks as the managers of their private pension plans). Nevertheless, it would appear that there is still a large gulf between aspirations and realities in this area (Astolfi and La Malfa, 1993). In particular, perhaps owing to the recent growth in mutual funds (56 percent in 1996, after a decline of 3 percent in 1995), some market participants appear to place great faith on the rapid development of managed savings instruments as a major source of income to banks, despite the extremely limited number of underlying assets traded in Italy, and the ease with which individuals can already buy them. For instance, government securities can now be efficiently and cheaply purchased at banks’ windows and postal offices, and although custody fees are lower than management fees, there is no apparent reason why banks should believe savers will be compelled to favor money-market mutual funds (e.g., offered by banks’ subsidiaries) over their electronic holding of government securities.57

124. Banks’ profits from equity-based mutual funds will also depend on a widening of the stock exchange, where trade remains heavily concentrated on a few utility and telecommunication companies. Offering funds based on foreign assets may be an alternative after EMU, but the comparative advantage of the main domestic banks in this area is not evident (in fact, such a strategy appears to be best tailored to small banks, which by acting as distributors of savings products originated at foreign banks could benefit from the economies of scales involved in these activities, while retaining customers that could buy other services from them). By contrast, there is little doubt about the feasibility of greater involvement in the distribution of insurance, as reflected in the industry forecast that in the period ahead banks will sell between one-third and one-half of life insurance policies, and perhaps capture some 10–20 percent of the home and car insurance markets. There is also some promise in consumer lending, as long as credit is granted on sound market principles and targeting (the ratio of consumer credit to household consumption is three times lower in Italy than in other large European countries, and the use of fee-generating credit and debit cards is negligible). Improvements in credit selection will be essential to avoid the adverse experience of the last decade, when following the doubling in the volume of consumer credit in 1986–90, delinquency rates shot up in the following years. Experience in other countries, where default rates of consumer loans are among the lowest in the credit market, suggest scope for improvements in this area.

125. On the corporate side, there is much scope for growth. Although in the 1970s large firms were heavily dependent on credit, self-financing has, as mentioned, increasingly been the norm for all firms—hence, the importance of commissions and fees as sources of income to banks. However, surveys conducted in the early 1990s showed that less than 10 percent of small- and medium-sized firms ever used services offered by banks. In recent years, there has been a modest increase in the supply of new accounting, foreign exchange, and hedging services by banks (often under the leadership of foreign banks).58 However, unless Italian banks develop a solid expertise in specific areas, firms may not be interested in deepening their relationship with them. They will not be be inclined to either move toward single-bank relationships or to admit domestic banks in their capital.

126. Investment banking, whose core activity was until the late 1980s under the virtual control of only one bank (Mediobanca), is a case in point. In the past, Mediobanca was instrumental in financing a key part of Italian industry. The segmentation of the banking market, however, hampered its ability to monitor firms, while restrictions in the capital markets (such as the ban of proxy votes) created disincentives to the flotation of firms’ stocks.59 These factors hindered the development of investment banking and contributed to the relatively small capitalization of the Italian stock exchange, despite the existence of more than a thousand manufacturing companies with more than Lit 50 trillion in capital, which are natural candidates for listing. In recent years, the weakness of Italian banks as investment banks was evidenced when foreign banks carved out a major role for themselves in the privatization process, despite the transformation undergone by Mediobanca in connection with the entry of foreign banks in its capital and the reinforcement of ties with its traditional main shareholders. Competition in this segment of banking will not abate, indicative of the challenges faced by most Italian banks in the more dynamic and unprotected segments of banking. Consolidation and privatization is likely, however, to help domestic banks assume a larger role in expanding the Italian capital market.60

127. When accompanied by a successful integration of operations, consolidation is also likely to improve the overall efficiency of the system.61 At the same time, care should be taken to avoid an excessive interlocking of large institutions, which could reduce transparency and hamper their governance (particularly in the case of exchanges between banks of similar sizes). The loss of transparency complicates not only the market evaluation of individual banks, but can raise some obstacles to their effective supervision (although the emphasis given by the banking law to the regulation and supervision of banking groups equips the Bank of Italy with a number of instruments in this area).

128. With privatization, the supervisory authorities are likely to move toward ever greater reliance on adequate capitalization as the main prudential tool. In this context, market discipline could be enhanced by the adoption of relatively strict structured early intervention and resolution rules. These call for prespecified rules governing when and how the authorities supervising banks benefiting from a deposit insurance scheme first may and then must impose sanctions, including the ultimate closure of the bank and auction of its assets.62 Such an approach is based on the belief that if the corrective actions (and sanctions) required at different levels of banks’ capitalization (adjusted for risk) are calibrated appropriately, bank owners will be willing either to recapitalize the bank when necessary, or forfeit the control of the bank before large losses accrue. The rationale for the adoption of this mechanism in the United States in 1991 was that deposit insurance is under federal responsibility and the mechanism tends to reduce the residual claims on the insurance fund. Although deposit insurance in italy is managed by the banking sector itself and penalizes riskier banks, it is well understood that its size (up to 0.8 percent of deposits) would not be sufficient to protect depositors in large banks.63

E. Overall Assessment

129. The comparatively weak performance of Italian banks during much of the current decade has resulted from a combination of cyclical factors, which have manifested themselves with different intensity across regions, and structural weaknesses. While the economic cycle clearly played a key role in the deterioration of asset quality in the recent past, and continued weak economic growth may still be exercising an impact, particularly in the South, the persistence of structural problems is perhaps the major factor now hampering the Italian banking sector as a whole from achieving a quick turnaround. There is a consensus that, despite the extent of local difficulties, current banking problems in the banking sector cannot be characterized as systemic. The authorities remain attentive to developments in the asset quality and profitability of banks, and have been vocal in stressing the need for structural changes in the sector.

130. Some of the weaknesses of the sector are illustrated by the scores attributed by the Bank of Italy to individual banks, based on data from the periodic reports filed by banks with the supervisors. In early 1996, the Bank of Italy judged that 167 banks (accounting for 16 percent of the system’s total assets) had serious anomalies that could impair their overall operation. The Bank also noted that one-third of the reporting banks showed risk management problems, and that one-fourth of them were recording an excessively low profitability.

131. On-site inspections carried out by the Bank of Italy in 1995 (both on a regular basis and in response to unusual findings in the periodic reports) pointed to a deterioration in the overall assessment of the institutions visited, as well as some chronic problems that are likely to have heightened the sensitivity of banks’ portfolios to fluctuations in economic activity. At the time, only about one-third of the Italian banks visited were judged to deserve a favorable assessment, while 36 percent received an overall unfavorable assessment (these results were somewhat worse than those in 1994). According to the Bank of Italy, these inspections underscored the persistence of shortcomings particularly in the granting and monitoring of loans. The practice of basing the granting of loans on (at times inadequate) guarantees, instead of on the intrinsic quality of the project being financed or on the borrower’s resiliency to changing fortunes in the economy, was found to be quite widespread. In the South, this problem was often compounded by interferences that affected the autonomy and objectivity of loan officers. Laxity in the collection of nonperforming loans was also prominent.

132. Moreover, while the problems in southern banks arising from organizational deficiencies and a weak economic environment appear to have been largely contained, persistent problems stemming from high operating costs, unprofitable portfolios, and little free capital appear still to plague a few large public banks elsewhere. This appears to be the case particularly where the costs entailed by acquiring a weaker institution were not offset by savings from a fast slim down of operational procedures and workforce. Although such problems have been minimized by the high level of liquidity prevailing throughout the banking system and the market power still enjoyed by traditional institutions, it is not inconceivable that financial packages may be required to put some of these banks on a sounder footing before they are privatized.

133. Excessive labor costs and over staffing continue to be a major problem for many domestic banks, and particularly in larger and ailing institutions. Wages are high, not only in relation to other countries, but also vis-à-vis the compensation earned by workers with similar skills in other sectors of the Italian economy. Overstaffing has been aggravated by the introduction of new technology, although it is generally recognized that Italian banks have lagged behind other European banks in this respect. The pace of personnel cuts in the past three years (2,500 positions) has clearly been insufficient. Workforce skills also suffer from a certain inadequacy to the new environment, as the proportion of middle managers is higher than in other countries, and careers have traditionally been based on seniority rather than performance. Nevertheless, the relative success achieved since the liberalization of branching in 1990 in transferring back-office workers to activities involving direct contact with the public constitutes an example of a positive response to the new challenges.

134. The intensification of competition in the run up of EMU will require Italian banks not only to reduce costs, but to innovate by expanding the range of product and services offered. The steps taken by several banks to position themselves to best benefit from the high savings rates of Italian households and the dynamism of Italian exporting firms, through the forging of alliances with foreign partners, and the acceleratation of privatization plans, are thus welcome. Consolidation and exit are also likely to take place in the period ahead.

135. The activity of supervisors in recent years has, in Italy as elsewhere, reflected an adjustment to the changes in supervisory approaches and methods that took shape during this period. From the early 1990s, prudential supervision began to shift from relying on the institutional control of single aspects of banks’ activities, toward a system based on market-oriented instruments, such as capital requirements and sector-wide rules on risk-taking. These changes undoubtedly added to the complexity of supervisors’ tasks, which were also rendered more demanding by the organizational transformation banks themselves were undergoing.

136. Evidence points to the existence, in Italy, of a timely reporting system satisfying international norms (see Appendix II): it has provided a substantial degree of transparency to the sector, as illustrated inter alia by the frequent and detailed publication of data on non-performing loans. The reporting system has nevertheless been judged to require some strengthening in order to cope fully with some peculiarities of Italian banking, such as the practice of multibank relationships on the part of firms. The Bank of Italy has accordingly initiated an upgrading of its reporting system and the associated databank on risks (centrale dei rischi). Supervisory activity also relies on frequent contacts between supervisors and bank officials, often focused on addressing specific problems identified in the review of banks’ accounts and reports. On-site inspections, whose frequency has—due to resource constraints—been less than the Bank of Italy itself would consider desirable (and perhaps overly tilted toward the multitude of small banks), have also been an important avenue for the diagnosis and correction of banks’ problems.

137. A question remains, however, about the promptness with which banks adopt the corrective measures requested by supervisors. In particular, public ownership—as in other countries where it predominates—may have hampered a quicker resolution of problems in certain cases. Moreover, where traditional and large banks are involved, there is, understandably, considerable political resistance to allowing a bank to fail. With the new impetus to privatization being given by the authorities, the difficulties posed by the current ownership structure are set to recede. Privatization will, inter alia, give full expression to the principles underpinning the 1993 banking law, i.e., those of equal treatment between banks and the recognition of their entrepreneurial nature.

138. On balance, current problems cannot be characterized as systemic: they have an important cyclical and regional component, and in several cases have stemmed from individual banks’ mismanagement. The supervisory setup has been able to address and contain the problems adequately, within the constraints posed by public ownership. All in all, Italian banks appear to have weathered the problems arising from the economic cycle and the transition underway since 1990. Nonetheless, this transition is incomplete: significant structural problems persist and need to be addressed, especially in view of the increasingly competitive environment facing the industry. Important challenges thus remain for the industry and its supervisors.

APPENDIX I Structure and Performance of the Italian Banking System: Further Details

139. The Italian banking system comprises 970 registered banks, of which six are foreign owned, and 52 are branches of foreign banks. In addition, since the Second EU Banking Directive was incorporated into Italian law in late 1995, the Bank of Italy has received 68 notifications of intent to offer cross-board banking services in Italy from EU banks. Among the domestic banks, the presence of the public sector is significant, with the Treasury or charitable foundations (fondazioni) governed by local authorities controlling several of the largest institutions and numerous medium-sized banks, which together account for 64 percent of outstanding loans.

140. The aggregate balance sheet of Italian banks amounts to about 120 percent of GDP, a proportion lower than in France or Germany. This reflects two somewhat contradictory features. On the one hand, households have historically maintained high saving rates and relied heavily on the banking system (almost 90 percent of total financial assets of the nonstate sector are in the form of deposits, certificates of deposit, and securities held in banks). On the other hand, firms have increasingly relied on self financing.

141. Generally, banking consolidation in Italy still lags that observed in other industrial countries. Despite the acquisition of several small banks by the seven largest banks,64 the process has involved mainly small (mutual) banks in the same regional area (the market share of the smallest 700 banks, although doubling in the period, still amounts to less than 4 percent). A modest acceleration in the process is, nevertheless, taking place. Sixty-four banks merged with, or were taken over by other institutions in 1995. This incipient consolidation has been fostered both by increases in competition and the economic slowdown. In recent years, about half of the mergers and acquisitions in the North and Center have involved at least one institution with some weakness, and 25 banks in the economically-depressed southern part of the country were taken over by banks from other regions (contrasting with only 18 in the whole of the 1980s).

142. The average solvency ratio of incorporated banks accepting short-term funds stood at 11 percent at end 1995; that of banks accepting long-term funds stood at 16.2 percent; and that of mutual banks stood at 14.7 percent.

143. The overall amount of loans classified as “large risks,” i.e., loans to customers whose borrowing from an individual bank or group exceeds 15 percent of the institution’s supervisory capital, comprises roughly one-sixth of the aggregate lending by Italian banks, despite a 25 percent decline in these risks in 1995. This trend is likely to continue due both to the phasing in of European directives65 and the expected consolidation of the sector, which will tend to increase the relative size of banks vis-à-vis their customers (one reason why large exposures were so prevalent in the past is that, because of the fragmentation of the sector, the size of Italian banks in relation to that of nonfinancial firms is lower than in partner countries).

144. The Italian banks largely meet the criteria of capital adequacy laid out by the Basle Committee. Less than 20 banks had shortfalls in respect of the minimum capital requirements. Three quarters of these shortfalls, amounting to Lit 2.4 trillion, were concentrated in the South. Nearly 300 banks had capital in excess of the minimum required; their “excess” capital added up to Lit 63 trillion. The increase in the capitalization of Italian banks in this decade has reflected mainly the ability to attract some fresh capital (in the case of public banks sometimes through the transfer by the government of stocks of companies in which the government has an interest), and the revaluation of assets in the early 1990s, in connection with the incorporation of banks. Retained earnings have played a minor role.

145. Supervisory capital increased by 5 percent in 1995 (to Lit 197 trillion), despite a reduction in the capital of southern banks due to losses not offset by capital injections. Tierone capital amounted to Lit 152 trillion (this figure excludes cooperative credit and central institutions). The amount of subordinated debt has also increased, as incorporated and cooperative banks have made widespread use of this instrument to raise fresh capital. The recapitalization needs of southern banks were also partially met with the subscription of debt.

146. The overall matching of the term structure of assets and liabilities of banks “accepting short-term deposits” has also changed since the new banking law eliminated most restrictions on asset holdings. The new regulations are estimated to have allowed these banks to increase the holdings of long-term assets by up to 60 percent (for a constant structure of liabilities, fixed investment, and stock holdings). Indeed, until 1995 the maturity of assets (spurred in part by a greater supply of government bonds) increased faster than that of liabilities. This imbalance was subsequently partially reversed due to an increase in the demand for medium-term CDs and bank-issued medium-term securities, a contraction in holdings of government bonds, and a slight pick up in short-term credit.

147. Several factors have driven the recent increases in labor costs at the firm level. Among them, the continued payment of bonuses despite the poor performance of banks (i.e., the de facto incorporation of these bonuses to the salary of most workers) and severance payments are noteworthy. Assicredito (1995, 1996) observes that increases at the firm level were higher among the largest and smallest banks, where there has been a reduction in personnel.

148. Banks estimate that about 40 percent of the face value of their bad loans (i.e., Lit 47 trillion) are not recoverable. The resulting shortfall has been provisioned, as required by prudential regulations. Because most provisions were not tax deductible, the amount of taxes paid by banks in 1991–95 remained stable, despite the steady decline in pretax profits in the period (Table 14). Capital losses due to the decline in bond prices were substantial in 1994 (amounting to about half of the total losses), and somewhat smaller in 1995 (as banks improved their hedging positions).

149. The largest banks have been among the least profitable (Table 17), posting losses in both 1994 and 1995. Medium size banks have fared better, especially in the North, where according to rating agencies, improved efficiency has permitted several regional savings banks to continue to make profits on their retail business and other core activities.

Table 17.

Italy: Selected indicators of Banks Classified by Size 1/

(In billions of lira, unless otherwise indicated)

article image
Source: Bank of Italy Annual Report.

Figures are obtained from a sample of 277 banks, which account for 94 percent of the deposits in the system; the number of banks in each category is indicated in parentheses.