The trading signals said to sell the Korea ETF and semiconductors this week while holding S&P and Nasdaq. They said to buy only half a position in Broadcom. Were these calls right? What should be done next week?

The System Did Not Panic-Sell Everything — It Cut Only What Was Broken

This week's signal action was about "selective severance." It cut themes, EM, and AI-specialized names where the trend broke, while holding core positions with ample cushion.

(Signal history, end-of-week 6/5, entry_ok · trail_buf)

The logic is clear. EWY's trail buffer was nearly exhausted at 0.45% — essentially at the stop line — while SMH, MSFT, and ROKT all broke their entry conditions (trend). The core (SPY, QQQ, XLK), by contrast, held trail buffers above 20%, with the trend never triggering a risk-off signal. This is textbook trend-following: cut when the trend breaks, hold when the cushion holds (AlphaMaven; QuantifiedStrategies, 2026).

The Following Week Graded the Calls Directly

As of today (June 12), the subsequent week's price action provides a direct scorecard.

EWY exit — correct. On the 6/10 CPI 4.2% shock, SK Hynix fell another ~-8% and Samsung Electronics -7.45% (CNBC, 2026-06-10). Cutting at a 0.45% trail buffer removed the highest-beta Korea exposure ahead of the BOJ/FOMC event week. That said, the Korea structural story remains intact — this is a re-entry candidate, not a permanent exit. The National Pension Service raised its domestic equity target from 14.9% to 20.8% (eliminating forced selling pressure, $1.07T AUM), and the probability of MSCI Developed Market watchlist inclusion is assessed at 60%+ for June (Bloomberg; Hana Securities, 2026).

SMH exit — right on the risk, early on the rebound. Cutting at 569.69 avoided the 6/10 additional decline, but the 6/11 semiconductor V-rebound (Micron +11%) was missed. This is the intrinsic trade-off of a trailing stop — it blocks the downside leg but cannot capture the V-snap. The system is not a tool for calling the bottom; it is a tool for keeping high-beta exposure flat until the trend reestablishes.

AVGO half-size buy (REDUCE ×0.5) — correct, and then one step further. The weekend discussion concluded "the growth structure is intact, but the short-term macro and technical damage was not sufficiently rebutted" — enter at 50% of normal size. AVGO's 6/10 retest at -5.12% (trail buffer 5.17%) validated that caution. And when the trail buffer recovered, the June 12 verdict was upgraded to BUY at full size.

"MA alignment + AI infrastructure seller structure intact + trail buffer recovered. Enter full size." — Signal discussion verdict (2026-06-12)

The half-size positioned the system to participate in the AI thesis while surviving the macro retest — and the system has now reconfirmed full conviction.

So What Should Be Done Next Week?

• Core (SPY · QQQ · XLK): Hold. Trail buffers above 20% never once signaled an exit. The trend is intact.
• EWY · SMH (exited): Do not buy back on a single day's rebound. Re-entry trigger is entry conditions (entry_ok) re-activating + catalyst confirmation — for EWY, that's the MSCI review (6/18, 6/23) and July NPS buying; for SMH, it's Micron's 6/24 HBM guidance.
• AVGO · MU (trimmed): The 6/12 AVGO full-size upgrade and MU's rebound confirm that the trim endured a round trip. Restore to normal weight gradually, as the trail buffer rebuilds above 15%.

Key takeaway: trend-following's selective exits look "wrong" on rebound days (missed Micron bounce) and "right" on shock days (avoided CPI 4.2%). The value of navigating the full event cluster (BOJ · FOMC · Micron) is not pinpointing the bottom — it is staying out of the most dangerous names (EWY · SMH) until the trend revives.

For detailed charts and signals on additional tickers, visit the /signals dashboard.