Equity Premium

What is the historical relationship between U.S. stock market earnings yield (E/P) and U.S. government bond yield (Y)? In their February 2018 paper entitled “Stock Earnings and Bond Yields in the US 1871 – 2016: The Story of a Changing Relationship”, Valeriy Zakamulin and Arngrim Hunnes examine the relationship between E/P Y over the long run, with focus on structural breaks, causes of breaks and direction of causality. They employ a vector error correction model that allows multiple structural breaks. In assessing causes of breaks, they consider inflation, income taxes and Federal Reserve Bank monetary policy. Using quarterly S&P Composite Index level, index earnings, long-term government bond yield and inflation data during 1871 through 2016, along with contemporaneous income tax rates and Federal Reserve monetary actions, they find that:

Keep Reading

A subscriber requested review of a finding that deviation of 10-year constant maturity U.S. Treasury note (T-note) yield from an intermediate-term linear trend predicts U.S. stock market return. Specifically, when weekly yield is more than one standard deviation of weekly trend divergences below (above) a weekly 70-week linear extrapolation, next-week S&P 500 Index return is on average unusually high (low). To confirm and test usefulness of this finding, we each week:

1. Perform a linear extrapolation of past T-note yields to forecast next-week T-note yield, but using a 52-week rolling window rather than a 70-week window. A 52-week lookback aligns with an annual inflation cycle, while a 70-week lookback seems arbitrary and may be snooped.
2. Calculate the difference between next-week actual and forecasted T-note yields.
3. Calculate the standard deviation of these differences over the 52-week rolling window.

We then segment weekly actual minus forecasted T-note yield differences into: those more than one standard deviation below forecasted yield (Below Lower); those between one standard deviation below and above forecasted yield (Between); and, those more than one standard deviation above forecasted yield (Above Upper). Next, we calculate next-week S&P 500 Index returns for these three segments. Limited by availability of weekly T-note yield data, return calculations commence January 1964. To check robustness of results, we also consider a recent subsample commencing January 2008. To test economic value of findings, we examine a Dynamic Weighted strategy that modifies a benchmark 60% allocation to SPDR S&P 500 (SPY) and 40% allocation to iShares Barclays 7-10 Year Treasuries (IEF), rebalanced weekly, to 80% SPY when T-note condition the prior week is Below Lower and 40% SPY when Above Upper. The strategy backtest commences with inception of IEF at the end of July 2002 and focuses on weekly return statistics, compound annual growth rate (CAGR) and maximum drawdown (MaxDD), ignoring rebalancing/reallocation frictions. Using weekly T-note yields (average of daily values measured on Friday) and contemporaneous S&P 500 Index levels since January 1962, and weekly dividend-adjusted levels of SPY and IEF since July 2002, all through January 2018, we find that: Keep Reading

Do technical trend trading/intrinsic momentum strategies work for widely used equity factors such as size (small minus big market capitalizations), value (high minus low book-to-market ratios), profitability (robust minus weak), investment (conservative minus aggressive) and momentum (winners minus losers)? In their January 2018 paper entitled “What Goes up Must Not Come Down – Time Series Momentum in Factor Risk Premiums”, Maximilian Renz investigates time variation and trend-based predictability of these five factors and the market factor. He first constructs price series for the six long-short factor portfolios. He then considers seven rules based on a short simple moving average (SMA) crossing above (bullish) or below (bearish) a long SMA measured in trading days: SMA(1, 20), SMA(1, 40), SMA(1, 120), SMA(1, 180), SMA(1, 240), SMA(20, 180) and SMA(20, 240). He also considers two intrinsic (absolute or time series) momentum rules based on change in price over the past 180 or 240 trading days (positive bullish and negative bearish). Motivated by prior research by others, he focuses on SMA(1, 180), daily price crossing its 180-day SMA. He measures trend-based statistical predictability of factor premiums and investigates economic value via a strategy that levers factor exposures between 0 and 1.5 using trend-based signals. Finally, he examines whether incorporating trend information improves accuracies of 1-factor (market), 3-factor (adding size and value) and 5-factor (further adding profitability and investment) models of stock returns. Using daily returns for the six selected U.S. stock market equity factors and for 30 industries during July 1963 through December 2015, he finds that: Keep Reading

How should an investor’s view of the equity risk premium vary with investment horizon? In the December 2017 update of their paper entitled “Volatility Lessons”, Eugene Fama and Kenneth French examine how the U.S. equity risk premium (difference in returns between the expected equity market return over some horizon and return on U.S. Treasury instrument of matched duration) varies across investment horizons ranging from one month to 30 years. To generate distributions of equity risk premiums for horizons longer than one month, they employ bootstrap simulations. Specifically, for each matched horizon/duration, they randomly draw 100,000 pairs of stock market and U.S. Treasury instrument returns with replacement from a base sample of monthly data, without or with an adjustment for uncertainty in associated equity premiums. They repeat analyses on three value portfolios (Market Value, Big Value and Small Value) and on a Small stock portfolio with no value tilt, defining size and value as follows: (1) big (small) stocks are U.S. listed stocks with market capitalizations above (below) the NYSE median; and, (2) value stocks are U.S. listed stocks with book-to-market ratios above the 70th percentile of those for NYSE stocks. All style portfolios are capitalization-weighted and rebalanced annually at the end of June. Using monthly U.S. stock returns and and U.S. Treasury instrument yields across durations during July 1963 through December 2016 (642 months), they find that: Keep Reading

“Usefulness of P/E10 as Stock Market Return Predictor” investigates whether P/E10 (or Cyclically Adjusted Price-Earnings ratio, CAPE) usefully predicts U.S. stock market returns over the long run. That analysis employs Robert Shiller’s data set, which defines P/E10 as inflation-adjusted S&P Composite Index level divided by average monthly inflation-adjusted 12-month trailing earnings of index companies over the last ten years. Do more timely country P/E10 series work for timing country stock markets and trading pairs of country stock markets? Within each country market, higher (lower) P/E10 suggests overvaluation (undervaluation). Across countries, variation in P/E10 gaps arguably indicates which country markets are relatively overvalued and undervalued. To investigate, we consider:

• P/E10 time series for Germany, Japan and the U.S. evaluated separately over available sample periods using DAX, Nikkei 225 and S&P 500 indexes, respectively. We also look at separately timing SPDR S&P 500 (SPY) and iShares MSCI Japan (EWJ).
• Japan P/E10 versus U.S. P/E10 for pair trading of SPY versus EWJ over the available sample period.

Using monthly data for the three P/E10s, the three associated stock market indexes, SPY, EWJ and 3-month U.S. Treasury bill (T-bill) yield as available during December 1981 through December 2017, we find that: Keep Reading

Is optimal stock anomaly exploitation long-only or long-short? If not long-short, does shorting the market rather than individual stocks work as well as shorting individual stocks? In his November 2017 paper entitled “How Do Short Selling Costs and Restrictions Affect the Profitability of Stock Anomalies?”, Filip Bekjarovski explores effects of short selling costs and constraints on the viability of exploiting seven U.S. stock anomalies: size, value, profitability, investment, momentum, accruals and net issuance. He constructs all anomaly portfolios via market capitalization weighting of stocks sorted into tenths (deciles). He measures portfolio alphas relative to the market excess return (1-factor). He considers long-only (long the top decile), conventional long-short (long the top and short the bottom deciles) and hybrid long-short (long the two highest alpha deciles, tilted toward the highest, while short the market). Anomaly portfolio rebalancing is annual for all except momentum (monthly). He analyzes effects of shorting costs based on an April 2017 proprietary snapshot of institutional stock borrowing fees. He specifically estimates a shorting cost threshold above which investors should switch between long-short and hybrid long-short exploitation methods. Using the stock borrowing fee snapshot and data required to construct seven anomalies from a broad sample of U.S. common stocks during July 1963 through December 2016, he finds that: Keep Reading

What is the smartest way (having the lowest prediction errors) to estimate market beta across stocks for the purpose of portfolio construction? In their November 2017 paper entitled “How to Estimate Beta?”, Fabian Hollstein, Marcel Prokopczuk and Chardin Simen test effects of different return sampling frequencies, forecast adjustments and model combinations on market beta prediction accuracy across the universe of U.S. stocks. Their primary goal is to identify optimal choices. They focus on a beta prediction horizon of six months. They consider past beta estimation (lookback) windows of 1, 3, 6, 12, 24, 36 and 60 months for daily data, 12, 36 and 60 months for monthly data and 120 months for quarterly data. They measure beta prediction accuracy based on average root mean squared error (RMSE) across stocks. Using returns for a broad sample of U.S. stocks during January 1963 through December 2015, they find that: Keep Reading

Is value investing particularly profitable when the price spread between cheap and expensive assets (the value spread) is extremely large (deep value)? In their November 2017 paper entitled “Deep Value”, Clifford Asness, John Liew, Lasse Pedersen and Ashwin Thapar examine how the performance of value investing changes when the value spread is in its largest fifth (quintile). They consider value spreads for seven asset classes: individual stocks within each of four global regions (U.S., UK, continental Europe and Japan); equity index futures globally; currencies globally; and, bond futures globally. Their measures for value are:

• Individual stocks – book value-to-market capitalization ratio (B/P).
• Equity index futures – index-level B/P, aggregated using index weights.
• Currencies – real exchange rate based on purchasing power parity.
• Bonds – real bond yield (nominal bond yield minus forecasted inflation).

For each of the seven broad asset classes, they each month rank assets by value. They then for each class form a hedge portfolio that is long (short) the third of assets that are cheapest (most expensive). For stocks and equity indexes, they weight portfolio assets by market capitalization. For currencies and bond futures, they weight equally. To create more deep value episodes, they construct 515 sub-classes from the seven broad asset classes. For asset sub-classes, they use hedge portfolios when there are many assets (272 strategies) and pairs trading when there are few (243 strategies). They conduct both in-sample and out-of-sample deep value tests, the latter buying value when the value spread is within its top inception-to-date quintile and selling value when the value spread reverts to its inception-to-date median. Using data as specified and as available (starting as early as January 1926 for U.S. stocks and as late as January 1988 for continental Europe stocks) through September 2015, they find that:

Keep Reading

Do investors underestimate the adverse import of large left tails for future stock returns? In their November 2017 paper entitled “Left-Tail Momentum: Limited Attention of Individual Investors and Expected Equity Returns”, Yigit Atilgan, Turan Bali, Ozgur Demirtas and Doruk Gunaydin investigate the relationship between left-tail risk and next-month returns for U.S. and international stocks. They measure left-tail risk at the end of each month via either of:

• Value-at-risk (VaR) – daily return of a stock at the first (VAR1) or fifth (VAR5) percentile of its returns over the past one year (250 trading days).
• Expected shortfall – average daily return of a stock for the bottom 1% (ES1) or bottom 5% (ES5) of its returns over the past year (250 trading days).

They then sort stocks into tenths (deciles) based on left-tail risk and examine variation in next-month average gross returns across deciles. Using daily prices and monthly firm characteristics and risk factors for U.S. stocks with month-end prices at least $5 during January 1962 through December 2014, they find that: Keep Reading

As described in “Quantifying Snooping Bias in Published Anomalies”, anomalies published in leading journals offer substantial opportunities for exploitation on a gross basis. What profits are left after accounting for portfolio maintenance costs? In their November 2017 paper entitled “Accounting for the Anomaly Zoo: A Trading Cost Perspective”, Andrew Chen and Mihail Velikov examine the combined effects of post-publication return deterioration and portfolio reformation frictions on 135 cross-sectional stock return anomalies published in leading journals. Their proxy for trading frictions is modeled stock-level effective bid-ask spread based on daily returns, representing a lower bound on costs for investors using market orders. Their baseline tests employ hedge portfolios that are long (short) the equally weighted fifth, or quintile, of stocks with the highest (lowest) expected returns for each anomaly. They also consider capitalization weighting, sorts into tenths (deciles) rather than quintiles and portfolio constructions that apply cost-suppression techniques. Using data as specified in published articles for replication of 135 anomaly hedge portfolios, they find that:

Keep Reading