Market Indicators
Tests of sentiment, breadth, economic and valuation indicators: which ones have predicted stock returns, and which have not.
- Articles
- 309
- Free to read
- 215
- Since
- 2004
- Latest
- Oct 2026
Start here
The articles in this guide that readers look for most.
- Chicago Fed NFCI as U.S. Stock Market Predictor Weekly changes in Chicago Fed NFCI show -0.13 correlation with next-week S&P 500 returns (-0.22 since 2000), with largest NFCI spikes preceding poor equity returns (Jan 1971–Apr 2020; in-sample).
- Predictive Power of Put-Call Ratios Put-call ratios show weak, inconsistent correlations with SPY future returns (R-squared 0.006 for best performer; 1995-2013; CBOE/ISE total/index/equity ratios over 5-63 day horizons; 68-trade strategy underperformed buy-and-hold).
- Jim Cramer Using the S&P Oscillator Cramer's S&P Oscillator calls underperformed, with "In" calls lagging average S&P 500 returns and "Out" calls outperforming (21-day horizons; Jun 2008–Oct 2019; 6 "In"/9 "Out" calls from CNBC search).
- Goldman Sachs Panic Index Proxy A VIX-percentile Panic Index Proxy (trailing 2-year rank, 0–100) shows weak lead-lag association with daily S&P 500 returns, with very high and very low quintiles followed by higher returns (Dec 2015–Nov 2025; in-sample; descriptive, not a…
- Testing the McClellan Oscillator and Summation Index McClellan Oscillator and Summation Index showed minimal predictive power for SPY returns with correlations near zero and R-squared statistics under 0.01 (Mar 1993–Jun 2012; NYSE breadth data).
Valuation
81 articlesCAPE, the Fed model, earnings yields and other valuation measures.
- Strictly Fundamental Valuation by AIs Machine-learning residual income valuations from accounting data alone yielded 6.3% gross annual alpha for a monthly, equal-weighted long-short valuation-to-price decile portfolio, concentrated in smaller stocks (U.S. firms; 1985–2024; gross of…
- CAPE Ratio (P/E10) Based on Index Component CAPEs Value-weighted Component CAPE predicted 10-year S&P 500 returns more accurately than conventional CAPE (out-of-sample R-squared 57% vs 47%; predictions 1974–2024, CAPE from 1964; constant-slope regressions; 10-year earnings history; descriptive, not…
- Permanently Lower Stock Market Earnings Yield? Aggregate U.S. enterprise value-based earnings yield (Integrated Macroeconomic Accounts) fell from 9.5% to 4.6% while free cash flow yield stayed near 3.6% (Jan 1952–Sep 2025; descriptive, explanatory; not a trading…
- Stock Market Valuation Perspectives The S&P 500 index level appears to outrun its operating earnings and dividends, and the Shiller-style smoothed earnings yield shows a downward linear trendline with high variability (quarterly S&P 500…
- Investors Expecting Earnings Growth and P/E Expansion? Change in earnings and change in P/E correlate negatively for 1- to 20-year intervals, including -0.50 at 10 years (S&P Composite Index, Shiller data, 1872–June 2025; descriptive return decomposition; not…
- Why Does P/E Vary So Much Across Stocks? What variables drive differences in price-to-earnings ratios (P/E) across U.S. stocks? In the September 2025 revision of their paper entitled "The Cross-section of Subjective Expectations: Understanding Prices and Anomalies", Ricardo…
Show all 81
- Stock Market Valuation Ratio Trends Descriptive; not a trading test: S&P 500 operating and as-reported trailing E/P and dividend yield are below their 1988–2025 averages (operating E/P forecast below 5.37% average; S&P forecasts through 2026;…
- Testing the Stock Market Earnings Yield-TIPS Yield Delta The S&P 500 operating earnings yield minus 10-year TIPS yield delta shows little evidence of usefully predicting next-month S&P 500 returns; quintile averages were non-monotonic (Jan 2003–Nov 2024; monthly lead-lag…
- CAPE Change Drivers What variables best explain increases and decreases in Cyclically Adjusted Price-to-Earnings ratio (CAPE or P/E10)? In their August 2024 paper entitled "Analyzing Changing 'Investor Exuberance': The Determinants of S&P Composite…
- Update on Real Earnings Yield and Future Stock Market Returns A strategy holding the S&P 500 when operating real earnings yield (O-REY) from two quarters earlier is positive, and cash otherwise, boosted gross CAGR by 2.5% versus buy-and-hold S&P 500…
- Impact of AI on Stock Valuations Firms in the top AI-exposure quintile within industries earned 0.40% average daily market-adjusted returns vs -0.25% for the bottom quintile after ChatGPT's release (Nov 15, 2022–Mar 31, 2023; descriptive; U.S.…
- Fed Model Nuance Adjusting Fed model variants (earnings yield vs 10-year Treasury yield ratio, difference, log difference) for regime shifts is associated with more statistically reliable S&P Composite return predictions at 1-, 3-,…
- Testing the Buffett Indicator Outside the U.S. Buffett Indicator strategy outperformed an equal-weighted portfolio of all 14 country stock indexes (10.5% vs 9.5% gross annualized return; 1973-2019; holds the seven countries with the highest model-predicted returns, reset…
- Stock Market Earnings Yield and Inflation Over the Long Run Lagged E/P minus inflation timing (hold S&P Composite Index when prior-year value is positive, else T-bills; annual) outperformed buy-and-hold S&P Composite Index with dividends (11.8% vs 11.2% gross CAGR; since…
- Valuation-based Stock Market Return Expectations Rowles and Mitchell's regression model predicts S&P 500 returns with increasing accuracy at longer horizons (R-squared 0.27 at 1-year vs 0.60 at 10-year; Jan 1871–Jun 2020; multi-variable regressions on P/E,…
- CAPE (P/E10) Version of Fed Model? Shiller et al. find U.S. CAPE rebounded from 23 to 32 by September 2020, with current CAPE-forecasted 10-year equity returns of 3% and equity-bond premium of 4% (Jan 1871–Sep 2020;…
- Fed Model Improvement? AIAE-based tactical allocation outperformed 60/40 and Fed model variants (14.3% CAGR; Aug 1977–Sep 2020; monthly rebalancing using 10-year equity risk premium forecasts from aggregate investor allocation ratios).
- Reversions from Stock Market Valuation Extremes Drive the Value Premium? Value premium concentrates during market extremes (3.42% monthly after undervaluation vs 1.70% after overvaluation; Jan 1962–Dec 2018; B/M-based market timing; equal-weighted U.S. stocks).
- COVID-19 Impacts on Stock Valuation COVID-19 earnings forecast revisions explained -12% average stock decline, but discount rate changes (rising from 10% to 13% in March, reverting to 10% by May) drove the V-shaped market evolution…
- Option Valuation How do market makers and sophisticated investors/traders determine option value? In his July 2019 essay entitled "Trading Volatility", Emanuel Derman outlines the history and shortcomings of option valuation as described…
- Neural Network Software Valuation of Fine Art Neural network software explained 75% of out-of-sample fine art price variation vs 5% for hedonic regression, though still lagged auctioneer estimates at 88% (985K training auctions 2008-14, 104K test auctions…
- Is CAPE Optimal for Market Valuation, and Useful? CAPE outperformed competing valuation ratios for market return prediction at 1-6 year horizons (Q3 1930–Q2 2017; Shiller data since Q1 1871), though authors caution against market timing due to unstable…
- Stock Market Timing Using P/E SMA Signals P/E SMA timing strategies underperformed buy-and-hold (9.4% vs 9.9% CAGR; 1886–2018; binary/scaled allocations on 5/10/15-year P/E SMAs; S&P Composite with 6-month earnings delay).
- Valuation of Crypto-assets Pagnotta and Buraschi argue network effects drive crypto-asset value, linking user count and privacy preference to price via demand-supply equilibrium (Bitcoin end-2017 calibration; theoretical model).
- Rise and Fall of the Fed Model? Zakamulin and Hunnes identify three regimes in earnings yield-bond yield relationships: E/P=2Y (1871-1929), no stable relationship (1930-1957), and E/P=Y Fed Model (1958-2016), with bond yields leading earnings yields during stable…
- Exploiting Investor Attention to P/E Earnings yield (E/P) hedge portfolios outperformed book-to-market portfolios in first 2 months after formation (0.71% vs 0.37% monthly returns; 1973-2015; value-weighted extreme deciles, monthly reform).
- Valuation-based Factor Timing Value spread timing modestly improves some factor portfolios but undermines diversified multi-factor portfolios, with timing costs likely eliminating gross advantages (1968–2016; HML/UMD/BAB factors).
- Long-run Predictive Power of Stock Market Dividend Yield Aggregate dividend yield reliably predicts stock market returns at annual and multi-year horizons across four centuries (1629-2015; Netherlands/UK/U.S.; in-sample and out-of-sample vs historical average).
- Enterprise Multiple Interactions with Other Stock Valuation Metrics High-mispricing portfolios (disagreement between enterprise multiple and other valuation metrics) outperformed low-mispricing portfolios (1.71% vs 0.34% monthly return; Jul 1972–Dec 2015; US stocks).
- Effects of Deflation on Stock Market Returns/Valuation Stock market P/E ratios decline more during high inflation (10.4) than mild deflation (15.5), with best nominal returns preceding mild deflation (Jan 1871–Feb 2016; Shiller data; descriptive analysis).
- Factor Portfolio Valuation and Timing of Factor Premiums Factor timing strategies based on value spreads added little to portfolios already exploiting value factors (Jan 1968–Jan 2016; long-short factor portfolios rebalanced monthly; gross returns).
- Comparing CAPE to Other Stock Market Valuation Ratios CAPE and price-to-book ratio similarly predict long-term returns, with lowest CAPE decile followed by 13.3% vs 13.6% real annual returns (17 countries; 1871-2015; 10-15 year forecasts).
- Abnormally Low Searching Equals Undervalued? Low-attention S&P 500 stocks (abnormally low Google search volume) outperformed the index (19.3% vs 5.2% gross annual return; Jan 2004–Oct 2014; weekly rebalancing, 122 stocks).
- Valuation/Trend Hedging of a Value and Momentum Stock Portfolio Value-momentum stock strategy with valuation/trend hedging outperformed S&P 500 (15.6% vs 10.0% annualized; 1964-2014; hedged portfolio suppressed drawdown to -27% vs -51%; monthly rebalancing of top 200 stocks).
- Effectiveness of Stock Valuation Based on Accounting Variables Fundamental valuation using 28 accounting variables outperformed a long-short hedge strategy (8.1% vs 5.3% annualized gross returns; equal vs value-weighted; monthly rebalancing; Jan 1977–Dec 2012; non-financial US stocks).
- Professional Equity Valuation Methods CFA Institute survey finds market multiples (led by P/E, EV/EBITDA) and discounted cash flow approaches dominate equity valuation practice (1,980 professionals; 2015; descriptive survey; not accuracy tests).
- Tactical U.S. Stock Market Allocations Based on Valuation Ratios Tactical allocation strategies based on D/P, P/E, and CAPE ratios generally matched or underperformed the 60-40 benchmark on a gross basis (1920-2014; annual/monthly rebalancing with 1SD/2SD triggers; S&P 500/T-bills).
- Real Commodity Prices as Valuation Aids Real commodity prices correlate negatively with 10-year future real returns across gold, S&P GSCI, crude oil, and corn (1965-2014; spot prices; descriptive; not a trading test).
- Composite Stock Market Valuation Model Jones finds MV/GDP outperforms P/E10 for predicting 10-year real equity returns (R-squared 0.52 vs 0.38; 1954–2014; in-sample), with composite model including demographics reaching 0.91 R-squared.
- Stock Screening Power of Long-term Average Valuation Ratios 10-year smoothed book-to-market outperformed other cyclically adjusted ratios (21.6% vs 16.6-19.1% CAGR for top decile with momentum overlay; 1962-2012; monthly rebalancing beats annual; high-momentum overlay adds 1%+ annually across all…
- Testing the Fed Model The Fed Model (S&P 500 forward earnings yield minus Treasury yield gap) shows practically no relationship with future stock returns (R-squared 0.005 for 6-month returns; Mar 1989–Jul 2013; in-sample test).
- Fed Model or P/E Model for Predicting Stock Market Corrections? BSEYD model variants outperformed P/E variants in predicting S&P 500 corrections of 10%+ (18 corrections, 25% average decline; 1962-2012; 16 model variants with rolling thresholds).
- 40-Year Valuation Ratio Horse Race EBITDA/TEV outperformed other valuation ratios with highest 17.7% annual return and 2.9% three-factor alpha vs E/M's 15.2% return with no alpha (cheapest 20% portfolios; July 1971–Dec 2010; NYSE stocks, annual…
- Required Yield Theory of Gold Valuation The Required Yield Theory model tracks gold prices with 12% average absolute error versus 16% for mining cost models (May 1979–May 2002; quarterly data; descriptive, not trading test).
- Fed Model Respecified? Timmer argues capital structure optimization drives earnings yield equilibrium with after-tax corporate bond yields rather than Treasury yields, with strong correlation emerging post-1982 (1968–2006; theoretical framework with empirical support).
- The Earnings Yield Anomaly Revisited Time-series E/P ranking (buy top 20%, sell bottom 30%) generated 0.17% monthly three-factor alpha vs cross-sectional E/P sorts showing no alpha (Jan 1962–Dec 2010; but alpha concentrated in 1990–2010 subsample).
- Interactions of Momentum, Valuation and Idiosyncratic Volatility Stock momentum is strongest for growth firms with high idiosyncratic volatility (1.17% vs 0.43% monthly return; 6-month formation/hold, long-short; 1965-2008 U.S. stocks) but reverses severely over years 2-5.
- Volatility and Valuation with High-frequency Trading Zhang finds high-frequency trading relates positively to stock price volatility and causes overreaction to fundamental news that subsequently reverses (1985-2009; U.S. stocks; regression analysis).
- Simple Versus Complex Valuation Metrics Complex valuation metrics (RIM and ROM) outperformed simple metrics (PE, Book Yield, Fair PE) with up to 5% annualized gross return advantage (Jan 1990–Apr 2010; ~450 MSCI Europe stocks; monthly…
- Valuation Metric Map and Critique The post analyzes five S&P 500 valuation metrics (lagged/forward E/P variants with/without inflation adjustments), finding current values indicate undervaluation but with limited explanatory power of 5-24% for future returns (Mar…
- Tobin Smith’s Fearless Forecasts Tobin Smith's forecasts underperformed a buy-and-hold benchmark with frequent dramatic opinion reversals undermining analysis depth (May 2000–March 2009; weekly WaveWire commentary graded against S&P 500 directional moves; large sample size).
- Preliminary Test of RYT Model Daily Valuations RYT Model daily valuations show no predictive power for next-day SPY returns (R-squared 0.00; Jul 2009–Mar 2010; correlation 0.06 close-to-close, 0.05 open-to-close; 160 daily observations).
- Long-term Trends and Short-term Variations in Valuation Ratios Decomposing S&P 500 valuation ratios into trend/cyclical components enhances predictive power, with cyclical components of earnings-price ratio showing statistically significant out-of-sample forecasting ability (1926–2008; quarterly data).
- Volatility and Valuation Ratios David and Veronesi find V-shaped relationships between valuation ratios and volatility, with stock P/E and volatility correlating negatively on average but positively during rare growth periods (1958–2008; theoretical model).
- Tobin’s q as a Stock Return Predictor? The only place Tobin's q shows up on CXOadvisory.com at present is...
- A Few Notes on The Dark Side of Valuation Damodaran argues valuation must anchor to cash flow generation capacity and risk-adjusted returns, linking return-on-capital/cost-of-capital spreads to sustainable growth assessment (2009 edition; conceptual framework).
- Required Yield Theory of Asset Valuation Faugere and Van Erlach's Required Yield Theory fits S&P 500 forward earnings yield with 88% adjusted R-squared versus 69% for Fed Model (1953–2006; quarterly data; theoretical framework linking real GDP…
- Valuation Forecasting Fly-off: Discounted Cash Flow vs. Comparables Optimized comparables valuation was more accurate than optimized DCF in forecasting equity values, while the textbook versions showed the reverse (89 DCF and 200+ comparables U.S. firms; valuation as of…
- Predictive Power of the Gap Between Stock Earnings Yield and T-note Yield ...investors may be able to use the Fed Model to enhance risk-adjusted returns via reduced portfolio volatility.
- Cash as a Valuation Indicator High excess cash holdings outperformed low cash holdings (0.4% monthly; 1960-2006; long-short hedge strategy, value-weighted, monthly rebalancing; non-financial U.S. firms).
- “It’s the P/E, Stupid!” Quarterly Shiller-based S&P Composite P/E correlates with Gallup presidential approval, with P/E changes associated with about 14% of approval variation (1950–Q3 2007; 231 observations; simple regression; descriptive, not a trading…
- Kicking the Body of the Fed Model ...investors seem to have left the Fed model for dead.
- Inflation as Fed Model Intermediator Bekaert and Engstrom find the S&P 500 dividend yield and 10-year Treasury yield correlate at 0.78 (Q4 1968–2007; quarterly data); the comovement is associated with expected inflation coinciding with high…
- Cash Flow Trumps Discount Rate in Stock Valuations? Cash flow news explains 16% of quarterly stock returns but 80% at seven-year horizons, with firm-level correlations rising from 0.25 to 0.79 across horizons (1985–2006; earnings forecast decomposition).
- Does the Fiscal Deficit Really Affect Asset Valuations? How do investors respond to the state of the U.S. federal fiscal deficit? In their August 2007 paper entitled "Fiscal Policy and Asset Markets: A Semiparametric Analysis", Dennis Jansen, Qi…
- The Logic of Valuation-motivated Short Sellers A model based on valuation-motivated short recommendations identifies future underperformers, with the top decile generating -0.8% vs bottom decile's 1.3% monthly risk-adjusted returns (out-of-sample 1990-1996; targets high-accrual, low book-to-market firms).
- Political Influences on Stock Valuation Levels U.S. stock valuations relative to present value models are associated with political factors: higher under Democratic Presidents, high approval and election years, lower after new military conflicts (1945–2005; 12 presidencies;…
- Enhancing the Value Premium Via P/E Analysis Enhanced P/E ratios using eight years of earnings history and adjusting for market/size/industry factors increased UK value premiums from 6% to 30%+ annually (1975-2004; extreme quintiles; annual rebalancing).
- A Fed Model Defense Clemens argues the Fed model (S&P 500 forward earnings yield minus 10-year Treasury yield) predicts relative stock-bond returns when yield spreads exceed ±1.5 standard deviations (Jan 1979–Aug 2006; theoretical framework).
- The Accuracies of Different Valuation Multiples (Ratios) European equity valuation multiples achieve median absolute errors under 30%, with two-year forward P/E outperforming other ratios (1996-2005; Dow Jones STOXX 600 and S&P 500; 50 multiples tested).
- Stock Valuation Indicator Fly-off ...over the last 80 years, a few price-normalized variables (price-dividend, price-earnings, price-output and price-consumption ratios) and the approximate consumption-aggregate wealth ratio have been the best stock market indicators.
- Which Financial Performance Measure Best Fits Stock Valuation? Forecasted earnings outperform cash flows and dividends as stock valuation predictors across all markets, with lowest mean/median valuation errors (1987-2004; U.S. and 10 international markets; cross-sectional analysis).
- Classic Paper: Company Valuation Methods Fernandez argues cash flow discounting methods are the only conceptually correct valuation approach, linking future cash flows and required equity returns to company value (2004 paper; theoretical framework).
- Focus on Return on Investment, Not P/E Sykes argues return on investment drives equilibrium stock returns, with investor return equaling earnings yield at equilibrium, linking earnings yield to core-value returns (1871-2004; theoretical model with large unexplained speculative…
- Last Nail in the Coffin of the Fed Model? Fed Model fails empirically across 20 countries, with projected P/E outperforming Fed Model predictions in 18 countries and actual P/E outperforming in all 20 (various periods ending June 2005).
- International Fed Model Test A long-term relationship among stock indexes, earnings and 10-year government bond yields exists for many of 13 countries, but bond yields do not set an equilibrium valuation (1973–2003; current earnings,…
- Fed Model Versus P/E Model Fed Model better explains market P/E behavior than mean-reversion model post-1960, with earnings yield correlating with 10-year Treasury yields (1880s–2005; Weigand-Irons structural break analysis).
- 50-Year Fed Model Meme? ...the Fed Model has worked pretty well starting about 1960, with interest rates since playing a key role in stock valuation.
- One Up on the Fed Model? In their June 2003 paper entitled "A General Theory of Stock Market Valuation and Return", Christophe Faugere and Julian Van Erlach contend that past stock returns are overstated and develop…
- Earnings Yield-Interest Rate Spread S&P 500 earnings yield-T-bill yield spread timing strategies outperformed buy-and-hold on risk-adjusted basis (higher Sharpe ratios; 1970-2000; switch to cash when spread below 10th percentile).
- Fed Model: Predictive or Not? ...the Fed Model is inferior to fundamental valuation in predicting long-term stock returns, but it may have some tactical value.
Sentiment
97 articlesSurveys, put-call ratios, fear and greed, and positioning.
- Testing a Proxy for the Goldman Sachs Panic Index Claude-estimated Goldman Sachs Panic Index values (pixel-reconstructed from a chart) showed high readings (above 8) followed by 1.2% average next-week S&P 500 returns vs -0.5% for readings below 4 (mid-Aug…
- AI-simulated CFO Sentiment GPT-5.4-simulated CFO optimism correlates with actual Duke-Federal Reserve CFO Survey responses (LLM mean 56.6 vs actual 60.1; regression coefficient over six standard errors from zero; 6,075 responses, 2002–2025; firm and…
- CFO U.S. Economic Sentiment and Stock Market Returns CFO average U.S. economic sentiment shows little ability to predict S&P 500 returns; correlations with next-four-quarter returns are very weak and negative, while change in sentiment shows no predictive link…
- How Are Sentiment-driven ETFs Doing? An equal-weighted average of three sentiment ETFs (BUZZ, MOOD, NEWZ; monthly rebalanced) underperformed VTI (1.0% vs 10.7% CAGR; MaxDD -51% vs -25%; sample periods vary, through Mar 2026), with all…
- Goldman Sachs Panic Index Proxy A VIX-percentile Panic Index Proxy (trailing 2-year rank, 0–100) shows weak lead-lag association with daily S&P 500 returns, with very high and very low quintiles followed by higher returns (Dec…
- Grok Sentiment Index? Weekly Grok Sentiment from X posts (engagement-weighted, ≥50 posts/week) showed no reliable lead-lag relationship with weekly S&P 500 returns; next-week S&P 500 return averaged 0.7% after positive sentiment vs 0.0%…
Show all 97
- Consumer Sentiment and Stock Market Returns Monthly changes in the University of Michigan Consumer Sentiment Index did not predict S&P 500 returns at any 1–6 month horizon; average rolling 36-month correlation with next-month return was -0.03…
- AAII Investor Sentiment as a Stock Market Indicator AAII net investor sentiment (percent bullish minus bearish) correlates weakly negatively with S&P 500 26-week future returns (-0.11; R-squared 0.01; July 1987–mid-July 2025; weekly data, in-sample), with the correlation near…
- Daily Global Investor Sentiment A risk-on/risk-off (RORO) index, the first principal component of daily changes in credit, equity volatility, liquidity and currency/gold proxies, is associated with a -0.8% contemporaneous USD equity return per one-standard-deviation…
- Extracting Sentiment Probabilities from LLMs Logit Extraction (continuous LLM sentiment probabilities) is associated with higher gross risk-adjusted long-short returns than discrete sentiment signals across LLMs and quantile thresholds (Jan 2010-Oct 2020; 216,837 headlines, 263 firms;…
- Fear as Treasuries Market Driver A TRMI-based bond Fear Index is associated with higher future 1-year excess returns on 2- to 5-year U.S. Treasuries over the 1-year rate (out-of-sample R-squared 0.22–0.30; Jan 1998–Dec 2022; expanding…
- Pattern Recognition Software Plus Confirming News Sentiment? A daily long-short portfolio of CNN-identified candlestick reversal patterns confirmed by news sentiment earned about 650% gross cumulative return in U.S. stocks (2003–2021) and 600% in Chinese A-shares (2014–2021); 1-day…
- Combining Financial Stress with AI News Sentiment to Time Stock Markets Dynamic SI News (monthly switch between a stress index and stress index × GPT-4 news sentiment) showed higher Sharpe and Calmar ratios and shallower drawdowns than volatility-scaled buy-and-hold across indexes…
- Validating CNN Fear and Greed Index as Return Predictor CNN Fear and Greed Index levels showed a slight, persistent negative association with future SPY returns, and next-day SPY returns did not vary systematically across deciles of daily F&G change…
- CNN Fear and Greed Index as Return Predictor The CNN Fear and Greed Index is associated with future returns of the S&P 500, Nasdaq Composite and Russell 3000, but not gold; coefficients are small, so timing frictions may…
- Combining Equity Market Stress and Sentiment Indications Dynamic SI+News (monthly switch between a stress-index signal and a stress-index×ChatGPT 4 news-sentiment signal) outperformed S&P 500 Buy-and-Hold (net Sharpe 0.81 vs 0.45; Jan 2005–Dec 2023; 0.02% turnover frictions; news-data…
- ChatGPT-generated Financial News Sentiment and NASDAQ Returns ChatGPT-based detrended cumulative sentiment strategy (long/short NASDAQ, 20-day lookback, 0.2% costs) beat NASDAQ buy-and-hold on Sharpe (0.88 vs 0.79), but returned only 1.2% vs 16.1% annualized (2010–2023; Bloomberg news).
- Interactions of Stock Mispricing and News Sentiment Overpriced stocks (bottom 30% of 11 mispricing signals) reacted about twice as strongly as underpriced stocks to positive and negative news days, then lost the positive-news gain over the next…
- Small Business Owner Sentiment and the U.S. Stock Market Monthly changes in the NFIB Small Business Optimism Index correlate positively with same-month and prior-month S&P 500 and Russell 2000 returns, but not with later returns (Jan 2003–Jan 2024; in-sample,…
- Predicting Short-term Market Returns with LLM-generated Market Sentiment ChatGPT and BARD news-sentiment scores are associated with next-day and next-week S&P 500 return reversals after pessimism increases, with about 2% gross annual CER gains over a control portfolio (daily…
- Using VIX and Investor Sentiment to Explain Stock Market Returns A joint high-VIX threshold (~80th percentile) and linear Baker-Wurgler sentiment model explains 4.7%, 10.8%, 19.2% and 28.7% of future S&P 500 excess return variation at 1, 3, 6 and 12…
- ChatGPT as Stock Sentiment Analyst ChatGPT headline sentiment scores correlated with next-day U.S. stock returns (0.23 overall; 0.65 in small stocks), and gross long-short results were strong but with large drawdowns (Oct 2021–Dec 2022; RavenPack…
- Testing SACEMS with Different Bull-Bear Lookback Intervals SACEMS Bull-Bear Combo (4-month lookback when the S&P 500 is above its 10-month SMA, 2-month when below) had higher gross CAGR than the 4-month baseline for EW Top 3 but…
- Trend Following Plus Relative Sentiment for Stocks-Bonds Allocation Over Nov 1994–Aug 2022, a 50-50 mix of trend following (10-month SMA, U.S. stocks/aggregate bonds) and relative sentiment (weekly COT/Sentix) thinned return tails versus each component, but relative sentiment alone…
- News Sentiment and Stock Market Returns Weekly Daily News Sentiment Index (SMA7) and its weekly change show hardly any relationship with next-week S&P 500 returns, and decile averages vary unsystematically (Jan 1980–Jul 2022; 222 weeks per…
- Dumb Money Confidence as a Stock Market Return Predictor Monthly SentimenTrader Dumb Money Confidence showed no useful relationship with next-month SPY total returns (R-squared 0.002; quintile averages unsystematic; Dec 1998–Jul 2022; 56–57 monthly observations per quintile; in-sample, descriptive; not…
- Panic Selling and Panic Sellers Panic selling (a ≥90% monthly account equity drop, ≥50% from trading) is rare: about 0.1% of investors at any time and 9% at least once, with 31% never reinvesting (653,455…
- Investor Sentiment as Measured by Social vs. Traditional Media Seeking Alpha article sentiment predicted next-day S&P 500 returns with about twice the power of Wall Street Journal sentiment for negative and pessimism indexes (-0.10% vs -0.04% per one-standard-deviation pessimism…
- Surprise in Short Interest as Stock Return Predictor Equal-weighted hedge portfolio long stocks with lowest short interest surprises, short highest surprises generates 0.43% gross monthly return (Mar 1980–Dec 2013; monthly rebalancing; U.S. stocks >$5).
- Relative Sentiment plus Machine Learning for Stock Market Timing Top 10 U.S. relative sentiment strategies (institutional vs individual Sentix expectations) outperformed benchmarks by 6.5-7% annually (Aug 2002-Sep 2019; 990 ML strategies tested; monthly rebalancing across SPY/bonds/cash).
- Deeply Learned Management Sentiment as Stock Return Predictor Deep learning-applied 10-K sentiment shows statistically significant but economically slim predictive power, with trading strategies generating insignificant 3-factor alpha (1994–2017; sentence-level analysis).
- Sentiment Indexes and Next-Month Stock Market Return Five sentiment indexes achieved 60-63% directional accuracy predicting next-month S&P 500 returns, with Improved Investor Sentiment generating 12.1% vs 7.1% gross annualized return (Jan 1983–Apr 2014; 120-month rolling regressions).
- Momentum and Bubble Stocks Bubble stocks (high shorting demand, low institutional ownership) show reversed momentum effects, with bubble winners generating -1.28% monthly 4-factor alpha versus -1.30% for bubble losers in year 1, then -0.75%…
- A Few Notes on Buy the Fear, Sell the Greed Connors presents seven behavioral quant strategies exploiting fear/greed extremes in stocks and ETFs, including RSI PowerZones, VXX shorting, and gap-down buying (gross returns through 2017; in-sample backtests with survivorship bias…
- Isolating Ends of Stock Booms and Panics? The study finds social media sentiment helps predict intraday mean reversion, with sentiment-weighted strategy generating 24.1% vs 20.6% gross annualized returns (2011-2014; 30-minute rebalancing on top 500 stocks by volume).
- AAII Stock Screens AAII screens averaged 12.1% gross CAGR versus 7.0-8.5% for benchmarks, but fell to 9.4% after 0.5% trading frictions (Jan 1998-Mar 2018; 60 screens vs SPY/NAESX/VTSMX; monthly equal-weight rebalancing).
- Social Trading Leader Overconfidence and Influence Self-enhancement biased social trading leaders underperformed unbiased leaders (-1.14% vs +0.70% raw returns) yet attracted 23% higher fund inflows (June 2012–Nov 2016; 3,519 portfolios; 360-day future returns).
- Financial Distress, Investor Sentiment and Downgrades as Asset Return Anomaly Drivers Stock/bond return anomalies concentrate in financially distressed firms during high sentiment periods, with overpriced distressed stocks underperforming underpriced by 96bp monthly (Jan 1986–Dec 2016; 12-anomaly composite ranking).
- Survey of Research on Investor Sentiment Metrics Baker-Wurgler sentiment index and other sentiment metrics show minimal predictive power for monthly U.S. stock returns, with R-squared ranging 0.000-0.122 (Jul 1965-Dec 2016; statistical analysis).
- Margin Debt as a Stock Market Indicator NYSE margin debt changes show 0.39 contemporaneous correlation with S&P 500 returns but zero predictive power at one-month horizon, with margin debt lagging stocks by 1-2 months (Jan 1959–Aug 2017;…
- Volatility Patterns as Bubble/Crash Indicators Volatility fails to predict bubble bursts, with two-thirds of 40 bubbles ending during low volatility periods; volatility spikes occur during crashes, not beforehand (Oct 1929–Sep 2011; event study).
- Aggregate Stock Option Put-Call Ratio as Market Return Predictor PC-OI (put-call open interest ratio) for high-volatility stocks predicted 3-month market returns with R-squared 0.12 but underperformed post-2008 (Jan 1996–Aug 2014; gross returns, long/short market timing).
- News Sentiment and Future Stock Returns The post argues automated news sentiment extraction generates statistically significant return predictability, with weekly hedge portfolios producing 3.75% during publication week and positive returns across subsequent 13 weeks (2003–2010; 900,754…
- Predicting Stock Index Reversals with Amareos Sentiment Indicators Amareos sentiment indicators using random forest algorithms achieved 75.1% accuracy identifying market bottoms but missed 53.3%, with strategy outperforming buy-and-hold on 4 hold-out indexes (2005-2016; 6-month forecast horizons).
- Blogger Sentiment Analysis Ticker Sense blogger sentiment negatively correlates with 4-week future S&P 500 returns (average -0.16 correlation, negative 78% of time; Jul 2006–May 2016; 496 weekly polls of 29 bloggers).
- Best Measure of Investor Sentiment? VIX exhibits predictive power for S&P 500 returns, with one standard deviation increase (4.24) predicting 0.70% next-month decline (Jan 1990–Dec 2015; in-sample regressions controlling for economic variables).
- Testing Sentdex Sentiment Trading Signals Sentdex sentiment signals underperformed buy-and-hold for AAPL (4.2% vs 16.4% CAGR with frictions) and BAC (9.8% vs 28.8% CAGR with frictions; Oct 2012–Jun 2015; buy +6 sentiment, sell negative, 0.1%…
- High-frequency Investor Sentiment and Stock Returns High-frequency sentiment changes predict SPY half-hour returns during final trading hours (1998-2011; TRMI sentiment data; gross performance before transaction costs and data fees).
- Mark Hulbert’s Nasdaq Newsletter Sentiment Index HNNSI shows negative correlation with S&P 500 returns, explaining 14-21% of variation over 5-63 trading days (Apr 2010-Sep 2015; 38 observations; contrarian sentiment indicator with significant clustering bias).
- Exploiting Crowdsourced Earnings Estimates and Stock Sentiments Combining negative Twitter sentiment with positive Estimize earnings surprises generates 0.87% three-factor alpha over 20 days post-announcement (Nov 2011–Dec 2014; gross returns; 16,840 earnings events).
- Exploiting Unusual Changes in Hedge Fund Holdings and Short Interest A hedge portfolio long top AHFSR quintile (abnormal hedge fund holdings minus abnormal short interest) generates 0.68% gross monthly excess return (1990–2012; quarterly rebalancing; broad US stocks excluding small/low-priced).
- Days-to-cover Short Interest as a Stock Return Predictor DTC hedge strategy (long low/short high days-to-cover deciles) outperformed short interest ratio strategy (1.19% vs 0.71% monthly equal-weighted; 1988-2012; monthly rebalancing; NYSE/AMEX/NASDAQ stocks).
- Aggregate Short Interest and Future Stock Market Returns Aggregate short interest outperformed historical average and 14 other predictors in out-of-sample tests (3.5-4.6% annualized certainty equivalent gains; 1990-2012; Short Interest Index vs S&P 500 excess returns; monthly rebalancing).
- Interaction of Sentiment and Liquidity with Stock Return Anomalies Stock return anomalies strengthen with high investor sentiment, with 80% showing larger magnitudes after above-median Baker-Wurgler sentiment months (Aug 1965–Dec 2011; 100 anomalies via long-short extreme decile portfolios).
- Exploiting Interaction of Hedge Fund Holdings and Short Interest A hedge portfolio long stocks with increasing hedge fund holdings plus decreasing short interest outperformed its short counterpart (2.5% vs -2.5% gross quarterly abnormal return; 2000-2012; quarterly rebalancing; 5,357 U.S.…
- Aggregate Short Interest as a Stock Market Indicator NYSE aggregate short interest changes weakly predict SPY returns (R-squared 2-3%; Jan 2002-Feb 2014), with biweekly timing strategies outperforming buy-and-hold but underperforming in second subperiod.
- Predictive Power of Sentiment Based on Retail Investor Trading Activity TD Ameritrade's IMX retail sentiment index shows weak predictive power for S&P 500 returns, with IMX explaining only 6% of return variation (Jan 2010–Aug 2013; R-squared 0.06; monthly data).
- Mark Hulbert’s Stock Newsletter Sentiment Index HSNSI shows weak contrarian predictive power for S&P 500 returns (-0.17 correlation, 3% R-squared at 5-day horizon; July 2002–Sep 2013; 287 observations, stronger correlation with past than future returns).
- Predictive Power of Put-Call Ratios Put-call ratios show weak, inconsistent correlations with SPY future returns (R-squared 0.006 for best performer; 1995-2013; CBOE/ISE total/index/equity ratios over 5-63 day horizons; 68-trade strategy underperformed buy-and-hold).
- ETF Short Interest and Future Returns ETFs with highest short interest ratios outperformed MSCI World Index by 0.58% monthly vs 0.71% for lowest ratios (Feb 2002–Dec 2011; 21 iShares single-country funds; 3,419 fund-months).
- Essence of Investor Sentiment The business cycle component of investor sentiment significantly predicts returns for 16 of 28 long-short strategies versus 3 for the residual component (1965–2010; cross-sectional stock return prediction tests).
- Gold Bubble? Yes Baur and Glover find statistically significant gold price bubbles during 1970-1980 and 2008-2012, with explosive price behavior driven by chartist dominance over fundamental demand (Jan 1970–Aug 2012; technical analysis).
- Investor Overconfidence and Trading Behaviors Merkle finds UK self-directed investors exhibit systematic overconfidence, overestimating returns by 4.3% and underestimating volatility by 2-4x, with overconfidence positively linked to turnover and risk-taking (617 investors; Jun 2008–Dec 2010;…
- Monthly News Sentiment Predicts Stock Market Returns? Uhl finds Thomson Reuters news sentiment outperformed LEI in predicting DJIA returns (negative sentiment strategy: 48% vs 8% buy-and-hold in 2010; monthly data 2003-2010; 12 out-of-sample trades).
- Gold Bubble? No Bialkowski et al. find no gold bubbles during 1978-2010, with commodity dividend models explaining over half of gold price variation (Nov 1978-Mar 2010; regime-switching tests on spot/futures prices).
- Stock Market Volatility by Bull-Bear Regime Volatility more consistently aligns with SMA-based regime classification than returns; 58% of extreme up/down days (>2σ) occur during bear regimes despite bears comprising only 31.7% of time (Jan 1954–Jul 2011;…
- Interaction of Investor Sentiment and Stock Return Anomalies Stambaugh et al. find 11 stock anomalies amplify following high sentiment periods, with 70% of hedge portfolio alpha occurring after above-median sentiment months (July 1965–Jan 2008; three-factor adjusted).
- Prelude to Panic? Stock market herding systematically precedes major panics (1985-2010; Russell 3000 sample; distribution analysis of daily co-movement), but findings suffer from potential data snooping with only 8 crisis events.
- Exploitability of Monthly Short Interest for Individual Stocks Heavy short interest predicts -0.7% monthly returns, but this concentrates in high-fee stocks where borrowing costs offset underperformance (Jan 2004–Dec 2008; NYSE-AMEX-NASDAQ sample matched on size/book-to-market/momentum).
- Total Fear Premium The Total Fear Premium (derived from Treasury yield depressions) strongly correlates with VIX during both pre-crisis and crisis periods, potentially offering a purer downside fear measure (Jul 2004–Feb 2010; daily…
- Sentiment from Google Insights and Return Continuation High Google search interest amplifies 4-week momentum continuation in S&P 500 stocks, with search-augmented winners continuing to advance and low-interest losers extending declines (214 stocks; 2004–2009; weekly returns).
- Bubbles: Ride, Watch or Play the Pop? Equal-weighted allocation to all industries in bubble mode generates 3%-8% annual abnormal returns (July 1926-December 2006; monthly rebalancing; 48 value-weighted industry indexes).
- Steve Sjuggerud’s Sentiment Steve Sjuggerud's sentiment-based U.S. stock market forecasts show mixed accuracy in a small sample (Oct 2005–2012; principal focus post-2008; contrarian sentiment approach with vague horizons).
- Technical Analysis of Market Bubbles and Anti-bubbles Li-Ren-Sornette's power-law model correctly identified S&P 500 bubbles ending in Oct 1987, Oct 1997, Aug 1998, and Q1 2000, while their anti-bubble model's rebound alarm correlated with actual rebounds (Jan…
- Fear of Disasters? Bollerslev and Todorov find large rare events induce a 5% nominal equity risk premium and account for over half the variance risk premium (1990-2008 S&P 500 futures; high-frequency empirical methods).
- When Market Sentiment Works Investor sentiment predicts low future returns for speculative stocks only during bull markets, with no predictive power in bear markets (1966-2005; equally weighted hedge portfolios of small-cap, unprofitable, growth, and…
- Robert Drach, Trading with 95% Confidence? Drach's Basic Timing Model Portfolio underperformed SPY (6.0% vs 6.3% CAGR; May 1995–Aug 2009; cash balance negatively correlated with S&P 500 returns at monthly/annual frequency).
- Purifying Stock Market Sentiment Indicators Purifying VIX of 18 price-dynamics variables raised profit factors of its trade signals versus random signals, while purification lowered profit factors for AAII, INV, HUL and PCR (five sentiment indicators;…
- Critically Delegating, or Fearfully Abrogating? ...evidence from this laboratory test suggests that individuals tend to accept without checking the risk-reward judgments of financial experts.
- Comparing German and American Investor Sentiment Indicators German sentiment (Sentix) positively predicts DAX returns while U.S. sentiment (AAII) loses predictive power post-2001 (German: Feb 2001–Jun 2008; U.S.: Jul 1987–Jun 2008; weekly survey data).
- De-Snooping Market Timing Rules Based on Fundamental and Sentiment Indicators Market timing rules based on fundamental and sentiment indicators showed no statistically significant outperformance versus S&P 500 buy-and-hold after correcting for data snooping bias (1980-2007; 6,792-8,280 simple rules tested).
- Do Stock Index Put-Call Ratio Trend Reversals Anticipate Trend Reversals of the Index Itself? S&P 500 put-call ratio trend reversals show no predictive power for index reversals (R-squared 0.02, correlation -0.13; Oct 2003–Jan 2008; 10-day normalized slopes, tested with various lags).
- Combining Short Interest and Analyst Recommendations Combining short interest with analyst recommendations generates hedge returns up to 9.6% per six months, with largest gains when short sellers disagree with positive analyst views (1994-2006; quarterly rebalanced portfolios).
- Investor Sentiment and Returns for Different Types of Stocks Smallest-decile stocks outperformed largest-decile stocks by an average 1.25% per month after low sentiment versus 0.27% after high sentiment (six-indicator PCA sentiment index; monthly decile sorts; 1962–2005; descriptive; not a…
- The Interplay of Short Interest and Institutional Ownership Short interest predicts negative returns and news, with heavily shorted stocks underperforming by 1.51% monthly; informativeness weakens with high institutional ownership (1988-2005; descriptive analysis).
- The Predictive Power of the Put-Call Ratio for Individual Stocks Low put-call ratio stocks outperformed high put-call ratio stocks by 0.4% next day and 1% next week (1990-2001; CBOE options; buy-to-open volumes only; effect concentrated in non-public detailed trading data).
- Whose Sentiment Matters, and for What Horizon? Institutional investor sentiment exhibits forecasting ability for six-month returns but not one-month, while private investor sentiment shows no predictive power at either horizon (6 markets; 2001-2006; sentix survey data).
- Fear Factor? VIX improves explanatory power when combined with Fama-French-Carhart four factors, with VIX increases associated with decreased equity premium and increased value premium (Feb 1993–Dec 2003; daily data; factor model analysis).
- Aggregate Investor Sentiment and Stock Returns Prior-month high aggregate sentiment (PCA index of six indicators) is associated with lower value-weighted market returns: -0.34% vs +1.18% average monthly when sentiment is >1 SD above vs below its…
- Screening for Fear When Portfolio Building Implied idiosyncratic volatility correlates positively with future stock returns, while realized volatility shows no predictive power (240 optionable stocks; Jan 1996–Jun 2005; cross-sectional regressions).
- Financial News Sentiment Predicts Stock Returns? News sentiment predicts next-day stock returns (21.1% annualized for long positive/short negative stories; daily rebalance on DJNS stories; 1980-2004 S&P 500), but transaction costs eliminate profitability.
- Measuring Company Management Sentiment Risk sentiment increases in 10-Ks predict stock underperformance, with long-short portfolios generating 10%+ annual excess returns (1994–2005; 34,180 filings; monthly rebalance on sentiment changes).
- Aggregate Analyst Sentiment in the Long Run Analyst upgrades to buy by brokers with the lowest percentage of buy recommendations outperformed upgrades by highest-buy-percentage brokers (50 basis points per month; Jan 1996–Jun 2003; 438K recommendations).
- Predicting the Past with Investor Sentiment Lowrisk.com investor sentiment correlates most strongly with DJIA performance over the prior three months rather than future returns, showing reactive rather than predictive behavior (May 1997–Dec 2004; Pearson correlations across…
- Expert Overconfidence? Professional traders and investment bankers showed greater overconfidence than advanced finance-student control groups, including overly narrow predicted ranges (33 traders, 90 investment bankers; task-based testing; descriptive; timeframe not stated).
- Does Investor Sentiment Give Trading Signals? ...just because there is a crowd of traders does not mean there is wisdom.
- Does Consumer Confidence Predict Stock Market Returns? Consumer confidence level is negatively associated with next-month returns, statistically significant only for Nasdaq and small-cap stocks, not S&P 500 stocks (Conference Board and University of Michigan measures; Fisher and…
Breadth and technical indicators
33 articlesAdvance-decline lines, oscillators and equal-weight comparisons.
- Do Equal Weight ETFs Beat Cap Weight Counterparts? Eight equal-weight ETFs underperformed their cap-weight counterparts on CAGR (average 10.4% vs 11.9%; none matched its benchmark; available sample periods through June 2026; monthly dividend-adjusted prices; RSP/SPY, QQEW/QQQ, others).
- Sector Breadth as Market Return Indicator Sector breadth (count of nine SPDR sector ETFs with positive past returns, 1–12 month lookbacks) shows very small correlations with next-month SPY return; breadths of 8–9 show modestly positive averages…
- Simple Ways to Beat Equal-weighted Stock Portfolios Past-return-enhanced equal-weighted portfolios (excluding low-momentum, high-volatility or low-Sharpe deciles) outperformed simple equal-weighted and cap-weighted portfolios (net Sharpe 0.54–0.67 vs 0.50 and 0.47; Apr 2002–Mar 2022; MSCI world stocks; monthly reformation;…
- RSI 14/Threshold 30 Applied to SPY with Fixed Holding Interval Daily and weekly RSI 14 oversold rules on SPY (buy on cross under/over 30, hold six months) offered little support over buy-and-hold SPY; daily rules are in SPY 58% of…
- Equal-weight vs. 19 Other Allocation Strategies Within and Across Asset Classes Across stocks, bonds, commodities and real estate, minimum variance strategies produced higher Sharpe ratios and lower VaR than equal weighting in multi-class portfolios, while mean-variance strategies did not consistently beat…
- Technical Indicator Model of Stock Returns A 14-indicator technical model (moving-average crossovers, momentum, on-balance volume) produced a long-short decile hedge portfolio with 4.22% equal-weighted monthly gross return vs 1.02% for the market portfolio (returns 1932–2020; gross…
Show all 33
- Testing Zweig’s Combined Super Model Zweig's Combined Super Model matched FFIDX CAGR while modestly improving Sharpe ratios via reduced volatility (Dec 1986–Mar 2020; monthly rebalancing between FFIDX and cash/bonds based on monetary indicators).
- Bollinger Bands: Buy Low and Sell High? Bollinger Bands strategies underperformed SPY buy-and-hold across all settings (best: 8.6% vs 10.3% CAGR; Apr 1993–Nov 2019; 21-day SMA, 0.5–2.5 std dev bands, SPY-cash switching with 0.2% frictions).
- Using RSI(2) to Trade Leveraged ETFs RSI(2) strategies on SSO underperformed buy-and-hold (7.7% vs 11.6% CAGR for 5-70 variant; 4.9% vs 11.6% for 10-90; Jul 2006–Oct 2019; oversold entry/overbought exit with SSO).
- Jim Cramer Using the S&P Oscillator Cramer's S&P Oscillator calls underperformed, with "In" calls lagging average S&P 500 returns and "Out" calls outperforming (21-day horizons; Jun 2008–Oct 2019; 6 "In"/9 "Out" calls from CNBC search).
- Combining RSI Range and RSI Momentum for Stocks RSI Bull Range-Momentum signals achieved 45-58% success rates with 1.84-2.40 profit/loss ratios across S&P 500 stocks (Jul 1998–Jun 2018; gross returns, trade-level analysis).
- Equal Weighting, Firm Age and Stock Returns Rebalanced equal-weighted portfolios outperformed bootstrapped equal-weighted portfolios (0.49%-1.23% annualized; July 1926-Dec 2016; age effect dominates size effect post-1976; 20k simulations).
- Mean-Variance Optimization vs. Equal Weight for Sectors and Individual Stocks MV-EW outperformed other 2-stage strategies across four metrics (0.83 vs 0.74 Sharpe ratio for EW-EW; Jan 1994–Aug 2017; Bayes-Stein MV for sectors, EW for stocks; U.S. sectors + largest cap…
- Can the Stock Market Have Bad Breadth? NYSE Up/Down Volume shows modest predictive power for monthly SPY returns, but the effect depends heavily on the 2008-2009 crash period (May 2003–Aug 2017; daily correlations and quintile rankings; descriptive…
- Conservative Breadth Rule for Asset Class Momentum Crash Protection VAA breadth momentum strategy outperformed benchmarks (10-16% vs 7-8% CAGR; out-of-sample 1993-2016; breadth-based crash protection with optimized thresholds; 12 global asset classes).
- Hard to Beat Equal Weighting? Equal weighting outperformed most sophisticated allocation strategies after accounting for data snooping and trading costs (0.05%-0.50% frictions; 1969-2014; White's Reality Check on 5,513 strategies across 7 equity datasets).
- Dual Momentum with Multi-market Breadth Crash Protection PAA2 (dual momentum with high crash protection) outperformed 60/40 (10.4% vs 8.2% CAGR; 1993–2015 out-of-sample; monthly rebalancing of top 6 risky assets from 13-ETF universe with breadth-based safe allocation; 0.1%…
- Multi-class RSI-based Dynamic Asset Allocation Dynamic RSI-based rotation to Top 5 assets within fixed allocations outperformed passive benchmarks (13.8% vs 11.2% CAGR for 50/30/20; 1991-2014; monthly equity, annual fixed income, quarterly hedge fund rotation; gross…
- Equal Weighting vs. All Feasible Long-only Mean-variance Optimals Equal weighting ranks middle-to-above middle against randomly specified long-only mean-variance optimal portfolios across 10 asset universes, performing similarly to cap-weighting (1963-2014; Sharpe ratios; annual rebalancing).
- Martin Zweig’s Four Percent Model Martin Zweig's Four Percent Model underperformed buy-and-hold across SPY, QQQ, IWM, and RSP despite reducing volatility and drawdowns (inception through Sep 2014; weekly signals on Value Line Index applied to…
- When Bollinger Bands Snapped Bollinger Bands volatility breakout strategy declined from 0.45% average daily gross buy-sell spread pre-1983 to 0.00% post-2001, suggesting market adaptation (14 international stock indexes; 1885-2014; gradual deterioration coinciding with rising…
- MPT Cannot Beat Equal Weight? Equal-weight portfolios outperformed long-short and long-only mean-variance optimal portfolios with similar/higher Sharpe ratios and much lower turnover (20 factor sorts and individual stocks; Jan 1963–Dec 2011; monthly rebal).
- Mean-Variance Optimizations Versus Equal Weight Mean-variance optimization beat equal weighting in 45 of 48 tests for European markets but only 4 of 48 tests for emerging markets (2002-2010; 96 implementations across 4 datasets; net Sharpe…
- Mean-Variance Optimization Versus Equal Weight Mean-variance optimization shows mixed results versus equal weighting across equity anomaly portfolios, with outperformance concentrated in size/value and size/momentum sets (Apr 1953–Dec 2010; Sharpe ratios; 5 asset sets).
- Testing the McClellan Oscillator and Summation Index McClellan Oscillator and Summation Index showed minimal predictive power for SPY returns with correlations near zero and R-squared statistics under 0.01 (Mar 1993–Jun 2012; NYSE breadth data).
- How to Beat Equal Weight Asset Allocation? Are there strategic asset allocation methodologies that reliably beat equal weight? In the February 2012 version of their paper entitled "Portfolio Optimization Using Forward-Looking Information", Alexander Kempf, Olaf Korn and Sven…
- Hindenburg Omens? Hindenburg Omens showed weak predictive power with market weakness typically peaking a quarter after signals, but 3 of 27 occurrences drove half the effect (1/3/86-8/13/10; S&P 500 weekly returns; very…
- Richard Moroney, Divining Dow Theory Richard Moroney's Dow Theory-based market forecasts showed mixed accuracy across 29 calls (2005-2012), with timing appearing reactive rather than predictive and recommended cash allocations too small for material impact.
- Out-of-Sample Tests of Bullish Regime 2-day RSI Signals TradingMarkets.com 2-day RSI rules showed mixed results in out-of-sample tests, with XLK averaging 0.25% per trade (70% win rate) and SPY averaging 0.69% per trade (83% win rate; Jan 2007–Mar…
- Combining RSI and MACD in Search of Concentrated Abnormal Returns Combined RSI/MACD signals underperformed individual RSI or MACD signals on SPY (DB-DS strategy 2.4% vs DB-SS 1.0% average round-trip gain; Jan 1993–Feb 2008; double-buy/sell vs buy/single-sell; 33-37 round trips).
- Simple Test of MACD Crossover as an Abnormal Returns Indicator MACD crossover strategy (12/26-day EMA difference crossing 9-day EMA with 25% filter) underperformed SPY buy-and-hold despite 1.3% average round-trip gains (Jan 1993-Feb 2008; 84 round trips, 45% market exposure).
- Simple Test of RSI as an Abnormal Returns Indicator RSI(14) oversold-overbought trading (30-70 thresholds) underperformed SPY buy-and-hold despite 2.5% average round-trip gains (Jan 1993–Feb 2008; 34 round trips, 29% time in market vs 71% in T-bills).
- Classic Research: Dow Theory Long Dead? Hamilton's Dow Theory editorials generated positive alphas and high Sharpe ratios (1902-1929; risk-adjusted analysis), but post-1929 trading frictions preclude large excess returns despite continued volatility reduction.
Economic indicators
98 articlesRecession signals, financial conditions, jobs, inflation and the yield curve.
- Yield Curve as a Stock Market Indicator 10-year T-note minus 3-month T-bill spread timing (monthly; SPY to cash when spread is over 1 SD below average) underperformed buy-and-hold SPY (CAGR 9.8% ITD, 8.9% rolling 20-year vs 10.8%;…
- Growth-inflation Regimes and Asset Class Returns Claude-generated charts of annual SPY, TLT, VNQ, GLD and DBC returns across four growth-inflation regimes show no obvious patterns; stagflation favored TLT and GLD but spans only four years (2006–2025;…
- BTC Interactions with GLD, CPI and EFFR Does bitcoin (BTC) return exhibit any exploitable leading or lagging roles with respect to gold (SPDR Gold Shares - GLD) return, change in the all-items consumer price index (CPI) or…
- Recent Interactions of Asset Classes with Inflation (CPI) Quarterly change in NSA all-items CPI correlates negatively with ETF returns two to three quarters later for multiple asset classes, with long-term Treasuries and gold differing (10 ETFs; Dec 2007–mid-Aug…
- Testing Wilshire 5000/GDP as Stock Market Predictor W5000/GDP level and its changes (quarterly, 2- and 5-year average change, slopes) show small, mostly negative correlations with future Wilshire 5000 returns up to eight quarters; quintile results are inconsistent…
- Kalshi vs. the Experts for Economic Variable Predictions Kalshi economic forecasts were comparable in accuracy to Survey of Market Expectations and Bloomberg consensus benchmarks, with a statistically significant improvement for headline CPI (2022–2025; high-frequency contracts; descriptive; no asset-trading…
Show all 98
- AI Panel Inflation Projections Is the evolving set of artificial intelligence (AI) platforms based on large language models interesting as economic forecasters? Are they monolithic, or diverse? As a simple exploration, we pose to…
- Economic Implications of U.S. Federal Debt Trajectory Auerbach and Gale project U.S. debt/GDP reaching 183% by 2054 under current legislation, and 199% if temporary tax/spending provisions become permanent (deficit and debt data 1965–2024; economic-model projections; descriptive, not…
- Recent Interactions of Asset Classes with Inflation (PPI) Change in not seasonally adjusted all-commodities PPI is associated with weaker future returns for multiple ETF asset classes two to three quarters later, while gold and long-term Treasuries differ (10…
- Gold Return vs. Change in M2 Monthly change in M2 shows no reliable relationship with gold returns, with near-zero lead-lag correlations inconsistent in sign and patternless next-month returns across M2-change deciles (spot gold; Dec 1974–Sep 2025;…
- GDP Growth and Stock Market Returns S&P 500 returns correlate positively with GDP, PCE and PDI growth one to three quarters ahead, while growth correlates very weakly negatively with later returns (quarterly and annual BEA data…
- Does M2 Lead Bitcoin or Gold? Monthly change in M2 shows only small correlations with future bitcoin and GLD returns at lead times of 1–6 months, and tercile sorts show no systematic relationship (Sep 2014–Feb 2025;…
- Mimicking Economic Expertise with LLMs GPT-4o mini persona simulations of Survey of Professional Forecasters respondents produced median forecasts closely aligned with human medians across economic variables (December 2024 paper; descriptive; no investment test; look-ahead bias…
- Public Debt, Inflation and the Stock Market Quarterly change in U.S. public debt correlates weakly with next-quarter S&P 500 return (Pearson 0.07; R-squared 0.007) and inflation (0.07; 0.005) (Jan 1966–Sep 2024; quarterly; in-sample; descriptive, not a trading…
- Consumer Inflation Expectations Predictive? University of Michigan expected annual inflation correlates 0.76 with next-year CPI inflation (R² 0.58) but 0.90 with current inflation; monthly changes do not predict SPY or TLT next-month returns (Jan…
- Money Supply (M1) and the Stock Market Monthly changes in seasonally adjusted M1 show near-zero correlations with S&P 500 returns at leads/lags of -6 to +6 months, and average returns two months later do not vary systematically…
- Money Supply (M2) and the Stock Market Monthly changes in M2 money stock show very small correlations with S&P 500 returns at leads of 0–6 months, and average returns do not vary systematically across M2-change deciles (Jan…
- Bitcoin Economics Hendrickson's March 2024 paper argues mainstream economic theory implies bitcoin could become a non-inflating currency but has not yet; crude valuation models range from $47,000 to $1,293,000 per bitcoin under…
- Economic Data Risk Premium In Short-term Options S&P 500 1DTE at-the-money call returns averaged 22.1% on economic data release days vs 2.6% on other days; at-the-money puts averaged -28.2% vs -1.6% (Jan 2012–early Dec 2023; midpoint quotes,…
- Expert Forecaster Inflation Forecasts Quarterly changes in Survey of Professional Forecasters 12-month inflation forecasts correlate -0.03 (GDP-based) and -0.23 (CPI-based) with changes in actual CPI inflation (GDP-based from Q2 1970, CPI-based from Q3 1981,…
- Economic Trend Following Economic Trend, a futures/forwards strategy trading trends in five economic fundamentals across equity, bond, currency and commodity assets, produced a 1.1 gross annualized Sharpe ratio, versus 1.4 for price trend…
- Treasury Yields and Inflation Lead-lag Monthly changes in 1- to 30-year Treasury yields (GS1–GS30) correlate modestly positively with CPI changes over the prior one to five months, with little predictive power from CPI beyond one…
- Leading Economic Index and the Stock Market U.S. Leading Economic Index changes showed no exploitable predictive power for SPY; release-to-release correlation was -0.06 (R² 0.00), and quintile returns were unsystematic (June 2002–Nov 2023; 257 as-released monthly observations;…
- Asset Class Reactions to Monthly Inflation Data Monthly change in U.S. CPI (all items) is associated with weakly negative correlations to equity ETF returns (SPY, IWM, EFA, EEM) 4-8 months later; all correlations are small (July 2002–Sep…
- Economic Policy Uncertainty and the Stock Market Monthly U.S. Economic Policy Uncertainty (EPU) Index levels and changes show no systematic relationship with next-month S&P 500 returns across quintiles; only the coincident correlation with EPU changes is negative…
- ISM Services PMI and Stock Market Returns ISM Services PMI level and monthly change show little relationship with next-month S&P 500 returns (correlations -0.03 and -0.09; R-squared 0.001 and 0.008; Jan 2008–Aug 2023; monthly data; descriptive, in-sample;…
- ISM Manufacturing PMI and Stock Market Returns ISM Manufacturing PMI level and monthly change show little association with next-month S&P 500 returns (correlations -0.06 and -0.01; Jan 1948–Aug 2023; in-sample, descriptive; not a trading test; PMI data…
- Firms that Keep Up with Inflation? Unexpected CPI inflation is associated with weaker 1-day U.S. stock returns for low-market-power firms (3.9% decline per 1% surprise) than for high-market-power firms (1.2%) (1977–2022; CPI releases vs. Bloomberg/Haver median…
- CPI-to-PPI Ratio and the Stock Market CPI/PPI changes show little predictive relationship with S&P 500 returns; monthly correlations are small and annual next-year returns are lower after both drops and jumps than in the middle tercile…
- CPI and Stocks Over the Short and Intermediate Terms Monthly U.S. total and core CPI changes show little association with S&P 500 returns (release-day and 5-day correlations near zero; quarterly correlation -0.12; Jan 1994–Mar 2023; 351 releases; in-sample, descriptive;…
- Weekly Economic Index and Asset Returns Weekly change in the Weekly Economic Index shows little or no correlation with next-week to four-week SPY or TLT returns, with no systematic decile pattern (weekly data, Jan 2008–mid-Feb 2023;…
- Retail Sales Growth and Stock Market Returns Monthly changes in seasonally adjusted and non-seasonally adjusted U.S. retail sales did not reliably predict next-month S&P 500 returns; average returns did not vary systematically across quintiles (Jan 1992–Oct 2022;…
- Stocks-Bonds Return Correlation and Inflation Monthly stock-bond return correlations were materially higher when inflation exceeded 5% for both FFIDX/FBNDX (Jan 1980–Sep 2022) and SPY/LQD (Jul 2002–Sep 2022); descriptive, small samples, not a trading test.
- Equity Factors Come and Go with Economic Regimes? Adjusting for the secular interest rate decline, 14 of 153 accepted equity factors (9%) were false positives and 63 (41%) robust, versus 100 of 1,395 candidate anomalies (7%) false positives…
- Asset Class and Factor Premium Performances Across Inflation Regimes Global equity and bond gross returns fall in high inflation, with stagflation equities averaging -7.1% nominal, while long/short value, momentum, low-risk and quality/carry premiums stay positive across regimes (1875–2021; gross…
- Debt-to-GDP Ratio and Investment Risk Premiums Higher U.S. government debt-to-GDP ratios are associated with higher future stock excess returns over T-bills (R-squared 0.11 at 1 year, 0.40 at 5 years; U.S. 1926–2020, 19 developed countries 1970–2018;…
- Economic Surprise Momentum Growth surprise nowcasts show short-term momentum and are associated with higher next-month stock index futures returns (gross Sharpe 0.50 global vs 0.36 for global equities; 1-factor alpha 2.1%; 191 variables,…
- Housing Starts and Future Stock Market/REIT Returns Monthly changes in U.S. housing starts show no reliable lead-lag relationship with S&P 500 or FTSE NAREIT All REITs returns; the strongest correlation is stock returns leading starts by two…
- Alternative Yield Discount (Inflation) Rates Relative to four inflation measures (CPI, core CPI, PCE, trimmed mean PCE), S&P 500 operating earnings yield averaged 5.29% vs 2.20%–2.64% inflation, while real yields correlate only modestly with future…
- Consumer Credit and Consumer Discretionary Sector Returns Monthly changes in U.S. consumer credit correlate weakly and negatively with XLY returns 3–6 months ahead; quarterly results are similar but not strongly differentiated (Dec 1998–Jan 2022; lead-lag correlations, terciles;…
- Recent Interactions of Asset Classes with Economic Policy Uncertainty Monthly changes in the Economic Policy Uncertainty Index correlate negatively with same-period equity ETF returns and show weak, noisy links to bond and real-asset returns (10 ETFs; Dec 2007–Dec 2021;…
- Leading Economic Index Exploitable for Sector Rotation? Leading Economic Index changes offer little support for sector rotation across nine SPDR sector ETFs; correlations with next-three-month returns average 0.00 (monthly change) and 0.01 (3-month average), and quintile results…
- Corporate Debt-to-GDP Ratio as a Stock Market Indicator Quarterly changes in the non-financial corporate debt-to-GDP ratio showed no systematic pattern with subsequent S&P 500 returns across quintiles, and the strongest correlation had R² of 0.07 (Q4 1951–Q2 2021;…
- Comparing the Sahm Indicator and the Yield Curve Yield curve inversion (3-month T-bill above 10-year T-note) and SMA10 signals avoided low-return S&P 500 months more effectively than the Sahm Indicator, which mainly reduced return variability (monthly returns, 739…
- Unemployment Rate and Stock Market Returns Monthly change in the U.S. unemployment rate (BLS, seasonally adjusted) is not associated with next-month S&P 500 returns (correlation 0.03; Jan 1948–Jun 2021); the annual-horizon correlation is 0.20 (in-sample; descriptive,…
- Consumer Credit and Stock Returns Consumer credit changes show minimal correlation with S&P 500 returns (-0.02 Pearson correlation, R-squared 0.001; Jan 1943–Mar 2021; monthly data with various lead-lag relationships tested).
- Facing Down Inflation Commodities delivered 7.4% average real returns during 8 U.S. inflation episodes versus -1.2% otherwise, while stocks (-7.7%) and bonds (-7.2%) underperformed (1926-2020; episodes defined as CPI >5% with duration rules).
- Real Bond Returns and Inflation T-note real returns correlate negatively with contemporaneous inflation levels and changes, with same-year inflation explaining 52% of real return variation (1872–2020; Shiller data; descriptive analysis).
- Diversifying across Growth/Inflation States of the Economy Adding a defensive economic factor overlay to a classic 60%/40% portfolio improved annualized return and Sharpe ratio while reducing maximum drawdown in-sample (Jan 2006–May 2020; perfect foresight modeling).
- Combining an All Weather Portfolio, Crash Protection and Unemployment Trend RAA strategy (unemployment trend + crash protection signals switching between All Weather-like risky and bond defensive portfolios) achieved 9.7% CAGR with 6.9% volatility and 4.9% max drawdown (Dec 1970–Nov 2020;…
- Combining Economic Policy Uncertainty and Stock Market Trend S&P 500 10-month SMA outperformed combined EPU-trend strategies, with adding Economic Policy Uncertainty signals failing to improve market timing performance (Jan 1993–Sep 2020; SPY vs T-bills).
- Combining Market Trend and Chicago Fed NFCI Signals Combining SMA40 trend signals with NFCI-based market timing does not improve performance versus NFCI alone (Feb 1993–Apr 2020; weekly SPY/cash switching with 0.1% friction; NFCI<Mean had superior drawdown/Sharpe).
- Exploiting Chicago Fed NFCI Predictive Power NFCI-based SPY timing strategies outperformed buy-and-hold SPY (11.8% vs 10.4% CAGR for conservative strategy; Feb 1993–Apr 2020; weekly rebalancing based on NFCI change thresholds; 0.1% switching costs assumed).
- Chicago Fed NFCI as U.S. Stock Market Predictor Weekly changes in Chicago Fed NFCI show -0.13 correlation with next-week S&P 500 returns (-0.22 since 2000), with largest NFCI spikes preceding poor equity returns (Jan 1971–Apr 2020; in-sample).
- Optimizing the Combination of Economic Growth and Price Trends Growth-Trend timing (unemployment rate SMA12 + S&P 500 SMA10) outperformed 60-40 (9.8% vs 8.5% CAGR; Jul 1981-Oct 2019; multi-asset variations reduce max drawdown but sacrifice return; 0.1% friction assumed).
- Combine Market Trend and Economic Trend Signals? Combining SPY SMA10 with unemployment rate SMA12 ("Either" signal) slightly outperformed buy-and-hold (10.7% vs 10.0% CAGR; 1948-2019; primarily driven by 2000s decade; 71-year sample).
- Short-term Equity Risk More Political Than Economic? The CTMSw news-based model outperformed historical average equity premium prediction (0.73 vs 0.47 net annualized Sharpe ratio; Jan 1999–Dec 2018; mean-variance portfolio with S&P 500/risk-free rate allocation).
- ISM PMI and Future Junk Bond Returns? PMI-based junk bond timing strategies did not materially reduce CAGR versus buy-and-hold despite lower volatility (VWEHX-cash/HYG-cash; 1986-2019/2007-2019; monthly switches on prior-month PMI change sign).
- Which Economic Variables Really Matter for Stocks? Sparse PCA identifies yields, housing, and optimism factors as key drivers of stock returns, with mimicking portfolios generating gross Sharpe ratios of 0.85, 1.02, and 0.53 respectively (Feb 1960–Jun 2018;…
- Unemployment Claims Reports and Near-term Stock Market Returns Weekly changes in unemployment claims show very small, erratic correlations with SPY returns across multiple timeframes (Jan 1993–Jul 2018; 1,330 weeks; release-day and next-week returns).
- Worldwide Long-run Returns on Housing, Equities, Bonds and Bills Housing outperformed equities on a risk-adjusted basis across 16 developed countries (7% real returns, 10% volatility vs 22%; 1870-2015; modeled housing returns with rental income; equal-weighted averages).
- Combining Market, Unemployment and Interest Rate Trends Three-signal timing strategy (SP500, unemployment, T-bill trends) outperformed SP500 (9.0% vs 8.2% CAGR; Jan 1970-Oct 2017; monthly rebalancing), but outperformance concentrated in 1974 and 2008 bear markets.
- Do Hedge Funds Effectively Exploit Real-time Economic Data? High economic data beta hedge funds outperformed low beta funds (5.5% annualized 7-factor alpha spread; 1994-2014; top vs bottom deciles of 2,224 equity hedge funds sorted on sensitivity to real-time…
- Interpreting Inverted Yield Curves as Economic Indigestion Parker argues yield curve inversions reflect economy information processing efficiency, linking an R/C ratio to equity bear market timing (1990-2016; theoretical with limited empirical testing).
- Economic Uncertainty as a Stock Return Factor Stocks with low economic uncertainty betas outperformed those with high uncertainty betas by 0.51% monthly equal-weighted (July 1972–December 2014; decile sorts on 60-month rolling regression betas).
- Finding Close Economic Substitutes for Stock Pairs Trading Constrained cointegration pairs trading outperformed conventional minimum distance strategy (1.0% vs lower returns; 1963-2013; close economic substitutes only; U.S. stocks).
- Enhancing Stock Market Prediction with Distilled Economic Variables Adding 118 distilled economic variables to conventional predictors modestly improved S&P 500 timing (62.1% vs 58.5% directional accuracy; 0.63 vs 0.49 Sharpe ratio; Jan 1967–Dec 2014).
- Economic/Market Factor Investing Heat Map The study finds economic/market factors explain most contemporaneous variations in asset class returns, but notes little research agreement on predictive power (1996-2013; descriptive factor analysis; not a trading test).
- Economic News Leaks to Some Traders? Russell 3000 stocks show consistent directional price movements 30 minutes before ISM Manufacturing and Construction Spending releases, with pre-release moves capturing most net changes (Jan 2013–Oct 2015; 1-min data; descriptive…
- Gold a Consistent Dynamic Inflation Hedge? Gold acts as an inflation hedge at short-term and long-run horizons across developed markets, with dynamic correlations below 0.19 but strengthening during high-inflation periods (1968-2014; wavelet analysis).
- Inflation-based Projection of the Price of Gold Erb and Harvey find gold's fair value at $823/ounce as of July 2015, with extreme range $349-$2,077 based on mean reversion of real gold prices to inflation (1975-2015; golden constant…
- Money Supply Growth and Future Stock Market Returns M2 money supply growth correlates negatively with future stock market returns at 1-10 year horizons, adding predictive power beyond dividend yield and other variables (Jan 1959–Dec 2012; in-sample/out-of-sample tests).
- Predicting Government Bond Term Premiums with Leading Economic Indicators LEI4 outperformed historical averages in predicting government bond term premiums, with stronger performance for U.S. and UK bonds than German and Japanese bonds (1990–2011 out-of-sample tests; inception-to-date regressions).
- Using Economic Fundamentals to Predict Currency Exchange Rates Economic fundamentals predict currency returns, with multi-currency hedge portfolios generating up to 6% gross annualized excess returns (carry trade 6%, real money growth 6%, GDP growth 3%; Q1 1974–Q3 2010;…
- Hedging Inflation with Commodities Commodities generally failed as inflation hedges, with only 15 of 45 individual commodities beating inflation (1960-2012; spot prices vs CPI; energy index most consistent).
- Real-time Economic Data and Future T-note Returns Economic variables show diminished predictive power for T-note returns when using real-time data versus revised data, with much predictability disappearing in out-of-sample tests (Mar 1982–Dec 2011; 68 monthly variables; 2-5…
- “Real” Assets and Inflation T-bills strongly correlate with inflation (0.70 correlation; 1952-2011), while TIPS portfolios show near-zero correlation (0.10; 1997-2011) due to real yield dynamics offsetting inflation protection.
- Dueling Consensus Forecasts of Economic Indicators Bloomberg consensus forecasts are more accurate than Briefing.com forecasts for economic indicators, with Bloomberg surprises showing higher explanatory power for S&P 500 futures returns (Jan 1998–Aug 2010; 59 indicators; high-frequency…
- Economic Announcements and VIX VIX declines by 0.4%-1.8% on U.S. economic announcement dates, with uncertainty relief independent of good/bad news (1994–2012; close-to-close changes for GDP, CPI, PPI, employment; R² = 0.00 for announcement surprise…
- Real Estate as Inflation Hedge Equity REITs matched inflation 65.8% of the time during high-inflation periods vs commodities at 70.4% (Jan 1978–Aug 2011; six-month intervals; frequency analysis of nominal returns vs CPI).
- Commodities as an Inflation Hedge S&P GSCI provided economically significant inflation hedging across four of five measures over various horizons, with Non-Precious Metals performing best (Jan 1982–Aug 2010; five hedging capacity measures).
- Exploiting Commodity “Yield Curves” Commodity strategies exploiting yield curve slope and stability outperformed long-only benchmarks (Jan 2001–Jul 2009; directional strategies using most stable curves; 23 energy, metals, agriculture, livestock).
- Housing Price Reversion to Trend Real housing prices revert to trend across US/UK/Japan (1960s-2009; 1.9%/5.5%/1.9% annualized real appreciation respectively), with US showing strongest reversion and 2009 prices at low end of range.
- Analyzing the Economic Value of Predictive Variable Trading Strategies Christie argues that portfolio-level tests overstate anomaly profitability, linking real-time implementation constraints to reduced net returns (standardized unexpected earnings example; theoretical analysis with empirical illustration).
- Hedging Against Inflation Commodities best hedge short-term inflation shocks (4-10% return increase per 1% inflation rise; 1927-2008), but equities prove worst long-term hedge with no meaningful recovery post-shock.
- Outperformance Based on Three Macroeconomic Indicators Dynamic allocation using credit standards, Baltic Dry Index changes, and 2-year CMS rate changes outperformed via SPX/T-note futures (17.2% excess return, 1.34 Sharpe; Jul 1993–Jun 2009; 36-month rolling regressions).
- Following the “Hot” Economic Indicators Adaptive switching between stocks and T-bills based on iteratively selected best-performing economic indicators outperformed buy-and-hold (1977-2005; monthly rebalancing based on 100 economic variables vs broad stock index).
- Asset Allocation Driven by Four Economic Phases Dynamic asset allocation using four-phase economic cycle outperformed static benchmark (0.4% annually constant-risk; 0.9% annually return-enhancing; 1948-2007; 8 asset classes, 4 indicators).
- Stock Market Returns and Inflation: Illusion or Regimes? Lee finds inflation-stock return relationships switched from positive pre-WWII to negative post-WWII, supporting a two-regime hypothesis over inflation illusion (1927-2007; U.S. data; regime analysis).
- Which Economic Data Announcements Matter? Nonfarm payrolls, GDP advance, and ISM manufacturing surprises generate economically significant asset price responses, with strongest impact on interest rates (Jan 1998–Jul 2007; intraday futures analysis).
- Effects of Macroeconomic News on Commodity Futures Commodity futures indexes react positively to inflation and real activity surprises during recessions but show minimal response during expansions (CRB Index and S&P GSCI; 1989-2005; event study methodology).
- Macroeconomic Shocks and the Stock Market Inflation shocks explain almost 33% of S&P 500 variation, with negative persistent effects on real stock prices and market trend (Aug 1952–Dec 2005; vector autoregression analysis).
- Industrial Production as a Predictor of Stock Returns Output gap (log industrial production deviation from linear-quadratic trend) is associated with future U.S. excess stock returns, with adjusted R-squared of 0.02, 0.05 and 0.11 at monthly, quarterly and annual…
- Growth Versus Value and the Yield Curve Growth stocks outperformed value when T-note/T-bill yield spread widened, contradicting Fisher's hypothesis (quarterly correlations 0.58 vs 0.41; Sep 2001-2007; longer 1989-2007 sample showed no relationship).
- Effects of Inflation Rate Trend and Volatility on Stock Returns CPI inflation trend and volatility show no meaningful relationship with S&P 500 returns (correlations 0.09 and 0.05 respectively; Jan 1990–Jun 2007; monthly data with 4-month lookbacks).
- Inflation, Monetary Policy and the Stock Market Stock market booms tend to commence when real output growth exceeds average and inflation is low, ending within months after inflation increases and monetary tightening (U.S./U.K./Germany; 20th century; descriptive analysis).
- Earnings, Inflation and Stock Returns S&P 500 quarterly returns show no correlation with concurrent earnings growth (0.03 correlation 1990-2022) or inflation (-0.01 correlation), failing to confirm negative earnings-returns relationship from 1970-2000 study.
- Should Equity Investors Hope for Good or Bad Economic Forecasts? ...equity investors should be contrarian when considering economic forecasts. Bad is good.