The Estonian EconomyMonthly newsletter from Swedbank’s Economic Research Departmentby Madis Aben, Annika Paabut No. 6 • 3 November 2011 High uncertainty induces savings and jeopardises consumption recovery Estonia experienced by far the highest spending and saving volatility in the EU during the boom and consequent crisis. The households’ saving rate dropped to -6.3% in 2006 and then climbed to +11.6% in 2009. Consumption loans were not enough to cover the excess of consumption over income during 2003-2006, so households had to liquidate their assets – for instance, to sell real estate to other sectors (foreigners and the enterprise sector) and to use mortgages for consumption purposes. Since the beginning of 2009, households have been reducing their loan stock to improve their net financial position. The saving and consumption behaviour of households is likely to be more stable in the future. Deleveraging will continue for some time, but the saving rate will probably come down slowly from its 2009 peak, if no new negative shocks are experienced. A new consumption boom with negative saving rates is very unlikely for the next several years.There has been much discussion lately about the loan payments are considered as savings. Therapid hike of the Estonian households’ saving rate saving rate is calculated as the ratio of savings toand its possible farther path. Amongst EU disposable income.countries, the volatility of the households’ savingrate in Estonia has been by far the highest during Saving and consumption behaviour duringthe recent boom and bust. Estonia has also the boom and bustsurpassed Latvia and Lithuania in that respect. We Income of households can either be consumed orherewith give an overview of the saving behaviour saved. If one rises, the other consequently falls,of households during the boom and consequent given that income stays constant. During the fivebust. successive years of rapid economic growth from 2003 to 2007, the households’ savings rate wasWith respect to household savings, the definition negative, reaching its lowest level (-6.3%) in 2009.behind it usually is households’ disposable income This was probably the result of the high level ofless their final consumption expenditure. confidence during that period: joining the EU inDisposable income is the net (after-tax) income of 2004, together with favourable economichouseholds from all sources, the most important developments, as evidenced by the continuallybeing wages and salaries, social transfers, and improving employment situation and rapidly growingcapital income (net interest payments, dividends,and entrepreneurial income). Final consumption wages. All this resulted in high consumerexpenditure comprises expenditure on goods and confidence and an overoptimistic belief thatservices. Purchases of semi-durable and durable incomes would increase at the current rate forever.goods (cars, home electronics, furniture, etc.) are This optimism from the consumer side was combined with the active marketing of loans andalso considered to be consumption, while credit facilities by the financial sector. The low realpurchases and capital repairs of dwellings are interest rates fuelled the housing boom, and rapidlyinvestment and, hence, not included in the rising real estate prices had a strong effect onconsumption measure. Moreover, final consumption consumers’ perceived wealth.expenditure also includes interest payments, while Economic Research Department. Swedbank AB. SE-105 34 Stockholm. Phone +46-8-5859 1000. E-mail: email@example.com www.swedbank.com Legally responsible publisher: Cecilia Hermansson, +46-8-5859 7720. Annika Paabut, +372 6 135 440. Elina Allikalt, +372 6 131 989.
The Estonian Economy Monthly newsletter from Swedbank’s Economic Research Department, continued Nr 6 • 3 November 2011Consumer confidence index and households’ savings rate such behaviour could not continue endlessly, as the 20 16 possibility of using future incomes to finance current consumption had to end some day, even without 10 12 the global crisis. 0 High volatility of saving rate has 8 unfavourable effects 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 -10 In the old member states of the EU, consumption is 4 -20 generally less volatile than GDP over the economic cycle. The situation has been the opposite in the 0 -30 Baltic countries. This is at least partly due to the low efficiency of automatic stabilizers. This means that -4 -40 the Estonian tax and benefit system does not smooth the households’ incomes over the economic -50 -8 cycle as well as it does in most other EU countries, Conf idence, pts (ls) Source: DG ECFIN, SE Sav ings rate, % (rs) or in the US. The tax system is less progressive and the social benefits are less generous in Estonia. The fluctuations in market incomes, whileHouseholds disposable income, consumption and savings encouraging participation in the labour market,(EUR mln) carry over more directly to the disposable income of 9,000 1,200 households. 1,000 8,000 2 800 Households’ saving rate (% of disposable income) 7,000 600 15 400 6,000 200 10 5,000 0 -200 4,000 5 -400 3,000 -600 2002 2003 2004 2005 2006 2007 2008 2009 2010 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 Sav ings (rs) Disposable income Source: SE Consumption expenditure -5The decline of households’ income during the -10recession, due, inter alia, to a steep drop in Estonia EU27 Latv ia Lithuaniaemployment and moderately falling wages, reduced Source: Eurostat, SE Finlandthe possibilities for consumption. This was furtherexacerbated by the rapid increase of the saving rate On the one hand, the consequent high volatility ofof households. In nominal terms, households’ private demand creates uncertainty for companiesconsumption expenditure dropped by 17% in 2009, offering goods and services to the local market. But,while the drop in disposable income was about 8%. on the other hand, this volatility also createsThe cumulative fall in real private consumption in uncertainty for the collection of tax revenues, as theEstonia was 25% from the pre-crisis peak to the Estonian tax system relies heavily on consumptionbottom at the end of 2009.1 It has to be taken into taxes. The share of consumption taxes in totalaccount, of course, that consumption growth was taxation peaked at 42.3% in 20063 – one of theunsustainably fast during the boom years prior tothe crisis, when consumption expenditure could notbe fully financed from current incomes. Therefore, 2 Saving rates for Estonia are from Statistics Estonia (because of a longer time series, which includes also 2010), for all other countries they are from Eurostat.1 3 The fall was of the same magnitude in Lithuania (25%) Taxation trends in the European Union, 2010 edition,and even bigger in Latvia (30%). Eurostat. 2 (5)
The Estonian Economy Monthly newsletter from Swedbank’s Economic Research Department, continued Nr 6 • 3 November 2011largest shares in the EU. This was equal to the ratio Loans and change in loan stockfor Romania and was smaller only than Bulgarias (% of income)(56.5%). By 2008, the first year of the recession, 100the share had dropped to 36.8%, the second-biggest drop (of 5.5 percentage points) in the EU 80after Latvia (5.7 percentage points). The EU27weighted average of the share of consumption 60taxes in total taxation has been close to 28% duringthe past decade. All in all, this means that the 40government’s goal of balancing the budget isheavily dependent on the consumption behaviour of 20households during each year of the economic cycle.Financing the deficit 0 2002 2003 2004 2005 2006 2007 2008 2009 2010As mentioned above, during the boom years, thehouseholds saving rate was negative, which, in -20turn, means that current consumption expenditure Source: EP, SB calculations Loans stock Loan stock growthwas higher than current income and had to befinanced from other sources, either by borrowing or 4by selling assets to other sectors. During 2003- Households’ savings and growth of consumer loans’ stock (EUR mln)2006, households’ consumption expenditureamounted to EUR 23.4bn, while disposable income 1000amounted to only EUR 21.9bn, creating a negative 800gap of EUR 1.5bn. During the same period, the 600increase of consumer loans’ and lease stock wasonly EUR 0.9bn. As this was not enough to finance 400the gap, households, in addition to consumer 200borrowing, had to sell assets to other sectors to 0finance their consumption. Moreover, a significant 2002 2003 2004 2005 2006 2007 2008 2009 2010part of the new housing loans was channelled to -200households, and these resources could also be -400used to finance consumption. This could happen -600because about one-half of the real estate contractsare made between two households. If such a -800purchase is financed by a mortgage, the money Consumer loans stock change Sav ings ICCends up in the hands of the household sector. It Source: SE, EP, Swedbank calculationsthen becomes that sectors decision whether toinvest it in financial assets or to consume the When, in the fourth quarter of 2008, the downturnmoney. started and labour income began to fall, deposits still continued to grow, albeit at a slowing rate. InHouseholds’ loan stock grew extremely rapidly aggregate, households did not use deposits toduring the boom years and peaked at almost one- smooth consumption throughout the crisis. This isthird of households’ disposable income in 2006. probably because those losing their jobs did notHouseholds started to decrease their loan stock at have significant liquid savings to compensate forthe beginning of 2009, but the loans-to-income ratio the sharp declines in income. Larger deposits,still continued to climb, peaking at 98% in 2009. As meanwhile, were probably held by those who alsoincome growth recovered by the end of 2010 and were better able to keep their jobs and incomeshouseholds continued to deleverage, the ratio during the downturn. By now, the growth rates ofstarted to decrease as well. The deleveragingprocess, however, is also affecting the overallsaving rate of households, as, by definition, the loanpayments (without interest payments) are 4accounted as savings. This figure illustrates the fact that, during 2003-2006, even the growth of the stock of consumption loansdid not cover the gap between consumption expenditure and disposable income. ICC is calculated as disposable income plus the growth of consumption loans’ stock minus consumption expenditure. 3 (5)
The Estonian Economy Monthly newsletter from Swedbank’s Economic Research Department, continued Nr 6 • 3 November 2011income and deposits are converging again, while expenditure has grown faster than income; thisthe loan stock continues to decrease at a marginally trend also shows up in a slightly declining savingslowing rate. rate.Loan, deposit, and labour income(annual growth, %) Labour income and consumption (annual growth, %) 70 30 60 25 50 20 40 15 30 10 20 5 0 10 2003 2004 2005 2006 2007 2008 2009 2010 2011 -5 0 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 -10 -10 -15 -20 -20 Source: SE, EP Deposits Loans Labor income Labour income Source: SE Consumption expenditure (nominal)Latest developments in income and Some improvement in consumption behaviourconsumption patterns might have been predicted already in 2009, asDevelopments in labour income are a good proxy consumer confidence started to improve quickly in 5for developments in total income. The share of the first half of that year; however, this improvementunemployment benefits (which to some extent did not show up in actual consumption until the firstsoftened the drop in the wage bill) at the aggregate quarter of 2010. The faster growth of expenditurelevel declined significantly from the first quarter of compared with that of income, starting only in 2010,2010 to the middle of 2011, as many of those who can partly be explained by the rising inflation ratewere eligible for unemployment insurance benefits (mostly food and housing costs), starting in thelost them due to the length of their unemployment.6 second half of 2010. Another reason was the hike inAt the same time, other social benefits (pensions, purchases of durables, which had been postponedchild benefits, etc.), have not changed much. during the years of severe crisis. Nevertheless, theCapital income is again quite unevenly distributed growth of consumption at the end of 2010 mightamongst different income groups, and its share in also be the result of the fear of an acceleration indisposable income was only around 10% in 2010. price increases after adoption of the euro (EstoniaSo, labour income is an acceptable proxy to joined the EMU on 1 January 2011). The growthdescribe the dynamics of total disposable income. rates of labour income and consumption are likely to converge in the near future if the overall recoveryDuring the boom years, nominal consumption of the economy stays on a stable track.expenditure actually rose hand in hand with labourincome. As confidence dropped at the very What could the future look like?beginning of the crisis, consumption was cut back There are reasons to believe that the saving andmore than income fell, implying an increasing consumption behaviour of households will be lesssaving rate. Concurrently, a deleveraging process volatile for the foreseeable future than in the past,began, as households started to decrease their loan as many of these households might have learnedstock. Since the beginning of 2010, consumption from the crisis years. This means that people should be better aware of the necessity of preparing financial buffers for possible drawbacks in incomes, and be more “rational and responsible” in their saving and consumption behaviour in both good5 Disposable income is calculated by Statistics Estonia and bad times.only on an annual basis, and the latest figures are for During the crisis, the Estonian households’ saving2010. Data on labour income are available quarterly. rate climbed from -6.3% in 2006 to 11.6% in 2009,6 Unemployment benefits are paid during up to one year. 4 (5)
The Estonian Economy Monthly newsletter from Swedbank’s Economic Research Department, continued Nr 6 • 3 November 2011before falling back to 9.6% in 2010. On the one A big risk for consumer confidence comes from thehand, the rates during the past two years have still-uncertain outcome of the resolution of thebeen quite comparable to those of advanced financial crisis in Europe. As the crisis is severe andeconomies (11.6% in Finland, for instance, in prolonged and no positive final solution has thus far2010). On the other hand, the saving rates of other been envisaged, the resulting uncertainty is likely toconverging economies are generally lower (9.4% in keep consumption distressed and savings relativelyLatvia and 7.9% in Lithuania in 2009). We are of the high, especially in those countries that experiencedopinion that the saving rate is likely to continue to more harsh declines in incomes.decrease slowly, as the labour market situation is All in all, households have probably learnedimproving and households’ expectations about their something from the recent overconsumptionfinancial well-being are getting brighter again. episode and, as a result, will keep theirThe probable slight decrease in the saving rate consumption expenditure more in line with currentfrom the current level is also supported by the income for the foreseeable future. In addition,reality that the average level of income is still much Estonia as a small economy with aging populationlower in Estonia than in the old EU member states will see pressure on its social system in the long run(64% of the per capita EU average in purchasing because the cohort leaving the labour market willpower standards in 2010). It is known from the be larger than the one entering it. This in turnrelevant literature that the propensity to consume indicates that the share of wage earners tofrom current income is usually higher at lower pensioners will decrease. The pension system wasincome levels; this holds both when comparing reformed in Estonia already in 1997, now consistingdifferent countries or different income groups within of three pillars. The first pillar is a pay-as-you-goa country. Moreover, recent history has shown that, system, i.e. the share of pensions that is covered bylike households in other young, relatively poor but the current revenues from the state budget. Therapidly growing economies, Estonian households second pillar is collected jointly – employees savehave put more weight on welfare from current 2% from their wage and the state contributesconsumption and worried less about the future. The another 4%. Savings to the second pillar aredifficult years following the boom have probably mandatory for the younger cohorts, but it was notreminded households of the possibility of negative yet set up to those who have reached the middlefuture developments and the need to be prepared age now. The third pillar is a voluntary savings fund.to compensate for sudden income losses in order to All in all, the pension system should support thesmooth consumption over economic cycles. idea of the life cycle theory of savings: people should be able to save during their prime workingOne more reason why the saving rate might fall age to finance their consumption expenditures aftermoderately is the steady increase in real estate retirement as well as the negative savings theyprices, which began in the second half of 2009. An might have had during younger age (student loansincrease in the perceived (housing) wealth tends to etc).improve the propensity to consume. On the otherhand, real estate prices are still significantly lowerthan during the mortgage boom years, so many Madis Abenhomeowners still have to deleverage to improvetheir balance sheets. This gives reason to believe Annika Paabutthat a new lending boom is very unlikely to financeconsumption in excess of income for the nextseveral years, and that the decrease in the savingrate will likely be slow.SwedbankEconomic Research Department Swedbank’s monthly newsletter The Estonian Economy is published as a service to ourSE-105 34 Stockholm customers. We believe that we have used reliable sources and methods in the preparationPhone +46-8-5859 1028 of the analyses reported in this publication. However, we cannot guarantee the accuracy firstname.lastname@example.org completeness of the report and cannot be held responsible for any error or omission in thewww.swedbank.com underlying material or its use. Readers are encouraged to base any (investment) decisions on other material as well. Neither Swedbank nor its employees may be held responsible forLegally responsible publisher losses or damages, direct or indirect, owing to any errors or omissions in Swedbank’sCecilia Hermansson, +46-8-5859 7720 monthly newsletter The Estonian Economy.Annika Paabut +372 6 135 440Elina Allikalt +372 6 131 989 5 (5)