AI rally forces Jupiter to sell TSMC, Samsung, MediaTek
Jupiter Asset Management’s Asia Equity Income portfolio manager Sam Konrad said the fund was forced to sell TSMC, Samsung and MediaTek on June 8. The trigger was not investment conviction, but portfolio concentration rules: AI-driven gains have made these chip stocks too large versus regional benchmarks.
The scale is extreme. Year-to-date, TSMC is up 52%, Samsung up 159% and MediaTek up 184%. Together, TSMC, Samsung and SK Hynix make up nearly one-third of the MSCI Asia Pacific ex-Japan Index. Locally, TSMC alone represents 41.5% of Taiwan’s TAIEX. Samsung and SK Hynix together account for 55% of South Korea’s KOSPI.
Konrad said the fund has been a “mechanical” forced seller, with nearly half its assets allocated to Taiwan and South Korea, where MediaTek had been the largest holding.
The article links this to an “Asian Magnificent 7” dynamic seen in the US in 2023–2024: when a few mega-cap winners dominate index weights, active managers hit rule-based limits just as passive flows keep pushing prices higher. It also notes strong inflows into Asian markets—about $510bn over five years, with roughly $127.5bn in the last six months—amplifying benchmark concentration.
Implication for investors: if the semiconductor cycle cools, index-linked exposures could react more violently because the benchmark is dominated by a handful of names, including TSMC. Active managers may rotate toward smaller, AI-adjacent supply-chain stocks (packaging, test equipment, components) to avoid concentration risk.
Neutral
This news is about forced selling inside Asia’s semiconductor equity benchmarks, driven by AI-led concentration rather than crypto-specific fundamentals. For traders, it mainly matters as a risk-sentiment signal.
When TSMC and other mega-cap semis become too large for active funds’ concentration limits, selling can occur mechanically. In the short term, that can add volatility risk to equity-linked sentiment and to broader “risk-on” positioning, especially if investors interpret it as stress at the top of the benchmark. In the longer term, the setup can reinforce a rotation: flows may shift from the most dominant names (including TSMC) toward smaller AI-adjacent suppliers to reduce benchmark and regulatory risk.
Parallels: the article draws comparison to the US “Magnificent 7” period, where passive inflows created extreme concentration and later, corrections could hit indices disproportionately. Historically, that kind of concentration can amplify drawdowns when the cycle turns, but it usually doesn’t directly dictate crypto prices. The likely crypto impact is therefore indirect—via liquidity expectations, equity volatility, and the risk premium—leading to a neutral stance rather than a clear bullish or bearish catalyst.