The average semi, not the mega-cap tape
SMHEW is a synthetic, daily-rebalanced equal-weight index of SMH's 25 semiconductor holdings — each name counts the same 4%, so it reads what the typical chip stock is doing rather than being dominated by NVDA/TSM/AVGO. Compared against cap-weighted SMH, their ratio is the concentration premium: when it rises, leadership is narrowing into the mega-caps; when it falls, the move is broadening across the average semi. SMHEW is a first-class symbol on the platform — history to 2010, and usable in the Spread Lab and MA-Query just like an ETF.
Concentration premium
50th
percentile of SMH/EW vs history
EW · YTD
+64.2%
vs SMH +54.6% — EW ahead
Basket breadth
4/25
constituents > their 50-day
EW vs 200-day
+18.1%
equal-weight trend stretch
Constituents
25
live · history to 2010
Premium rich → ride mega-caps (long SMH / short EW)
| Fwd | Median | % up | n |
|---|
| 5d | +0.08% | 64% | 39 |
| 10d | +0.46% | 59% | 39 |
| 21d | +0.16% | 54% | 39 |
| 42d | +0.67% | 69% | 39 |
| 63d | +1.04% | 72% | 39 |
Premium cheap → ride avg semi (long EW / short SMH)
| Fwd | Median | % up | n |
|---|
| 5d | +0.07% | 55% | 94 |
| 10d | +0.11% | 55% | 94 |
| 21d | +0.28% | 57% | 94 |
| 42d | +0.27% | 59% | 94 |
| 63d | +0.30% | 55% | 94 |
Premium now at the 29th trailing-2y percentile. Events de-clustered ≥21 sessions. The premium trends, it doesn't revert: a positive median here means the stretch persists, so riding the current leader has paid and fading it has lost.
Constituent scorecard — ranked by 3-month return · RS = relative to the equal-weight basket
SMHEW = daily-rebalanced equal-weight of 25 SMH holdings (validated to the source workbook's EW index to 1e-13). Each day earns the mean return of the constituents trading that day, so history runs back as far as the older names allow (2010, 25 names today). Concentration premium = SMH ÷ SMHEW. RS 3M = a name's 3-month return minus the basket's. Adjusted closes, yfinance. Educational research, not advice.