Log = comparable erasA P/E of 20 and a dot-com/hyper-growth P/E of 400 only sit on ONE readable chart in log scale — the default here. The axis still reads the actual multiple.
Valuation vs sizeThe P/E chart asks how expensive each name is; the market-cap chart asks how big it has become. A rising cap on a falling P/E = earnings out-running the price (value building); a rising P/E on a flat cap = multiple expansion (hope building).
Hot/cold tape = who's stretchedThe line charts show exact paths; the tapes make it glanceable which names are red-hot (expensive / largest vs their own history) and which are washed-out blue right now.
Cross-sectional rankSwitch a tape to cross-sectional rank to see, each month, which of the 12 was the most expensive / largest relative to the others — rotation across the mega-cap complex.
Isolate one nameClick any line to isolate it (the others dim, benchmarks-style is N/A here); click empty space to clear. Click a tape date column to rank every stock at that month.
Gaps are honestWhere P/E is undefined (negative earnings) the line simply breaks. Younger names (Tesla, Meta, Alphabet) start partway across — never stretched to fake a longer history.
Residual-z = valuation vs size decouplingThe residual-z strip automates the read-through above: it regresses each month's P/E move on its market-cap move and colors the leftover — red = P/E ran richer than size implies (multiple expansion beyond size), blue = P/E ran cheaper than size (earnings out-running the price). Coupled months are pale.
Seasonality = individual months, no averageThe seasonality panel shows each calendar month's return distribution across all years as a rail (every year is a tick, black pin = most recent). There is no single seasonal-average line by design — the spread and skew are the point, and an average would hide them.