Working Papers
Stock Prices Under Pressure: How Tax and Interest Rates Drive Seasonal Variation in Expected Returns (with J. Kang, T. Pekkala, and C. Polk), link
ABSTRACT: We show that interest rates drive mispricing at the turn of a tax period as investors face the trade-off between selling a temporarily-depressed stock this period and selling next period at fundamental value, but with tax implications delayed accordingly. We confirm these patterns in US returns, volume, and individual selling behaviour as well as in UK data where tax and calendar years differ. At quarter-end, the trade-off is only present following recessions, consistent with the tax code. We then link a significant portion of the variation in the risks and abnormal returns of size, value, and momentum to tax-motivated trading.
Term Structure(s) of the Equity Risk Premium (with L. Gomes), under revision, link
ABSTRACT: UNDER REVISION By simultaneously using dividend and variance swap data, we show how the term structure of the equity risk premium varies over time and how its shape is affected by liquidity risk premia. The term structure is always positively sloped, while funding liquidity premia and betas explain the high unconditional returns for all dividend claims. Alphas for short-dated dividend claims become negative, whereas alphas for long-dated claims seem to be positive. The term structure slope varies positively with the market risk premium, but it is never negative relative to the first contract – due to the nearly zero risk premium in the first maturity – and rarely hump-shaped in some empirical models. We demonstrate how the maturity term structure – the risk premium for dividend strips with different maturities – is connected to both the horizon term structure – linked to the variance swap term structure – and various funding liquidity measures. The risk premium is on average increasing with investment horizon, while the 𝑚𝑎𝑡𝑢𝑟𝑖𝑡𝑦 risk premium depends primarily on the short-horizon risk premium, implying that short-horizon investors are the marginal ones. All our results hold in the US, the UK, Europe and Japan. All these facts are consistent with, for instance, a long-run risk model with jump risks.
Gambling, Risk Appetite and Asset Pricing (with C. Carvalho, D. Cordeiro and E. Zilberman), link
ABSTRACT: A measure of the propensity to gamble in casinos constructed without any asset price data provides relevant information for asset pricing. It improves the fit of conditional asset pricing models such as the conditional CAPM, explains cross-sectional differences in future returns for portfolios sorted on various characteristics, and helps forecast market and portfolio excess returns out of sample. The relationship between risk appetite and asset prices appears to be mainly explained by simultaneous changes in risk and risk premia, but some results suggest that our measure may also capture changes in investor sentiment.
Forecasting Large Realized Covariance Matrices: The Benefits of Factor Models and Shrinkage (with M. Medeiros and D. Brito), link
ABSTRACT: We propose a model to forecast very large realized covariance matrices of returns, applying it to the constituents of the S\&P 500 on a daily basis. To deal with the curse of dimensionality, we decompose the return covariance matrix using standard firm-level factors (e.g. size, value, profitability) and use sectoral restrictions in the residual covariance matrix. This restricted model is then estimated using Vector Heterogeneous Autoregressive (VHAR) models estimated with the Least Absolute Shrinkage and Selection Operator (LASSO). Our methodology improves forecasting precision relative to standard benchmarks and leads to better estimates of the minimum variance portfolios.
Pre-FOMC Announcement Relief (with V. Martello), link
ABSTRACT: We show that the pre-FOMC announcement drift in equity returns occurs mostly in periods of high market uncertainty or risk premium. Specifically, this abnormal return is explained by a significant reduction in the risk premium (implied volatility and variance risk premium) prior to the announcement, but only when the risk premium is high, e.g., when it is above its median. Likewise, the magnitude of the FOMC Cycle and other related patterns varies with uncertainty and risk premium. Market uncertainty measures are persistent and are not related to policy uncertainty or expectations. Markets become only marginally stressed in the days prior to the announcement and changes in uncertainty appear to be of lower frequency. We also explain why recent studies suggest that the pre-FOMC drift might have disappeared in the past decade, as this moderation is due to time variation that was also present in older data. Additionally, CAPM only works on FOMC dates when the risk premium is high, e.g., implied vol above its prior median level. The results are robust to different samples and measures of risk premium and uncertainty.
Sentiment, Electoral Uncertainty and Stock Prices (with C. Carvalho and E. Zilberman), link
ABSTRACT: We study the effect of a huge sports sentiment shock, unrelated to economic conditions or government actions, on stock market outcomes. After Brazil's 7-1 humiliating defeat to Germany in the 2014 World Cup, which is likely to be one of the largest sports sentiment shocks ever, the stock market went up. We provide evidence of two opposing effects on stock prices. One is the usual negative effect due to the investor sentiment channel documented in the literature. This effect was, however, overwhelmed by the arguably rational response of investors to voters' sentiment. In particular, the 7-1 defeat was perceived by stock market participants as a political shock affecting the upcoming close presidential election. To decompose these two effects, we devise an empirical strategy that allows us to compute the component of daily returns associated with political news.
Dividend Factors and the Cross-Section of Stock Returns (with N. Doskov and T. Pekkala), under revision, link to old version
ABSTRACT: We propose a new set of tradable aggregate risk factors that help us understand the cross-section of stock returns. We argue that the true stochastic discount factor is a combination of aggregate return factors that drive equity market returns. Hence, we consider new factors using data such as market dividend swaps and market volatility futures. In the particular case of value and size portfolios, we find that differences in expected returns can be explained by a single-factor projection of the discount factor that loads only on a dividend growth return factor constructed with market dividend swap data. Hence, value and small capitalization stocks have higher expected returns due to their exposure to dividend growth returns implying that growth risks (dividend growth news and/or expected return news associated with dividend growth) are the only source of their risk premia. A tradable dividend level factor and a volatility-based factor are also priced in the cross-section of other stock portfolios sorted on dividend yield, earnings yield and cash-flow-to-price.
Work in Progress
Loan Fee Dispersion and the Cross-Section of Stock Returns (with F. Barbosa, M. Bonomo and L. Mota)
Currency Returns and Interest Rate Slopes (with P.H . Castro)
Hedging Momentum Crashes (with C. Garcia and M. Medeiros)
Short-Term Momentum and Reversals (with C. Garcia and M. Medeiros)
When Stock Returns are Predictable (with I. Honda and M. Medeiros)
Market Predictability and Betas (with P.H. Castro and C. Polk)
Intertemporal Substitution without Consumption Smoothing (with C. Carvalho and M.V. Castro)
Value Premium, Growth Expectations and Extrapolation (with K. Oliveira)
Currency Risk Premia and Equity Prices (with P.H . Castro)
Market Capacity of Alpha Strategies (with I. Honda)
Very Old Stuff
The Excess Comovement of International Stock Markets in Bad Times: A Rational Expectations Equilibrium Model (with P. Veronesi), 2002, link
ABSTRACT: We present an intertemporal, rational expectations equilibrium model where the cross-sectional covariances and correlations of international market returns increase during bad times, as a consequence of an endogenous increase in the uncertainty about the global economy. We assume that the drift rates of the fundamental processes of international economies are jointly affected by an unobservable global business cycle indicator. We show that as investors strive to learn the state of the global economy, their uncertainty fluctuate, thereby affecting the cross-covariances and correlations of asset returns. Excess comovement during bad times is so obtained as a reflection of higher uncertainty. When estimated with data on seven major countries, the model is able to replicate well the historical pattern of international average covariances and correlations.
Predictable Dividends and Returns , 2002
ABSTRACT: The conventional wisdom is that the aggregate stock price is predictable by the lagged price-dividend ratio, and that aggregate dividends follow approximately a random-walk. Contrary to this belief, this paper finds that variation in the aggregate dividends and price-dividend ratio is related to changes in expected dividend growth. The inclusion of labor income in a cointegrated vector autoregression with prices and dividends allows the identification of predictable variation in dividends. Most of the variation in the price-dividend ratio is due to changes in expected returns, but this paper shows that part of variation is related to transitory dividend growth shocks. Moreover, most of the variation in dividend growth can be attributed to these temporary changes in dividends. I also show that the price-dividend ratio (or dividend yield) can be constructed as the sum of two distinct, but correlated, variables that separately predict dividend growth and returns. One of these components, which could be called the expected return state variable, predicts returns better than the price-dividend ratio does.
Stock Prices Under Pressure: How Tax and Interest Rates Drive Seasonal Variation in Expected Returns (with J. Kang, T. Pekkala, and C. Polk), link
ABSTRACT: We show that interest rates drive mispricing at the turn of a tax period as investors face the trade-off between selling a temporarily-depressed stock this period and selling next period at fundamental value, but with tax implications delayed accordingly. We confirm these patterns in US returns, volume, and individual selling behaviour as well as in UK data where tax and calendar years differ. At quarter-end, the trade-off is only present following recessions, consistent with the tax code. We then link a significant portion of the variation in the risks and abnormal returns of size, value, and momentum to tax-motivated trading.
Term Structure(s) of the Equity Risk Premium (with L. Gomes), under revision, link
ABSTRACT: UNDER REVISION By simultaneously using dividend and variance swap data, we show how the term structure of the equity risk premium varies over time and how its shape is affected by liquidity risk premia. The term structure is always positively sloped, while funding liquidity premia and betas explain the high unconditional returns for all dividend claims. Alphas for short-dated dividend claims become negative, whereas alphas for long-dated claims seem to be positive. The term structure slope varies positively with the market risk premium, but it is never negative relative to the first contract – due to the nearly zero risk premium in the first maturity – and rarely hump-shaped in some empirical models. We demonstrate how the maturity term structure – the risk premium for dividend strips with different maturities – is connected to both the horizon term structure – linked to the variance swap term structure – and various funding liquidity measures. The risk premium is on average increasing with investment horizon, while the 𝑚𝑎𝑡𝑢𝑟𝑖𝑡𝑦 risk premium depends primarily on the short-horizon risk premium, implying that short-horizon investors are the marginal ones. All our results hold in the US, the UK, Europe and Japan. All these facts are consistent with, for instance, a long-run risk model with jump risks.
Gambling, Risk Appetite and Asset Pricing (with C. Carvalho, D. Cordeiro and E. Zilberman), link
ABSTRACT: A measure of the propensity to gamble in casinos constructed without any asset price data provides relevant information for asset pricing. It improves the fit of conditional asset pricing models such as the conditional CAPM, explains cross-sectional differences in future returns for portfolios sorted on various characteristics, and helps forecast market and portfolio excess returns out of sample. The relationship between risk appetite and asset prices appears to be mainly explained by simultaneous changes in risk and risk premia, but some results suggest that our measure may also capture changes in investor sentiment.
Forecasting Large Realized Covariance Matrices: The Benefits of Factor Models and Shrinkage (with M. Medeiros and D. Brito), link
ABSTRACT: We propose a model to forecast very large realized covariance matrices of returns, applying it to the constituents of the S\&P 500 on a daily basis. To deal with the curse of dimensionality, we decompose the return covariance matrix using standard firm-level factors (e.g. size, value, profitability) and use sectoral restrictions in the residual covariance matrix. This restricted model is then estimated using Vector Heterogeneous Autoregressive (VHAR) models estimated with the Least Absolute Shrinkage and Selection Operator (LASSO). Our methodology improves forecasting precision relative to standard benchmarks and leads to better estimates of the minimum variance portfolios.
Pre-FOMC Announcement Relief (with V. Martello), link
ABSTRACT: We show that the pre-FOMC announcement drift in equity returns occurs mostly in periods of high market uncertainty or risk premium. Specifically, this abnormal return is explained by a significant reduction in the risk premium (implied volatility and variance risk premium) prior to the announcement, but only when the risk premium is high, e.g., when it is above its median. Likewise, the magnitude of the FOMC Cycle and other related patterns varies with uncertainty and risk premium. Market uncertainty measures are persistent and are not related to policy uncertainty or expectations. Markets become only marginally stressed in the days prior to the announcement and changes in uncertainty appear to be of lower frequency. We also explain why recent studies suggest that the pre-FOMC drift might have disappeared in the past decade, as this moderation is due to time variation that was also present in older data. Additionally, CAPM only works on FOMC dates when the risk premium is high, e.g., implied vol above its prior median level. The results are robust to different samples and measures of risk premium and uncertainty.
Sentiment, Electoral Uncertainty and Stock Prices (with C. Carvalho and E. Zilberman), link
ABSTRACT: We study the effect of a huge sports sentiment shock, unrelated to economic conditions or government actions, on stock market outcomes. After Brazil's 7-1 humiliating defeat to Germany in the 2014 World Cup, which is likely to be one of the largest sports sentiment shocks ever, the stock market went up. We provide evidence of two opposing effects on stock prices. One is the usual negative effect due to the investor sentiment channel documented in the literature. This effect was, however, overwhelmed by the arguably rational response of investors to voters' sentiment. In particular, the 7-1 defeat was perceived by stock market participants as a political shock affecting the upcoming close presidential election. To decompose these two effects, we devise an empirical strategy that allows us to compute the component of daily returns associated with political news.
Dividend Factors and the Cross-Section of Stock Returns (with N. Doskov and T. Pekkala), under revision, link to old version
ABSTRACT: We propose a new set of tradable aggregate risk factors that help us understand the cross-section of stock returns. We argue that the true stochastic discount factor is a combination of aggregate return factors that drive equity market returns. Hence, we consider new factors using data such as market dividend swaps and market volatility futures. In the particular case of value and size portfolios, we find that differences in expected returns can be explained by a single-factor projection of the discount factor that loads only on a dividend growth return factor constructed with market dividend swap data. Hence, value and small capitalization stocks have higher expected returns due to their exposure to dividend growth returns implying that growth risks (dividend growth news and/or expected return news associated with dividend growth) are the only source of their risk premia. A tradable dividend level factor and a volatility-based factor are also priced in the cross-section of other stock portfolios sorted on dividend yield, earnings yield and cash-flow-to-price.
Work in Progress
Loan Fee Dispersion and the Cross-Section of Stock Returns (with F. Barbosa, M. Bonomo and L. Mota)
Currency Returns and Interest Rate Slopes (with P.H . Castro)
Hedging Momentum Crashes (with C. Garcia and M. Medeiros)
Short-Term Momentum and Reversals (with C. Garcia and M. Medeiros)
When Stock Returns are Predictable (with I. Honda and M. Medeiros)
Market Predictability and Betas (with P.H. Castro and C. Polk)
Intertemporal Substitution without Consumption Smoothing (with C. Carvalho and M.V. Castro)
Value Premium, Growth Expectations and Extrapolation (with K. Oliveira)
Currency Risk Premia and Equity Prices (with P.H . Castro)
Market Capacity of Alpha Strategies (with I. Honda)
Very Old Stuff
The Excess Comovement of International Stock Markets in Bad Times: A Rational Expectations Equilibrium Model (with P. Veronesi), 2002, link
ABSTRACT: We present an intertemporal, rational expectations equilibrium model where the cross-sectional covariances and correlations of international market returns increase during bad times, as a consequence of an endogenous increase in the uncertainty about the global economy. We assume that the drift rates of the fundamental processes of international economies are jointly affected by an unobservable global business cycle indicator. We show that as investors strive to learn the state of the global economy, their uncertainty fluctuate, thereby affecting the cross-covariances and correlations of asset returns. Excess comovement during bad times is so obtained as a reflection of higher uncertainty. When estimated with data on seven major countries, the model is able to replicate well the historical pattern of international average covariances and correlations.
Predictable Dividends and Returns , 2002
ABSTRACT: The conventional wisdom is that the aggregate stock price is predictable by the lagged price-dividend ratio, and that aggregate dividends follow approximately a random-walk. Contrary to this belief, this paper finds that variation in the aggregate dividends and price-dividend ratio is related to changes in expected dividend growth. The inclusion of labor income in a cointegrated vector autoregression with prices and dividends allows the identification of predictable variation in dividends. Most of the variation in the price-dividend ratio is due to changes in expected returns, but this paper shows that part of variation is related to transitory dividend growth shocks. Moreover, most of the variation in dividend growth can be attributed to these temporary changes in dividends. I also show that the price-dividend ratio (or dividend yield) can be constructed as the sum of two distinct, but correlated, variables that separately predict dividend growth and returns. One of these components, which could be called the expected return state variable, predicts returns better than the price-dividend ratio does.