Research-backed writing on financial seasonality, machine learning in markets, and the methodology that separates a real edge from a coincidence.
Most published win rates arrive without the losing picks they were computed from. TradeWave keeps a public daily-pick ledger where all 75 picks since March 17, 2026 sit in the open, the 13 losers marked in red beside the wins. Here is what that page shows and how to check it yourself.
Read article ->TradeWave's daily AI pick keeps a public ledger, and this is the honest read of it: 64 resolved picks since March 17, 2026, an 80% win rate, and 13 losses sitting in plain view. Here is what counts as a win, how the number is computed, and why we publish the rows that went against us.
Read article ->Any service can quote a win rate. A real track record survives five checks: picks timestamped before the outcome, a stated denominator, visible losses, a defined win condition, and stats computed live from the record. Here is the standard, and how TradeWave's public ledger measures against it, losses included.
Read article ->It is July of a midterm year, historically the only stretch of the four-year election cycle where the S&P 500 loses ground. Eighty days from now, on September 27, the strongest broad-market seasonal window in nearly a century of data opens. Here is what the calendar actually says, measured, with the receipts on screen.
Read article ->If you tested 100 random strategies, the best one would look like genius. That's not a hypothetical - it's the math. Here are the five most common ways backtests overstate future returns, and how to defend against each.
Read article ->Ken French's 1980 paper showed Mondays returned negative on average while Fridays returned positive. The pattern was real, robust, and is now mostly gone. Here's the autopsy.
Read article ->Bouman and Jacobsen's 2002 study found Nov-Apr returns dramatically higher than May-Oct across 37 markets. Two decades of follow-up research has tested the claim against every reasonable counter-explanation. Here's what holds up - and what to actually do with it.
Read article ->Rozeff and Kinney's 1976 finding that small-cap stocks earn outsized January returns triggered 50 years of work. The mechanism (tax-loss harvesting) is well understood. The interesting question now is whether the anomaly survived its own discovery.
Read article ->Tree-based models can extract real predictive information from financial data that linear factors miss. Deep nets help less than the marketing implies. Here's the honest map of the territory.
Read article ->Stocks that beat earnings keep rising for weeks after the announcement; stocks that miss keep falling. The pattern was first documented in 1968. It has resisted being arbitraged away for over half a century. Here's why.
Read article ->Yale Hirsch's Stock Trader's Almanac claims year three of every U.S. presidential cycle has averaged 14% returns. The number is real. The interpretation is where investors get into trouble.
Read article ->A 10-year average return that's positive 8 out of 10 years feels like a real edge. Run the math: it isn't, by itself. Here's the toolkit for reading patterns honestly - sample size, multiple testing, regime breaks, and what 'statistical significance' really tells you.
Read article ->Heston and Sadka's 2008 paper found that stocks with high returns in a given calendar month tend to outperform again in the same month for years afterward. The combined seasonal-plus-momentum pattern is stronger than either alone.
Read article ->Sector-level seasonal patterns are unusual among market anomalies: they have plausible economic mechanisms behind them. Driving season, retail holidays, harvest cycles, and weather patterns each leave traceable footprints in equity prices. Here's the map.
Read article ->