The signal discussion said to buy only half of the Korea ETF (EWY), and it fell more than 7% in a week. Did buying only half matter, and should I sell if I'm holding it now?

Half-size entry only reduced the size of the loss and did not change the odds

If you entered at the 10-05 close of $191.46, you were down -7.92% by the 10-08 close. Someone who planned 10 million won and bought full size lost about 790,000 won, and at half size about 400,000 won. With SKIP, the loss is zero. Half size cut the loss amount in half, but the probability of hitting the stop was unchanged. This week EWY fell -2.64%, -1.45% and -4.03%, about 8% in three days. The trail buffer of 6.87% at the time of the discussion was too small to survive three days at this week's volatility.

Structural cause: the stop sat at the same level as the 50-day line

The theme group's trail is 20% from the high. The stop is $176.71, 80% of EWY's high of $220.89. MA50 at the time of the discussion was $176.65. The stop and the 50-day moving average were effectively the same price. In an uptrend, a single pullback to the 50-day line is a common correction, but here a common correction becomes a liquidation. The Bear's point that the close's 8.6% gap from MA50 was larger than the 6.87% trail buffer was exactly this problem, and in fact the 10-08 close of $176.29 fell below both MA50 ($179.01) and the stop.

The report took the lesson that with a trail buffer under 7%, standing aside beats cutting size. This can be turned into a more concrete rule. If the stop price is at or above MA50, treat REDUCE as SKIP. In that case, reducing size does not lower the probability of being stopped out in a normal correction.

Only Korea fell on its own

The same week, Taiwan's weighted index rose +1.73% through 10-08, while the KOSPI fell -5.39% and EWY -7.63%. The won strengthened 1.29% against the dollar, which actually favored EWY in dollar terms. Neither AI chip demand worries nor the exchange rate can explain this gap. The cause is Korea-specific flows.

• Reports said that on the day of Samsung Electronics' preliminary results, semiconductor ETF rebalancing, options expiry and the end of share buybacks coincided (Business Today, Korea Times, 2026-10-08). Foreign investors were net sellers of about 6.47 trillion won over four sessions.
• Samsung Electronics' operating profit of 107.4 trillion won beat consensus, but revenue was reported at about 195 trillion won (foreign media roundup, 2026-10-08). That is about 3% below the FnGuide consensus of 201.7 trillion won (Money Today, 2026-10-05). An earnings beat paired with a revenue miss is the same shape as the July second-quarter release, when the stock fell 6.9% on the revenue shortfall.

Flow events usually have an end. But that is a reason to re-enter later, not a reason to ignore the stop and hold on.

What to do now

• If you hold it, follow the system's Exit call. The 10-09 close of $177.24 returned to 0.24% above the stop, but MACD (0.40) is well below the signal line (1.69) and the close is below both MA10 and MA50. A 0.24% buffer is not a reason to hold. The reaction to Korea's 10-12 reopening will be reflected in EWY on the next trading day.
• There is no reason to rush back in. While the high of $220.89 is the reference, EWY must be at $198.80 or higher for the trail buffer to exceed 10%. That is about 12% above the current price. Re-entering in the $180-190 range would repeat this week's thin buffer.
• The re-entry checklist has three items. EWY closes back above MA50 ($179.33) and MACD crosses above the signal line again, foreign investors turn net buyers for two days or more, and the stop price sits sufficiently below MA50.
• For someone who is long-term positive on Korean semiconductors, buying in tranches in won outside the signal system is a separate decision. Even then, set the stop criterion in advance so that it does not overlap with the 50-day line.

You can see detailed charts and signals for other tickers on the /signals dashboard.