5

No: 2015-16 / 08 September 2015

RESEARCH NOTES IN ECONOMICS

Consumer Confidence Indices and Financial Volatility1

Hatice Gökçe Karasoy

Real economic activity data, specifically private consumption, is published with a considerable
Abstract: time lag. This makes indicators with shorter publication lags quite popular for policy makers as
well as the public. In this regard, the confidence indices which are released monthly are followed closely; and their
signals on private consumption and economic activity are interpreted carefully. This note investigates the factors
that affect confidence indices. It is well-established that confidence indices are sensitive to developments in
financial markets. This analysis shows that they are also affected by financial volatility. The domestic volatility
indicators of financial variables such as exchange rate, interest rate, and BIST (stock exchange), and VIX index
which essentially represent external financial volatility deteriorate consumer sentiment. Moreover, it is found that
confidence indices are affected much more from inflation dynamics compared to the economic activity in the short
run. The strong relation between confidence and volatility points out that the policies aimed at stability in financial
markets may also improve consumption tendency.

Reel iktisadi faaliyete ilişkin veriler, özellikle de özel tüketim, oldukça gecikmeli olarak açıklanmaktadır.
Özet: Bu yüzden daha zamanlı açıklanan göstergelere politika yapıcılar ve kamuoyu tarafından önem
atfedilmektedir. Aylık olarak açıklanan tüketici güven endeksleri bu bağlamda yakından takip edilmekte, tüketime
ve iktisadi faaliyete ilişkin verdikleri sinyaller dikkatle yorumlanmaktadır. Bu not güven endekslerinin hangi
faktörlerden etkilendiğini sorgulamaktadır. Finansal değişkenlere hassaslığı bilinen güven endekslerinin finansal
oynaklığa da duyarlı oldukları gösterilmiştir. Döviz kuru, faiz oranı, BIST endeksine ilişkin oluşturulan yurt içi
finansal oynaklık göstergeleri ile yurt dışı finansal oynaklığı temsil eden VIX endeksi, güven endekslerini olumsuz
yönde etkilemektedir. Bunun yanında, güven endeksleri kısa vadede ekonomik aktiviteden çok enflasyon
gelişmelerinden etkilenmektedir. Güven ve oynaklık arasında bulunan güçlü ilişki finansal piyasalarda istikrarı
hedef alan politikaların, tüketim eğilimi açısından da önemli olduğunu vurgulamaktadır.

1
I would like to thank Mahmut Günay, Utku Özmen, Çağlar Yüncüler and the anonymous referee for their invaluable comments
and suggestions.

Research Notes in Economics 1

 5
No: 2015-16 / 08 September 2015

1. Introduction

Real economic activity data are published with a considerable time lag. This makes
timely data at higher frequency valuable to get an insight of current economic situation.
Consumer Confidence Indices (CCIs) are among such indicators and they are followed
closely by analysts, policy makers, as well as the public. Analysts tend to relate the changes
in consumer confidence to macroeconomic indicators such as private consumption demand,
economic growth or inflation. Hence, understanding the factors affecting CCIs is important for
better interpretation of confidence data releases. This note aims to document the factors that
affect short term fluctuations in CCIs, with a special emphasis on financial volatility. To this
end, we select four different indices as measures of consumer sentiment: Overall CCI from
CNBC-e and from TURKSTAT-CBRT surveys, as well as Propensity to Consume Index
(PCI) and Consumer Expectations Index (CEI) from the CNBC-e survey.

There is a vast literature on the relation between consumer confidence indices and
economic activity and macroeconomic indicators. Generally, the focus is on the predictive
power of CCIs for private consumption.1 Even though measuring the forecast power is
important from the viewpoint of a policy-maker, investigating the drivers of consumer
confidence is also a valuable endeavor. Considering the large fluctuations in CCIs that
Turkey has experienced recently, we study whether such fluctuations signal merely changes
in economic activity, private consumption and whether they are also sensitive to other
shocks. We conjecture that not only certain macroeconomic and financial indicators affect
confidence but also perceived distress and volatility are quite significant for short term CCI
fluctuations. Using foreign exchange (FX) and Borsa Istanbul (BIST) Index volatility, Figure 1
and Figure 2 present the monthly relation between changes in CNBC-e consumer confidence
index and financial volatility.2 Figures show that higher financial volatility is associated with
lower consumer sentiment.

1
There are a number of studies that investigate the forecast power of confidence indices for private consumption for different
countries. See for example Bram and Ludvigson (1998) and Ludvigson (2004) for US, Dion (2006) for Euro Area, Karasoy and
Yüncüler (2015) for Turkey, and Golinelli and Parigi (2004) for a cross country comparison.
2
See Section 2 for calculation of these indicators.

Research Notes in Economics 2

 5
No: 2015-16 / 08 September 2015

Figure 1. CNBC-e Consumer Confidence Figure 2. CNBC-e Consumer Confidence

Index and FX Volatility* Index and BIST Volatility*
15 CNBC-e Monthly Change 2.2 15 CNBC-e Monthly Change 2.5

10 FX Basket Volatility (Right Axis) 1.7 10 BIST Volatility (Right Axis) 2

1.5
5 1.2 5
1
0 0.7 0
0.5
-5 0.2 -5
0
-10 -0.3 -10
-0.5
-15 -0.8 -15 -1
-20 -1.3 -20 -1.5
03-10
06-10
09-10
12-10
03-11
06-11
09-11
12-11
03-12
06-12
09-12
12-12
03-13
06-13
09-13
12-13
03-14
06-14
09-14
12-14
03-15

03-10
06-10
09-10
12-10
03-11
06-11
09-11
12-11
03-12
06-12
09-12
12-12
03-13
06-13
09-13
12-13
03-14
06-14
09-14
12-14
03-15
Source: CNBC-e, CBRT. Source: CNBC-e, BIST.
*Volatility measure is author’s own calculation.. *Volatility measure is author’s own calculation.

Many studies present evidence on the co-movement of consumer confidence and

financial variables. Among earlier studies, Katona (1975) states that; consumption is affected
by confidence through individuals’ perception about their future financial conditions. Indeed,
Lemmon and Portniaguina (2006) and Otoo (1999) show that consumer sentiment is very
closely related to US stock market developments. Çelik (2010) and Çelik et al. (2010) model
consumption as a function of financial variables and production in a long-term setup for
Turkey. Similarly, Kandır (2006) and Görmüş and Güneş (2010) also focus on the link
between consumer confidence and stock market. Notably, these studies focus on the long-
term relationship between confidence and financial variables for Turkey, ignoring the
potential effect of financial volatility and/or financial distress on consumer confidence. With
this motivation, this note aims to provide evidence with a short term perspective on how
various macroeconomic and financial factors and volatility affect the consumer sentiment in
Turkey. Any detrimental effect of financial volatility on confidence will provide evidence in
favor of Mishkin’s (1978) “liquidity hypothesis,” which states that during times of financial
distress, confidence worsens, as a result; agents prefer to hold liquid assets and that’s why
the durable good consumption declines.

There are many studies showing how uncertainty affects economic activity via
households’ and firms’ expectations. From the households’ expectations perspective, Carroll
(1996) and Romer (1990) show that higher uncertainty leads agents to build up a buffer stock
of savings. Benito (2004) finds that perceived uncertainty leads to delays in purchases of
durable consumption goods. Bachman et al. (2013) point out the “wait and see” mechanism
for firms in highly volatile times. When confidence falls, firms postpone investment plans and
this eventually leads to a drop in overall economic activity. This note contributes to the
literature by providing evidence of a strong relation between households’ confidence and
financial volatility.

Research Notes in Economics 3

 5
No: 2015-16 / 08 September 2015

2. Data

2.1 Consumer Confidence Indices in Turkey

We use monthly consumer confidence surveys of CNBC-e and Turkish Statistical
Institute (TURKSTAT) conducted in collaboration with Central Bank of the Republic of Turkey
(CBRT). CNBC-e survey has been conducted monthly since January 2002 and consists of 5
questions. First two questions assess the respondent’s personal condition and expectations.
Following two questions evaluate the expectations with regard to the current and future state
of the Turkish economy. The fifth question measures current durable good consumption
tendency of consumers. There are two sub-indices of CCI: Consumer’s Expectations Index
(CEI), calculated from question 2 and question 4, and Propensity to Consume Index (PCI),
calculated using the fifth question only.

TURKSTAT-CBRT Survey aims to measure the present situation assessments and future
period expectations of consumers' on personal financial standing and general economic
course. CCI is calculated from four questions on financial and general expectations,
expectations on unemployment and probability of saving.3 Both CNBC-e and TURKSTAT-
CBRT CCIs are evaluated between 0 and 200, where a value above 100 (below 100) refers
to optimistic (pessimistic) consumer confidence.

2.2 Macroeconomic Data and Volatility Indicators

Non-farm unemployment rate and Consumer Price Index (CPI) (2003=100) are obtained
from TURKSTAT. Non-farm unemployment data is available from 2005, whereas CPI is
available from 2003 and monthly changes are computed with seasonally adjusted data. To
proxy economic activity and aggregate demand, we generate monthly GDP figures by using
Fernandez (1981) methodology. The last GDP observation is for the first quarter of 2015. For
real stock price index, we deflate the nominal BIST 100 Index by CPI. Daily US Dollar/TL and
Euro/TL parities and foreign exchange basket are obtained from Bloomberg. For the interest
rate, we use the personal loan rate, which is announced weekly, obtained from the Banking
Regulation and Supervision Agency (BRSA). The reason for using personal loan rate is that
it is easily observed and followed by households. Moreover, it generally co-moves with other
interest rates. Daily and weekly data are aggregated to obtain monthly figures as simple
averages.

3
The survey started to be applied in accordance with the Joint Harmonized European Union Programme of Business and
Consumer Surveys in 2012. A back-casting method using Reg-ARIMA model has been applied in order to produce the historical
index values of the new Consumer Confidence Index for the 2004-2011 period.

Research Notes in Economics 4

 5
No: 2015-16 / 08 September 2015

Financial volatility is calculated for the foreign exchange basket, personal loan rate, and
natural logarithm of BIST Index. First, within-month standard deviations are calculated using
daily or weekly data, and then the series are standardized, which makes interpretation of the
regression coefficients easier. The VIX index measuring the implied volatility of S&P 500
index options is reported by the Chicago Board Options Exchange. Using this index, we will
be able to capture the effect of perceived external volatility on financial markets. Moreover
controlling this variable is also important for identifying the effect of domestic volatility
indicators on confidence, since one would expect external volatility to affect both.

3. Regression Models

3.1 Simple Model

In this study, we focus solely on the short-term changes in confidence indices. The long-
run relation of the level of indices and certain macroeconomic variables are presented in
Çelik (2010) and Çelik et al. (2010). At the simplest setup, one may relate the changes in
consumer sentiment to the changes in economic activity and expect some persistence. We
take month-on-month changes of GDP as a proxy for economic activity.4 Thus our simple
model is given in Equation 1.5

∆𝐶𝐶𝐼𝑡 = 𝛼0 + ∑𝑘𝑖=0 𝛽𝑖 ∆𝐺𝐷𝑃𝑡−𝑖 + ∑𝑚

𝑖=1 𝛾𝑖 ∆𝐶𝐶𝐼𝑡−𝑖 + 𝜀𝑡 . (1)

We see that, in this very simple model, consumer confidence is related to its own lags but
does not depend on economic activity indicator (Table 1 in Appendix, columns 1-4).

3.1 Macroeconomic Indicators

Next, following the literature, we add inflation and non-agricultural unemployment rate as
explanatory variables. Even though monthly GDP and some lags of CCIs are found to be
insignificant above, we keep these variables in the model. In what follows, we will preserve
only statistically significant variables in the regressions. It seems that among these
macroeconomic indicators, only inflation matters for confidence. Inflation has a negative
influence on all four types of indices. Notably, especially for PCI (which essentially measures
propensity to consume for durable goods) higher inflation leads to considerable fall in the

4
One may think that instead of using monthly GDP, other economic activity indicators can be used such as industrial
production. We repeat all the regressions with industrial production index and its sub components such as consumption,
investment, durable or non-durable consumption good production and conclude that our main results are robust.
5
The lags which are significant at 10 percent level are kept in regressions. We tried 4 lags at most.

Research Notes in Economics 5

 5
No: 2015-16 / 08 September 2015

index value. After adding these macroeconomic indicators, own lags of consumer sentiment
are not significant except for TURKSTAT-CBRT (Table 1 in Appendix, columns 5-8).

3.2 Financial Indicators

It is not that surprising to observe that consumer confidence is not related to economic
activity or unemployment. Literature documents that the short-term movements in the
confidence are closely related to financial indicators. With this concern, we add the first
difference of the log real BIST 100 index, the first difference of the interest rate for personal
loans, and the change in foreign exchange basket. We omit the GDP and unemployment rate
and keep only variables and lags that are statistically significant following general to specific
(GETS) methodology.6

Results show that financial variables are indeed closely related to consumer
confidence.7 The foreign exchange basket is highly significant for all four indices (Table 1 in
Appendix, columns 9-12). Personal loan rate is statistically significant for PCI and for
TURKSTAT-CBRT. Meanwhile, BIST is statistically significant for CNBC-e and CEI. As
expected, both the rise in interest rate and a depreciation of Turkish lira deteriorate
confidence. A rise in the FX basket is especially detrimental for the PCI index. Positive
changes in the BIST index, on the other hand, improve CNBC-e and CEI CCIs. Note that
even if the financial variables are added to the regressions, inflation stays to be important for
short-term fluctuations in CCIs.

3.3 Financial Volatility

The main goal of this note is to investigate the relation between financial volatility and
confidence level. To this end, we include our volatility measures into the analysis. It is
possible to include the volatility indicators to the regressions one by one or all three at the
same time. We report the results for both cases and continue to keep only statistically
significant variables. One concern is the inclusion of both the change and the standard
deviation of financial variables as explanatory variables. The reason for such a concern is
that some of the volatility information is already embedded into the changes. Therefore,
instead, we include volatility indicators and control for the levels of financial variables. As
reported in Table 2 in Appendix, domestic volatility matters for consumer confidence. In
these specifications, the changes in VIX index are statistically significant as well, suggesting

6
One should note that even if we do not omit and run our regressions with these variables, the findings do not change.
7
At this point, a discussion of simultaneity is necessary. This note constitutes an initial attempt to detect the relation between
financial variables and confidence indices; bi-directional relations should be studied in a VAR type of model which is postponed
as future work.

Research Notes in Economics 6

 5
No: 2015-16 / 08 September 2015

that external volatility is also one of the drivers of the changes in confidence indices. For
CNBC-e CCI, all three domestic volatility indicators are statistically significant in addition to
VIX. For CEI, FX and personal loan volatility indicators are statistically significant. Notably,
for PCI and TURKSTAT-CBRT loan volatility does not affect confidence. For TURKSTAT-
CBRT, only FX volatility is statistically significant. In addition, for PCI the coefficients of FX
and BIST volatility are larger compared to the other CCIs which suggests that the impact of
volatility is much worse for durable good consumption tendency. The observed differences in
different consumer confidence indices may be emanating from sample differences between
the two surveys.

Instead of running regressions with volatility indicators one by one, we may also include
all at once in order to differentiate separate effects of volatility measures. Once we add these
three variables to the regressions, we see that VIX is again highly statistically significant for
all. Coming to domestic volatility indicators, controlling for VIX, FX volatility and personal loan
rate volatility remain statistically significant for CNBC-e CCI. For CEI, personal loan rate and
for PCI, BIST volatility indicators are significant (Table 3 in Appendix).

4. Discussion and Conclusion

The above analyses show that consumer confidence indices in Turkey are mainly
sensitive to inflation, VIX index, and various domestic volatility indicators in the short run. The
findings suggest that there is a close relationship between financial volatility and consumer
sentiment. The result that confidence is affected by volatility indicators may reveal an extra
transmission mechanism in distressful times suggesting that households’ consumption and
saving decisions are sensitive to the presence of uncertainty in the financial markets.
Moreover, our results further enhance the importance of financial stability policies in the
aftermath of the financial crisis. The strong relation between confidence and volatility shows
that policies aiming at stability in financial markets also improve consumption tendency.

The findings on the VIX index trigger further discussions. It might be the case that the
VIX index captures the “news” effects. To be explicit; even if the consumers do not follow
foreign financial markets directly, the news on increased global volatility provided by media,
may inform households. Alternatively, it may simply reflect some part of domestic volatility,
when there is a co-movement between domestic and external volatility. The indices’ different
responses to volatility indicators might also be related to this issue. For example in Table 3 in
Appendix, it seems that TURKSTAT-CBRT CCI does not respond to domestic volatility
indicators but it responds to VIX index. Actually if we exclude VIX and consider domestic
volatility measures, we see that FX volatility becomes statistically significant for this index.

Research Notes in Economics 7

 5
No: 2015-16 / 08 September 2015

Even though the long term concurrent relation is well presented with various VECM
models, the simultaneity might be investigated for the short-run as well. Some researchers
(see for example Otoo (1999) and Starr (2012)) point out that confidence and financial
variables along with other macroeconomic indicators affect each other contemporaneously.
Desroches and Gosselin (2002) conduct a threshold analysis and show that especially during
highly volatile periods, consumer confidence indices’ explanatory and forecast power
increases. They suggest that economists and policy makers should pay attention to
consumer confidence especially in times of economic or political uncertainty. This kind of
VECM or VAR type analysis which includes domestic and external volatility would give an
idea on causality and transmission mechanisms; therefore present a clearer picture. These
are postponed as future work.

Research Notes in Economics 8

 5
No: 2015-16 / 08 September 2015

Appendix

TABLE 1: Regressions with macro-economic and financial variables

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
CNBC-e CNBC-e CNBC-e TURKSTAT- CNBCe CNBCe CNBCe TURKSTAT- CNBCe CNBCe CNBCe TURKSTAT-
CCI CEI PCI CBRT CCI CCI CEI PCI CBRT CCI CCI CEI PCI CBRT CCI
GDP - MoM -0.262 -0.0764 -0.0514 -0.152 -0.274 0.0226 -0.380 -0.112
(-0.71) (-0.22) (-0.10) (-1.44) (-0.75) (0.07) (-0.72) (-0.95)
L.GDP - MoM -0.437 -0.527 -0.464 0.174 -0.337 -0.354 -0.0762 0.167
(-1.17) (-1.45) (-1.11) (1.65) (-0.93) (-0.98) (-0.20) (1.52)
Non-Agriculture -1.785 -1.060 2.184 -0.384
Unemp. Rate-change (-0.83) (-0.52) (0.72) (-0.52)
CPI - MoM -4.120*** -2.986*** -6.139*** -1.522*** -2.241** -3.152* -0.915**
(-3.24) (-2.85) (-2.78) (-3.66) (-2.34) (-1.78) (-2.38)
D. Real BIST Index - 22.27* 22.01**
logged (1.96) (2.54)
D.Personal Loan Rate -1.623* -0.397***
(-1.92) (-4.17)
D.FX Basket -44.81*** -31.91*** -85.54*** -17.63***
(-3.06) (-2.96) (-3.72) (-5.00)
L.CCI 0.144 0.0324 0.0488 0.264*** 0.0963 0.0238 0.0164 0.234** 0.143*
(1.46) (0.32) (0.62) (2.67) (0.93) (0.20) (0.21) (2.38) (1.74)
L2. CCI -0.273** -0.216* -0.0939 -0.143 -0.149 -0.0783 -0.0564 -0.158* -0.164*
(-2.59) (-1.90) (-1.17) (-1.51) (-1.50) (-0.67) (-0.78) (-1.72) (-1.94)
Constant 0.0548 0.00678 -0.0886 -0.228 2.382** 1.659* 3.227** 0.771** 1.240 -0.156 1.737 0.521*
(0.08) (0.01) (-0.09) (-1.05) (2.42) (1.82) (2.32) (2.52) (1.50) (-0.31) (1.18) (1.88)
Observations 156 156 156 132 122 122 122 122 122 146 122 122
Adjusted R2 0.0834 0.0551 -0.00763 0.0579 0.102 0.0390 0.0474 0.160 0.373 0.204 0.275 0.338

Estimation period: January 2002-June 2015 for (1), (2) and (3). January 2004-June 2015 for (4), March 2005-June 2015 for (5)-(8), (9), (10), (12); January 2003-June 2015 for (11)
Dependent variable is change in CCI varies in regressions 1 to 12.
L denotes lag of that variable.
MoM implies monthly percentage change.
t statistics in parentheses
* p<0.1, ** p<0.05, *** p<0.01

Research Notes in Economics 9

 5
No: 2015-16 / 08 September 2015

TABLE 2: Regressions with volatility indicators

(FX) (Loan (BIST) (FX) (Loan (BIST) (FX) (Loan (BIST) (FX) (Loan Rate) (BIST)
Rate) Rate) Rate)
CNBCe CNBCe CNBCe CNBCe CNBCe CNBCe CNBCe CNBCe CNBCe TURKSTAT- TURKSTAT- TURKSTAT-
CCI - CCI - CCI - CEI - CEI - CEI - PCI - PCI - PCI - CBRT CCI - CBRT CCI - CBRT CCI -
change change change change change change change change change change change change
CPI - MoM -3.613*** -4.205*** -3.898*** -2.460** -2.926*** -2.741*** -4.724** -5.846*** -5.549*** -1.265*** -1.476*** -1.476***
(-3.27) (-3.61) (-3.33) (-2.40) (-2.94) (-2.76) (-2.53) (-2.84) (-2.64) (-3.30) (-3.75) (-3.75)
D.VIX -0.320** -0.452*** -0.377*** -0.336*** -0.342*** -0.625*** -0.443*** -0.0644* -0.138*** -0.138***
(-2.39) (-4.78) (-3.87) (-3.47) (-3.50) (-4.28) (-3.00) (-1.68) (-4.10) (-4.10)
Real BIST Index - -7.937**
logged (-1.99)
L. CCI - change 0.161* 0.181* 0.181*
(1.79) (1.82) (1.82)
FX Basket -1.340* -1.345*** -3.757** -0.709***
Volatility - (-1.89) (-2.79) (-2.50) (-2.87)
Standardized
Personal Loan -0.913** -0.964***
Rate Volatility - (-2.33) (-3.90)
Standardized
BIST Index -1.531** -3.433***
Volatility - (-2.01) (-3.00)
Standardized
Constant 1.806** 2.075** 2.603*** 1.160 1.339 1.390 83.77** 2.926** 4.001*** 0.619** 0.785*** 0.785***
(2.04) (2.35) (2.76) (1.28) (1.59) (1.64) (2.05) (2.14) (2.64) (2.13) (2.70) (2.70)
Observations 122 122 122 122 122 122 122 122 122 122 122 122
Adjusted R2 0.205 0.192 0.201 0.0999 0.134 0.117 0.163 0.131 0.170 0.257 0.197 0.197

Estimation period: March 2005-June 2015.

The regressions are run by domestic volatility indicators one by one as shown above of each regression in parenthesis.
Controlling levels.
Dependent variable is CCI varies in regressions 1 to 4.
L denotes lag of that variable, D denoted first difference of that variable.
MoM implies monthly percentage change.
t statistics in parentheses
* p<0.1, ** p<0.05, *** p<0.01

Research Notes in Economics 10

 5
No: 2015-16 / 08 September 2015

Table 3: Regressions with all three domestic volatilities, added at once

(1) (2) (3) (4)
CNBCe CCI CNBCe CEI CNBCe PCI TURKSTAT-CBRT CCI
CPI - MoM -3.789*** -2.926*** -5.549*** -1.476***
(-3.38) (-2.94) (-2.64) (-3.75)
D.VIX -0.316** -0.336*** -0.443*** -0.138***
(-2.34) (-3.47) (-3.00) (-4.10)
FX Basket Volatility - Standardized -1.327*
(-1.87)
Personal Loan Rate Volatility - Standardized -0.897** -0.964***
(-2.58) (-3.90)
BIST Index Volatility - logged - Standardized -3.433***
(-3.00)
L. CCI 0.181*
(1.82)
Constant 1.761* 1.339 4.001*** 0.785***
(1.97) (1.59) (2.64) (2.70)
Observations 122 122 122 122
Adjusted R2 0.214 0.134 0.170 0.197

Estimation period: March 2005-June 2015

Dependent variable is CCI varies in regressions 1 to 4.
L denotes lag of that variable, D denoted first difference of that variable.
MoM implies monthly percentage change.
t statistics in parentheses
* p<0.1, ** p<0.05, *** p<0.01

References:

Bachmann, R., Elstner, S., & Sims, E. R. (2013). Uncertainty and Economic Activity:
Evidence from Business Survey Data. American Economic Journal: Macroeconomics,
5 (2).

Benito, A. (2004). Does job insecurity affect household consumption? Bank of England
Working Paper No. 220.

Bram, J. and Ludvigson, S. C. (1998). Does Consumer Confidence Forecast Household

Expenditure?: A Sentiment Index Horse Race, FRBNY Economic Policy Review,
June.

Carroll, C (1996). Buffer-stock saving and the life cycle / permanent income hypothesis.
NBER Working Paper No. 5788.

Çelik, S. (2010). An Unconventional Analysis of Consumer Confidence Index for the Turkish
Economy. International Journal of Economics and Finance Studies, 2 (1).

Çelik, S., Aslanoğlu, E., and Uzun, S. (2010). The Relationship between Consumer
Confidence and Financial Market: Variables in Turkey during the Global Crisis. Topics
in Middle Eastern and North African Economies, electronic journal, Volume 12, Middle
East Economic Association and Loyola University, Chicago.
Desroches, B., and Gosselin, M. A. (2002). The Usefulness of Consumer Confidence Indices
in the United States. Bank of Canada Working Paper. 2002-22.

Research Notes in Economics 11

 5
No: 2015-16 / 08 September 2015

Dion, D. P. (2006). Does Consumer Confidence Forecast Household Spending? The Euro
Area Case. MPRA Paper, No. 911.
Fernandez, R. (1981). A Methodological Note on the Estimation of Time Series. The Review
of Economics and Statistics, 63 (3).

Golinelli, R. and Parigi, G. (2004). Consumer Sentiment and Economic Activity: A Cross-
Country Comparison, Journal of Business Cycle Measurement and Analysis, 1 (2).

Görmüş, Ş. and Güneş, S. (2010). Consumer Confidence, Stock Prices and Exchange
Rates: The Case of Turkey. Applied Econometrics and International Development, 10
(2).

Kandır, S. Y. (2006). Tüketici Güveni ve Hisse Senedi Getirileri İlişkisi: İMKB Mali Sektör
Şirketleri Üzerinde Bir Uygulama. Çukurova Üniversitesi Sosyal Bilimler Enstitüsü
Dergisi, 15 (2).

Karasoy, H. G. and Yüncüler, Ç. (2015). The Explanatory Power and the Forecast
Performance of Consumer Confidence Indices for Private Consumption Growth in
Turkey. Research and Monetary Policy Department Working Paper Series, No. 15/19,
Central Bank of the Republic of Turkey.

Katona, G. (1975). Psychological Economics. Elsevier.

Lemmon, M., & Portniaguina, E. (2006). Consumer Confidence and Asset Prices: Some
Empirical Evidence. Review of Financial Studies, 19 (4).

Ludvigson, S. C. (2004). Consumer Confidence and Consumer Spending, Journal of

Economic Perspectives, 18 (2).

Mishkin, F. S. (1978). Consumer Sentiment and Spending on Durable Goods. Brookings

Papers on Economic Activity. 1.

Otoo, M. W. (1999). Consumer Sentiment and the Stock Market (No. 1999-60). Board of
Governors of the Federal Reserve System (US).

Romer, C (1990). The Great Crash and the Onset of the Great Depression. Quarterly Journal
of Economics. 105 (3).

Starr, M. A. (2012). Consumption, Sentiment, and Economic News. Economic Inquiry, 50 (4).

Research Notes in Economics 12

 5
No: 2015-16 / 08 September 2015

The CBT Research Notes in Economics is a refereed publication prepared by Central Bank of Turkey with the purpose of
disseminating results from research on the Turkish economy with a particular emphasis on macroeconomic developments
and issues relevant for monetary policy, and thereby contributing to discussions on economic issues in a timely fashion. The
views and opinions expressed in the CBT Research Notes in Economics do not necessarily represent the official views of the
Central Bank of Turkey and its staff. A written permission must be obtained to reprint the entire text of notes published in
CBT Research Notes in Economics series. For questions and comments:

Editör, Ekonomi Notları, TCMB İdare Merkezi, İstiklal Cad, No: 10, Kat:15, 06100, Ulus/Ankara/Türkiye.
E-mail: ekonomi.notlari@tcmb.gov.tr

Research Notes in Economics 13