What Drives Changes in Business and Consumer Sentiment?

What Drives Changes in Business and Consumer Sentiment? HWANG Sang-Yeon Research Fellow, Samsung Economic Research Institute Week ly Insight I. Limit of Real GNI to Measure Business and Consumer Sentiment Many people in Korea are saying that economic statistics, especially GDP growth, is beguiling since they can hardly feel the state of the economy as good as the growth figure suggests. Professional economists have been trying to provide explanations for the gap between the growth figure and the way consumers or businesses feel about the economy. Most of them refer to another growth statistics named GNI. Real GNI is deemed to represent the purchasing power of economic participants, hence their feeling about the current state of the economy. Indeed, growth in real gross national income (GNI) has Gap between the Growth Rates of Real GDP and Real GNI 4.00 been outstripped by real gross domestic product (GDP) since the mid 1990s. In particular, the gap between the two indicators has progressively widened since the end of 2000. The average gap increased from 1.2 percentage points during the period of 1996-1999 to 1.8 percentage points during the period of 2000-2006. In particular, the gap expanded to roughly 3 percentage points in 2000, 2005 and 2006. In terms of the total gap, the two indicators had a differential of 68 trillion won in 2006, with real GDP at 759.2 trillion won and real GDI at 691.1 trillion won. The increasing growth gap is commonly attributed to worsening terms of trade. 1 The terms-of-trade index plunged from 138.5 points in 1995 to 73.2 in 2006. 3.51 (Unit : % ) 3.00 2.96 2.71 2.00 1.00 0.00-1.00 0.92-0.09-0.45-0.70 0.50 0.24-0.40 0.08-0.13-0.29-0.61 1.39 1.93 1.44 0.08 1.22 1.02 0.82 0.00-2.00-1.50-1.63-1.87-2.03-3.00 '81 '82 '83 '84 '85 '86 '87 '88 '89 '90 '91 '92 '93 '94 '95 '96 '97 '98 '99 '00 '01 '02 '03 '04 '05 '06 Source: Bank of Korea. 1The terms of trade refers to represents the volume of goods that can be imported per one export unit. The index = (the export unit value index/the import unit value index) x 100. 10

June 11, 2007 A sharp fall in export prices between 1996 and 2002, coupled with a rise in import prices, led to a deteriorating terms of trade. To be specific, the export unit price index dropped from 161.8 in 1995 to 83.1 in 2002. On the contrary, the import unit price index shot up from 87.5 in 2002 to 126.3 in 2006. However, real GNI statistics has several limitations to measure consumer or firm s feeling about the current state of the economy. Let s see why. First, even if real GNI 2 is generally considered to be a measure of consumers and firms overall purchasing power, it cannot adequately reflect expanding purchasing power as a result of improved productivity because the statistics only considers the terms of trade based on relative prices. For example, even if net barter terms of trade worsen due to reduced prices (a product of higher productivity) and accordingly lower export prices, GNI will be negatively affected, which will be accepted as lower purchasing power of firms and consumers, hence worse sentiment of them. However, cost reduction due to productivity improvement, eventually strengthening corporate profitability (meaning firms stronger purchasing power) despite falling export prices, can lead to wage increases, which in turn leads to stronger consumers purchasing power. This example suggests that real GNI is too abstract to completely capture economic reality as experienced by consumers and firms. Second, GNI has a limit in estimating each economic participant s sentiment towards the current state of the economy. This is because it aims to estimate purchasing power of the economy as a whole, regardless if it s at the individual or firm level. However, it is often the case that the economic sentiment of household and corporate side substantially diverges. This situation occurred in 2006 when real GNI increased 2.3% overall, but household incomes rose 3.1% and firms real operating surplus decreased 6.6%. For 2006, therefore, the real GNI understated households overall purchasing power, and overstated that of firms. Additionally, the real GNI cannot adequately reflect qualitative changes in the real economy. Thus, measuring the each economic participant s sentiment toward the economy needs to take into account a range of factors from income inequality, life insecurity to changes in the industrial structure, alongside with the income level reflecting the change in purchasing power. 2 It is calculated by adding gains or losses from trade and Korean nationals net factor income from abroad following changes in the trade terms to real GDP. An exchange ratio of imported goods to exported goods. II. Key Economic Sentiment Indicators Consumer Sentiment Indicators Korea s major public economic institutes developed their own consumer economic sentiment indicators to gauge how consumers view the economy: the Samsung Economic Research Institute s (SERI) Consumer Attitude Index, the Bank of Korea s (BOK) Consumer Sentiment Index, Korea National Statistical Office s (KNSO) Consumer Evaluation and Expectation Index. These indices survey how consumers feel about the economy, as well as present and future expectations regarding their standard of living. Scores over the benchmark mean that consumers see the economy and their living conditions getting better. Industrialized nations such as the US, European Union (EU) and Japan have also consumer economic sentiment indicators. In the US, the Conference Board s Consumer Confidence Index (CCI) and the University of Michigan s Consumer Sentiment Index (MCSI) are used most often. Both consist of the index representing the present economy and living conditions and the index reflecting the future economy and living conditions. In the EU, the Business and Consumer Survey uses the Consumer Confidence Indicator, which is different from the US indicators in that it surveys the likelihood of saving for the next year. In Japan, the Cabinet Office uses the Consumer Confidence Survey. Business Sentiment Indicators The business economic sentiment is measured by Korea s seven institutes including, most prominently, the BOK, the Federation of Korean Industries (FKI) and the Korea Chamber of Commerce & Industry (KORCHAM). Headline indices are the BOK s Business Survey Index (BSI) about the business condition, the FKI s BSI about the overall economic performance and the KORCHAM s BSI about economic outlook. When each index goes over the benchmark of 100, there are more companies positive about the present and future economic situations than negative ones. This is true of other advanced nations like the US, Japan and the EU. In the US, the Institute for Supply Management (ISM) releases a monthly economic outlook report for the manufacturing and non-manufacturing sectors. The SIM applies the Purchasing Managers Index (PMI) to manufacturing. The PMI is weighted with the following variables; new orders received (30%), production (25%), employment (20%), delivery speed of suppliers (15%) and inventory (10%). In Japan, the Bank of Japan has used the Tankan Index, an index that Samsung Economic Research Institute 11

measures perceived business conditions since 1957. The index questions businesses on their economic assessment, quarterly economic conditions, the 6-month and yearly performance and outlook on the number of newly employed (in June and December). In the EU, the European Commission figures out the composite business economic sentiment by using the Economic Sentiment Indicator. The indicator combines 4 the industry survey index of manufacturing, services, retail, construction and with the consumer confidence index. III. Factor Analysis of Economic Sentiment Indicators Entering 2007, general economic sentiment has turned more optimistic. Most consumer sentiment indicators measuring present and future economic conditions bottomed out in the fourth quarter of 2006, expanding into the first two quarters in 2007. For example, SERI s Consumer Attitude Index rose to 48.5 in the second quarter of 2007, a record since the 49.3 registered in the second quarter of 2006. KNSO s Consumer Evaluation Index also reached 87.4 in April 2007, the highest since March 2006 when it recorded 90.1. BOK s Consumer Sentiment Index reached 103 in the first quarter of 2007, surpassing the 100 benchmark after the last two quarters subpar figures. Business sentiment has also improved gradually since the third quarter of 2006. The FKI s BSI on the overall economic performance posted 105.8 in April 2007, far above the 94.7 registered in the first quarter of the year. The peak and trough (representative of optimism and pessimism) of the BSI has risen since the third quarter of 2006. The BOK s BSI has improved since the third quarter of 2006 as well. 1. Why are Consumer and Business Sentiment Improving? Although virtually all consumer and business sentiment indicators have shown improvement, the reasons underpinning such improvements are difficult to pinpoint. The BOK and KNSO do not analyze why consumers attitudes have changed. Their BSI questions difficulties in management, but not the reason for either an increase (decrease) of the BSI, either. Therefore, more 4 Each index for manufacturing, services, consumers, wholesale/retail and construction is combined according to its own weight (40%, 30%, 20%, 5% and 5%). accurate measurement of the economic sentiment is needed based on the economic data on what have driven changes in the sentiment. Consumer Sentiment: Growth Rate of Real Income In order to figure out why consumer sentiment has risen, SERI ran a series of regressions with the BOK s Consumer Sentiment Index as the dependent variable along with a number of independent variables. 5 Overall, the regression result suggests that real disposable income (the total income minus non-living expenditure such as taxes), the income inequality level (the ratio of highest to lowest income quintiles, 6 the Misery Index and stock prices are the core variables affecting consumer sentiment. Rising real disposable income and stock prices are significant in positively affecting consumer sentiment, with higher income inequality and the Misery Index negatively affecting consumer sentiment. Real disposable income also played a key role in the recent recovery of consumer economic sentiment. Between the last quarter of 2006 and the first quarter of 2007, the ratio of highest to lowest income quintiles expanded from 4.8 times to 5.6 times; the Misery Index also rose 0.2 percentage points to 5.5% due to a rising unemployment rate. Nevertheless, in the first quarter of 2007, real disposable income grew at an annual rate of 7.1%, more than offsetting negative impacts of higher income inequality and unemployment and leading to the higher-than-benchmark consumer sentiment index of 103. Business Economic Sentiment: Higher Stock Prices and Robust Sales Using the same method, main variables determining the business sentiment were identified by using the FKI s BSI as the dependent variable. An array of management indices was examined in order to identify possible independent variables. 7 As a result, stock pric- 5 Every possible variable was considered as independent variables: real disposable income, stock prices, the Misery Index (the unemployment rate + inflation), the ratio of highest to lowest income quintiles, real estate prices, the employment rate, the ratio of debt to financial assets, the national burden ratio, interest margin, the tax burden ratio, foreign exchange rates and the proportion of educational expenses. 6 One of the inequality gauge that is the division of the incomes of the highest 20% income earners by those of the lowest 20%. 7 Sales, foreign exchanges rates, the bill default ratio, stock prices, the operating profit/sales ratio, the business Gini coefficient*, corporate tax rates, debt/equity ratio, dependency on borrowed money, the proportion of marginal companies**, 12

June 11, 2007 es, sales and per-capita labor cost, in that order, were the main variables that affected business economic sentiment (See the Index for regression results). Stock prices were strongly correlated with positive business sentiment. On the contrary, per-capital labor cost was negatively correlated with business sentiment as rising labor costs naturally led to a reduction in profits. IV. Conclusions and Implications There is a mutually complementary relationship between economic sentiment and changes in the real economy. On the one hand, both consumer and business sentiment are directly related to changes in the real economy. Meanwhile, the real economy is also affected by how individuals and firm view the economy s future prospects. Economist A. C. Pigou even argued, The varying expectations of business men... and nothing else, constitute the immediate cause and direct causes or antecedents of industrial fluctuations." To invigorate the vitality of the national economy, therefore, it is important to recognize the complementary nature of this relationship. Indeed, if firms and individuals have optimistic sentiment towards the economy, consumption levels will increase. The BOK s Consumer Sentiment Index illustrates the Granger causality 8 with private consumption. The BOK s BSI on the business condition was found to result in facility investment increase over two quarters as well. Granger causality in both cases is statistically significant. 9 When it comes to the business sentiment, growing sales and stock prices drive changes in the sentiment. Thus, enhancing the business sentiment needs creating an environment that will help sales and profit generation. To this end, regulations on business activities should be lifted, debt be managed more flexibly and the labor and management relations be stabilized. It seems, however, that firms should be able to contain wage increases within labor productivity growth. Unit labor cost is one negative factor for the business sentiment. Korea s unit labor cost increased at an annual average rate of 2.8% during the period of 2000-2006, higher than those in more industrialized or comparable nations. During the same period, for example, the US unit labor cost rose merely 2.0% and Japan s even decreased 2.1%. The increased unit labor cost has encroached firms profitability thereby suppressing business sentiment. In order to enhance consumer sentiment, it will be far better for job creation to lead the income generation of consumers rather than through wage increase. Of course, both wage increase and job creation can raise incomes of consumers. However, considering the presently high unit labor cost, it is hard to expect for firms to raise wages further. Therefore, only sustained income generation through job creation will work to boost the consumer sentiment. Hence, again, the need for deregulation effort by the government. Furthermore, at the current income level, expanding disposable income by lowering tax and quasi-tax burdens which have been rapidly increasing these days, will contribute to the consumer sentiment. SERI per-capita labor cost, the polarization index between large companies and small- and mid-sized enterprises (SMEs), the polarization index*** between IT firms and non-it firms, facility investment and inventory. * The Gini coefficient that is calculated to figure out the inequality structure of operating profit, based on the operating profit data of companies listed on the securities market and those registered on the KOSDAQ market. ** Marginal companies have net interest cover rates ((operating profit/interest expense)x100) falling short of 100. *** The polarization index measures the divergence of operating profit rates among companies in each category from one (absolute equality). 8 Granger causality is a statistical technique developed by Economic Nobel laureate Clive Granger. It means that a change in one variable have a statistically significant effect on other variable over time. 9 At the 99% confidence level. Samsung Economic Research Institute 13

Analysis of Factors Affecting the Consumer Economic Sentiment (2Q 1996-4Q 2006) Variables Regression Coefficient Standard Error t-statistics Constant 2.862*** 0.490 5.847 Growth Rate of Real Disposable Income 0.724*** 0.239 3.036 Ratio of Highest to Lowest Income Quintiles -0.349* 0.174-2.005 Growth Rate of the Misery Index -0.135** 0.067-2.034 Growth Rate of the Real Stock Prices (KOSPI) 0.100*** 0.027 3.656 Consumer Sentiment Index of the Previous Quarter 0.375*** 0.107 3.505 R2 = 0.86 Note: 1. Disposable income is from the KNSO data (year-on-year growth rate). 2. The Misery Index = the unemployment rate (of the people who sought work but could not work for over one hour for one week) + consumer price inflation (seasonally adjusted growth rate on a yearly basis). 3. Stock prices (realized adjusted for consumer prices) are the KOSPI index (year-on-year growth rate). 4. Real estate prices are the national apartment price index realized into consumer prices (year-on-year growth rate). 5. All variables excluding income inequality are in the form of algebraic difference. 6. *, **, *** mean that they are significant statistically at the level of significance of 10%, 5% and 1% each. 7. R2 is coefficient of determination meaning explanatory power. Source: Bank of Korea, Korea National Statistical Office, Kookmin Bank. Analysis of Factors Affecting the Business Economic Sentiment (1Q 2000-4Q 2006) Variables Regression Coefficient Standard Error t-statistics Constant -0.033 0.020-1.654 Real Stock Prices 0.878*** 0.168 5.230 Real Sales 0.607** 0.228 2.657 Per-capita Labor Cost -0.193*** 0.059-3.252 R2 = 0.58 Note: 1. Sales are realized adjusted for producer prices and stock prices adjusted for consumer prices. 2. Per-capita labor cost = the total labor cost / the number of the employed. It is presented at the current level adjusted for inflation. 3. Every explanatory variable is a year-on-year growth rate. 4. *, **, *** mean that they are significant statistically at the level of significance of 10%, 5% and 1% each. 5. R2 is coefficient of determination meaning explanatory power. Source: Bank of Korea, Korea Investors Service. Inc. 14