Are consumer sentiment shocks state-dependent?

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 paper develops a real business cycle model with five types of fundamental shocks and one “equity sentiment shock” that captures fluctuations driven by animal spirits. The representative agent's perception that movements in equity value are partly driven by sentiment turns out to be close to self‐fulfilling. I solve for the sequences of shock realizations that allow the model to exactly replicate the observed time paths of US consumption, investment, hours worked, the stock of physical capital, capital's share of income, and the S&P 500 market value from 1960.Q1 onwards. The model‐identified sentiment shock is strongly correlated with survey‐based measures of US consumer sentiment. Counterfactual scenarios with the model suggest that the equity sentiment shock has an important influence on the paths of most US macroeconomic variables.

This paper develops a real business cycle model with five types of fundamental shocks and one "equity sentiment shock" that captures animal spirits-driven fluctuations. The representative agent's perception that movements in equity value are partly driven by sentiment turns out to be close to self-fulfilling. I solve for the sequences of shock realizations that allow the model to exactly replicate the observed time paths of U.S. consumption, investment, hours worked, the stock of physical capital, capital's share of income, and the S&P 500 market value from 1960.Q1 onwards. The model-identified sentiment shock is strongly correlated with survey-based measures of U.S. consumer sentiment. Counterfactual scenarios with the model suggest that the equity sentiment shock has an important influence on the paths of most U.S. macroeconomic variables. [The first version of this paper was July 4, 2018.]

We estimate the dynamic causal effects of consumer sentiment shocks in the US. We identify autonomous changes in survey evidence on consumer confidence using fatalities in mass shootings as an instrument. We find the instrument to be significant for an aggregate index of consumer expectations and also back up the identification scheme with micro evidence that exploits the geographical variation in mass shootings. Sentiment shocks have real macroeconomic effects. A negative sentiment shock is recessionary: It sets off a persistent decline in consumer confidence and induces a contraction in industrial production, private sector consumption and in the labour market, while having less evident nominal effects. Finally, sentiment shocks explain a non-negligible part of the cyclical fluctuations in consumer confidence and real macroeconomic aggregates.

Measuring news sentiment

This paper demonstrates state-of-the-art text sentiment analysis tools while developing a new time-series measure of economic sentiment derived from economic and financial newspaper articles from January 1980 to April 2015. We compare the predictive accuracy of a large set of sentiment analysis models using a sample of articles that have been rated by humans on a positivity/negativity scale. The results highlight the gains from combining existing lexicons and from accounting for negation. We also generate our own sentiment-scoring model, which includes a new lexicon built specifically to capture the sentiment in economic news articles. This model is shown to have better predictive accuracy than existing, “off-the-shelf”, models. Lastly, we provide two applications to the economic research on sentiment. First, we show that daily news sentiment is predictive of movements of survey-based measures of consumer sentiment. Second, motivated by Barsky and Sims (2012), we estimate the impulse responses of macroeconomic variables to sentiment shocks, finding that positive sentiment shocks increase consumption, output, and interest rates and dampen inflation.

This paper demonstrates state-of-the-art text sentiment analysis tools while developing a new time-series measure of economic sentiment derived from economic and financial newspaper articles from January 1980 to April 2015. We compare the predictive accuracy of a large set of sentiment analysis models using a sample of articles that have been rated by humans on a positivity/negativity scale. The results highlight the gains from combining existing lexicons and from accounting for negation. We also generate our own sentiment-scoring model, which includes a new lexicon built specifically to capture the sentiment in economic news articles. This model is shown to have better predictive accuracy than existing, “off-the-shelf”, models. Lastly, we provide two applications to the economic research on sentiment. First, we show that daily news sentiment is predictive of movements of survey-based measures of consumer sentiment. Second, motivated by Barsky and Sims (2012), we estimate the impulse responses of macroeconomic variables to sentiment shocks, finding that positive sentiment shocks increase consumption, output, and interest rates and dampen inflation.

ABSTRACTWe show that market sentiment shocks create demand shocks for risky assets and a systematic risk for assets. We measure a market sentiment shock as the unexpected portion of the University of Michigan Consumer Sentiment Index’s growth. This shock prices stock returns in arbitrage pricing theory framework at 1% after controlling for market, size, value, momentum, and liquidity risk factors. Its premium lowered the implied risk aversion by 97.9% to 11.46 between 1978 and 2009 in our sentiment consumption-based capital-asset-pricing model. Merton’s [1973. “An Intertemporal Capital Asset Pricing Model.” Econometrica 41: 867–887]. intertemporal capital-asset-pricing model reconfirms our finding that this market sentiment shock is a systematic risk factor that provides investment opportunities.

This paper studies the dynamic effects of the fiscal policy shock on private activity using an array of vector autoregressive models for the post-war U.S. data. We are particularly interested in the role of consumer sentiment in the transmission of fiscal stimulus. Our major findings are as follows. Private spending fails to rise persistently in response to government spending shocks, while they exhibit persistent and significant increases when the sentiment shock occurs. Employing not only linear but also nonlinear state-dependent VAR model estimations, we show that the government spending shock generates consumer pessimism in all phases of business cycle resulting in subsequent decreases in private activity, which ultimately weakens the effectiveness of the fiscal policy. Our counterfactual simulation exercises confirm the important role of sentiment in propagating fiscal stimulus to private spending.

The Market Sentiment Premium: A Sentimental Consumption Approach