Machine Learning Can’t Pick Winning Funds. But It Can Help You Avoid Losers
A 2023 Journal of Financial Economics paper claimed machine‑learning models could generate a 2.4% net annual alpha by selecting long‑only mutual funds. A 2025 replication discovered a coding error that gave the algorithms future returns, creating a classic look‑ahead bias. After correcting the error, the outperformance vanished, with returns falling 1.37‑1.42 percentage points and no statistical significance. The study also suffered from survivorship bias, undermining its original conclusions.
The Curious Case of “Dead” Factors
Recent data overturns the prevailing narrative that value and size premiums are dead. From October 2020 through early 2026, the Fama‑French U.S. small‑value index delivered a 19.7% annual return versus 13.8% for the broad market, a 6‑percentage‑point premium that was mirrored in...
Private Credit Risk: Look Past the Default Rate
Private credit has surged into the spotlight, yet most commentary fixates on default rates. A new April 2026 paper by Aksia and Calamos argues that the true risk lies in principal loss, not the loan’s label. The authors demonstrate that low...
Beyond the Hype: How AI Adoption Is Quietly Reshaping Corporate Efficiency
A new FTSE Russell and Lloyd’s List Intelligence study links corporate AI mentions to measurable efficiency gains, with the correlation turning sharply positive after 2015 and peaking in 2024. Researchers used NLP to count AI‑related terms in transcripts and filings,...
After Brutal Run Of Underperformance, Is The Value Premium Back?
Brian Chingono’s Verdad research paper examines whether the long‑awaited value premium has returned after a 14‑year “value famine” (2007‑2020) that saw value stocks lag growth by roughly 5.7% annually. By pairing historical valuation spreads with forward five‑year return vintages across...
When Analysts Get It Wrong: Expectation Bias, Market Inefficiency, and What It Means for Investors
A February 2026 study by Cheng Gao, Siyuan Ma and Peixuan Yuan introduces an Expectation Bias (EB) metric that predicts analyst forecast errors using only publicly available firm data via an XGBoost model. Sorting stocks by EB generates a long‑short...
Does Academic Research Actually Give Investors an Edge? A New Study Says Probably Not
Researchers Andrew Chen, Alejandro Lopez‑Lira, and Tom Zimmermann compared 212 peer‑reviewed stock‑return predictors with roughly 29,000 mechanically generated accounting ratios. In out‑of‑sample tests, the two approaches performed almost identically, with only a modest edge for top‑journal papers and for studies...
Private Credit: The Market’s Quiet Stabilizer
A March 2026 study by Franz Hizen, Giorgio Mondini, Paul Rintamäki and Sascha Steffen finds private credit has become a counter‑cyclical source of corporate financing. Using PitchBook data from 2005‑2024, they show the private‑credit share of new debt rose from about 20% in 2008 to...
Beyond The Hype: What Really Drives IPO Prices?
A new SSRN study of 848 U.S. IPOs from 2009‑2019 finds that first‑day price jumps are driven more by market overpricing than by underwriters’ deliberate underpricing. By benchmarking each offering against industry peers using price‑to‑sales multiples, the authors separate intrinsic...
AI Isn’t the Boom, or the Bust, Many Expect. Here’s What That Means for Investors
A March 2026 survey of 560 economists, AI insiders and professional forecasters found that 61% expect moderate or rapid AI progress by 2030, yet the median projection for annual GDP growth remains modest at 2.5%, closely tracking historical trends. The...
How Looking Back One Year Can Transform Your Small Cap Returns
A new Bridgeway Capital Management paper argues that the small‑cap premium appears muted because the current definition mixes recently demoted large caps with fresh market entrants. By separating stocks that have only recently fallen into the small‑cap universe from true...
The S&P 500 Bump That Doesn’t Last Stocks Added to the Index Get a Short-Lived Boost but Often Lag Comparable...
A new study by Sandifer, Smith, and Impink finds that while S&P 500 additions enjoy a short‑term price surge, they underperform comparable non‑index peers over the long run. The research expands beyond the usual days‑or‑weeks window, tracking performance for several...
When Stock Market Valuations Actually Matter: The Power of Extremes
Javier Estrada’s February 2026 paper analyzes 150 years of U.S. market data to test when valuation multiples best forecast 10‑year real returns. The study finds that extreme values—top and bottom 25%—of dividend yield, earnings yield, and CAPE yield deliver far...
The Hidden Cost of Convenience: Why Balanced Mutual Funds May Be Costing You Hundreds of Thousands
A new study of 1,260 balanced mutual funds managing about $1.6 trillion finds they consistently lag low‑cost index portfolios. Researchers examined 32 years of performance across four equity‑allocation tiers and discovered lower returns, weaker risk‑adjusted metrics, and negative net alpha after...
Rethinking Exit Strategies: How Machine Learning Can Boost Anomaly Returns
A January 2026 study by Nitin Kumar, Nagpurnanand Prabhala and Ravi Ranjan shows that using machine‑learning to time exits dramatically improves classic anomaly portfolios. By applying random convolutional kernels to 15‑day return windows, the model selects optimal exit dates for...