A decomposition of the value premium

Important findings:

• The value characteristics of a stock (defined by book value to market value) can change due to two levers: (1) the change in the market value, (2) the change in the book value.
• The complete value premium can be explained by changes in the market value.
• Therefore, there are value stocks that do not receive a value premium (e.g. neutral market valuation with high book value) as well as growth stocks that receive a value premium (e.g. under-valued stocks despite low book value).

Source:

A network approach explaining the connections between asset prices

Important findings:

• The authors present a novel network approach to explain the comovements and dependencies of asset prices.
• Explanatory factors serve as an input to describe these connections. This way, the relative importance of factors can be disentangled and changes in the network structure over time can be analyzed

Source:

Alternative weighting schemes for bonds

Important findings:

• Alternative weighting schemes that do not weigh assets according to their market capitalization also work for corporate bonds.
• The more inefficient and volatile the market the higher its outperformance potential.
• A noise-in-price model can serve as an explanation.

Source:

An explanation for momentum and reversal

Important findings:

• This paper proposes a theory to jointly explain time-series momentum and reversal.
• The key assumption is that new information is not distributed instantly, but flows at an increasing rate.

Source:

An optimized value approach

Important findings:

• In academic research, value is commonly measured as book value to market value (B/P), while in practice earnings to market value (E/P) are also regularly used.
• This paper shows that a combination of both approaches outperforms each of the single approaches. Moreover, sector constraints help to improve performance.

Source:

Anomalies and company news

Important findings:

• Returns on equity market anomalies are 50% higher on corporate news days and even six times higher in earnings announcement days.
• Potential explanations for this observation are dynamic risk, mispricing due to biased expectations, or data mining.
• The results of this paper hint towards mispricing due to biased expectations.

Source:

CO2 emissions as proxy for productivity

Important findings:

• A low carbon ratio (CO2 emissions relative to revenues) corresponds to a higher future profitability and higher equity returns. This observation also holds for industries that are barely impacted by CO2 taxes or other forms of pollution regulation.
• The relationship is much more fundamental in the sense that the carbon ratio serves as a proxy for overall productive efficiency. The less inputs are required in production (which nearly all use some form of energy and thus emissions) to generate a certain amount of revenues, the more efficiently the business is run.
• It is seemingly not the CO2 reduction per se that is impacting the performance, but the general productive efficiency which is approximated by the carbon ratio.

Source:

Interessenkonflikte von bankabhängigen Fonds

Wichtige Erkenntnisse:

• Diese Studie analysiert die Performance von Aktienfonds, welche von den Asset-Management-Divisionen von Universalbanken verwaltet werden.
• Bankunabhängige Fonds erzielen gegenüber angeschlossenen Fonds eine Outperformance von 92 Basispunkten pro Jahr.
• Interessenkonflikte stellen eine mögliche Erklärung dar, da die Underperformance vor allem dort auftritt, wo Aktien von Kreditnehmern der Bank übergewichtet werden.
• Angegliederte Manager unterstützen damit die Kreditvergabe der Banken und sichern damit ihre eigene Karriere zu Lasten der Fondsinvestoren.

Quelle:

Crash Sensitivity as explanation for the momentum premium

Important findings:

• This paper shows that the classical momentum premium can be to a large extent explained by the crash-sensitivity of stocks in the momentum portfolio.
• This observation is not limited to the U.S., but holds across 23 international equity markets.

Source:

De facto seniority of corporate bonds

Important findings:

• The position in the maturity structure of an issuer has a substantial influence on the credit risk of a specific bond.
• Bonds that are due relatively late in the maturity structure have higher credit risks, larger credit spreads and a higher correlation towards equities.
• Therefore, bond-specific credit risk scores are important, whereas rating agencies perform their ratings on an issuer level.

Source:

Differences between time series and cross sectional strategies

Important findings:

• There are two groups of strategies to forecast returns: (1) strategies that time a single asset (time series strategies) and (2) strategies that select attractive assets relative to other assets (cross sectional strategies).
• While cross-sectional strategies are market neutral (long-short) in their pure form, time series strategies have a time-varying long-exposure.
• For individual stocks, both strategy types mainly differ due to this time-varying long exposure but work similarly well besides that.
• Across asset classes, however, cross-sectional strategies achieve a superior performance.

Source:

Downside risk as driver of the equity risk premium

Important findings:

• Most of the equity risk premium is a compensation of bearing downside risk. Upside risk, in contrast, is barely compensated.
• Option-based strategies are therefore expected to yield a substantial underperformance in the long term.

Source:

Dynamic factor investing

Important findings:

• The authors analyze how factors can be optimally combined in a portfolio dynamically. To do so, they use time series as well as cross-sectional data to derive asset allocation signals.
• While time series signals fail to significantly add value, cross-sectional signals lead to strong performance improvements.

Source:

Effects of increased benchmarking

Important findings:

• Benchmarking leads to a higher effective risk aversion, a reduced willingness to speculate and to obtain information.
• A higher number of benchmark oriented investors reduces the information content of market prices, which leads to decreased prices of risky asset and an increased volatility.
• At the same time, it opens up more opportunities for non-benchmarked investors to outperform benchmarked investors.

Source:

Equity financing risk

Important findings:

• This paper analyzes the impact of liquidity reserves, R&D investments and equity issuances on expected returns in the presence of equity financing risk.
• High cash reserves relative to R&D investments increase the slack of companies, reduce the equity financing risk and thereby the expected returns.
• Equity issuances reduce the equity financing risk and the expected returns decrease concomitantly.
• Cash reserves and equity issuances have no impact on expected returns for companies with large internal resources.
• A factor that harvests the equity financing risk achieves considerable excess returns which cannot be explained by common factor models but can itself subsume the Fama & French (2015) investment factor.

Source:

Estimating expected returns based on a large number of firm characteristics

Important findings:

• This paper presents a new approach to estimate expected returns from a large number of company characteristics.
• This approach is based in 26 characteristics and outperforms other approaches to estimate expected returns.
• Moreover, this approach shows that many asset price anomalies have commonalities.

Source:

Factor premia and the yield curve

Important findings:

• Existing research mainly describes the yield curve using the three parameters level, slope and curvature.
• This paper shows that common factors in other asset classes such as value, momentum and carry are related to the parameters of the yield curve.
• The value factor can subsume the level, while momentum and carry subsume the slope and the curvature.
• Moreover, the factors value, momentum and carry contain additional information regarding economic growth, inflation and the Cochrane & Piazzesi (2005) bond factor.

Source:

Factor strategies and transaction costs

Important findings:

• This study analyses the number of company specific characteristics that are jointly significant in explaining the cross-section of equity returns.
• They find that the number of relevant characteristics increases if transaction costs are taken into account.
• Even though this finding may seem contra-intuitive at first glance, it is due to a combination of several characteristics reducing the turnover and therefore reducing transaction costs.
• The presence of transaction costs is one reason why it may be warranted to consider a larger number of characteristics in asset pricing models.

Source:

Forecasting tail risk

Important findings:

• It is possible to reduce tail risks without sacrificing much return for it.
• Forecasting future skewness of returns in particular considerably benefits investors.
• This approach helps improving classical diversification, which reduces volatility, but often fails in crises.

Source:

Global Market Inefficiencies

Important findings:

• The authors use a new methodology to determine the fair values of 25’000 stocks in 36 countries.
• A trading strategy based on the deviations from the fair value delivers a significant outperformance, particularly in the Emerging Markets.
• The alphas are correlated with transaction costs, but exceed them.
• Thus, global markets are inefficient.

Source:

Hierarchische Risikoparität als neuer Gewichtungsansatz

Wichtige Erkenntnisse:

• Dieses Forschungspapier präsentiert einen neuen Portfolio-Gewichtungs-Ansatz, die sogenannte hierarchische Risikoparität.
• Gegenüber der klassischen Risikoparität ist diese deutlich stabiler und kann mit unsaubereren Daten besser umgehen.
• Daneben gelingt es der hierarchischen Risikoparität das Portfoliorisiko gegenüber der klassischen Risikoparität zu reduzieren.

How to measure trends?

Important findings:

• Two common trend measures are time-series momentum and moving-average crossovers.
• This paper shows that both measures are empirically as well as theoretically very similar.
• In the general form, both approaches are mathematically equivalent, where past prices and returns as well as the time horizon are decisive drivers.

Source:

Idiosyncratic momentum pays off globally

Important findings:

• Idiosyncratic momentum is momentum that has been adjusted for market movements.
• It reduces the risk compared to the classical momentum strategy and shows a sizeable alpha.
• These results are not limited to the U.S. equity market, but hold for 21 additional international markets.
• In particular, it holds for Japan, where classical momentum does not seem to work.

Source:

Improving the Shiller P/E for forecasting equity returns

Important findings:

• The cyclically adjusted Shiller P/E performs well in predicting long term equity returns. It, however, fails for short term predictions.
• The authors find that this is due to the normal level von the Shiller P/E varying with the macroeconomic conditions (real interest rates, inflation).
• Adjusting the P/E for the current macroeconomic conditions, it succeeds in predicting short term returns.

Source:

Instable results in academic research papers

Important findings:

• In scientific journals, there is a prevailing pressure to produce significant results. The so called p-value is primarily used for this purpose.
• Particularly the large amount of different tests and other data mining approaches severely limits the usefulness of p-values.
• Therefore, the author presents guidelines for a robust, transparent research culture in financial economics.

Source:

Investment, momentum and reversal

Important findings:

• Equity markets exhibit momentum in the medium term (3-12 months) and reversal in the long term (3-5 years).
• A potential explanation for this pattern is the inherent delay in firm investment.
• Winners only continue to win, when there is subsequent investment. In contrast, losers only continue to lose, when there is subsequent disinvestment.

Source:

Latent factors in the equity market

Important findings:

• This paper presents a new approach to estimate latent asset pricing factors that explain equity prices in the time series as well as the cross-section.
• According to this approach, there are five important factors to explain equity returns: the broad equity market (market beta), a value factor, a momentum factor, a profitability factor as well as a Sharpe ratio factor which loads heavily on reversal.
• Many further equity characteristics do not help explaining equity returns.

Source:

Longterm reversal in industry performances

Important findings:

• Companies in loser industries yield a substantial outperformance compared to companies in winner industries over the next five years.
• This effect is a decisive driver behind long-term reversal in equities.
• Industry reversal exists in all phases of the business cycle and is more pronounced in industries with high valuation uncertainty.

Source:

Macroeconomic determinants of stock market betas

Important findings:

• This paper decomposes the market beta into several cyclical components with different frequency.
• Especially the surplus-consumption-ratio as well as the default premium have a significant influence on the equity market beta.

Source:

Market impact of large transactions

Important findings:

• Most studies that estimate the market impact of large transactions consider each transaction separately.
• Neglecting the interactions in the order flow can lead to substantial under-estimation of transaction costs. Potential contagion effects can also not be captured.
• This paper proposes a model to alleviate this concern.

Source:

Misvaluation as explanation of the profitability factor

Important findings:

• Macroeconomic risks can only partially capture the profitability premium. A misevaluation factor based on sentiment, however, contributes substantially in explaining the profitability premium.
• The profitability premium is primarily present in companies whose profitability is underestimated in times of high sentiment. Companies with high profitability and low valuation are more likely to be underestimated by analysts and they show price jumps after earnings announcements.

Source:

Momentum in company fundamentals

Important findings:

• Momentum not only exists in returns, but also for company fundamentals.
• Twin momentum – the combination of price momentum and fundamental momentum – considerably improves on each of the two momentum variants alone.

Source:

Momentum spillover from equities to corporate bonds

Important findings:

• This paper finds momentum spillovers from equities to corporate bonds, i.e. companies that had high recent equity returns will show high corporate bond returns.
• This momentum strategy shows distinct structural as well as time-varying default risks.
• Taking the company specific momentum (residual momentum), these risks can be substantially reduced and the Sharpe ratio can be doubled.

Source:

New approaches for hedging equity risk

Important findings:

• Traditional hedging approaches for equity portfolios such as put options, US treasuries, gold and credit protection are expensive and / or unreliable.
• The authors propose two strategies to protect against large drawdowns: (1) a time-series momentum strategy (similar to many CTAs) and (2) a long-short strategy on the equity quality factor.

Source:

Optimized rebalancing of multi asset portfolios

Important findings:

• Weights of a multi asset portfolio shift over time due to market movements.
• Rebalancing brings the weights back to the strategic asset allocation.
• In between rebalancings, investors deviate from the strategic asset allocation and incur timing bets regarding when to rebalance.
• Using a short option overlay, the basis risk can be reduced and investor can concurrently harvest a volatility risk premium.

Source:

Popularity as return driver

Important findings:

• In efficient markets, systematic risks are undesirable and therefore require a return premium.
• In the behavioral finance world, behavioral biases affect asset prices.
• Popularity unifies both of these perspectives by pricing assets according to their rational (risk based) or irrational (emotional) desirability.
• Characteristics that are permanently undesirable require a price discount respectively a return premium.

Source:

Profitability shocks and the size effect

Important findings:

• Many studies find that the size effect disappeared after the early 1980s.
• The paper shows that this disappearance is due to negative profitability shocks of small companies while there were positive shocks for large companies.
• When adjusting for these profitability shocks, a robust size effect remains.

Source:

Quality controls resurrect the size effect

Important findings:

• Recently the size factor has been widely attacked. It is supposedly too weak, too instable, weakening since its discovery, concentrated among microcaps and in January as well as internationally instable.
• Controlling for quality, however, these criticisms can be alleviated.
• After controlling for quality, the size factor performs on par with other anomalies such as value or momentum.

Source:

Residual Momentum in Japan

Important findings:

• Residual momentum is momentum that has been adjusted for the Fama & French factor exposures.
• In contrast to classical momentum, it also works in Japan.
• While momentum returns show reversals in the long term, which necessitates regular portfolio adjustments, residual momentum does not show reversals.
• Investor underreaction is the most likely explanation for residual momentum.

Source:

Reversals of Alphas

Important findings:

• Assets with low realized CAPM-alphas yield positive excess returns compared to assets with high realized CAPM-alphas.
• Even though the CAPM-alpha is mis-specified in a multi-factor-world, it is widely used for performance analysis in practice. This gives fund managers the incentive to shift portfolios towards higher realized alphas. This shift leads to an undervaluation of stocks with low realized alphas.
• This behavior is consistent with three possible explanations: (1) investors are leverage constrained, (2) managers are benchmarked towards the CAPM, (3) the market overreacts to extreme values of realized alphas.
• Academic research seemingly does not only diminish anomalies, but can also create new anomalies.

Source:

Risk-adjustment of time series momentum

Important findings:

• This paper proposes to adjust time-series momentum strategies by their volatilty to balance their risk exposure.
• This strategy leads to an exposure to market beta, value and momentum, while at the same time considerably reducing the turnover.

Source:

Short-term underreaction of equity markets

Important findings:

• Following jumps in U.S. stock prices, there is short-term underreaction.
• A strategy that goes long in stocks with positive upward jumps while shorting stocks with negative jumps yields a significant outperformance over the 1-3 month horizon.
• Limited investor attention offers one potential explanation.

Source:

Social media and equity prices

Important findings:

• Social media influences financial markets. Often the relationship remains relatively unclear and only few hedge funds make use of this kind of data.
• The sentiment of tweets can predict daily returns above the information contained in the Fama-French five factor model.
• The authors claim to have discovered a social media factor as sixth factor that complements the existing Fama-French five factor model.

Source:

Survivorship bias in peer group comparisons

Important findings:

• Survivorship bias, the focus on surviving securities, leads to considerable biases in backtests of investment strategies or peer group comparisons.
• Backtesting strategies will look too good if only surviving securities are used. Peer group comparisons, in contrast, look too bad, as the fund is only compared to surviving and thus successful funds.
• The authors propose a new method to correct for this effect when calculating performance based measures for peer group comparisons.

Source:

Systemic risk from short volatility strategies

Important findings:

• Short volatility strategies are widely popular at the moment.
• These strategies bet on the volatility staying low in the future and thus are equivalent to some type of financial insurance against market turbulence.
• Therefore, these strategies contain a high systematic risk and can lead to market crashes, if many investors want to exit those strategies at the same time.
• Besides directly writing options, many other strategies have an implicit exposure to short volatility risk such as carry trade strategies, risk parity strategies and volatility targeting strategies.

Source:

Tail risk sensitivity of individual stocks

Important findings:

• This study measures the sensitivity of equities towards large market downturns, the so called tail beta.
• Stocks with high tail betas lose 2-3x as much in crises compared to stocks with low tail betas.
• Even though equities with high tail betas are considerably more risky, there appears to be no return premia for holding them.

Source:

The cyclically adjusted price-earnings-ratio

Important findings:

• The cyclically adjusted price-earning-ratio (CAPE) receives much attention as well as criticism for predicting equity returns.
• The authors propose diverse technical improvements for a more robust estimation of the CAPE. In particular, enhancing the CAPE with other stable accounting measures like cash-flows and revenues seems warranted.
• Comparing the current CAPE level with the historical mean does not make much conceptional sense as there is no steady-state level.
• Moreover, using CAPE as market timing tool requires caution and is better suited for analyzing the relative returns of equities compared to other asset classes.

Source:

The effect of accruals on equity returns

Important findings:

• Accruals are non-cash components of earnings and can be easily manipulated by shifting claims between periods. To gain more reliable estimates of profitability, cash-flows are often used.
• Expected returns rise in profitability, but decrease in accruals. Measuring profitability while excluding accruals leads to higher expected returns.
• Defining profitability correctly (without accruals), there is no need to have a separate accruals factor.

Source:

The excess return of equities and bonds is not guaranteed

Important findings:

• The authors test excess returns for equities and bonds over bills in 20 different countries over a period of more than a century.
• The excess returns are only significant in less than half of the cases.
• Globally, the reward for higher risk is not set in stone.

Source:

The financial intermediation premium

Important findings:

• Companies that borrow money from highly leveraged financial intermediaries have higher expected returns than companies with low-leverage lenders.
• This return difference cannot be explained by other company characteristics. However, it is driven by the company exposure towards the financial sector.
• The dispersion of the leverage of financial intermediaries forecasts macroeconomic aggregates.

Source:

The impact of ETFs on capital markets

Important findings:

• Over the last 25 years, the ETF market has grown manifold and now has a considerable impact on the capital markets as a whole.
• While ETFs facilitate price discovery, they cause volatility that is not driven by economic funds as well as influence the correlation structure of returns.
• Finally, ETFs heavily impact market liquidity.

Source:

The influence of hormones on financial risk-taking

Important findings:

• This paper discusses biological reasons for risk taking behavior. The focus is on the role of steroid hormones (the endocrine system).
• Hormones have a significant influence on trader performance. High cortisol levels increase the perception of risk, while high testosterone levels lead to increased focus on opportunities and increased risk-taking behavior.
• Bull and bear markets and the concomitant excessive optimism or pessimism could be influenced by hormone-induced shifts in confidence and risk preferences.
• A higher endocrine diversity of financial market participants could increase the stability of the financial system.

Source:

The interaction of short term reversal and momentum

Important findings:

• Short-term trends heavily depend on intermediate momentum.
• Short-term reversal strategies work best for intermediate momentum losers.
• Momentum strategies work best for short-term trend winners.

Source:

The lottery effect as an explanation of the beta anomaly

Important findings:

• The beta-anomaly refers to the unexpectedly high return of low beta stocks, while high beta stocks suffer from lower than expected returns.
• This paper shows that this anomaly can be explained well by the so called lottery effect, i.e. that people are willing to give up returns for a minor chance for a high payoff.
• This effect is concentrated among high beta stocks with low institutional ownership.

Source:

The popularity of minimum volatility strategies

Important findings:

• Minimum volatility strategies gained in popularity over the recent years.
• Compared to the overall market, they remain minuscule and many active mutual funds overweight high volatility stocks.
• A neutral positioning of these active mutual funds would lead to a much stronger shift towards minimum volatility than all current investments.
• Moreover, minimum volatility strategies are not overvalued compared to historical standards and the performance is consistent with the expectations for the respective market regime.

Source:

The roll yield

Important findings:

• The so called roll yield which occurs while rolling futures positions particularly in commodity markets is an often misunderstood concept.
• It is not an actual cash flow during the roll. Gains and losses accrue only through price changes during the holding period of futures and not during the roll.
• The roll yield, however, describes the difference between futures prices and spot prices.

Source:

The value of volatility management

Important findings:

• Recent studies show that volatility managed portfolios yield a higher Sharpe ratio.
• This paper shows that this observation only holds for equities and credit, but is negligible for fixed income, currencies and commodities.
• However, tail risks can be significantly reduced across all asset classes.
• Thus, volatility management significantly increases the attractiveness of portfolio risk and return characteristics.

Source:

Trends over different time horizons

Important findings:

• In the equity markets, depending on the time horizon, there are three distinct types of trends. In the short-term, there is reversal; in the medium term we can observe momentum, while in the long-term, there is reversal once more.
• A trend factor that captures these three trends simultaneously, yields a considerable improvement compared to the three trends taken separately.

Source:

Uncertainty of beta estimates

Important findings:

• Depending on the time period, the market proxy or the factor model, beta estimates can vary widely.
• It shows that equities whose beta estimates vary more and thus encompass a higher uncertainty have higher returns than equities with narrower beta estimates.
• This phenomenon is independent of the absolute level of the systematic risk / the market beta.

Source:

Valuation as an approach to factor timing

Important findings:

• The authors analyze value as a measure to predict future returns.
• Currently, equity factors are not overvalued from a valuation standpoint.
• Value timing is highly correlated with a static value strategy, but tends to be less stable.
• Therefore, value timing disappoints in a multi-factor context that already includes the value factor in the portfolio due to reduced diversification.

Source:

Value Timing

Important findings:

• The value spread helps predicting the return of value strategies on equities, commodities, currencies, bonds and equity indices.
• Asset class specific as well as common components of the value spread contribute equally to the forecasting performance.
• The common component is highly correlated with classical risk factors such as dividend yield, leverage or illiquidity.

Source: