Korean ETF buyers put 5.35 trillion won into cash and US index funds that averaged 0.83% in a month

The comparison is a little unfair to the people who moved the money. In Koscom's ETF CHECK figures, the 0.83% is the average for the 10 funds that drew the most cash ^[8]. The 26% to 32% belongs to the 10 best performers among all ETFs, with leveraged and inverse funds excluded ^[10]. One figure samples where money went. The other is the top of the whole table. A fairer benchmark is the KOSDAQ index itself, up 11.11% on the month ^[12]. The four largest recipients took 3.37 trillion won, about 63% of the top-10 total ^[1][2], and earned roughly 0.56% weighted by inflow ^[3]. The weighted figure sits below the simple average because the biggest single inflow went to a money market fund earning 0.28% ^[2].

Annualized, the cash side looks less like an error. KODEX Money Market Active and KODEX CD Rate Active both returned 0.28% ^[2][4], about 3.4% a year ^[4], and between them drew 2.25 trillion won ^[5]. A holder who wants won available next week was paid a cash rate for keeping it there. That holder gave up the index gain and lost nothing on the position.

The US index money is harder to explain. TIGER U.S. S&P 500 lost 0.59% ^[3] in a month when KODEX U.S. Nasdaq 100 gained 3.85% ^[5], a 4.44-point gap between two US index funds ^[8]. Sedaily's report does not explain it. Together those two funds absorbed 1.12 trillion won and earned about 1.13% weighted ^[6], and Sedaily attributes the US buying to habit ^[16]. The allocation also shows what this money skipped. Samsung Electronics stalled on worries that earnings had peaked and on persistent foreign selling ^[14], and every won in the top 10 inflows went to US markets or safe assets ^[7].

The supplier rally happened without that money. DS KOSDAQ Active's 32.36% ^[9] is about 2.9 times the index's move ^[7], and Sedaily says the main KOSDAQ ETFs more than doubled the index's gain ^[13]. Trading was close to absent, what the market calls an "empty house", and targeted buying went into names with proven technology ^[15].

Third-quarter earnings season ^[18] can settle the gap in either direction. Results that confirm the factory utilization recovery would pull some of the cash sitting in two large money market active funds ^[17] toward the suppliers, and the flows would turn out to have trailed the returns by a month. Weak results after a rally on thin trading would hand the better month to the cash holders and their 3.4%. Or neither happens, and the inflow table and the return table keep diverging.

I think the gap says more about how thin trading in the Korean market is than about investor error. A group of supplier funds can double the index on targeted buying while most new money earns 0.28% a month. The case against that view is the parked cash itself. If the money market balance falls through earnings season while the KOSDAQ holds above 893, the divergence was a one-month lag.