Assessing the Damage
October 28, 2009
By Daniele Antonucci | London
Summary and Conclusions
Like other advanced economies, Italy appears to be past the worst of the global recession. But how quickly and to what extent will the country recover? In this report, we focus on Italy's medium-term prospects and on the outlook for the main drivers of potential growth, i.e., labour productivity and the labour force. We reach three main conclusions:
• First, the damage on both labour productivity and the labour force is likely to be substantial. We estimate that Italy's potential growth rate will be negative this year and the next, and remain at 1% between 2011 and 2014 - below our pre-crisis 1.2% estimate.
• Second, this implies a smaller - but still very large - output gap relative to the pre-crisis baseline, pointing to a period of disinflation ahead. Although we don't see outright deflation as a real possibility, we do see mounting disinflationary pressures.
• Third, without structural reforms aimed at raising potential growth, it will be difficult to bring down the public debt to an acceptable level. Based on our simulations, debt/GDP would decline only in our bull-case scenario of growth higher than 3%.
The main takeaway for investors is that extrapolating into the future the pre-crisis growth rate of potential GDP might be too far-fetched. If the recession lowered the pace at which the Italian economy can sustainably expand, as we believe, the rate of growth of aggregate corporate profits, at least over the long term, will be lower too.
Slowly Climbing Out of a Deep Hole
After the slump at the turn of the year, the outlook for the Italian economy has improved. Indeed, the latest data have been encouraging and - with Germany and France having already emerged from recession in 2Q - brought some comfort that economic growth might have come back into positive territory in Italy too, perhaps as soon as in 3Q. For example, business sentiment has picked up considerably from March's record low, firms' assessment of demand is now on an upward trend and the level of inventories looks close to ‘normal'.
The latest hard indicators are positive too, especially on three fronts. First, industrial production rose by a very substantial 7%M in August; even assuming a full payback in September, the chances are that industrial production will rise by about 5%Q in 3Q, quite a turnaround after five quarterly contractions in a row. Second, new car registrations surged over the summer, courtesy of the car scrapping incentives - which are likely to be extended. Third, exports to non-EU countries rose by 17.4%M in September and our forecast for global demand points to a further improvement to come.
Morgan Stanley's proprietary GDP indicator points to clear upside risks for economic activity in the short term. Taken at face value, this indicator suggests that the Italian economy expanded by a very strong 1.2%Q in 3Q - far above the latest published consensus forecast of 0.5%, in our view.
Morgan Stanley's GDP indicator is an econometric tool that provides a timely estimate of GDP growth by synthesising the information contained in several leading and coincident indicators - ranging from industrial production and orders to consumer confidence and the yield curve. Unlike most other econometric tools, this indicator has fully captured the extent of the downturn. Its ability to capture big swings on the downside suggests that it should not fail to do so on the upside too.
The upshot is that we now expect the Italian economy to expand by 1.2% next year compared to our previous forecast of 0.6%. What's more, we predict a milder contraction this year, to the tune of 4.5%, compared to our previous forecast of -5.1% (see Much Better on Bottom-Up, October 14, 2009).
However, this is not a V-shaped recovery. With the labour market still under severe stress, there seems to be little prospect for a significant reduction in households' saving rate to fund their spending. Accordingly, we expect the current period of anaemic private consumption growth to extend well into next year. What's more, although the annual rate of decline in domestic demand should slow in the coming months, an actual increase compared to a year earlier is still some way off.
In all, the short-term outlook appears to have improved, but the economic situation remains precarious: it will take until the end of 2014 for the level of GDP to regain the lost ground and reach the peak of 1Q08 - assuming a rate of growth of 1.2% from 2011 (which we estimate to be Italy's pre-crisis potential growth rate).
What's the Economy's Productive Capacity?
So, the short-term outlook has improved, though Italy will probably lag behind its European neighbours next year. But what do its medium-term prospects look like? Of course, this question is always important, but it is particularly relevant now as the most severe post-war recession has left a deep scar on the economy's productive capacity.
To see why this matters, let's focus on the main traits of a textbook recession: demand falls short of supply, causing high unemployment, as well as falls in consumption, production and investment. But when the recovery arrives, the resources left lying idle are brought back into production, making the economy grow faster than normal until it reaches the point where it would have been without the recession, thus running again at full potential.
However, if the shortfall in demand persists long enough, it can do lasting damage to supply, i.e., to the economy's productive capacity, and reduce the level of potential output or even its rate of growth. When this happens, the economy will never regain its lost ground compared to the baseline where recession had not occurred, even after demand accelerates during the subsequent recovery.
This is crucial for investors because their expectations for a country's economic performance - and hence in aggregate corporate profits - will be disappointed if they simply extrapolate into the future the pre-crisis growth rate of potential GDP. If the recession lowered the pace at which an economy can sustainably expand, the rate of growth of aggregate corporate profits, at least over the long term, will be lower too.
How can a persistent shortfall in demand hurt the economy's productive capacity? This can happen through four channels:
• First, if the period of joblessness is excessively long, the skills of the unemployed will atrophy, partly because of the lack of training on the job. When demand strengthens, they will struggle to find a new job, or a job providing the same earning power. In turn, this may result in an increase in the so-called structural or long-term unemployment.
• Second, if firms perceive that the lower level of activity will last for a long time, they will cut investment aggressively, thus ceasing to add to the capital stock. And they may even start scrapping some of it, either voluntarily or because they go out of business. Either way, the result is the same: the capital stock will shrink.
• Third, an impaired or over-regulated financial system might result in a higher cost of capital - perhaps because of competition for limited funding opportunities or tighter lending standards - encouraging firms to use less capital per unit of output. Accordingly, the economy's productive capacity might be negatively affected.
• Fourth, governments might decide to shield firms and workers from foreign competition through restrictions on foreign trade - such as tariffs on imported goods or restrictive quotas - or on foreign takeovers of local firms. This might delay or complicate the restructuring of the economy or encourage firms to relocate elsewhere.
We will discuss the above-mentioned channels in the following sections. What matters here is that an assessment of their impact is crucial to gauge Italy's medium-term prospects. In the IMF's latest World Economic Outlook, for example, an analysis of the impact of 88 banking crises over the past four decades shows that, seven years after a bust, an economy's level of output was on average almost 10% below where it would have been without the crisis - a huge gap.
Thinking Years Ahead - Not Decades
We have just seen that there are good reasons to think that Italy's potential output might have been hit by the recession, at least on theoretical grounds. Is there any empirical evidence?
To set the stage, it is worth clarifying what sort of timeframe we have in mind. Many people think of potential growth (or the rate at which the economy can expand without creating inflationary pressures) as a very long-term concept.
One simple way to approximate it is to average the rate at which the economy expanded over a period of decades or, alternatively, to use statistical smoothing techniques to remove short-term fluctuations from the GDP data. Another approach is the so-called production function framework, in which output growth is modelled as a function of the factors of production (labour and capital inputs) and technological change (for a discussion on the various methods see, for example, Furceri and Mourougane, 2009; or Cahn and Saint-Guilhem, 2007).
Italy is suffering from chronically low economic growth - regardless of the chosen methodology - by a large body of research (see, for example, the OECD's latest Economic Survey of Italy). In particular, potential growth has steadily declined from approximately 4% in the 1970s to less than 1.5% before the turmoil. In addition, since the start of this decade, Italy has expanded by a mere 1.2% on average, far slower than the euro area average of 2.0% and than its own post-war norm.
However, while this type of analysis does have its own value, what is most relevant at this juncture is where the economy is currently operating relative to its potential productive capacity, and at what pace it can grow over the next few years, say three to five. This is relevant for investors because, even if the economy is past the worst of the global recession, as long as there is slack in the economy, supply bottlenecks don't arise and firms operating in competitive markets do not have much pricing power. In this environment, upward price pressures are muted and policy can remain expansionary without the risk of fuelling inflation.
This cyclical phase is a ‘sweet spot' for risky assets because policy remains at recession-combating levels while investors see looming recovery. We will therefore focus on the next few years rather than on the decades ahead, and discuss whether and how the recession has taken its toll on Italy's productive potential.
One problem with this approach is that it is relatively easy to think of potential growth in terms of long-term developments in factors like demography, technology or education - which are unlikely to make a significant difference over the timeframe considered. However, in trying to assess a country's productive potential in one individual year, or in thinking about the impact of recent events, it is easy to confuse cyclical with structural developments: the former depend on demand-side factors; the latter - which are those that affect productive capacity - depend on supply-side factors.
For example, a decline in foreign and domestic spending has caused a cyclical drop in Italy's fixed investment; but as and when demand recovers, these dynamics will reverse. However, constraints on the supply of credit might have caused firms to cut back their investment by more than the decline in demand alone would imply, thus reducing the economy's potential growth.
Ultimately, it all comes down to the growth rates of labour productivity (or the output produced per worker) and the labour force - the two determinants of potential GDP growth in the medium term. Let's examine each one in turn, and think about how the economic and financial turmoil might have affected them.
Labour Productivity Growth - Further Reduction
Since the start of this decade, the annual growth rate of output per worker has averaged a meagre 0.3% in Italy - down from 0.9% in the 1990s and 1.2% in the 1980s. This is quite low by international standards (in fact, one of the lowest among the advanced countries) and even weaker than the far-from-impressive euro area average of 0.8% during the same period.
Of course, the various labour market reforms - such as the so-called ‘Treu package' or the ‘Biagi law' - that introduced new types of part-time and temporary contracts might explain, at least in part, Italy's poor labour productivity record. With the unemployment rate falling from 11% in 1999 to less than 7% last year, and many of the new part-time and temporary workers employed in some of the least productive segments of the services sector, it is no wonder that output per worker is so low.
From this perspective, low labour productivity might just be symptomatic of an economy in transition. The problem for potential growth arises when growth in both labour productivity and the labour force is low, as is the case in Italy.
But why is labour productivity so low? Contrary to popular belief, Italy's poor productivity growth record does not reflect a lack of capital or training. The rise in the capital stock per worker over the past decade has exceeded that of Germany and Spain. Further, spending on education, as a share of GDP, is only a little below the euro area average.
Italy's disappointing performance seems to reflect tightly regulated product markets, as well as a complex legal system and a slow judiciary system, which can make it difficult to conduct business.
This is not to say that Italy has not implemented important reforms. After all, the OECD's product market regulation indicator has improved markedly over the past ten years, reflecting the fact that liberalisation has continued. In particular, the two so-called ‘Bersani laws' (named after the minister who introduced these new norms) helped to remove obstacles to competition in retail trade and professional services, among others. However, this is clearly not enough to bolster growth. Extending the scope of liberalisation to raise productivity and growth needs to continue. According to the OECD, completing the liberalisation process could increase per capita GDP by 2.5%. Regrettably, a third ‘Bersani law' never saw the light, and the services sector is still too sheltered from competition.
In any case, Italy's structural features are unlikely to change radically over a period of just a few years. Therefore, they are less relevant over the timeframe considered (three to five years). Hence, in what follows we abstract from them and focus instead on three recession-related factors that might hit labour productivity over the next few years: capital scrapping, financial constraints and the potential expansion of the public sector.
1. Capital Scrapping
The first effect that the global financial turmoil could have had on labour productivity relates to capital stock destruction - such as machinery and equipment - as an increasing number of firms struggled to deal with the drop in demand and eventually failed. Bankruptcies in Italy have risen sharply since mid-2007. The likely impact of this surge in corporate failures on the stock of capital is difficult to quantify, but we can get an idea by looking at past episodes of economic weakness.
For example, following the recession in 1993, the annual growth rate of the capital stock dropped from 4.7% to 0.8% over five years. Although bankruptcies in the current downturn are not yet as high as they were in 1993, they are rising very quickly. Indeed, the number of corporate insolvencies was 8,800 last year, up from 6,062 in 2007. The chances are that this number will increase by about 30% this year and a further 15% next year, according to estimates from credit insurance companies such as Euler Hermes.
Of course, there is no one-to-one correspondence between capital scrapping and bankruptcies; if a firm defaults on its debt, it does not necessarily mean that its capital will be scrapped altogether. But the fact is that the current downturn is the deepest since records began in 1970. Therefore, it seems plausible that the annual growth rate of the capital stock will slow even more than in 1993.
In its Regional Economic Outlook, the IMF estimates that the capital stock contributed 1pp to potential GDP growth between 1995 and 2005, over which period it grew by about 1.4% per annum. Assuming that this rate falls to a quarter of its previous level, or about 0.35%, we might expect the effect on potential growth to fall from 1pp to 0.25pp. In other words, this would imply a reduction in potential growth of around three-quarters of a percentage point per annum over the next few years - a substantial decrease.
One note of caution is that the relationship between the rate of growth of the capital stock per worker and GDP is affected by long and varying lags. No two recessions are completely alike. This is crucial because assuming a milder slowdown in the capital stock, the impact on potential growth would be smaller. For example, after the recession in 1975, the growth rate of the capital stock per worker almost halved after five years. But we believe that the mark left by the recent downturn will ultimately be bigger. In all, while there is a wide range of equally plausible outcomes around our assumptions on the likely evolution of the capital stock, the main message remains that the slowdown should be quite significant.
2. Financial Constraints
The second effect on labour productivity relates to the damage that the recession imparted on the financial industry. Clearly, the overall damage to the financial industry is not known yet. But judging from the available information, this will exert both direct and indirect effects on the labour productivity of the whole economy:
• Direct effect - With activity in the financial industry having already fallen sharply and tighter regulation very much on the cards, some businesses will struggle. Therefore, it is unlikely that the financial industry will achieve the strong growth rates of the recent past for some time. Hence, its contribution to the overall economic activity will be smaller.
• Indirect effect - Higher interest rates and tighter lending standards might prevent firms from investing. The average interest rate charged on loans to non-financial corporations is currently at its record low and has declined by 350bp since the beginning of 2008, thus facilitating investment. Firms might therefore struggle once this rate becomes less favourable. In addition, lending standards, at least so far, remain very tight.
Although it is difficult to quantify the impact of the above-mentioned factors with some degree of precision, it is clear that the financial sector's restructuring - perhaps because of regulatory changes - is likely to lead to a reduction in its share of total activity, and to a slowdown in the rate of growth of labour productivity. Naturally, this would prompt productivity growth in the broader economy to fall.
The European Commission estimates that the financial crisis has caused a 1.2pp increase in firms' financing costs, which will peter out only gradually as conditions improve over the next four years or so. This will reduce productivity growth by 0.1pp per year over the next four years, the Commission estimates. However, we suspect that the impact could be even bigger because tighter lending standards are likely to negatively affect the feasibility or timeliness of investment decisions and lead to a further reduction in efficiency.
3. Expanded Public Sector
The third way in which the recession could have damaged productivity growth relates to the effects of an expanded public sector. This could take two different forms: higher public consumption and investment - both through discretionary fiscal spending and the so-called ‘automatic stabilisers' - set off by the recession or interventions to support, for example, the banking sector.
Productivity growth in the public sector has been far lower than in the private sector; at times, it has even been negative. Accordingly, the rising share of government spending in the economy could reduce aggregate productivity growth. What's more, economic theory suggests that the public sector is less efficient in the allocation of capital relative to the private sector.
The impact of an expanded public sector on overall labour productivity is likely to be negligible in Italy, where - unlike in other European countries - the financial and economic downturn has not triggered a significant increase in discretionary public spending. This was partly due to limited margins of fiscal manoeuvre because Italy entered the recession with a debt/GDP ratio in excess of 100%. Furthermore, there was no need to arrange any major bank bailout in Italy.
One reason why Italy could not put more financial resources into productive use is the high interest rate burden. For example, last year Italy's interest payments on its debt amounted to €80.9 billion, or 5.1% of GDP, against a euro area average of about 3%. Having said that, our forecasts for government consumption and investment suggest that the public sector's share of total expenditure will remain broadly unchanged through 2011. In other words, this factor is unlikely to affect labour productivity growth to a great extent.
So, what do the three above-mentioned factors imply for labour productivity growth? The main takeaway is that the recession might have caused the rate of growth of labour productivity to decline outright this year and the next. What's more, the chances are that financial constraints will exert a further modest downward impact through 2014. Hence, we expect very subdued labour productivity growth over the next five years.
Capital stock destruction is likely to be the main driver of lower labour productivity growth, especially over the next two years or so, we think.
Labour Force Growth - Falling into Negative Territory
Apart from labour productivity, the other driver of potential growth is the labour force, which is driven by both short-to-medium term and long-term dynamics. In the context of the timeframe considered (three to five years), the latter are unlikely to change radically. So, in this section we focus on the former and look at three recession-related factors that could hit the labour force over the next few years: migration flows, atrophying professional skills, and a potential reduction in the working-age population that is willing to work.
1. Migration Not a Big Issue
As birth rates have fallen, Italy's working-age population is already in decline and, similar to most other European countries, this trend looks set to worsen.
Of course, increased inward migration could provide at least a partial offset. However, with the region's borders set to remain closed to immigrants from Eastern Europe until 2012, we doubt that this will have any significant effect over the period considered.
Conversely, one recession-related demographic factor that could have negative repercussions for Italy's potential growth over the next three to five years is if immigrants leave the country. This would result in at least a temporary reduction in potential labour force growth. However, as immigrants sometimes work in less-productive occupations than natives, the pick-up in average productivity might offset the negative impact on the labour force from their departure.
2. Unemployment Boosted by Atrophying Skills
Those who have lost their jobs might see their skills become obsolete, making them unable to find employment even once the recovery begins. This would cause a one-off increase in the rate of unemployment that is consistent with stable inflation (known as the natural or structural rate of unemployment, NAIRU).
Italy's unemployment is rising, with 347,000 jobs lost since the trough in unemployment in 2Q07. This is not a very large number compared to the job losses in the other major European countries, but will translate, at least in part, into a pick-up in long-term unemployment and the NAIRU. Indeed, a typically European feature of the labour market is its inertia: after an increase in unemployment, not everyone finds a new job when the cycle turns, thus putting a floor on structural unemployment.
In other words - as the OECD shows in its latest Economic Outlook - following severe downturns in the major European economies over recent decades, even once output has returned to potential, there has been a rise in unemployment typically proportional to the severity of the recession.
For example, we illustrate the increase in the unemployment rate from the quarter when the output gap was closest to zero prior to a severe downturn, to the quarter when the output gap was again closest to zero following the downturn (only downturns where the cumulative annual output gap exceeded two percentage points were considered). The main takeaway is that, after a major downturn, unemployment is generally higher in Europe even after the subsequent recovery. This does not seem to be the case in the US.
Such effects could arise because workers that remain unemployed for a long period may become less attractive to employers due to declining human capital or because the intensity of their job search diminishes. Hence, the long-term unemployed put less downward pressure on wages and inflation, thus contributing to the persistence of unemployment.
The evidence from past economic downturns in Italy is consistent with this view. Our illustration shows that unemployment remained high for a very long time after the recession in the early 1990s.
In other words, not only has actual unemployment increased, but the natural rate of unemployment has trended higher as well. The OECD's estimate of the NAIRU rose by around 0.8pp in the three years following the recession in 1993.
To what extent structural unemployment will rise is of course an open question. This downturn is deeper, though the reforms of recent years may have increased the flexibility of the labour market. This suggests that it might be easier to get people back into work as and when the economy recovers. On balance, we think that the NAIRU might rise by 0.7pp this year. Moreover, the downward trend in the NAIRU has probably already slowed last year and is unlikely to resume in 2010.
3. Reduced Willingness to Work
The third way in which the recession could have hit labour force growth is if the participation rate (or the share of the working-age population that is willing to work) has fallen. Italy's participation rate has been on a steady upward trend since the inception of the euro area, partly reflecting the impact of various labour market reforms. However, despite this appreciable increase, the gap with the euro area average remains broadly unchanged.
What's more, this trend has probably stalled, perhaps because of potential workers' disillusionment about the state of the economy. In addition, some workers might even have opted for, or been forced into, early retirement. There is a chance that the participation rate will stagnate this year and the next. This might have knocked around 0.1pp off potential labour force growth in 2008 and 0.5pp in the subsequent two years.
The upshot is that labour force growth will likely be negative in 2009 and flat in 2010, courtesy of an increase in structural unemployment and a drop in the participation rate - but it might be higher than would otherwise be the case from 2011, as the fall in the participation rate is reversed. In the report, we summarise our view on the likely impact of the recession on Italy's labour force growth and show how each of the three factors just discussed affects the pre-crisis baseline scenario.
To Sum Up
Bringing together all the factors that we analysed separately in the previous sections, the main takeaway is that potential output will shrink this year, courtesy of an outright fall in the rate of growth of both labour productivity and the labour force. Meanwhile, the potential growth rate could remain in the red in 2010 and rebound to a modest 0.8% the following year. After that, we expect Italy's potential growth to accelerate gradually.
Implications for Inflation - Mind the Output Gap
Significant slack pointing to negative core inflation? So, what does this mean for the inflation outlook? In 1Q08, when the Italian GDP reached its peak, the economy was operating beyond its capacity, with the so-called output gap well into positive territory. But GDP dropped by 6.5% since then. If potential growth had remained at about 1.2%, spare capacity would have surged to about 6.3% of GDP by now. And our forecasts for GDP to drop by 4.5% this year, rise by just 1.2% in 2010 and expand at around the same pace in 2011 would leave the output gap close to 8% of GDP.
This would have profound implications for price pressures in the economy. Taken at face value, the relationship suggests that core inflation might drop by as much as three full percentage points from the current rate of +1.9% to -1.1% at the end of 2011. In addition, with spare capacity failing to close, core inflation might just keep falling and falling.
Not if Productive Capacity Has Been Damaged!
But if the economy's productive capacity has been damaged by the recession, the output gap might not be as big as these calculations suggest, implying weaker deflationary pressures. And if the factors limiting productive potential persist, any spare capacity will be worked off more quickly than would otherwise be the case.
Of course, there is a lot of uncertainty surrounding any estimate of the output gap. After all, this variable is unobservable and there is a lot of ‘guesswork' involved. Bearing these caveats in mind, our finding suggests that the post-recession output gap might be consistent with core inflation roughly at zero - not negative. This means that Italy is likely to go through a prolonged period of disinflation, but will escape outright deflation. Still, the amount of slack in the economy affects the change in inflation; so, as long as the output gap persists, inflation will continue to fall.
Of course, different assumptions would have led to a different forecast for core inflation. For example, labour productivity growth might surprise on the upside in the near term because mounting job losses might boost the rate of growth of output per worker, all else being equal. In addition, the gyrations in commodity prices are always the wildcard for the (headline) inflation forecast.
Implications for Public Finances - How Sustainable?
Lower potential growth has implications for the sustainability of public finances too. As an illustrative example, we focus on the evolution of one key variable: the debt/GDP ratio. The behaviour of this ratio is driven by the dynamics of the numerator (the stock of public debt) and the denominator (nominal GDP). The time path of the stock of debt D is given by the standard debt equation:
D(t) = -[T(t) - G(t)] + [1+i(t)] x D(t-1) + a(t) 
where D(t) is the stock of debt at time t, i(t) is the interest rate paid on this debt, T(t) - G(t) is the primary surplus - defined as government/tax receipts T minus government spending (excluding interest payments) G - and a captures one-off factors, e.g., privatisations and asset sales. Similarly, the dynamic of nominal GDP (NGDP) could be modelled using the following equation:
NGDP(t) = [1+n(t)] x NGDP(t-1) 
where n is the rate of nominal GDP growth (the sum of real GDP growth and the GDP deflator growth).
In the absence of any contribution from the primary balance (i.e., T(t) - G(t) = 0) and one-off factors (i.e., a(t)=0), we can see from equations  and  that the debt over GDP ratio would be driven by the difference between the rate of (nominal) GDP growth n and the implied interest rate paid on the public debt i.
In this case, if n is larger than i, then the ratio of debt over GDP will fall over time, and vice versa when i is larger than n. Overall, the interest rate i(t) that the government pays on its debt seems to be rather stable in Italy, and is relatively close to 5%. Hence, one way to ensure a downward path in the debt over GDP ratio - all else being equal - would be to maintain the growth rate of nominal GDP above 5% in the long term.
With GDP deflator growth averaging 2.5% since 1999, the rate of growth of real GDP would need to remain above 2.5% so that n remains larger than i. Unfortunately, Italy's economic growth has been far more sluggish than required. The good news, however, was that the primary balance (i.e., government borrowing or lending excluding interest payments) was in surplus, providing a partial offset. But with the primary balance likely to be negative this year and in equilibrium at best next year, the fiscal outlook will be more challenging.
We illustrate the three different scenarios for the debt/GDP ratio using equations  and . The starting point for the debt/GDP ratio is equal to the figure for 2008. After that, we project the debt/GDP ratio through 2050, assuming a constant real interest rate of 2.5%:
Base case - potential growth in line with pre-crisis experience (real GDP growth = 1.2%, inflation = 2.5% and real interest rate = 2.5%. This implies nominal GDP growth = 3.7% and nominal interest rate = 5.0%).
Bull case - potential growth boosted by structural reforms (real GDP growth = 3.0%, inflation = 3.0% and real interest rate = 2.5%. This implies nominal GDP growth = 5.5% and nominal interest rate = 5.5%).
Bear case - potential growth permanently damaged by recession (real GDP growth = 0.5%, inflation = 1.0% and real interest rate = 2.5%. This implies nominal GDP growth = 1.5% and nominal interest rate = 3.5%).
Our study shows that the impact on the debt/GDP ratio of different paths for potential growth and inflation are quite remarkable.
In the base case, Italy's public finances would continue to deteriorate and the debt/GDP ratio would increase from 105.8% in 2008 to 118.8% in 2050. The increase would be much bigger in the bear case, with the ratio climbing to 147.9%. In the bull case, the ratio would stabilise and eventually start declining - albeit at a very moderate pace. Risks to these projections run in both directions. Clearly, a return to a positive primary surplus will help to stabilise the debt burden. However, ageing-related expenditure might increase the ratio. For example, in its recent Sustainability Report, the European Commission simulates the path of the ratio for all the European Union member countries through 2060. By then, the ratio would reach 205.9% in Italy.
However, as alarming as this might sound, most other countries would be in a much worse fiscal position. For example, by 2060 the debt/GDP ratio would be 318.9% in Germany, 431.3% in France and 766.6% in Spain. Indeed, the European Commission's assessment shows that the long-term sustainability risk to Italy's public finances is medium - the same category for Germany and France. Conversely, the long-term sustainability risk to the Irish, Spanish, Greek and Dutch public finances is high.
Naturally, these projections are highly uncertain. It is unlikely that bond markets would keep financing government debt amounting to a multiple of the GDP of the respective countries - or that governments would maintain their policies unchanged in the presence of ever-increasing debts. Still, these simulations highlight the need for deep structural reforms to bring potentially exploding public finances under control.
In all, the importance for Italy to push ahead with supply-side reforms - which could lead to an increase in potential growth - should be quite clear. This would ensure that the rate of GDP growth, n, remains consistently above that of the interest rate that Italy has to pay on its debt, i. Of course, the ‘growth dividends' generated by a strengthened economy will take time to emerge even under a swift reform agenda. In the meantime, the bond market will keep Italy under the spotlight.
In all, there are three main takeaways for financial markets:
• First, extrapolating into the future the pre-crisis growth rate of potential GDP might be too far-fetched. If the recession lowered the pace at which the Italian economy can sustainably expand, as we believe, the rate of growth of aggregate corporate profits, at least over the long term, will be lower too.
• Second, Italy is likely to go through a prolonged period of disinflation but will escape outright deflation. Indeed, if its productive capacity has been damaged by the recession, the output gap might not be as big as would otherwise be the case, implying weaker deflationary pressures.
• Third, the country needs deep structural reforms in order to bring down its debt/GDP ratio. Only by increasing substantially its potential growth rate will the country ensure that the rate of nominal GDP growth remains consistently above that of the interest rate that Italy has to pay on its debt.
Fedspeak: Roadmap for the Exit
October 28, 2009
By Richard Berner | New York & David Greenlaw | San Paulo
The FOMC next meets on November 3-4. In anticipation, market attention is riveted on every nuance of Fed language and actions. That's hardly surprising, as the case for today's ‘sweet spot' in risky assets rests on short rates staying anchored, liquidity remaining abundant and prospects improving for growth, however modest. Fed officials won't change policy any time soon. But they know that effective communication is more critical than ever as they craft the exit strategy from an ultra-accommodative stance. Market participants want the Fed to clarify its views about the outlook and about the circumstances in which the Fed will unwind its stance. That will help to reduce uncertainty about the implications of the exit process for financial markets.
By reducing uncertainty, Fed officials can actually improve the effectiveness of monetary policy. Three areas are critical: First, officials can update the baseline outlook and risks to it, and how they will react to changes. Second, they can map out their exit game plan, continuing to identify the tools they will use and the sequence for deploying them. Third, they can say what they do and do not know about the exit process, given that the Fed and market participants are in uncharted waters.
Outlook risks. As noted in the minutes from the September FOMC meeting, the Fed's baseline outlook has probably improved since late June, when the central tendency for growth over the four quarters of 2010 was a range of 2.1-3.3%, and the Fed projected core inflation to run about 1.5%. The revisions will probably reflect generally improving incoming data and a further easing in financial conditions.
Despite our conviction that the recovery will be sufficiently strong as well as sustainable, we'd be the first to admit that both are uncertain. Underlying vehicle and housing demand remains unclear following the expiration of ‘cash-for-clunkers' and the first-time homebuyer tax credit. With payrolls still declining and income growth weak, consumer spending strength is in doubt. And contributions to growth from capex, net exports and the government are uncertain. There is even more dispersion around the inflation outlook: Economic slack is unprecedented, while monetary stimulus, rising commodity prices and a weaker dollar might boost inflation expectations and inflation. Even the extent to which inflation has fallen is unclear: The Fed's preferred inflation gauge - the core PCE price index - rose only 1.3% in the year ended in August, but removing a sharp decline in the so-called non=market component of PCE prices yields a rate of 1.7%.
Clarifying the reaction function. Given that uncertainty, clarity on how officials will react to changes in the outlook will help market participants understand the roadmap for policy. Many will expect the Fed to tighten sooner if it boosts its growth outlook. However, even if its revised outlook for growth improves to match ours, with ‘core' inflation low and declining, we think that officials will keep policy accommodative through mid-2010. An ongoing improvement in financial conditions may promote a gradual unwinding of the quantitative/credit easing that the Fed implemented to offset the credit crunch.
Nonetheless, the current circumstances suggest that the existing guidance on interest rates - "economic conditions are likely to warrant exceptionally low levels of the federal funds rate for an extended period" - is probably due for a change. A softening of the language at either the November or December FOMC meetings would be entirely consistent with our expectation that the first phase of the exit strategy (whether you call it reining in excess reserves, shrinking the balance sheet, or unwinding QE) will commence in 1H10 and that rate hikes will follow in 2H. Unlike in 2004, this language is highly conditional on economic conditions. By spelling out how much economic conditions have changed and what will be the response, the officials can help market participants understand the Fed's preferences and reactions. The "exceptionally low" language describes the stance of policy rather than its direction, so it is conceivable that it could still be applied to a 1% or even 2% funds rate target - especially with some small tweaks to the wording.
Tools to use. While there appears to be some disagreement regarding the timing and the triggers for exit, all Fed officials seem to agree that the exit strategy has two basic components: 1) shrinking the volume of excess reserves and 2) raising the policy rate. Reverse repurchase agreements (RRPs) will be used to address the former. But there are several unanswered questions: How and when will the Fed use reverse RPs to drain excess reserves, given the massive size of the job? Will it need other tools such as accepting term deposits or selling assets? Will paying interest on reserves give it control over the funds rate?
Here is some background. Thanks to aggressive expansion by the Fed, excess reserves - the surplus held by banks at the Fed over required reserves - currently total US$987 billion. We estimate that the excess will grow to about US$1.2 trillion by early 2010. The expected increase reflects three factors: 1) the ongoing impact of the large-scale asset purchases (LSAPs); 2) a looming slowdown in the pace of unwinding other Fed liquidity support facilities (the unwinding has provided significant offset to the LSAPs to this point but simply does not have much more room to go); and 3) the winding down of the Treasury's Supplementary Financing Program (SFP). Thus, far from exiting any time soon, the Fed will actually be adding more quantitative easing unless it takes offsetting action.
This means that an important near-term issue is whether the Fed passively accepts more QE or attempts to offset the growth that should soon appear in the balance sheet and bank reserve data reports. The hawks on the FOMC would appear to have a powerful argument if they want to push the issue: "OK, we understand the rest of you don't want to exit yet, but if the recession is over, why are we doing the opposite of exiting?" The counter argument from the doves will be that draining operations might send a confusing signal to financial markets, and as long as the reserves are merely being stockpiled in cash accounts at the banks, what does it matter if more are added? Indeed, this seemed to be the prevailing sentiment at the September FOMC meeting, according to the recently released minutes. Thus, while it will be interesting to watch this debate play out in coming weeks as excess reserve balances soar to new highs, we don't sense that the Fed is about to start draining reserves. At this point, our best guess is that draining operations won't commence until early 2010.
Whenever it does decide to start draining, the Fed will quickly enter uncharted territory. Although officials point out that reverse RPs are a standard component of the Fed's toolkit, the largest previous operations ever conducted with primary dealers were on the order of US$25 billion - and that was only for a relatively brief interval last autumn when the Fed was still attempting to sterilize the impact of all the new liquidity support facilities. Communications are critical here too: For example, the Fed has been conducting test reverse RPs, and to insure that investors understand the difference between policy and operational testing, the New York Fed issued a clarifying press release.
Fed officials have actually outlined three approaches to draining reserves: reverse RPs, term deposit accounts and asset sales. We doubt that the Fed intends to sell assets any time soon, since such action would probably be quite disruptive to markets - this is more of a long-run option. Moreover, while offering term deposits would appear to be a useful means of draining significant volumes of reserves, this is an untested innovation that carries some legal and technical complexities. The Fed seems to be moving forward most aggressively with the reverse RP option. Initially, there was some concern that dealer capacity for large tri-party reverse RPs was quite limited due to balance sheet constraints, so the Fed signaled that it might have to conduct operations directly with large investors, such as money market funds. However, it now appears that dealers may be able to absorb a much larger volume of operations than previously believed, which would avoid the need for the Fed to deal with a new set of counterparties. Since this matter is expected to be a topic of discussion at the upcoming FOMC meeting, we are likely to get additional clarity on the technical aspects of how reverse RP operations fit into the overall exit strategy following that session.
Finally, looking further ahead, the Fed has another tool - interest on excess reserves (IOER) - that, if effective, would enable officials to raise the policy rate without significant reserve draining. If the Fed hikes that rate, it raises the cost of borrowing; when banks can hold excess reserves on deposit with the Fed, they won't lend at rates below the IOER. Yet there are some institutions (in particular, the GSEs) that are not authorized to receive IOER, so there are leakages in the system. And at very low rates, IOER may not function properly. Therefore, the open market desk at the NY Fed may have to use the other tools noted above to execute the tightening as instructed by the FOMC.
Uncertain impact of exit on markets. Beyond when exit will start and how it will work, there are several questions about the impact of such policies on financial markets. Will banks deploy their cash assets into securities or loans, reducing market rates and expanding credit supply? Will the interest rate on excess reserves become the policy rate, at least for a while? What will be the impact on market funding rates of conducting large reverse RP operations?
The September FOMC minutes note that the Fed staff has examined the impact of very high reserve balances for bank balance sheet management and the economy. In tandem with the substantial volume of excess reserves, large banks that report weekly hold nearly US$1.2 trillion in cash assets on their balance sheets, or more than double the year-ago level. The staff believes that as banks grow more comfortable with the economic outlook, those now holding these balances for liquidity-management purposes could redeploy them into securities or loans. That shift in the asset mix would narrow spreads or increase credit availability; instead of boosting bank liabilities and the ‘money' multiplier, this would increase monetary stimulus through the asset side of banks' balance sheets.
Finally, we suspect that large reverse RP operations could put upward pressure on financing rates - with general collateral RP rates possibly trading well above fed funds. We will be exploring this issue in greater detail over the course of coming days because it opens the door to the possibility that the fed funds rate might lose its status as the main barometer of monetary policy.
Political constraints? Effective Fed communications will also help the FOMC navigate the current political environment. There is a deeply held concern that the Fed will be unable to execute an exit strategy due to political constraints. That's not surprising, given that criticism of the Fed is at its highest level in 70 years, according to some historians. And there will always be some politicians who criticize the Fed when it embarks on a tightening campaign.
Nonetheless, we believe that such concerns are overblown. Since the days of Arthur Burns, the Fed has largely managed to run policy independently, and there is a new and very powerful argument in favor of independence. Since many place the blame for the financial and economic crisis of recent years squarely on an overly lax monetary policy coming out of the last recession, it's going to be especially difficult to criticize the Fed for exiting this time around. In addition, Chairman Bernanke's campaign to communicate directly with the public at large has strengthened the Fed's hand and probably contributed to President Obama's decision to nominate him for a second term.
In the end, it's certainly conceivable that the Fed could delay exit longer than it should (just as it is possible that it moves too soon). But we are convinced that if the Fed was to wait too long, it would be because it misread the economy, not because of political influences.