Same universe. Six and a half times the rebalances. The gap is where mean reversion lives.
Run the same 30 stock Nifty 200 equal weight basket through both cadences for five years. The fee column looks the same on the brochure. The compounding column does not. Quarterly hands back 3.6 pp of CAGR and prints 2.6 pp more drawdown in the down year. Both numbers are the cost of latency, the line item that does not show up next to TER.
Cadence ledger . 5Y on the same 30 stock universe
Metric
Quarterly
Two-week
5Y end on Rs 10L
37.43L
39.85L
Max drawdown
-26.4pct
-22.7pct
2025 calendar return
-9.4pct
-6.8pct
Two ways the same universe ends differently
Median top-weight drift
6.2pp
1.4pp
Runner overweight in window
8.6pct
4.2pct
Laggard underweight
1.6pct
2.7pct
Implicit single name cap
drift not capped
capped at 2W
What the gap does to a single name . 3.33 pct equal weight start, no trim
Path inside the gap
After 10 trading days
After 63 trading days
Top runner up 60 pct vs basket flat
3.5 pct
8.6 pct
Steady name up 5 pct vs basket up 4
3.34 pct
3.36 pct
Laggard down 30 pct vs basket flat
3.0 pct
1.6 pct
Drift is a silent overweight, not a free upside. The quarterly book holds its runner all the way up to a 2.6x equal weight before the next trim, and holds its laggard all the way down to half. The 2W book never gives the runner that ceiling or the laggard that floor. The cadence is doing the trimming the brochure forgot to advertise.
Numbers illustrative. Backtest engine same equal-weight Nifty 200 universe, no slippage, 0.20 pct round-trip cost per leg. Quarterly = 4 rebalances per calendar year on first trading day of Jan / Apr / Jul / Oct. Two-week = fortnightly on Monday open. 5Y window through Apr 2026. Real-world tax friction (STCG vs LTCG) discussed in 13 May "2W rebalance after-tax math" post and not netted here. Past performance . backtest only . not a guarantee.