| Symbol | Name | Sector▾ | Cohort▾ | Mkt Cap▾ | $ Volume▾ | Amihud▾ | C-S Spread▾ |
|---|---|---|---|---|---|---|---|
| Loading… | |||||||
The cohorts are a proxy, not a small-cap index. "Small-Cap" here is the bottom quintile (~100 names) of the S&P 500 by market cap, and "Mega-Cap" is the top quintile (~100 names) of the same index. Neither is the Russell 2000, S&P SmallCap 600 or any dedicated small-cap benchmark. There is no free small-cap constituent list, so the cohorts use the S&P 500's own weekly metadata (name, sector, shares outstanding, market cap, from Sector Beeswarm's meta.json). Even the smallest S&P 500 company is worth billions and is more liquid than a typical small-cap, so read "small-cap" on this page as "smallest within the S&P 500."
Liquidity metrics. Dollar volume is adjusted close × share volume. The Amihud (2002) illiquidity ratio (Amihud, "Illiquidity and Stock Returns: Cross-Section and Time-Series Effects," Journal of Financial Markets) is |daily return| ÷ dollar volume, scaled to price impact per $1M traded. A stock that moves a lot on little dollar volume scores high. The Corwin-Schultz (2012) spread estimator (Corwin & Schultz, "A Simple Way to Estimate Bid-Ask Spreads from Daily High and Low Prices," Journal of Finance) infers the bid-ask spread from two consecutive days' high/low ranges, with no intraday quotes needed. The formula can produce a negative spread in low-volatility stretches, which is a known property of the estimator. This page sets those to zero.
21-day rolling window. All three metrics are 21-trading-day (about one calendar month) rolling averages in the charts and the table. Daily dollar volume and the Amihud ratio are very noisy day to day (one block trade can move them by an order of magnitude), and a rolling average is the standard way to smooth that. Cohort series are the equal-weighted average of each ticker's rolling value across the ~100 names in the cohort, over the trailing ~2 years (~504 trading days). Each weekly run recomputes everything from scratch.
The regression. The chart tests whether liquidity gets worse in months when small-caps underperform. Each point is one calendar month. X is that month's small-cap vs. large-cap relative return, from the same month-end IWM/SPY ratio logic as Small Cap vs. Large Cap Spread. Y is that month's change in the Small/Mega Amihud ratio (this month-end minus last month-end). A negative correlation means liquidity deteriorates when small-caps underperform: months of small-cap weakness (negative X) pair with a widening liquidity gap (positive Y). A positive correlation means the opposite. As on Factor Analysis and Small Cap vs. Large Cap Spread, Pearson (OLS) and Spearman rank correlation are shown side by side, so a couple of extreme months can't carry the result. A low r² just means one variable explains little of the other.
Refreshes weekly (Saturday), after the cohort membership (Sector Beeswarm's meta.json) refreshes. It runs weekly, not daily, because the ~200-ticker history pull is heavy and cohort membership changes slowly.