Skip to main content

Advertisement

SpringerLink
Log in

On the econometric modelling of consumer sentiment shocks in SVARs

  • Published:
Empirical Economics Aims and scope Submit manuscript

Abstract

This paper applies recently developed methods for modelling systems of I(0) and I(1) variables to SVARs of consumer sentiment. We first model the shock associated with the structural equation for the I(0) consumer sentiment variable as having a permanent effect on the I(1) variables. Here it appears to convey news about future productivity. The contribution of the accumulated consumer sentiment shock to the permanent component of consumption and GDP increases substantially from 2000 to 2007, a finding we relate to recent work on boom–bust productivity episodes. We then model the sentiment shock as having a transitory effect on the I(1) variables. Here it appears to convey little news and is best thought of as an ‘animal spirits’ shock unrelated to productivity. The impact responses suggest that ‘animal spirits’ are not important in either model.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3

Similar content being viewed by others

Notes

  1. In a related contribution, Barsky and Sims (2011) identify news as the shock orthogonal to the reduced-form shock in TFP that best accounts for unexplained variation in measured TFP over a long horizon. Their key result is that a good news shock is associated with an increase in consumption and a decline in output, hours and investment on impact, which is more consistent with the findings in standard DSGE models. Their result is in contrast to the findings of Beaudry and Portier (2006), Beaudry et al. (2010), and Beaudry and Lucke (2009). These authors identify news in SVAR models that include TFP and stock prices and find that a good news shock leads to positive responses of macroeconomic variables on impact (i.e. positive co-movement), a finding more difficult to reconcile with standard DSGE models.

  2. The reduced-form VAR representation is then said to be non-invertible. BLL provide a proof to show that the non-invertibility result holds in general DSGE models with partially informative signals.

  3. By the absence of the noise shock, we mean that the variance of the noise shock is zero so that the signalling variable is fully informative about movements in permanent productivity. In this case, the reduced-form VAR representation is invertible.

  4. Barsky and Sims report that the results from their VARs are nearly identical when they impose a cointegrating relation between consumption and GDP. In these VARs, all the shocks are still permitted to have permanent effects on consumption and GDP.

  5. It is natural to set the SVAR up this way because cointegration implies that a permanent-transitory decomposition of the structural shocks associated with the I(1) variables exists. This is due to the Granger Representation theorem.

  6. The question is: “Looking ahead, which would you say is more likely—that in the country as a whole we’ll have continuous good times during the next 5 years or so, or that we will have periods of widespread unemployment or depression, or what?”.

  7. A value of 100 indicates neither optimism nor pessimism, while a value of 130 means that the fraction of responses reflecting optimism about the future is above the fraction reflecting pessimism by 30 % points.

  8. The correlation between E5Y and, respectively, E12M, PFE and ICE is 0.91, 0.85 and 0.96.

  9. The structure of Eq. (2) comes from the result shown in Pagan and Pesaran (2008) that the error-correction variable appears in differenced form in any structural equation that has a permanent shock.

  10. Note that in Eq. (2), \(a_{21}^0 E5Y_t +a_{21}^1 E5Y_{t-1}\) can be expressed as \(a_{21}^0 \Delta E5Y_t +(a_{21}^0 +a_{21}^1)E5Y_{t-1}\). The long-run zero restriction is \(a_{21}^0 +a_{21}^1 =0\), leaving only the difference of consumer sentiment to appear in Eq. (6).

  11. The other rows of \(\varGamma (1)\) have elements that are all zero because the effects of the shocks on I(0) variables are transitory.

  12. In the bootstrap of the data generating process (the structural equations for the respective model), the number of replications was 5000.

  13. We decided to test the over-identifying restriction that the sentiment shock has a zero long-run impact on GDP. It is a test of the null hypothesis that the sum of the contemporaneous and lagged coefficients on consumer sentiment in the GDP equation is zero. The t test rejected this null at less than the 1 % significance level.

  14. A test of the restriction that the coefficients on the contemporaneous variables in the consumer sentiment equation are jointly zero was rejected by an F test at less than the 1 % significance level. The sentiment equation does not reduce to the one of Model A.

  15. The contribution of the inflation and real interest rate shocks to the forecast error variance in a variable in Model A is the same as in Model B, because both models identify these shocks the same way.

  16. A slightly noticeable quantitative difference appears here when either of the other sentiment measures is used. The contribution of the sentiment shock becomes somewhat larger and that of the GDP shock somewhat smaller at long horizons. For example, for E12M, the sentiment shock contributes 55 % to the FEV of GDP at 60-quarters, while the GDP shock contributes around 35 %.

  17. The contribution is noticeably larger at short horizons when either E12M or ICE is used instead of E5Y, but it is less than 25 % at 1 quarter and falls quickly to be less than 10 % by 8 quarters.

  18. When the regressions are augmented still further with favourable/unfavourable news about the government, the coefficient on unfavourable government is highly significant and many of the economic news variables become statistically insignificant. It may be that unfavourable economic news is being condensed as unfavourable news about the government.

  19. The model in Cao and L’Huillier (2014) is an open-economy real business cycle model, while that in L’Huillier (2012) is a New Keynesian model. The correspondence between the two model types is explored in Cao et al. (2014).

  20. These contributions are the same as those in Model A because the real interest rate and inflation shocks are the same in both models.

References

  • Barsky RB, Sims ER (2012) Information, animal spirits, and the meaning of innovations in consumer confidence. Am Econ Rev 102(4):1343–1377

    Article  Google Scholar 

  • Barsky RB, Sims ER (2011) News shocks and business cycles. J Monetary Econ 58(3):273–289

    Article  Google Scholar 

  • Beaudry P, Dupaigne M, Portier F (2010) The international propagation of news shocks. Oxford University and Toulouse School of Economics, Mimeo, New York

    Google Scholar 

  • Beaudry P, Lucke B (2009) Letting different views about business cycles compete. In: Acemoglu D, Rogoff K, Woodford M (eds) NBER Macroeconomics Annual 2009, vol. 24(1), pp. 413–455

  • Beaudry P, Portier F (2014) News-driven business cycles: insights and challenges. J. Econ. Lit. 52(4):993–1074

    Article  Google Scholar 

  • Beaudry P, Portier F (2006) News, stock prices and economic fluctuations. Am Econ Rev 96(4):1293–1307

    Article  Google Scholar 

  • Blanchard OJ, L’Huillier J-P, Lorenzoni G (2013) News, noise, and fluctuations: an empirical exploration. Am Econ Rev 103(7):3045–3070

    Article  Google Scholar 

  • Cao D, L’Huillier J-P (2014) Technological revolutions and the three great slumps: a medium-run analysis. Mimeo, New York

    Google Scholar 

  • Cao D, L’Huillier J-P, Yoo D (2014) The New Keynesian model and the small open economy RBC model: equivalence results for consumption. Mimeo, New York

    Google Scholar 

  • Cochrane JH (1994) Shocks. Carnegie–Rochester conference series on public policy 41:295–364

    Article  Google Scholar 

  • Fernald J (2012) A quarterly, utilization-adjusted series on total factor productivity. Federal Reserve Bank of San Francisco, Working Paper 2012–19

  • Fisher LA, Huh H-S, Pagan AR (2015) Econometric methods for modelling systems with a mixture of \(I(1)\) and \(I(0)\) variables. J. Appl. Econ. doi:10.1002/jae.2459

    Google Scholar 

  • L’Huillier J-P (2012) Did the US consumer overreact? A test of rational expectations. Econ Lett 116(2):207–209

    Article  Google Scholar 

  • Pagan AR (2013) Patterns and their uses. National Centre for Econometric Research Working Paper No. 96

  • Pagan AR, Pesaran MH (2008) Econometric analysis of structural systems with permanent and transitory shocks. J Econ Dyn Control 32(10):3376–3395

    Article  Google Scholar 

  • Shapiro MD, Watson MW (1988) Sources of business cycle fluctuations. In: Fischer S (Ed.) NBER Macroeconomics Annual 1988, vol. 3, pp. 111–148

Download references

Author information

Authors and Affiliations

  1. Department of Economics, Macquarie University, Sydney, Australia

    Lance A. Fisher

  2. School of Economics, Yonsei University, 50 Yonsei-ro, Seodaemun-gu, Seoul, 120-749, Republic of Korea

    Hyeon-seung Huh

Authors
  1. Lance A. Fisher
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Hyeon-seung Huh
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Hyeon-seung Huh.

Additional information

We thank Adrian Pagan for bringing this topic to our attention and for helpful discussion, and two anonymous referees for useful feedback. Any errors or omissions are our own. The first author gratefully acknowledges financial support from the Australian Research Council (Grant DP120102239). The second author thanks the Ministry of Education of the Republic of Korea and the National Research Foundation of Korea (NRF-2015S1A5A2A01010187) for funding.

Rights and permissions

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Fisher, L.A., Huh, Hs. On the econometric modelling of consumer sentiment shocks in SVARs. Empir Econ 51, 1033–1051 (2016). https://doi.org/10.1007/s00181-015-1038-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00181-015-1038-4

Keywords

JEL Classification

Search

Navigation