Fundamental Valuation

What is the state of cyclically adjusted price-earnings ratios (CAPE, P/E10 or Shiller PE), stock index level divided by average real earnings over the past ten years, across country equity markets worldwide? In his October 2013 paper entitled “What the Shiller PE Says About Global Equity Markets: Update 2013”, Joachim Klement updates expected returns for equity markets around the world based on P/E10 (see “Predictive Power of P/E10 Worldwide”). He adjusts P/E10 for economic conditions for each country via regression of P/E10 versus real GDP growth, real per capita GDP growth, real interest rate and inflation. Using stock index level, P/E10 and economic data for 20 developed and 18 emerging equity markets as available through September 2013, he finds that: Keep Reading

Which long-term smoothed (10-year average) stock valuation metric works best to screen stocks? In the end-of-September 2013 version of their paper entitled “On the Performance of Cyclically Adjusted Valuation Measures”, Wesley Gray and Jack Vogel compare the abilities of five valuation ratios expressed as yields to predict U.S. stock returns, as follows: 

1. 10-year average real earnings to market capitalization (CA-EM)
2. 10-year average real book values to market capitalization (CA-BM)
3. 10-year average real earnings before interest, taxes, depreciation and amortization to total enterprise value (CA-EBITDA/TEV)
4. 10-year average real free cash flow to total enterprise value (CA-FCF/TEV)
5. 10-year average real free gross profits to total enterprise value (CA-GP/TEV)

They adjust for inflation using the monthly U.S. Consumer Price Index (CPI). Each year on June 30 (with accounting data lagged to ensure availability), they sort stocks into tenths (deciles) based on each ratio. They then calculate monthly decile returns over the next 12 months based on equal initial weights, with either annual (next June 30) or monthly decile reformation. For monthly reformation, they recalculate market capitalizations (but not valuation ratios) each month. They also test a stock return momentum overlay, each month splitting each valuation ratio top decile into high-momentum and low-momentum halves, with momentum defined as cumulative return from 12 months ago to one month ago. They do not account for trading frictions or taxes. Using stock prices and accounting data for U.S. firms with at least 10 years of the required accounting data and market capitalizations above the 40th percentile for NYSE-listed stocks, and contemporaneous CPI data, during 1962 through 2012, they find that: Keep Reading

Do book value-to-price ratio (B/P) and earnings-to-price ratio (E/P) indicate reward-for-risk opportunities at the country level worldwide? In their September 2013 paper entitled “Risky Value”, Atif Ellahie, Michael Katz and Scott Richardson investigate relationships among these valuation ratios, earnings growth and future returns at the country level for 30 countries over the past two decades. They construct monthly country-level valuation and earnings growth outlooks from capitalization-weighted firm fundamentals and earnings forecasts. They then relate these measures to country capitalization-weighted stock market future excess returns (relative to local risk-free rates), with the return measurement interval commencing four month after fundamentals are available. They replace negative country-level E/P values with zero. Using monthly firm-level fundamentals and stock data, as well as macroeconomic forecasts, for 30 countries during March 1993 through June 2011 (6,600 country-month observations), they find that: Keep Reading

The guiding belief of the Fed Model of stock market valuation is that investors use a Treasury note (T-note) yield as a benchmark for the expected (forward) earnings yield of the stock market. When the gap between the forward earnings yield and the T-note yield is positive (negative), stocks are relatively attractive (unattractive), and investors bid stocks up (down) to restore yield balance. Does evidence justify this belief? To investigate, we relate the month-end gap between the S&P 500 1-year forward operating earnings yield and the 1-year T-note yield to future returns for the S&P 500 index. We calculate the 1-year forward operating earnings yield from the Earnings Forecast and the level of the S&P 500 Index. Using end-of-month data for the two yields over the period March 1989 (limited by availability of an input variable for the Earnings Forecast) through July 2013 (over 24 years), we find that: Keep Reading

Can investors rely on overvaluation signals from the market price-earnings ratio (P/E) and the Fed Model to predict major stock market corrections? Which model works better? In their July 2013 paper entitled “Does the Bond-Stock Earning Yield Differential Model Predict Equity Market Corrections Better Than High P/E Models?”, Sebastien Lleo and William Ziemba test the power of eight bond-stock earnings yield differential model (BSEYD) variants and eight market P/E model variants to predict stock market corrections. They specify the bond yield for the BSEYD model as that of the 10-year U.S. Treasury note (T-note). They define a stock market correction as a decline of 10% or more within one year. They specify the 16 model variants based on: (1) either BSEYD, the natural logarithm of BSEYD, P/E or the natural logarithm of P/E; (2) either current year or rolling 10-year average stock market earnings; and, (3) either of two ways of calculating the threshold for extreme overvaluation. Both methods of setting the extreme overvaluation threshold for the 16 indicators are out-of-sample based on indicator average and standard deviation over a rolling one-year historical window. They measure success of model variants based on both the proportion of signals followed by corrections within two years and, conversely, the proportion of crashes preceded by signals within the past two years. Using daily S&P 500 Index levels, S&P 500 earnings data and daily T-note yields during 1962 through 2012, they find that: Keep Reading

How well does stock screening research translate into performance? In the mid-1990s, James O’Shaughnessy identified “cornerstone value” and “cornerstone growth” as best-of-breed equity investment strategies. The former emphasizes dividends among large-capitalization stocks, and the latter momentum/earnings growth for a broader universe. Based on Standard and Poor’s Compustat data, he found that the value (growth) strategy returned 15% (18%) per year during 1952-1994, compared to 8.3% for the S&P 500 Index. He implemented these two strategies in late 1996 via mutual funds and publicized them in early editions of his book What Works on Wall Street: A Guide to the Best-Performing Investment Strategies of All Time. He subsequently sold the mutual funds (which apply slightly different portfolio formation rules from those specified in the original research) to Hennessy Funds in 2000, where they survive as the Hennessy Cornerstone Value Fund (HFCVX) and the Hennessy Cornerstone Growth Fund (HFCGX). Do these funds outperform simpler exchange-traded funds (ETF) that track their respective benchmarks funds: iShares Russell 1000 Value Index (IWD) for HFCVX and iShares Russell 2000 Index (IWM) for HFCGX? Using monthly total returns for HFCVX, HFCGX, IWD and IWM during May 2000 (inception of the ETFs) through July 2013, we find that: Keep Reading

Does the carry trade concept provide a useful framework for valuation of securities within and across all asset classes? In their July 2013 paper entitled “Carry”, Ralph Koijen, Tobias Moskowitz, Lasse Pedersen and Evert Vrugt investigate expected return across asset classes via decomposition into “carry” (expected return assuming price does not change) and expected price appreciation. They measure carry for: global equities; global 10-year bonds; global bond yield spread (10-year minus 2-year); currencies; commodities; U.S. Treasuries; credit; equity index call options; and equity index put options. Their measurements of carry vary by asset class (based on: futures prices for equity indexes, currencies and commodities, modeled futures prices for global bonds, U.S. Treasuries and credit; and, option prices for options). They further decompose carry returns into passive and dynamic components. The passive component is the return to a hedge (carry trade) portfolio designed to capture differences in average carry returns across securities, and the dynamic component indicates how well carry predicts future price appreciation. Finally, they determine the conditions under which carry strategies perform poorly across all asset classes. Using monthly price/yield data for multiple assets within each class as available during January 1972 through September 2012, they find that: Keep Reading

What initial retirement portfolio withdrawal rate is sustainable over long horizons when, as currently, bond yields are well below and stock market valuations well above historical averages? In their June 2013 paper entitled “Asset Valuations and Safe Portfolio Withdrawal Rates”, David Blanchett, Michael Finke and Wade Pfau apply predictions of bond yields and stock market returns to estimate whether various initial withdrawal rates succeed over different retirement periods. They define initial withdrawal rate as a percentage of portfolio balance at retirement, escalated by inflation each year thereafter. They simulate future bond yield as a linear function of current bond yield with noise, assuming a long-term average of 5% and bounds of 1% and 10%. They simulate future U.S. stock mark return as a linear function of Cyclically Adjusted Price-to-Earnings ratio (CAPE, or P/E10), the ratio of current stock market level to average earnings over the last ten years, assuming P/E10 has a long-term average of 16.4 with noise (implying average annual return 10% with standard deviation 20%). They simulate inflation as a function of bond yield, change in bond yield, P/E10 and change in P/E10 with noise. They assume an annual portfolio management fee of 0.5%. They run 10,000 Monte Carlo simulations for each of many initial withdrawal rate scenarios, with probability of success defined as the percentage of runs not exhausting the portfolio before the end of a specified retirement period. Using initial conditions of a government bond yield of 2% and a P/E10 of 22 as of mid-April 2013, they find that: Keep Reading

Does adding fundamental firm quality metrics to refine stock sorts based on traditional value ratios, book-to-market ratio (B/M) and earnings-to-price ratio (E/P), improve portfolio performance? In his 2013 paper entitled “The Quality Dimension of Value Investing”, Robert Novy-Marx tests combination strategies to determine which commonly used quality measures most enhance the performance of value ratios. He considers such quality metrics as Piotroski’s FSCORE, earnings accruals, gross profitability (GP) and return on invested capital (ROIC). His general test approach is to reform capitalization-weighted portfolios annually from stocks sorted at the end of each June according to value ratios and quality metrics for the previous calendar year. He uses the 1000 largest (2000 next largest) stocks by market capitalization to represent large (small) stocks. He considers both long-only (long the top 30%) and long-short (long the top 30% and short the bottom 30%) portfolios. He also considers the incremental benefit of incorporating stock price momentum based on return over the previous 11 months with a skip-month (11-1) into stock selection. He estimates trading frictions based on calculated turnover and effective bid-ask spreads. Using stock prices and associated firm fundamentals during July 1963 through December 2011, he finds that: Keep Reading

Do equity sectors have exploitably measurable relative value? In his February 2013 paper entitled “Equity Sector Rotation via Credit Relative Value” (the National Association of Active Investment Managers’ 2013 Wagner Award winner), Dave Klein outlines a long-only strategy that ranks Standard & Poor’s Select Sector SPDR exchange-traded fund (ETF) based on relative value. The strategy seeks to exploit a belief that sector valuations increase (decrease) differently as macro credit risk falls (rises). Specifically, the strategy each week: (1) regresses the weekly unadjusted price for each sector ETF versus the weekly option-adjusted spread for the Bank of America/Merrill Lynch High Yield B (HY/B) credit index over the last six months to determine a best-fit (fair value) line; (2) ranks sector ETFs from cheapest to most expensive (percentage below or above their respective fair value lines) based on the prior-day HY/B index value; and, (3) forms long-only, equally weighted portfolios of the cheapest one (Top 1) to eight (Top 8) ETFs. He asserts that a six-month regression is long enough to discover the relationship between ETF price and credit index, and short enough to ignore inflation and dividends and “forget” major market disruptions. He uses SPDR S&P 500 (SPY) and an equally weighted portfolio of all nine sector ETFs (All 9) as benchmarks. An alternative strategy substitutes 3-month U.S. Treasury bills (T-bills) for any of the cheapest ETFs in a portfolio that are still expensive (above their respective fair value lines). He also considers a relaxation of weekly portfolio reformation/rebalancing. Using weekly levels of the sector ETFs (both unadjusted and adjusted for dividends), the HY/B credit index and the T-bill yield during July 1999 through December 2012, he finds that: Keep Reading