Equity Premium

A subscriber proposed four alternative ways of timing the U.S. stock market based on simple moving averages (SMA) of the market price-earnings ratio (P/E), as follows:

1. 5-Year Binary – hold stocks (cash) when P/E is below (above) its 5-year SMA.
2. 10-Year Binary – hold stocks (cash) when P/E is below (above) its 10-year SMA.
3. 15-Year Binary – hold stocks (cash) when P/E is below (above) its 15-year SMA.
4. 5-Year Scaled – hold 100% stocks (cash) when P/E is five or more units below (above) its 5-year SMA. Between these levels, scale allocations linearly.

To obtain a sample long enough for testing these rules, we use the monthly U.S. data of Robert Shiller. While offering a very long history, this source has the disadvantage of blurring monthly data as averages of daily values. How well do these alternative timing strategies work for this dataset? Using monthly data for the S&P Composite Index, annual dividends, annual P/E and 10-year government bond yield since January 1871 and monthly 3-month U.S. Treasury bill (T-bill) yield as return on cash since January 1934, all through August 2018, we find that: Keep Reading

Is there a best way for investors to measure firm profitability for global stock selection? In their August 2018 paper entitled “Constructing a Powerful Profitability Factor: International Evidence”, Matthias Hanauer and Daniel Huber investigate which measure of firm profitability best predicts associated stock returns. They consider six measures: return on equity; gross profitability; operating profitability calculated in two ways; cash-based operating profitability (excluding accruals); and, cash-based gross profitability (also excluding accruals). They construct a long-short profitability factor for each measure and test its power to predict stock returns both standalone and in combination with other kinds of factors (market, size, book-to-market, momentum, investment and accruals) and the other profitability factors. Using monthly returns and annual accounting data for non-financial common stocks in 49 countries (excluding the U.S.) during July 1989 through June 2016, they find that: Keep Reading

How, and how well, do institutional equity traders manage global stock trading frictions? In the April 2018 draft of their paper entitled “Trading Costs”, Andrea Frazzini, Ronen Israel and Tobias Moskowitz examine the real-world trading frictions of a large trader. They define trading frictions as the difference in results between a theoretical portfolio with zero frictions and a practical tracking portfolio with frictions. They account for all components of trading frictions: broker commissions, bid-ask spreads and price impacts of trading. They record market price at trade initiation, volume traded and execution price for each share traded, as well as type of trade (buy long, buy-to-cover, sell long or sell short). They describe how frictions vary by trade type, stock characteristics, trade size, time and exchange. Based on preliminary findings, they devise and test out-of-sample a price impact model based on market conditions, stock characteristics and trade size calibrated to actual U.S. and international trades. Using $1.7 trillion of orders and trade execution data from a large institutional money manager spanning 21 developed equity markets during August 1998 through June 2016, they find that: Keep Reading

Do stock dividends exhibit exploitable risk premiums? In their July 2018 paper entitled “A Model-Free Term Structure of U.S. Dividend Premiums”, Maxim Ulrich, Stephan Florig and Christian Wuchte construct a term structure of the dividend risk premium and test strategies to time this premium at specific horizons. They specify dividend risk premium as the spread between:

• Expected dividend growth rate based on analyst 1-year and 2-year S&P 500 dividend forecasts, extended by analyst 5-year earnings growth estimates assuming constant future payout ratio.
• Expected dividend growth rate derived from equity index put and call option prices across different maturities.

They model an S&P 500 dividend capture portfolio for a given horizon as: long an S&P 500 Index put option of maturity matching the horizon; short an index call option of same maturity and strike price; long the index; and, short the money market in an amount matched to the option strike price. They test two strategies for capturing this premium at a 12-month horizon: (1) each month (last trading day) reform and hold the dividend capture portfolio; or, (2) each month reform and hold the dividend capture portfolio only when the dividend risk premium is positive (analyst-estimated dividends are higher than options-implied dividends). They model the risk-free rate/money market rate across horizons using the U.S. Dollar Overnight Index Swap rate for one day to 10 years. For the S&P 500 Index, they assume annual expense ratio 0.07% and 0.01% average bid-ask spread. For options, they estimate trading frictions with actual bid-ask spreads. Using S&P 500 Index/options and analyst forecast data as specified during January 2004 through October 2017, they find that:

Keep Reading

From a purely statistical perspective, how many factors are optimal for explaining both time series and cross-sectional variations in stock anomaly/stock returns, and how do these statistical factors relate to stock/firm characteristics? In their July 2018 paper entitled “Factors That Fit the Time Series and Cross-Section of Stock Returns”, Martin Lettau and Markus Pelger search for the optimal set of equity factors via a generalized Principal Component Analysis (PCA) that includes a penalty on return prediction errors returns. They apply this approach to three datasets:

1. Monthly returns during July 1963 through December 2017 for two sets of 25 portfolios formed by double sorting into fifths (quintiles) first on size and then on either accruals or short-term reversal.
2. Monthly returns during July 1963 through December 2017 for 370 portfolios formed by sorting into tenths (deciles) for each of 37 stock/firm characteristics.
3. Monthly excess returns for 270 individual stocks that are at some time components of the S&P 500 Index during January 1972 through December 2014.

They compare performance of their generalized PCA to that of conventional PCA. Using the specified datasets, they find that: Keep Reading

Is there an exploitable way to predict when short-term stock market return will be negative? In his June 2018 paper entitled “Predictable Downturns”, Carter Davis tests a random forest regression-based forecasting model to predict next-day U.S. stock market downturns. He uses the value-weighted return of a portfolio of the 10 U.S. stocks with the largest market capitalizations at the end of the prior year minus the U.S. Treasury bill (T-bill) yield as a proxy for excess market return. He employs a two-step test process:

1. Use a rolling 10-year historical window of 143 input variables (economic, equity factor, market volatility, stock trading, calendar) to find when the probability of negative portfolio daily excess return is at least 55%.
2. Calculate whether the average portfolio gross excess return of all such days is in fact significantly less than zero.

He corrects for data snooping bias associated with the modeling approach. He further investigates which input variables are most important and tests a market timing strategy that holds the 10-stock portfolio (T-bills) when predicted portfolio return is negative (non-negative) as specified above. Using data for the input variables and returns for test portfolio stocks during July 1926 through July 2017, he finds that: Keep Reading

The many factor-based indexes and exchange-traded funds (ETFs) that track them now available enable investors to construct multi-factor portfolios piecemeal. Is such piecemeal construction suboptimal? In their July 2018 paper entitled “The Characteristics of Factor Investing”, David Blitz and Milan Vidojevic apply a multi-factor expected return linear regression model to explore behaviors of long-only factor portfolios. They consider six factors: value-weighted market, size, book-to-market ratio, momentum, operating profitability and investment(change in assets). Their model generates expected returns for each stock each month, and further aggregates individual stock expectations into factor-portfolio expectations holding all other factors constant. They use the model to assess performance differences between a group of long-only single-factor portfolios and an integrated multi-factor portfolio of stocks based on combined rankings across factors. The focus on gross monthly excess (relative to the 10-year U.S. Treasury note yield) returns as a performance metric. Using data for a broad sample of U.S. common stocks among the top 80% of NYSE market capitalizations and priced at least $1 during June 1963 through December 2017, they find that: Keep Reading

Does conventional reward-for-risk wisdom about the long-run performance of the U.S. stock market translate to the typical stock? In the May 2018 update of his paper entitled “Do Stocks Outperform Treasury Bills?”, Hendrik Bessembinder compares the performance of the typical U.S. stock to that of the 1-month U.S. Treasury bill (T-bill) over monthly, annual, decade and life-of-stock horizons. He also performs simulations to gauge the effectiveness of holding just one stock and of diversifying across portfolios of five, 25, 50 and 100 stocks. Using monthly total (dividend-reinvested) returns for 25,967 U.S. common stocks while listed during July 1926 through December 2016, he finds that: Keep Reading

Which factor models of stock returns are currently best? In their June 2018 paper entitled “q⁵“,  Kewei Hou, Haitao Mo, Chen Xue and Lu Zhang, introduce the q⁵ model of stock returns, which adds a fifth factor (expected growth) to the previously developed q-factor model (market, size, asset growth, return on equity). They measure expected growth as 1-year, 2-year and 3-year ahead changes in investment-to-assets (this year total assets minus last year total assets, divided by last year total assets) as forecasted monthly via predictive regressions. They define an expected growth factor as average value-weighted returns for top 30% 1-year expected growth minus bottom 30% 1-year expected growth, calculated separately and further averaged for big and small stocks. They examine expected growth as a standalone factor and then conduct an empirical horse race of recently proposed 4-factor, 5-factor (including q⁵) and 6-factor models of stock returns based on their abilities to explain average return differences for value-weighted extreme tenth (decile) portfolios for 158 significant anomalies. Using monthly return and accounting data for a broad sample of non-financial U.S. common stocks during July 1963–December 2016, they find that:

Keep Reading

“Is There Really an Size Effect?” summarizes research challenging the materiality of the equity size effect. Is there a counter? In their June 2018 paper entitled “It Has Been Very Easy to Beat the S&P500 in 2000-2018. Several Examples”, Pablo Fernandez and Pablo Acin double down on the size effect via a combination of market capitalization thresholds and equal weighting. Specifically, they compare values of a $100 initial investment at the beginning of January 2000, held through April 2018, in:

• The market capitalization-weighted (MW) S&P 500.
• The equally weighted (EW) 20, 40, 60 and 80 of the smallest stocks in the S&P 1500, reformed either every 12 months or every 24 months.

All portfolios are dividend-reinvested. Their objective is to provide investors with facts to aid portfolio analysis and selection of investment criteria. Using returns for the specified stocks over the selected sample period, they find that:

Keep Reading