Momentum Reversal Research Articles

Hedge Fund Strategies: A non-Parametric Analysis

We investigate why top performing hedge funds are successful. We find evidence that top performing hedge funds follow a different strategy than mediocre performing hedge funds as they accept fewer risk factors that mostly anticipate the troubling economic conditions prevailing after 2006. Holding alpha performance constant, top performing funds mostly avoid relying on passive investment in illiquid investments but earn risk premiums by accepting market risk. Additionally, they seem able to exploit fleeting opportunities leading to momentum profits while closing losing strategies thereby avoiding momentum reversal.

Momentum, Reversal, and the Firm Fundamental Cycle

We link momentum and long-run return reversal to the cyclic behavior of firm fundamentals, which are represented by a fundamental index that summarizes succinctly and efficiently a broad range of business activities at firm level. In responding to repeated unanticipated positive (negative) shocks in fundamentals, investors continue to raise (lower) prices for winner (loser) firms, yielding momentum. However, due to the cyclicality of firm fundamentals, the unanticipated positive (negative) shocks decrease in magnitude overtime and eventually reverse, generating the reversal pattern. In addition, we find that firm fundamentals can explain stronger momentum in microcap stocks, and a long/short decile portfolio based on the firm fundament index outperforms the popular momentum portfolio substantially by doubling the Sharpe ratio and does not suffer crashes.

Understanding Momentum and Reversal

Stock momentum, long-term reversal, and other past return characteristics that predict future returns also predict future realized betas, suggesting these characteristics capture time-varying risk compensation. We formalize this argument with a conditional factor pricing model. Using instrumented principal components analysis, we estimate latent factors with time-varying factor loadings that depend on observable firm characteristics. We show that factor loadings vary significantly over time, even at short horizons over which the momentum phenomenon operates (one year), and this variation captures reliable conditional risk premia missed by other factor models commonly used in the literature. Our estimates of conditional risk exposure can explain a sizable fraction of momentum and long-term reversal returns and can be used to generate even stronger return predictions.

Redundant Information and Predictable Intraday Returns

This paper examines how well investors distinguish between genuinely novel private information and information that already is priced (labeled information). We derive a structural model of stock price returns that identifies investors’ non-Bayesian weighting of redundant information distinctly from information asymmetry, transaction costs, and serially correlated liquidity trader demand. We estimate this model using five-minute, 12-minute, and 30-minute returns and find that, on average, investors behave as if over 47 percent of the information content in the immediately prior price change is private information. These results suggest an information-processing mechanism that drives momentum and mean reversion in intraday returns.

Alpha Momentum and Alpha Reversal in Country and Industry Equity Indexes

Do past alphas predict future country and industry returns? Examination of equity indexes from 51 stock markets between 1973 and 2018 allows us to demonstrate new return patterns in the cross-section of country and industry returns. Past short-term (long-term) alphas positively (negatively) predict future returns. These phenomena can be translated into effective international equity allocation strategies, producing economically and statistically significant raw and risk-adjusted returns. The profitability is robust to many considerations, including alternative alpha models, the effect of trading costs, different holding periods, or subsample and subperiod analyses. Also, the alpha momentum and reversal subsume their return-based counterparts. Finally, the alpha-based patterns are particularly pronounced following bull markets and across the markets characterized by high arbitrage constraints, supporting the behavioral explanation of the anomalies.

Testing extensions of Fama & French models: A quantile regression approach

This research compares twelve different factor models in explaining variations in U.S. sector returns between Nov. 1989 and Feb. 2014 using the quantile regression approach. Specifically, these models rely on the Fama and French three- and five-factor models (1993 and 2015), adding other explanatory factors such as nominal interest rates, and its components: real interest and expected inflation rates, the Carhart (1997) risk factor for momentum and for momentum reversal and the Pastor and Stambaugh (2003) traded liquidity factor. Additionally, for robustness, this paper includes the alternative Amihud (2002) illiquidity measure, as well as analyses the validity of the proposed models, depending upon the stage of the economy, by distinguishing between recession and expansion sub-periods. This paper concludes that the most complete model shows the highest explanatory power. Furthermore, the extreme quantile 0.1 of the return distribution shows the best results in all the factor models, and some sectors, such as Industrials and Financials, consistently evidence more statistically significant coefficients and thus higher Adj. R2 values in the total sample as well as in recession and expansion sub-periods. In contrast, the Utilities sector consistently shows the lowest explanatory power for all the models, quantiles and periods.

Monotonic Effects of Characteristics on Returns

This paper considers the problem of modeling a firm’s expected return as a nonlinear function of its observable characteristics. We investigate whether theoretically-motivated monotonicity constraints on characteristics and nonstationarity of the conditional expectation function provide statistical and economic benefit. We present an interpretable model that has similar out-of-sample performance to black-box machine learning methods. With this model, the data provide support for monotonicity and time variability of the conditional expectation function. Additionally, we develop an approach for characteristic selection using loss functions to summarize the posterior distribution. Standard unexplained volume, short-term reversal, size, and variants of momentum are found to be significant characteristics, and there is evidence this set changes over time.

International Real Estate Review

We examine the relation between risk and returns in the U.S. residential housing market. We find that the risk of house price changes and the magnitude relative to the risk of income changes vary with economic conditions. We measure the excess risk of house price changes by adjusting for the risk of income changes and economic variables associated with the real estate and financial sectors of the economy, and find a significant and positive relation between house price changes and excess risk. We also find that excess risk has significantly adverse effects on the short-run momentum and long-run reversal of house price changes across metro areas, thus implying that excess risk induces price rigidity and helps to explain for the serial correlations in price changes in the U.S. single-family housing market.

Deep Learning and the Cross-Section of Stock Returns: Neural Networks Combining Price and Fundamental Information

Traditional empirical finance research uses hand-engineering and trial-and-error to look for the anomalies in the cross-section of stock returns. In this paper, we take advantage of deep learning and utilize both the price and fundamental information to separate stocks’ winners from losers. For the first model, we used a 2-layer Long Short-Term Memory (LSTM) neural network with past 80 days’ return information as inputs to predict the next day’s return and find a before-trading-cost monthly return of 29.58% with a t-statistic of 26.81. The return of the stocks shows a strong short-term reversal pattern. For the second model, we design a novel 2-layer LSTM and Multi-layer Perceptron (MLP) hybrid neural network and utilize the monthly return and annual accounting data to predict the returns in the next month. We achieve a monthly return of 2.37% with a t-statistic of 8.97 before trading cost from 1993 through 2017. We use TAQ intraday data to explicitly estimate the trading cost and find that profits of the daily trading strategy in the first model turn negative. However, the trading strategy utilizing both the price and fundamental information in the second model keeps significantly positive with a monthly return of 1.57% and t-statistic of 6.03 after the trading cost. We show that the 2-layer LSTM and MLP hybrid model performs better than the MLP-only and the hand-engineering momentum and short-term reversal double sort trading strategy.

Understanding Momentum and Reversals

Stock momentum, long-term reversal, and other past return characteristics that predict future returns also predict future realized betas, suggesting these characteristics capture time-varying risk compensation. We formalize this argument with a conditional factor pricing model. Using instrumented principal components analysis, we estimate latent factors with time-varying factor loadings that depend on observable firm characteristics. We show that factor loadings vary significantly over time, even at short horizons over which the momentum phenomenon operates (one year), and this variation captures reliable conditional risk premia missed by other factor models commonly used in the literature. Our estimates of conditional risk exposure can explain a sizable fraction of momentum and long-term reversal returns and can be used to generate even stronger return predictions.

Reversal of orbital angular momentum arising from an extreme Doppler shift

The linear Doppler shift is familiar as the rise and fall in pitch of a siren as it passes by. Less well known is the rotational Doppler shift, proportional to the rotation rate between source and receiver, multiplied by the angular momentum carried by the beam. In extreme cases the Doppler shift can be larger than the rest-frame frequency and for a red shift, the observed frequency then becomes "negative." In the linear case, this effect is associated with the time reversal of the received signal, but it can be observed only with supersonic relative motion between the source and receiver. However, the rotational case is different; if the radius of rotation is smaller than the wavelength, then the velocities required to observe negative frequencies are subsonic. Using an acoustic source at [Formula: see text]100 Hz we create a rotational Doppler shift larger than the laboratory-frame frequency. We observe that once the red-shifted wave passes into the "negative frequency" regime, the angular momentum associated with the sound is reversed in sign compared with that of the laboratory frame. These low-velocity laboratory realizations of extreme Doppler shifts have relevance to superoscillatory fields and offer unique opportunities to probe interactions with rotating bodies and aspects of pseudorelativistic frame translation.

Buy-Side Competition and Momentum Profits

We show that a new measure of buy-side competition explains momentum profits. The momentum quintile spread is 1.11% when competition is low and negligible when competition is high. Better alphas are attained with superior Sharpe and Sortino ratios, no negative skewness and in more investible strategies featuring value-weighted portfolios and large capitalization stocks. Stock characteristics traditionally related to momentum do not explain our results. Tests based on long-term reversals, the trading patterns of funds, their style peers, distant funds, and retail investors suggest that slow information diffusion explains the large momentum spreads and momentum reversals in low competition markets.

Momentum and Reversal in Financial Markets with Persistent Heterogeneity

This paper investigates whether short-term momentum and long-term reversal may emerge from the wealth reallocation process taking place in speculative markets. We assume that there are two classes of investors who trade long-lived assets by holding constantly rebalanced portfolios based on their beliefs. Provided beliefs, and thus portfolios, are sufficiently diversified, all investors survive in the long-run and, due to waves of mispricing, the resulting equilibrium returns exhibit long-term reversal. If, moreover, asset dividends are positively correlated, investors’ profitable trades become positively correlated too, thus generating short-term momentum in equilibrium returns. We use the model to replicate the performance of the Winners and Losers portfolios highlighted by the empirical literature and to provide insights on how to improve upon them. Finally, we show that dividend positive autocorrelation is positively related to momentum and negatively related to reversal while diversity of beliefs is positively related to both momentum and reversal.

Semiclassical analysis of jm → j'm' transitions in rotationally inelastic collisions in cell experiments.

Recent quantum calculations of rotationally inelastic collisions of NaK (A1Σ+) with He or Ar in a cell experiment are analyzed using semiclassical approximations valid for large quantum numbers. The results suggest a physical interpretation of jm → j'm' transitions based on the vector model and lead to expressions that explicitly involve the initial and final polar angles of the angular momentum of the target molecule. The relation between the polar angle θ and the azimuthal quantum number m links the semiclassical results for the change in polar angle (θ → θ') to quantum results for an m → m' transition. Analytic formulas are derived that relate the location and width of peaks in the final polar angle distribution (PAD) to the K-dependence of the coefficients dK(j, j'), which are proportional to tensor cross sections σK(j → j'). Several special cases are treated that lead to final PADs that are approximately Lorentzian or sinc functions centered at θ' = θ. Another interesting case, "angular momentum reversal," was observed in the calculations for He. This phenomenon, which involves a reversal of the direction of the target's angular momentum, is shown to be associated with oscillatory behavior of the dK for certain transitions. Finally, several strategies for obtaining the dK coefficients from experimental data are discussed.

BM(book-to-market ratio) factor: medium-term momentum and long-term reversal

To explain medium-term momentum and long-term reversal, we use the difference between the optional model and the CAPM model to construct a winner-loser portfolio. According to the CAPM model’s zero explanatory ability with respect to stock market anomalies, we obtain an anomaly interpretative model. This study shows that this anomaly interpretative model can explain stock market perceptions and medium-term momentum. Most importantly, BM is a critical factor in the model’s explanatory ability. We present a robustness test, which includes selecting new sample data, adding new auxiliary variables, changing sample years, and adding industry fixed effects. In general, the BM effect does have considerable explanatory power in medium-term momentum and long-term reversal.

Enhanced configurational sampling with hybrid non-equilibrium molecular dynamics-Monte Carlo propagator.

Molecular dynamics (MD) trajectories based on classical equations of motion can be used to sample the configurational space of complex molecular systems. However, brute-force MD often converges slowly due to the ruggedness of the underlying potential energy surface. Several schemes have been proposed to address this problem by effectively smoothing the potential energy surface. However, in order to recover the proper Boltzmann equilibrium probability distribution, these approaches must then rely on statistical reweighting techniques or generate the simulations within a Hamiltonian tempering replica-exchange scheme. The present work puts forth a novel hybrid sampling propagator combining Metropolis-Hastings Monte Carlo (MC) with proposed moves generated by non-equilibrium MD (neMD). This hybrid neMD-MC propagator comprises three elementary elements: (i) an atomic system is dynamically propagated for some period of time using standard equilibrium MD on the correct potential energy surface; (ii) the system is then propagated for a brief period of time during what is referred to as a "boosting phase," via a time-dependent Hamiltonian that is evolved toward the perturbed potential energy surface and then back to the correct potential energy surface; (iii) the resulting configuration at the end of the neMD trajectory is then accepted or rejected according to a Metropolis criterion before returning to step 1. A symmetric two-end momentum reversal prescription is used at the end of the neMD trajectories to guarantee that the hybrid neMD-MC sampling propagator obeys microscopic detailed balance and rigorously yields the equilibrium Boltzmann distribution. The hybrid neMD-MC sampling propagator is designed and implemented to enhance the sampling by relying on the accelerated MD and solute tempering schemes. It is also combined with the adaptive biased force sampling algorithm to examine. Illustrative tests with specific biomolecular systems indicate that the method can yield a significant speedup.

Experimental optimisation of power for large arrays of cross-flow tidal turbines

Abstract As commercial scale tidal energy devices are shortly to be deployed in the first arrays, the knowledge of how different array layouts perform is a key and under-examined field. Here, the Momentum Reversal Lift (MRL) turbine, developed by the University of Exeter, is deployed in five different array layouts utilising up to 15 devices. The use of dynamic turbines allows the inclusion of analysis of the effects of flow direction in the wake. The layouts investigated explore the effect of lateral and stream-wise turbine spacings as well as differences between staggered and in-line layouts on power. The staggered array with decreased streamwise spacing is shown to have the highest total power per ‘footprint’ area among the layouts tested. For the staggered arrays, increased downstream separation had little effect on total power generated, while decreasing the lateral spacing below 2 rotor diameters decreased the power. The in-line arrays showed a lower power per device but similar total power. It was also shown that increased in-flow into a turbine didn't necessarily lead to an increased power extraction. The decrease in power with a decrease in streamwise spacing is in-line with theoretical and CFD predictions.

Large-eddy simulations of dense-gas dispersion within a high-Reynolds number turbulent boundary layer

Large-eddy simulations of flows over a backward-facing step with release of four different gases with increasing densities have been performed. The results have been analysed with particular emphasis on the part of the flow field sufficiently far downstream where the local effect of the source dynamics is no longer dominating the flow field. The dense gas plumes maintain high concentrations close to the ground and become approximately twice as shallow in comparison with a neutrally buoyant plume. The vertical mixing is significantly reduced close to the ground, and both momentum and scalar flux reversal are found in the two densest cases, indicating negative shear and buoyancy production of turbulence kinetic energy. Examination of the gradient Richardson number indicates that all dense gases are significantly affected by buoyancy. Interestingly, a narrow layer within the plumes is identified where the impact of stratification is expected to be reduced. It is found that relaminarization is most likely occurring and that fluctuations found might originate from interfacial wave modes rather than from turbulence. The large-scale structures are seen to resemble those found in the plane channel flow close to the wall and further away from it. In the middle layer, significant differences are found due to buoyancy effects. The large-scale structures become more pancake-shaped with large-scale vorticity almost exclusively about the spanwise direction in the dense gas cases.

The Distribution of Cross Sectional Momentum Returns

The cross sectional momentum (CSM) strategy, considered in Jegadeesh and Titman (1993), is a trading strategy in which assets with highest past returns are equally weighted and held long while assets with lowest returns are also equally weighted but held short. Although anomalous returns from such momentum strategies, indicating persistence or reversal in the relative performance of underlying assets, have been the subject of numerous empirical studies, very little appears to be known about the distributional properties of the CSM returns. Most of the known results in this regard, such as those obtained in Lo and MacKinlay (1990), Jegadeesh and Titman (1993), Lewellen (2002), and Moskowitz, et al. (2012), are limited to expected values and the first order autocorrelations in the returns of alternative momentum strategies in which the assets are weighted not equally, but instead in proportion to their returns over the ranking period. In this paper, we derive the density of CSM returns in analytic form, along with moments of all orders, under a reasonably general set of assumptions on the underlying asset returns. In particular, if the asset returns over the ranking and holding periods are independent then the density of the CSM returns is shown to be a mixture of univariate normals. We also obtain the density and the moments of equally weighted long only portfolios consisting of assets from an arbitrary return percentile band that we refer to as quantile portfolios.

Risks and rewards for momentum and reversal portfolios

Rational asset pricing implies a positive relation between the expected risk-adjusted return and the volatility of a factor-mimicking portfolio. The relation for the momentum portfolio is weak after its return is adjusted for the risks associated with the market return, the size factor, and the book-to-market factor. However, the relation is significantly positive and captures most of the average return on the momentum portfolio after the return is adjusted for the market return and the risk associated with the short-term reversal portfolio return. The result supports the hypothesis that there is a common factor underlying both momentum and short-term reversal. The dynamics of the factor loadings and the correlation structure of the underlying factors have important implications for the risk prices associated with the factor-mimicking portfolios and the risk–return trade-off for momentum and reversal portfolios.