Research
Analytical notes and tools built on Market Runners data. Each study pairs our market pipeline with statistical analysis — the full methodology sits behind every write-up.
Prices 227 US banks both by the market and by a model that strips out growth expectations. The market sits ~37% above hope-free tangible value. Of 24 banks below the line, 19 are microcaps or mutual-holding-company thrifts — structural discounts, not mispricings.
Regresses 2,068 stocks against 10 macro and style drivers to give each name a fingerprint. Median systematic R² is 0.22, so ~78% of the average name is idiosyncratic — but the fingerprints still surface cross-sector twins, names that have left their sector, and genuine portfolio diversifiers.
Blends how a stock behaves (PCA-denoised market-residual returns) with what the company is (fundamentals and sector structure) across 2,068 names. Recovers peers the sector label misses, and a 25-way clustering of the universe.
Decomposes an asset’s price history into five independent axes — trend, quality, volatility, phase and relative strength — then composes them into one plain-language label. Scored across 2,881 assets over 1,037 sessions; 60% of the universe is choppy rather than trending.
Identifies 10 macroeconomic regimes from 22 years of monthly data via hierarchical Ward clustering on a PCA-trimmed 10-feature panel. Validated with bootstrap stability, feature perturbation, NBER alignment, and Markov transition analysis.