Hungary: Selected Issues
Author: International Monetary Fund
Publication Date:
26 Jul 2007
eISBN:
9781451818086
Language:
English
Keywords:
ISCR; CR; Hungary; enterprise; consumption growth; income hypothesis; consumption behavior; income shock; log consumption; income growth; Productivity; Exports
Download PDF (420.1 KB)

This Selected Issues paper on Hungary reports that the public enterprises may pose significant fiscal risks on account of their quasi-fiscal activities and contingent liabilities. More than 85 percent of the economy is in private hands. According to the Privatization Act, assets may remain in long-term state ownership if they belong to a national public utility provider or are considered to be of strategic importance for the national economy or defense. Capital-intensive and labor-intensive enterprises remain as state property.

Abstract

This Selected Issues paper on Hungary reports that the public enterprises may pose significant fiscal risks on account of their quasi-fiscal activities and contingent liabilities. More than 85 percent of the economy is in private hands. According to the Privatization Act, assets may remain in long-term state ownership if they belong to a national public utility provider or are considered to be of strategic importance for the national economy or defense. Capital-intensive and labor-intensive enterprises remain as state property.

II. Could Hungary’s Growth Deceleration Persist? Inferring Productivity Trends from Consumption Volatility18

A. Introduction

32. Recent research differentiates business cycles in emerging and developed markets.19 The key question raised in this research is the nature of shocks, which, in turn, leads either to cyclical or trend reactions. Shocks, though not directly observable, cause fluctuations in consumption, income, investment and trade balance. Importantly, the size and permanence of the fluctuations differ in different types of economies. In emerging economies, the shocks to income itself tend to be persistent, leading to volatility in income growth that is twice that in developed markets. Consumption is even more volatile than output, which leads also to sizeable changes in imports and hence to a deterioration of the trade balance during booms.

33. The higher volatility of consumption in emerging markets has been interpreted as implying that consumers view income shocks to be of a relatively “permanent” nature. As such, while consumers maintain (or smooth) consumption in the face of a shock, they also adjust their consumption levels to the new information revealed by the shock. Similarly, investment and net exports shift in response to anticipated future output.

34. In this paper, the finding is that Hungary has features of developed and emerging markets. First, Hungarian consumers behave in much the same manner as consumers elsewhere. They respond to transitory income changes but are also forward-looking and change their consumption behavior when the future outlook changes. This behavior is consistent with the permanent income hypothesis, where “permanent” is a horizon that may be 3-4 years long (see Carroll 2001). Second, Hungary’s income or output volatility is low, and is comparable to that of advanced countries. Third, however, its consumption volatility is relatively high and is particularly high in relation to its income volatility. In combination then with the finding that Hungarian consumers are forward-looking, the implication is that the high consumption volatility is a response to shocks that have a relatively permanent character. It is in this sense that Hungary is most like an emerging market: shocks tend to have long-lasting effects on income and output growth. Finally, the trade balance is countercyclical and net export volatility is relatively high, in line with that of other emerging markets.

35. This paper is motivated in part by the slowdown in Hungary’s GDP growth relative to regional peers since 2005, a slowdown that turned more emphatic in 2006. While the 2006 deceleration is related to an ongoing fiscal adjustment, the objective of this paper is to understand the extent to which the recent slowdown in Hungarian growth is likely to persist reflecting a more permanent negative shock to productivity growth.

36. There has, at the same time, been a sharp deceleration in Hungary’s consumption growth. As noted, consumption growth can be particularly informative in gauging the perceptions of consumers about the future. The literature on permanent income hypothesis and precautionary savings represents two converging strands explaining consumption behavior. Empirical evidence on determinants of consumption growth shows that income uncertainty plays a role besides current income growth (Carroll, 1992). In other words, in response to a negative shock to output growth, consumers adjust their consumption growth downwards not only in response to the current lower growth in income but also to negative perceptions about the future, in the belief that the current conditions are going to persist.

37. Because the relatively short time series makes it difficult to precisely estimate the permanent and transitory components of productivity, Aguiar and Gopinath (2007) have proposed using the consumption volatility (and other moments such as the correlation between net exports and output) to infer the relative importance of permanent shocks. We repeat their exercise for a large number of countries, including a number of new members of the European Union. Given the high ratio of consumption to income volatility, the implication is that the permanent component of productivity shocks is relatively high in Hungary (between two-thirds and 100 percent of the shock tends to be permanent), which is significantly higher than in the Czech Republic or Poland. Consumption volatility, in turn, is associated with net-exports volatility among the countries—such association could imply that the source of permanent shocks could be related to terms-of-trade shocks arising from shocks to external competitiveness. Unless such structural shocks are corrected by policy initiatives, the low Hungarian output growth could persist in the near term.

38. The rest of this paper is organized as follows. An overview of Hungary’s recent growth experience is followed by an analysis of the determinants of its consumption growth. After a brief summary of the recent analytical procedures for identifying productivity processes from moments of macro aggregates such as consumption and net exports, some stylized facts about business cycle moments are shown for Hungary and other countries. Finally, the paper provides a quantitative measure of the persistence of productivity shocks.

B. Hungary’s Recent Growth Performance and the Role of Consumption

GDP growth and consumption dynamics

39. It is useful to consider three growth phases in Hungary since 2000. From 2000-2003, Hungary grew relatively strongly (above the average of Czech Republic, Poland, and Slovak Republic or the CE3). Growth slowed down in 2003 and 2004 but remained at about the level of regional peers. Since 2005Q2, Hungary’s growth rate has increasingly fallen behind, even as the other new member states and the Euro Area accelerated (Figure 1).

Figure 1.
Figure 1.

GDP Growth, 2000-06 1/

(year-on-year percent change)

Citation: IMF Staff Country Reports 2007, 251; 10.5089/9781451818086.002.A002

1/ CE3 refers to Czech Republic, Poland and Slovak Republic; Baltic refers to Estonia, Latvia and Lithuania.

40. In each of these three growth phases, the role of private consumption has been important and informative (Figure 2). In the first phase, consumption growth followed in the wake of surging real wage growth and buoyant consumer confidence about low future unemployment (Figure 3). In the second phase, falling growth in real wages pulled down consumption growth, even though investment growth strengthened and expectations about unemployment improved between 2003Q3 and 2005Q1.

Figure 2.
Figure 2.

Hungary: Contributions to GDP Growth, 2000-06 1/

(Year-on-year, in percent)

Citation: IMF Staff Country Reports 2007, 251; 10.5089/9781451818086.002.A002

Source: Hungarian Central Statistical Office.1/ The revised methodology for the calculation of GDP, in which the aggregation is done based on weights of the previous year, rather than on the weights of the base year, implies that the sum of the components does not exactly match GDP (see Box 1-1 of the Quarterly Report on Inflation (Update), Magyar Nemzeti Bank, February 2007)
Figure 3.
Figure 3.

Consumption and Real Wages 1/

Citation: IMF Staff Country Reports 2007, 251; 10.5089/9781451818086.002.A002

1/ Annual percent change in gross real wage and actual final consumption expenditure by households.

41. The third phase saw a decreasing contribution of consumption growth to overall growth. Household consumption growth fell from 3.8 percent in 2005 to 1.2 percent in 2006, a much larger shift than the change in real GDP growth from 4.2 percent to 3.9 percent over the same period. By end-2006, annualized consumption had stopped growing. This decrease was associated with falling growth in real wages and a markedly deteriorating consumer confidence about future employment prospects. With falling investment and a depreciating exchange rate in mid-2006, the trade surplus increased.

Role of consumption

42. The literature explaining consumption behavior is made up of two strands. The first is the permanent income hypothesis, which implies that current spending is determined by “permanent income”, or the “expected level of income in the very near term” (Friedman, 1957, 1963, Carroll, 2001). Distant future labor income is uncertain and it is difficult to borrow against such income due to capital market imperfections. Permanent income hypothesis implied that the marginal propensity to consume out of transitory shocks is about a third, and not close to 1 as was assumed in Keynesian models. The second strand is the precautionary savings motive that says that impatient consumers will save to build up a “buffer stock” of wealth to tide future income uncertainty (Carroll, 1992). If actual cash on hand is below the ‘target’ wealth, precautionary savings motive will outweigh impatience and the consumer will try to build wealth back toward the target.

43. If there is uncertainty in future labor income of impatient consumers, the behaviors of consumers under permanent income hypothesis and precautionary savings motive are indistinguishable. In fact, Friedman had “acknowledged the importance of precautionary motives induced by uncertainty of labor income” (Carroll, 2001). Precautionary savings and liquidity constraints are, in turn, connected: constrained consumers have the same behavior as unconstrained consumers with a precautionary motive—in the first, consumers are refused credit and in the second, consumers have a self-imposed reluctance to borrow. If consumers are in either of these two scenarios then consumption growth can be strongly tied to current income growth.

44. Econometric evidence supports the role of uncertainty in determining consumption growth. For example, Campbell and Mankiw (1989) find future income uncertainty playing a role in current consumption growth with the latter connected to current income growth. The test on the role played by uncertainty is based on a regression of consumption growth on income growth and unemployment expectations. Carroll (1992) attributes the persistently low consumption growth in the United States during the 1990-91 recession to a higher probability of unemployment in the future.

45. Econometric analysis shows that consumption growth in Hungary is affected by future uncertainty as well as by current income growth. A regression similar to Campbell and Mankiw’s shows that consumption growth reacts to real wage growth and to unemployment expectations (Table 1).20 Consumption growth is regressed (and estimated by Ordinary Least Squares) on two lags, real wage growth and expectations about future unemployment. A decrease in the real wage growth by 1 percentage point would decrease consumption growth by 0.27 percentage point on impact, and (to the extent consumption growth is smoothed) by 0.55 percentage point over time. Deterioration of expectations about employment prospects would also have an adverse effect on current consumption. A 10 percentage point increase in unemployment expectations would decrease consumption growth by 0.8 of a percentage point on impact and by 1.6 percentage points over time.

Table 1.

Determinants of Consumption Growth, 1999Q2 to 2006Q4 1/

(Dependent Variable: Consumption Growth, in annual percent change)

article image

Ordinary Least Squares estimates, 1999Q2—2006Q4, with Newey-West standard errors, ** (+) implies significance at 1 percent (10 percent).

46. The analysis shows that Hungarian consumers are forward-looking and adjust their consumption to perceptions about the future. In particular, if there is a sense that the current conditions would persist at least for another year, consumers hold back on consumption growth. To the extent labor income uncertainty has increased, consumption growth has decreased to build up a higher target wealth as a buffer stock or an insurance against such uncertainty, if we believe in a world with precautionary savings. Consumption growth slowed less than real wages, but more than real GDP growth did. The overall effect is an average of that of various types of consumers in the economy—some are rule-of-thumb ones consuming their income every period and some are able to borrow (from banks) to smooth consumption. The continuing strength of growth in household credit is evidence of such smoothing; however, credit growth did not accelerate in 2006 as in the previous years.

47. The current slowdown in consumption growth could, therefore, reflect an underlying slowdown in permanent income growth. Consumption behavior driven by permanent income hypothesis in a world with uncertainty or by precautionary savings behavior seems to point to one observable characteristic—a perception of the current negative shock persisting into the future—that is making consumers wary.

48. The extent of persistence of negative shocks can be identified from correlations and volatilities of key macroeconomic aggregates. Aguiar and Gopinath (2007) employ a methodology that uses a real business cycle (RBC) model and matches its implications on correlations and volatilities of key variables with their empirical counterparts to extract the underlying productivity parameters. Estimates of the underlying productivity process would throw light on the persistence of negative shocks in Hungary.

49. The remaining sections describe the methodology, compare volatilities and correlations in Hungarian macroeconomic data with those in other countries, and quantify the extent to which output is driven by a volatile productivity trend.

Figure 4.
Figure 4.

Consumption and Unemployment Expectations 1/

Citation: IMF Staff Country Reports 2007, 251; 10.5089/9781451818086.002.A002

1/ Consumers’ expectations about unemployment in the next 12 months indicates the difference, in percentage points, between the percent of survey responders who expect unemployment rate to increase and those who expect it to decrease in the next 12 months.

C. Identification of Productivity Shocks in the Recent Literature

50. Given the short time series in emerging markets, a trend-cycle decomposition of productivity shocks is likely to be imprecisely estimated. Recent research suggests an alternative approach to understanding business cycles in these countries. Using a theoretical RBC model in which output is driven by both transitory and permanent technology shocks, Aguiar and Gopinath (2007, AG henceforth) derive business cycle features of both emerging and developed small open economies. These features are in terms of moments within business cycle frequencies: relative volatilities of consumption and income, the volatility of trade balance, and the correlation of the trade balance with income, among others. The theoretical moments are driven by the relative volatilities of the permanent and the transitory components of the productivity process, and these relative volatilities are different between emerging and developed countries.

51. In emerging markets, the permanent component of productivity is much more volatile than the transitory component, rendering the trend to be more volatile than that in developed markets. Shocks to trend growth are therefore the primary source of fluctuations in these countries. Accordingly, optimizing agents respond to income shocks depending upon whether they believe the shock to persist. If the economy is hit with a negative income shock, and agents believe that there is an even larger (negative) effect on future output, then consumption responds more than income, increasing savings and reducing the trade deficit. However, if the shock is believed to be transitory, then savings will decrease and the trade deficit will reduce by a smaller amount. In the data, if there is a large response of consumption to income accompanied by a large change in net exports, then the standard business cycle model will identify the underlying shock as a change in trend.

52. The relative volatilities of transitory and permanent shocks are related to the theoretical moments of macroeconomic aggregates in the business cycle models. There are two types of shocks to productivity—shocks to its level and shocks to its growth.21 To see the relationship between the relative importance of these two shocks and macroeconomic moments, the sensitivity of theoretical moments to assumptions about the relative volatility of these two types of shocks is replicated from AG (Figure 9). An increase in the relative variance of the trend shocks is positively related to the volatility of consumption, investment and net exports relative to output at business cycle frequencies. Furthermore, correlations of filtered consumption, investment, and net exports with filtered income show that the one with net exports is most sensitive to the relative importance of shock to productivity growth.

D. Some Stylized Facts About Business Cycle Moments in Emerging Markets

53. This section looks at a set of stylized facts about the behavior of macroeconomic aggregates from emerging and developed country clusters. In particular, it locates Hungarian business cycle moments and compares them to those in other emerging and developed markets. The analysis follows the methodology in AG to extract moments from seasonally adjusted data on real consumption, income, investment, and net exports. Keeping the developed country samples from AG, we take a larger set of emerging markets than in AG. National accounts data for Hungary spans 1995Q1 to 2006Q4.22

Stylized facts

(1) Emerging markets tend to have higher volatility of cyclical output and output growth than developed markets (Table 2 and Figure 5). The lower volatility of output comes either from lower incidence of shocks or better management of monetary and fiscal policies in developed countries. The clusters for developed and European emerging markets are close to each other, with a few exceptions. Some developed countries are highly susceptible to commodity price shocks—Norway and New Zealand are such cases—that render a relatively higher output volatility. In emerging Europe, domestic policies may not be as well managed or constrained because of currency boards or a high level of euroization.

Table 2.

Emerging and Developed Markets Moments 1/

article image

The series for each country are deseasonalized using the x12 command in Eviews. The income (Y), consumption (C) and investment (I) series were logged and HP-filtered using a smoothing parameter of 1600. Net exports to income (NX/Y) were HP-filtered similarly. For growth rates, the unfiltered series were used. A ‘σ’ denotes standard deviation, and a ‘ρ’ denotes correlation coefficient. For the correlations, ρ(X,X’) denotes the first autocorrelation coefficient, where X’ denotes one-lag of X.

Figure 5.
Figure 5.

Relative Volatilities of Output, Output Growth, and Consumption 1/

Citation: IMF Staff Country Reports 2007, 251; 10.5089/9781451818086.002.A002

1/ “Sigma” denotes the standard-deviation of macroeconomic aggregates: C (log consumption), Y (log GDP), NX/Y (Net Exports to GDP), all at business cycle frequencies or HP-filtered from quarterly data. “Cor” denotes correlation coefficient. See Appendix for country and region codes.

(2) Emerging markets have higher volatility of consumption relative to that of output (Table 2 and Figure 5). The presumption is that when a shock to output is thought to be more permanent (leading to changes in future output), consumers respond to it by adjusting consumption much more than changes in current output, according to permanent income hypothesis. The developed countries are mostly below or just at the 45-degrees line.

(3) The volatility of consumption is very tightly linked with that of net exports (Figure 6). Consumption fluctuates almost one-to-one with net exports. This could be due to the nature of shocks affecting exports: shocks to commodity prices affecting the terms of trade or shocks to competitiveness are more of a structural or permanent nature. Some countries, such as a few in Latin America, have much higher consumption volatility than that of net exports, possibly due to other structural domestic shocks.

Figure 6.
Figure 6.

Relative Volatilities of Output, Consumption and Net Exports 1/

Citation: IMF Staff Country Reports 2007, 251; 10.5089/9781451818086.002.A002

1/ “Sigma” denotes the standard-deviation of macroeconomic aggregates: C (log consumption), Y (log GDP), NX/Y (Net Exports to GDP), all at business cycle frequencies or HP-filtered from quarterly data. “Cor” denotes correlation coefficient. See Appendix for country and region codes.

(4) The trade balance deteriorates with an income shock and the extent of deterioration is higher for emerging markets (Figure 7). This observation is closely associated with the interpretation of permanent shocks through consumption (see (2)). As consumers respond to a permanent shock by adjusting consumption more than one-for-one with current output, the trade balance deteriorates with decrease in private savings. Thus trade balance is more countercyclical in emerging markets.

Figure 7.
Figure 7.

Correlations of Output and Net Exports 1/

Citation: IMF Staff Country Reports 2007, 251; 10.5089/9781451818086.002.A002

(5) Relative consumption volatility is higher in countries with low financial depth (Figure 8). Developed countries, with higher credit/GDP ratios have lower consumption volatility compared to emerging countries. This observation could support the view that the presence of liquidity constraints in emerging economies make consumers cut back on consumption when there is a negative income shock, irrespective of whether the consumers view it as permanent or transitory. However, many emerging markets have higher consumption volatility than their level of financial depth may suggest about liquidity constraints.

Figure 8.
Figure 8.

Financial Depth and Relative Consumption Volatility

Citation: IMF Staff Country Reports 2007, 251; 10.5089/9781451818086.002.A002

1/ “Sigma” denotes the standard-deviation of macroeconomic aggregates: C (log consumption), Y (log GDP), NX/Y (Net Exports to GDP), all at business cycle frequencies HP-filtered from quarterly data. “Cor” denotes correlation coefficient.
Figure 9.
Figure 9.

Sensitivity of Moments and Relative Size of the Random Walk Component to the Relative Size of Shocks to Productivity

(Replication of Aguiar and Gopinath (2007, Figure 4) 1/

Citation: IMF Staff Country Reports 2007, 251; 10.5089/9781451818086.002.A002

1/ The random walk component is estimated by varying the ratio of σg2 / σz2 (Sig_g/Sig_z), keeping all other parameters the same. The estimates for Mexico and Canada are taken from AG who compute them using sig C and sig Y.

How does Hungary compare?

54. Given the stylized facts outlined above, Hungary seems to be enjoying both developed and emerging market features (Table 2 and Figures 5-8). The volatilities and correlations in key macroeconomic aggregates qualitatively match those in recent work done on central and eastern European countries (Benczúr and Ratfai, 2007):

  • Hungary enjoys the lowest output volatility among emerging markets. This observation is in contrast to the findings in AG that “emerging market economies have a business cycle twice as volatile as their developed counterparts.” The cyclical volatility of output is only 0.79, much smaller than the emerging markets average of 2.40. This low volatility is also mirrored in a low volatility of overall output growth—in this case Czech Republic and Hungary share more stable output growth than other emerging European markets (Figure 5).

  • Hungary has the highest consumption volatility relative to income volatility among European emerging markets. While part of this ‘excessive’ relative volatility is explained by the low output volatility, a large part of it is still due to high consumption volatility. For instance, Hungary has almost twice the consumption volatility as Czech Republic. The latter is nearer the developed markets cluster, whereas Hungary is the highest among the range of countries with similar output volatilities (Figure 5).

  • The trade balance in Hungary is countercyclical, which is a distinguishing feature of emerging markets. In contrast, the trade balance in developed markets is almost procyclical. The correlation of the trade balance with output at business cycle frequencies is -0.23 in Hungary. This is similar to Poland, but quite opposite to Czech Republic (procyclical) and much lower than Slovakia (Figure 7).

  • Volatility of net exports is within the emerging markets average in Hungary. High trade balance volatility is tightly correlated with consumption volatility, with Hungary almost on the 45-degrees line (Figure 6).

  • Consumption volatility in Hungary is much higher than what the level of financial depth would suggest (Figure 8).

Discussion

55. The set of stylized facts reveal some systematic differences between the emerging and the developed market groups. Yet, there are countries within each of the two groups that display features of the other group. In particular, Hungary stands out among emerging markets in its low volatility of income (and income growth). At the same time, it has other moments that are much more in line with its emerging market counterparts.23

56. Most notable among these are the volatility of consumption relative to that of income and volatility of net exports. Highly volatile net exports could be symptomatic of terms of trade shocks—commodity exporters and countries that are very open to trade would be more prone to higher volatility of net exports. The East Asian group and some of the Latin American countries display a strong correlation between net exports volatility and income volatility. Perhaps these countries’ dependence on natural resources and commodities with large fluctuations in price explain the higher volatility of their income. Norway, an oil exporter, has emerging market features in this respect. The central eastern European countries, on the other hand, rely more on manufactured exports or might not be as open to trade as their Asian or Latin American counterparts. Their trade balance volatility on average is, therefore, smaller (2.4) than the average for the Asian (3.2) or the Latin American (2.75) groups.

57. The calculated moments together with the discussion on Hungary’s growth experience could help us make certain inferences about the nature of productivity shocks experienced by the country. The high relative volatility of consumption would have us believe that shocks to trend productivity are relatively more important in Hungary than shocks to the cyclical part of productivity. This means that a negative shock to output is more likely to depress productivity growth than just its level. Hungary’s trade-balance volatility (and its close association with high consumption volatility) could reflect shocks to competitiveness (possibly due to structural rigidities in the domestic labor market or the nature of labor taxes), that are viewed to be more of a permanent nature by consumers.

58. The extent to which shocks affect the productivity growth versus its level is informative about the permanence of output shocks. A high stochastic component or permanence would imply that a shock, however small, would have a long lasting effect on output growth. In the next section, we provide rough estimates of the importance of this stochastic or random walk component for various countries.

E. Calculations of the Random Walk Component of the Productivity Process

59. The productivity process in a real business cycle (RBC) is driven by a trend, Γt, and a stationary component, zt.24 The shocks to the growth of productivity contribute to the stochastic trend of productivity. Specifically, the trend is the cumulative product of productivity growth shocks. Productivity growth, gt, has a long-term mean, μg, and variance of shocks, σg2. The stationary component follows an AR(1) process with variance of shocks given by σz2.

(1)zt=ρzzt−1+ɛtz,|ρz|<1, ɛtz∼iidN(0,σz2)
(2)Γt=egtΓt−1
(3)gt=(1−ρg)μg+ρggt−1+ɛtg, |ρg|<1, ɛtg∼iidN(0,σg2)

The log of the solow residual is a sum of the trend or a random walk and the transitory components. The importance of trend shocks in the productivity process is the variance of trend-growth relative to the overall variance of the productivity growth.

Random⁢ walk⁢ component=a(σg2/σz2)b+c(σg2/σz2)

60. The parameters, a, b, and c are combinations of various parameters (other than σg2 and σz2) in the productivity processes ((1)–(3)) and the share of labor in the production function. Thus, keeping all other parameters constant, the random walk component is an increasing function of σg2 / σz2, or the relative importance of shocks to trend-growth.

61. The full RBC model, when solved, has implications for moments of the income, investment, consumption and the net exports processes. These moments are in terms of parameters that constitute a, b, and c, σg2 / σz2, and other model parameters. To see how the theoretical moments relate to σg2 / σz2, and to the random walk component, one of the diagrams from AG is replicated here after adding the random walk component (Figure 9). It shows that both σg2 / σz2 and the random walk component are positively related to the moments involving volatility. Higher volatilities of investment, consumption and net exports are associated with higher random walk components. Among the moments involving correlations, the correlation of net exports and output have a strongly negative relationship with the random walk component. Thus highly volatile consumption could reflect a very high random walk component, as does a strongly counter-cyclical trade balance.

62. Given the short time series, the random walk component of the productivity process can be identified matching some of the data moments listed in the previous section with the theoretical moments derived from the RBC model. For example, we can pick σg2 / σz2 so that the theoretical moment, σ(c)/σ(y), could be matched to its empirical counterpart for a particular country (Table 3).25 A range of the random walk components is derived for each country by varying σg2 / σz2 and two moments—the relative volatility of consumption and the correlation of the trade balance and output.

Table 3.

Random Walk Components Implied by Two Types of Moments 1/

article image

When the estimate for random walk component is greater than 1, we report it as 1. NX/Y refers to net-exports/GDP, C to consumption, Y to GDP, all at business cycle frequencies; ρ refers to correlation and σ to standard deviation.

Discussion

63. Hungary has a high random walk component, higher than two-thirds. The two moments yield a range for the random walk component. While the relatively higher consumption volatility deliver an extremely high random walk component, the relatively low countercyclicality of net exports imply a lower random walk component. Consumption volatility is an especially informative moment, in terms of distinguishing the set of emerging and developed markets (Table 2). Although the correlation of net exports to output is also a sensitive moment, its empirical counterpart does not distinguish emerging and developed markets as much as consumption volatility does. Hungary, therefore, is more likely to behave like other emerging economies rather than developed ones.

64. Very few countries among the emerging European group strike out as “obviously emerging.” As in the example used in AG, Mexico stood out as a country with very obvious emerging market features with a tight range (of almost 1) of random walk components derived from various moments. Only Croatia (and to some extent Slovakia) stand out as such among the emerging European group. In contrast, Czech Republic could pass as a developed country.

F. Conclusions

65. Hungary has some emerging market features that point towards the nature of income shocks it faces. In particular, a very high volatility of consumption relative to income suggests that rational consumers perceive shocks to income as being more permanent. This implies that a negative income shock is more likely to prolong the period of low output growth. This is in contrast to Czech Republic and Poland. In comparison with other emerging markets, Hungary’s income volatility is very low. Such low volatility of income combined with a relatively high volatility of consumption could suggest that although shocks to income are small and possibly infrequent, they have a long-lasting effect.

66. In particular, Hungary’s trade-balance volatility (and its close association with high consumption volatility) could reflect shocks to competitiveness (possibly due to structural rigidities in the domestic labor market or the nature of labor taxes), which are viewed to be more permanent by consumers. Even though these shocks may be small—rendering a low volatility of overall output—they are long-lasting and create uncertainties that depress consumption growth and leads to high consumption volatility. Unless structural policies are implemented to correct such shocks, the effect of these shocks could persist in the near term.

Appendix 1. Country Codes, Data Sample, and Sources

article image

LATAM stands for Latin America; EASIA for East Asia; EM EUR for Emerging Europe; DEVD for Developed countries.

AG stands for Aguiar and Gopinath (data is available at http://www.economics.harvard.edu/faculty/gopinath/papers/datapage.html); DX for dX time from EconData; PSO for Polish Statistical Office.

References

  • Benczúr, Péter and Attila Ratfai, 2007, “Economic Fluctuations in Central and Eastern Europe: The Facts,” CEU Economics Working Paper No. 9/2004.

    • Search Google Scholar
    • Export Citation

  • Campbell, John Y., and N. Gregory Mankiw, 1989, “Consumption, Income, and Interest Rates: Reinterpreting the Time Series Evidence,” NBER Macroeconomics Annual 1989, edited by Olivier Blanchard and Stanley Fischer. Cambridge, Mass.: MIT Press.

    • Search Google Scholar
    • Export Citation

  • Carroll, Christopher D., Robert E. Hall, and Stephen P. Zeldes, 1992, The Buffer-Stock Theory of Saving: Some Macroeconomic Evidence, Brooking Papers on Economic Activity, Volume 1992(2).

    • Search Google Scholar
    • Export Citation

  • Carroll, Christopher D., 2001, A Theory of the Consumption Function, with and without Liquidity Constraints, The Journal of Economic Perspectives, Volume 15(3), Summer.

    • Search Google Scholar
    • Export Citation

  • Friedman, Milton A., 1957, A Theory of the Consumption Function. Princeton: Princeton University Press.

  • Friedman, Milton A., 1963, “Windfalls, the ‘Horizon,’ and Related Concepts in the Permanent Income Hypothesis,” in Measurement in Economics, Carl Christ et al., eds. Stanford: Stanford University Press.

    • Search Google Scholar
    • Export Citation

  • Gopinath, Gita and Mark Aguiar, 2007, Emerging Market Business Cycles: The Cycle is the Trend, Journal of Political Economy, Volume 115(1), February 2007.

    • Search Google Scholar
    • Export Citation

  • Garcia-Cicco, Javier, Roberto Pancrazi, and Martin Uribe, 2006, Real Business Cycles in Emerging Countries? NBER Working Paper No. 12629, October

    • Search Google Scholar
    • Export Citation
18

Prepared by Srobona Mitra, who wishes to thank Abdul Abiad, Gita Gopinath, Daniel Leigh, Ashoka Mody, and Martin Uribe for helpful discussions, and seminar participants at the Magyar Nemzeti Bank during the Article IV mission for useful comments.

20

The variable ‘unemployment expectations’, compiled by Eurostat, is based on a consumer survey that measures consumers’ expectations of unemployment in the next 12 months. The variable indicates the difference in percentage points between the percent of survey responders who expect the unemployment rate to increase and those who expect it to decrease. The scale is different from Carroll (1992), which reproduces Campbell and Mankiw’s regression using unemployment expectations in the United States—the fraction of households who believe unemployment will increase minus the fraction who believes it will decrease.

21

Productivity is modeled as comprising of a stationary process with volatility of shocks given by σz, and a stochastic trend with volatility of its growth rate given by σg.

22

See Appendix I for data sources and sample sizes for all the countries.

23

Czech Republic stands out among the emerging market group in having all its moments point towards a developed country.

24

The underlying production function is Cobb-Douglas using labor and capital as inputs.

Same Series

Other IMF Content

Other Publishers

Cover IMF Staff Country Reports
Hungary: Selected Issues
Author: International Monetary Fund
Volume/Issue:
Volume 2007: Issue 251
Publisher:
International Monetary Fund
ISBN:
9781451818086
ISSN:
1934-7685
Pages:
59
DOI:
https://doi.org/10.5089/9781451818086.002

GDP Growth, 2000-06 1/

(year-on-year percent change)

Hungary: Contributions to GDP Growth, 2000-06 1/

(Year-on-year, in percent)

Consumption and Real Wages 1/

Consumption and Unemployment Expectations 1/

Relative Volatilities of Output, Output Growth, and Consumption 1/

Relative Volatilities of Output, Consumption and Net Exports 1/

Correlations of Output and Net Exports 1/

Financial Depth and Relative Consumption Volatility

Sensitivity of Moments and Relative Size of the Random Walk Component to the Relative Size of Shocks to Productivity

(Replication of Aguiar and Gopinath (2007, Figure 4) 1/