Equity Premium

In their July 2019 paper entitled “Momentum-Managed Equity Factors”, Volker Flögel, Christian Schlag and Claudia Zunft test exploitation of positive first-order autocorrelation (time series, absolute or intrinsic momentum) in monthly excess returns of seven equity factor portfolios:

1. Market (MKT).
2. Size – small minus big market capitalizations (SMB).
3. Value – high minus low book-to-market ratios (HML).
4. Momentum – winners minus losers (WML)
5. Investment – conservative minus aggressive (CMA).
6. Operating profitability – robust minus weak (RMW).
7. Volatility – stable minus volatile (SMV).

For factors 2-7, monthly returns derive from portfolios that are long (short) the value-weighted fifth of stocks with the highest (lowest) expected returns. In general, factor momentum timing means each month scaling investment in a factor from 0 to 1 according its how high its last-month excess return is relative to an inception-to-date window of past levels. They consider also two variations that smooth the simple timing signal to suppress the incremental trading that it drives. In assessing costs of this incremental trading, they assume (based on other papers) that realistic one-way trading frictions are in the range 0.1% to 0.5%. Using monthly data for a broad sample of U.S. common stocks during July 1963 through November 2014, they find that: Keep Reading

Is the Office of Financial Research Financial Stress Index (OFR FSI), described in “The OFR Financial Stress Index”, useful as a U.S. stock market return predictor? OFR FSI is a daily snapshot of global financial market stress, distilling more than 30 indicators via a dynamic weighting scheme. The index drops and adds indicators over time as some become obsolete and new ones become available. Unlike some other financial stress indicators, past OFR FSI series values do not change due to any periodic renormalization and are therefore suitable for backtesting. To investigate OFR FSI power to predict U.S. stock market returns, we relate level of and change in OFR FSI to SPDR S&P 500 (SPY) returns. Using daily and monthly values of OFR FSI and SPY total returns during January 2000 (OFR FSI inception) through June 2019, we find that:

Keep Reading

Do most stocks worldwide beat the risk-free rate of return? In their July 2019 paper entitled “Do Global Stocks Outperform US Treasury Bills?”, Hendrik Bessembinder, Te-Feng Chen, Goeun Choi and John Wei  compare returns of individual global common stocks to that of 1-month U.S. Treasury bills (T-bills). They screen stock price data for obvious errors and filter/correct accordingly. For delisted stocks with no delisting return available, they set the final return to -30%. Using monthly returns with reinvested dividends in U.S. dollars for 17,505 U.S. and 44,476 non-U.S. stocks across 41 other countries (25 developed and 16 emerging) and monthly T-bill yield during 1990 through 2018, they find that:

Keep Reading

Do sentiment indexes usefully predict U.S. stock market returns? In his May 2018 doctoral thesis entitled “Forecasting Market Direction with Sentiment Indices”, flagged by a subscriber, David Mascio tests whether the following five sentiment indexes predict next-month S&P 500 Index performance:

1. Investor Sentiment – the Baker-Wurgler Index, which combines six sentiment proxies.
2. Improved Investor Sentiment – a modification of the Baker-Wurgler Index that suppresses noise among input sentiment proxies.
3. Current Business Conditions – the ADS Index of the Philadelphia Federal Reserve Bank, which combines six economic variables measured quarterly, monthly and weekly to develop an outlook for the overall economy.
4. Credit Spread – an index based on the difference in price between between U.S. corporate bonds and U.S. Treasury instruments with matched cash flows. (See “Credit Spread as an Asset Return Predictor” for a simplified approach.)
5. Financial Uncertainty – an index that combines forecasting errors for large sets of economic and financial variables to assess overall economic/financial uncertainty.

He also tests two combinations of these indexes, a multivariate regression including all sentiment indexes and a LASSO approach. He each month for each index/combination predicts next-month S&P 500 Index return based on a rolling historical regression of 120 months. He tests predictive power by holding (shorting) the S&P 500 Index when the prediction is for the market to go up (down). In his assessment, he considers: frequency of correctly predicting up and down movements; effectiveness in predicting market crashes; and, significance of predictions. Using monthly data for the five sentiment indexes and S&P 500 Index returns during January 1973 through April 2014, he finds that: Keep Reading

Are mean-variance (MV) strategies preferable for allocations to asset classes and equal-weight (EW) preferable for allocations to much noisier individual assets? In their May 2019 paper entitled “Horses for Courses: Mean-Variance for Asset Allocation and 1/N for Stock Selection”, Emmanouil Platanakis, Charles Sutcliffe and Xiaoxia Ye address this question. They focus on the Bayes-Stein shrinkage MV strategy, with 10 U.S. equity sector indexes as asset classes and the 10 stocks with the largest initial market capitalizations within each sector (except only three for telecommunications) as individual assets. The Bayes–Stein shrinkage approach dampens the typically large effects of return estimation errors on MV allocations. For estimation of MV return and return covariance inputs, they use an expanding (inception-to-date) 12-month historical window. They focus on one-month-ahead performances of portfolios formed in four ways via a 2-stage process:

1. MV-EW, which uses MV to determine sector allocations and EW to determine stock allocations within sectors.
2. EW-EW, which uses EW for both deteriminations.
3. EW-MV, which uses EW to determine sector allocations and MV to determine stock allocations within sectors.
4. MV-MV, which uses MV for both deteriminations.

They consider four net performance metrics: annualized certainty equivalent return (CER) gain for moderately risk-averse investors; annualized Sharpe ratio (reward for risk); Omega ratio (average gain to average loss); and, Dowd ratio (reward for value at risk). They assume constant trading frictions of 0.5% of value traded. They perform robustness tests for U.S. data by using alternative MV strategies, different parameter settings and simulations. They perform a global robustness test using value-weighted equity indexes for UK, U.S., Germany, Switzerland, France, Canada and Brazil as asset classes and the 10 stocks with the largest initial market capitalizations within each index as individual assets (all in U.S. dollars). Using monthly total returns for asset classes and individual assets as specified and 1-month U.S. Treasury bill yield as the risk-free rate during January 1994 through August 2017, they find that: Keep Reading

How does news flow interact with short-term stock market return? In their April 2019 paper entitled “Forecasting the Equity Premium: Mind the News!”, Philipp Adämmer and Rainer Schüssler test the ability of a machine learning algorithm, the correlated topic model (CTM), to predict the monthly U.S. equity premium based on information in news articles. Their news inputs consist of about 700,000 articles from the New York Times and the Washington Post during June 1980 through December 2018, with early data used for learning and model calibration and data since January 1999 used for out-of-sample testing. They measure the U.S. stock market equity premium as S&P 500 Index return minus the risk-free rate. Specifically, they each month:

1. Update news time series arbitrarily segmented into 100 topics (with robustness checks for 75, 125 and 150 topics).
2. Execute a linear regression to predict the equity premium for each of the 100 topical news flows.
3. Calculate an average prediction across the 100 regressions.
4. Update a model (CTM^Sw) that switches between the best individual topic prediction and the average of 100 predictions, combining the flexibility of model selection with the robustness of model averaging.

They use the inception-to-date (expanding window) average historical equity premium as a benchmark. They include mean-variance optimal portfolio tests that each month allocate to the stock market and the risk-free rate based on either the news model or the historical average equity premium prediction, with the equity return variance computed from either 21-day rolling windows of daily returns or an expanding window of monthly returns. They constrain the equity allocation for this portfolio between 50% short and 150% long, with 0.5% trading frictions. Using the specified news inputs and monthly excess return for the S&P 500 Index during June 1980 through December 2018, they find that:

Keep Reading

Have long-short alternative beta (style premium) strategies worked well in practice? In their February 2019 paper entitled “A Decade of Alternative Beta”, Antti Suhonen and Matthias Lennkh use actual performance data to assess alternative beta strategies across asset classes from the end of 2007 through the end of 2017, including quantification of fees and potential survivorship bias in public data. Specifically, they form three equal volatility weighted (risk parity) composite portfolios of strategies at the ends of each year during 2007-2016, 2007-2011 and 2012-2016. Each portfolio includes all the strategies launched during the first year and then adds strategies launched each following year at the end of that year. When a strategy dies (is discontinued by the offeror), they reallocate its weight to surviving strategies within the portfolio. They also create two additional portfolios for each period/subperiod that segregate equities and non-equities. They further evaluate alternative beta strategy diversification benefits by comparing them to conventional asset class portfolios. Using weekly post-launch excess returns in U.S. dollars for 349 reasonably unique live and dead alternative beta strategies offered by 17 global investment banks, spanning 14 styles and having at least one year of history during 2008 through 2017, they find that:

Keep Reading

Does widespread investor acceptance of the capital asset pricing model (CAPM) of stock returns drive undervaluation of stocks with low past alphas? In his February 2019 paper entitled “The Unintended Impact of Academic Research on Asset Returns: The CAPM Alpha”, Alex Horenstein examines whether such acceptance distorts the U.S. stock market. Specifically, he each year at the beginning of January reforms a betting against alpha (BAA) hedge portfolio that is long (short) stocks with alphas lower (higher) than the median based on monthly returns over the past five years. He then weights stocks according to their respective alpha ranks, rescales the long and short sides separately to have market beta 1.0 and holds for one year. He analyzes performance of this portfolio and eight widely accepted equity factors (size, value, momentum, profitability, investment, short-term reversal, long-term reversion and betting against beta) during three subperiods: (1) pre-CAPM era (1932-1964); (2) CAPM era (1965-1992); and, (3) smart beta era (1993-2015). Using total returns for a broad sample of U.S. common stocks and returns for eight accepted equity factors during January 1927 through December 2015, he finds that:

Keep Reading

A subscriber asked about an annually rebalanced portfolio of 50% stocks and 50% trend following managed futures as recommended in a 2014 Greyserman and Kaminski book [Trend Following with Managed Futures: The Search for Crisis Alpha], suggesting Equinox Campbell Strategy I (EBSIX) as an accessible managed futures fund. To investigate, we consider not only EBSIX (inception March 2013) but also a longer trend following hedge fund index with monthly returns back to December 1999. This alternative “is an equally weighted index of 37 constituent funds…designed to provide a broad measure of the performance of underlying hedge fund managers who invest with a trend following strategy.” The correlation of monthly returns between this index and EBSIX during April 2013 through February 2019 is 0.84, indicating strong similarity. We use SPDR S&P 500 (SPY) as a proxy for stocks. Using annual returns for EBSIX during 2014-2018 and for the trend following hedge fund index and SPY during 2000-2018, we find that: Keep Reading

What are potential monthly returns and alphas from applying machine learning to pick stocks? In their February 2019 paper entitled “Machine Learning for Stock Selection”, Keywan Rasekhschaffe and Robert Jones summarize basic concepts of machine leaning and apply them to select stocks from U.S. and non-U.S. samples, focusing on the cross-section of returns (as in equity factor studies). To alleviate overfitting in an environment with low signal-to-noise ratios, they highlight use of: (1) data feature engineering, and (2) combining outputs from different machine learning algorithms and training sets. Feature engineering applies market/machine learning knowledge to select the forecast variable, algorithms likely to be effective, training sets likely to be informative, factors likely to be informative and factor standardization approach. Their example employs an initial 10-year training period and then walks forecasts forward monthly (as in most equity factor research) for each stock, as follows:

• Employ 194 firm/stock input variables.
• Use three rolling training sets (last 12 months, same calendar month last 10 years and bottom half of performance last 10 years), separately for U.S. and non-U.S. samples.
• Apply four machine learning algorithms, generating 12 signals (three training sets times four algorithms) for each stock each month, plus a composite signal based on percentile rankings of the 12 signals.
• Rank stocks into tenths (deciles) based on each signal, which forecasts probability of next-month outperformance/underperformance.
• Form two hedge portfolios that are long the decile of stocks with the highest expected performance and short the decile with the lowest, one equal-weighted and one risk-weighted (inverse volatility over the past 100 trading days), with a 2-day lag between forecast and portfolio reformation to accommodate execution.
• Calculate gross and net average excess (relative to U.S. Treasury bill yield) returns and 4-factor (market, size, book-to-market, momentum) alphas for the portfolios. To estimate net performance, they assume 0.3% round trip trading frictions. 

They consider two benchmark portfolios that pick long and short side using non-machine learning methods. Using a broad sample of small, medium and large stocks (average 5,907 per month) spanning 22 developed markets, and contemporaneous values for the 194 input variables, during January 1994 through December 2016, they find that: Keep Reading