The Overlooked EM ETF (XCEM) Is Up 38% YTD — What the S&P 500 Took Five Years to Match in Just Five Months

Columbia EM Core ex-China ETF (XCEM) is up about 38% year to date through June 3, 2026, versus SPY up roughly 11%, with a $50,000 allocation to XCEM now worth about $19,000 more than the same amount in the S&P 500. The outperformance is tied to exposure to Taiwan, South Korea, and India, plus a weaker U.S. dollar and AI-driven semiconductor demand, while avoiding China-related policy risk. The article also notes low retail chatter but improving institutional ownership and a 16% expense ratio advantage versus some ex-China peers.

Analysis

XCEM’s outperformance is less a broad EM thesis than a factor stack: semis + India + weaker USD + no China. That combination is unusually powerful because it compresses three independent beta sources into one vehicle, but it also means the trade is more crowded at the underlying factor level than the fund’s retail obscurity suggests. The second-order winner is the upstream Asia supply chain—TSM, memory, and select Indian financials—while the loser is any EM basket that still dilutes exposure into China, where capital has to fight policy overhang and less favorable sentiment.

The key risk is not mean reversion in “EM” but reversal in the two variables doing most of the work: DXY and hyperscaler capex. A 3-5% dollar bounce can erase a meaningful chunk of local-equity gains for U.S. holders over a short window, and semis are the fastest channel for disappointment if AI spending shifts from build-out to digestion. That makes this a months-not-years setup: the next TSM monthly print and guidance commentary from the Taiwan/Korea supply chain matter more than macro headlines.

Consensus is probably underestimating how much of XCEM’s move was simply a clean expression of avoiding China at the exact moment China stopped being “cheap enough” to ignore. But the market may also be overpaying for the idea that ex-China EM is a durable structural winner; if China stimulus finally bites, relative performance can mean-revert fast because investors will rotate back toward the benchmark weight, not just into the best businesses. In that sense, XCEM is a regime trade with a regime-dependent exit.

Key Decisions for Investors

• Go long XCEM vs. EMXC for 1-3 months as a low-friction relative-value expression of the same theme; XCEM’s cheaper fee profile plus cleaner country mix gives it the edge if the dollar stays soft, but cap exposure at 1-1.5% NAV because the trade is factor-crowded.
• Pair long TSM / short BABA or KWEB on a 2-6 week horizon to isolate the winner from China-policy noise; this captures the market’s current preference for real AI manufacturing cash flows over stimulus-dependent beta.
• Buy a modest upside call spread in TSM or TSM-linked exposure into the next monthly revenue/guidance window; risk/reward is attractive if AI capex remains intact, but define risk tightly because semis will gap on any sign of demand digestion.
• Use DXY strength as the stop-loss trigger for XCEM longs: if DXY reclaims trend resistance or US rates reprice higher, reduce exposure quickly, since the currency leg can unwind faster than the equity leg can adapt.
• For longer-horizon accounts, prefer XCEM over broad EM for a 6-12 month allocation, but fund it by trimming richer India-only exposure; XCEM’s diversification still keeps the trade anchored to the same growth complex without paying up for single-country valuation risk.