Fundamental Valuation

Given the conflicting evidence about the import of the size effect, is there a way investors can extract a reliable premium from small stocks? In their January 2015 draft paper entitled “Size Matters, If You Control Your Junk”, Clifford Asness, Andrea Frazzini, Ronen Israel, Tobas Moskowitz and Lasse Pedersen examine whether controlling for firm quality mitigates the following seven unfavorable empirical findings that the size effect:

1. Is weak overall in the U.S.
2. Has not worked out-of-sample and varies significantly over time.
3. Only works for extremely small stocks.
4. Only works in January.
5. Only works for market capitalization-based measures of size.
6. Is subsumed by illiquidity.
7. Is weak internationally.

They control for quality using a Quality-Minus-Junk (QMJ) factor based on profitability, profit growth, safety and payout. They use a portfolio test approach, ranking stocks into value-weighted tenths (deciles) each month to examine differences among stocks sorted by factor. Focusing on returns and factor metrics for a broad sample of U.S. common stocks during July 1957 (when quality metrics become available) through December 2012 and for 23 other developed country stock markets during January 1983 through December 2012, they find that: Keep Reading

Can investors usefully apply stock quality metrics to entire country stock markets? In his December 2014 paper entitled “Country Selection Strategies Based on Quality”, Adam Zaremba investigates whether quality metrics effectively predict country stock market index performance. He also examines whether (1) quality-size and quality-value double sorts enhance country-level value and size strategies; and, (2) high-quality markets offer a hedge during times of market distress. He considers six quality metrics: accruals, cash (cash divided by total assets), profitability (return on assets), leverage (total assets divided by common equity), payout (dividends as a fraction of income) and turnover (dollar volume of trading divided by market capitalization). Firm metric aggregation weightings are those used in constructing respective country indexes. After lagging the time series by three months to avoid a look-ahead bias, he forms capitalization-weighted portfolios of country markets by ranking them into fifths (quintiles) based on quality metric sorts. He identifies times of market distress based on: the spread between U.S. LIBOR and U.S. Treasury bill yields; VIX; the spread between U.S. corporate BBB bond and 10-year U.S. Treasury note yields; and, the spread between U.S. Treasury 10-year and 2-year note yields. Using stock market index returns and accounting data in U.S. dollars across 77 country stock markets during February 1999 through September 2014 as available, and contemporaneous market distress indicator values, he finds that: Keep Reading

Is firm profitability trend, or momentum, a useful indicator of future stock returns? In their December 2014 paper entitled “The Trend in Firm Profitability and the Cross Section of Stock Returns”, Ferhat Akbas, Chao Jiang and Paul Koch investigate the relationship between trend in firm profitability and stock returns, while controlling for level of profitability. They calculate gross profit quarterly as sales minus cost of goods sold, divided by total assets. They specify level of profitability as average gross profit over the past eight quarters. They specify trend in profitability as linear slope over the past eight quarters. They employ assumptions that ensure public availability of all data at the time of measurement, including a skip-month between portfolio formation and holding period. Using firm characteristics and returns for a broad sample of U.S. common stocks during January 1977 through December 2012, they find that: Keep Reading

A subscriber suggested evaluation of average investor allocation to stocks as “The Single Greatest Predictor of Future Stock Market Returns”. For this evaluation, we test simple ways to time the broad U.S. stock market using the quarterly time series for average U.S. investor allocation to stocks as provided in the article. We assume that the dates in this series are the first days of measured quarters. Using this quarterly series and the contemporaneous S&P 500 Index during December 1951 through September 2014, and quarterly dividend-adjusted returns for SPDR S&P 500 ETF (SPY) and contemporaneous 13-week U.S. Treasury bill (T-bill) yields during March 1993 through September 2014, we find that: Keep Reading

Do changes in consensus analyst earnings forecasts that disagree with contemporaneous stock returns signal exploitable mispricings? In their November 2014 paper entitled “To Follow or Not to Follow – An Analysis of the Profitability of Portfolio Strategies Based on Analyst Consensus EPS Forecasts”, Rainer Baule and Hannes Wilke investigate the power of a variable that relates consensus earnings forecast momentum to stock price momentum to predict stock returns. Specifically, the variable is the ratio of (1+change in consensus earnings forecast) to (1+stock return) over the last six months. Their consensus earnings forecast metric is a rolling average of consensus estimates for the current and next years weighted according to proximity of the current-year forecast to the end of the firm’s fiscal year (for example, three months before the end of the fiscal year, the rolling 12-month metric is 3/12 of the forecast for the current year plus 9/12 of the forecast for next year). They measure predictive power via a portfolio that is each month long (short) the fifth of stocks with the highest (lowest) last-month variable values. They evaluate both raw excess portfolio performance (relative to the risk-free rate) and four-factor portfolio alpha (adjusting for market, size, book-to-market and momentum factors). They limit the stock universe to the widely covered and very liquid components of the S&P 100 Index. Using monthly analyst consensus earnings forecasts and total returns for S&P 100 stocks during February 1978 through December 2013 (a total of 278 stocks listed for at least one month), they find that: Keep Reading

How do the major components of U.S. stock market performance behave over time? In his October 2014 paper entitled “Long-Term Sources of Investment Returns and a Simple Way to Enhance Equity Returns”, Baijnath Ramraika decomposes long-term returns from the U.S. stock market (as proxied by Robert Shiller’s S&P Composite Index) into four components:

1. Dividend yield
2. Inflation
3. Real average change in 10-year earnings (E10)
4. Change in the Cyclically Adjusted Price-Earnings ratio (CAPE, or P/E10)

He further segments this decomposition by decade. Using his decomposition by decade for 1881 through 2010 (13 decades), we find that: Keep Reading

Is Piotroski’s FSCORE really as effective at picking stocks as indicated in the original study, which screens stocks with high book-to-market ratios to isolate those that will eventually provide high returns? In their April 2014 paper entitled “Implementability of Trading Strategies Based on Accounting Information: Piotroski (2000) Revisited”, Sohyung Kim and Cheol Lee re-examine the FSCORE testing methodology. Their broader goal is to investigate the effect of a problematic research approach used in many accounting-based studies. Specifically, they evaluate the difference in investment outcomes between:

• Accounting studies that typically accumulate returns for individual stocks starting three or four months after their respective fiscal year-ends.
• Finance studies that synchronize return accumulation with a common annual starting date (regardless of the fiscal year-ends of individual firms) via periodically reformed portfolios with a specified weighting scheme (such as equal weighting).

Accounting study assumptions may confound real-time position weighting due to lack of timely information on how many stocks should be in the portfolio, thereby incorporating look-ahead bias. The study replicates the original Piotroski methodology with equal position weighting, and then makes adjustments to synchronize return calculations by each year reforming a hedge portfolio that is long (short) the equally weighted stocks with high (low) FSCOREs based on a common starting date and most recently available firm fundamentals. Using monthly data for a broad sample of U.S. common stocks with enough information to calculate FSCOREs during 1976 through 2007, they find that:

Keep Reading

Does compounding rules tend to improve the performance of stock-picking strategies? In the October 2014 draft of their paper entitled “Does Complexity Imply Value, AAII Value Strategies from 1963 to 2013”, Wesley Gray, Jack Vogel and Yang Xu compare 13 stock strategies labeled as “Value” by the American Association for Individual Investors (AAII) to each other and to a simple “low-price” value strategy. The simple strategy each year selects the tenth of stocks with the highest Earnings Before Interest, Taxes, Depreciation and Amortization-to-Total Enterprise Value ratios (EBITDA/TEV), excluding financial firms. To ensure liquidity, they focus on stocks with market capitalizations above the NYSE 40th percentile. To ensure real-time availability of inputs, they lag firm accounting data. To assess performance consistency, they consider three subperiods: July 1963 through December 1980; January 1981 through December 1996; and, January 1997 through December 2013. Because portfolios are equally weighted, they include the S&P 500 equal-weight total return index (S&P 500 EW) as a benchmark. Using stock price and firm accounting data for a broad sample of U.S. common stocks during July 1963 through December 2013, they find that: Keep Reading

Does adjusting an asset allocation retirement glidepath according to a stock market valuation metric such as Shiller’s cyclically adjusted price-earnings ratio (CAPE ratio or P/E10) improve the outcome? In their September 2014 paper entitled “Retirement Risk, Rising Equity Glidepaths, and Valuation-Based Asset Allocation”, Michael Kitces and Wade Pfau investigate the interaction of pre-determined allocation glidepaths and P/E10 valuation based on long-run U.S. historical data. They consider the following strategy alternatives:

• Fixed equity allocations of either 45% or 60%.
• Declining (accelerated declining) equity glidepaths that start retirement at 60% stocks and reduce the allocation by 1% (2%) per year.
• Rising (accelerated rising) equity glidepaths that start retirement at 30% stocks and increase the allocation by 1% (2%) per year.
• A standalone dynamic valuation-based strategy with baseline 45% equity, raised (lowered) to 60% (30%) at the beginning of any year for which P/E10 is less than (greater than) 67% (133%) of its inception-to-date median. (See the chart below.)
• Unbounded and bounded combinations of declining or rising glidepaths and the dynamic valuation-based strategy, adding (subtracting) 15% from the equity glidepath at the beginning of any year for which P/E10 indicates undervaluation (overvaluation). Bounded combinations constrain equity allocation to a minimum 30% and a maximum 60%.

They consider both short-term bills (six months to a year) and long-term bonds (10-year) as complements to equities. They use overlapping 30-year intervals to approximate retirement outcomes. They focus on worst-case maximum sustainable real (inflation-adjusted) withdrawal rate over the 30-year retirement interval as the main strategy performance metric. Withdrawals occur at the beginning of each year, with the residual portfolio then rebalanced to target allocations. They assume withdrawals pay the taxes. Using Robert Shiller’s monthly data for U.S. stock market returns, associated P/E10, short-term bill yields (six-month commercial paper/one-year U.S. Treasury notes) and long-term bond yields (10-year U.S. Treasury notes or equivalent) during 1871 through 2013, they find that: Keep Reading

Which firm accounting measures best predict future stock returns? In the August 2014 version of their paper entitled “Are Cash Flows Better Stock Return Predictors than Profits?”, Stephen Foerster, John Tsagarelis and Grant Wang investigate the power of enhanced cash flow measures to predict stock returns. They first devise procedures for transforming indirect cash flow and income statements into estimates of cash flow directly available to stockholders (see the table below). They then compare the ability of these measures and of alternative cash flow/profit/income measures to predict stock returns via hedge portfolios that are each month long (short) the tenth of stocks with the best (worst) values of each measure. They scale all measures either by total assets or by market value of equity. They consider both value-weighted and equal-weighted hedge portfolios. They use the real-time S&P 1500 (excluding financial firms) as their stock universe to ensure investability. Using monthly accounting data lagged by four months and monthly stock returns for the specified set of firms during October 1994 through December 2013, they find that: Keep Reading