Wealth Manager Sheds $5.3 Million Worth of Data Center ETF, According to Latest SEC Filing

Centurion Wealth Management sold 216,719 shares of DTCR in Q1 2026, an estimated $5.30 million transaction that reduced its stake to 80,137 shares valued at $1.92 million. The position fell from 1.4% to 0.4% of AUM and dropped out of the fund’s top five holdings, reflecting portfolio repositioning rather than a material change in the ETF’s fundamentals. DTCR remains supported by AI-related digital infrastructure demand and lower rates, with the ETF up 72.9% over the past year and yielding 0.88%.

Analysis

The sale reads less like a view on data-center fundamentals and more like a position-sizing decision after a sharp rerate. That matters because passive infrastructure exposure has become a crowded “duration-plus-AI” trade; when allocators trim, it often reflects a desire to reduce factor overlap rather than a call on the secular thesis. In that sense, the better read-through is not bearish on the asset class, but a warning that incremental capital is becoming more selective and less willing to pay any price for perceived AI adjacency.

Second-order winners are the operating assets with pricing power and the cleanest power/land pipelines. DLR and EQIX should continue to absorb a disproportionate share of institutional demand because they are the most credible public proxies for scarce, mission-critical capacity, while APLD is the more reflexive beneficiary if investors keep chasing higher-beta AI infra exposure. The weaker link is anything dependent on a perpetual cap-rate compression narrative: if rates stop falling, the cash-flow duration embedded in REIT multiples gets exposed quickly, and the crowding unwind can be sharper than the fundamental deterioration.

The contrarian point is that the move may be underdone on the downside for the ETF wrapper but overdone as a signal for the underlying theme. ETF selling can mechanically lower flows without implying any change in the supply-demand imbalance for data-center power and space, which still looks tight over a 12-24 month horizon. The immediate risk is not demand collapse; it is multiple compression if real yields stabilize or if AI spending rotates from broad infrastructure names toward more differentiated compute and networking winners.

For risk/reward, the best setup is to own the scarce-asset operators and fade the basket. A simple expression is long EQIX/DLR versus short DTCR or a broader REIT-duration proxy, because the ETF dilutes the strongest pricing assets with more cyclical exposure. If rates back up 50-75 bps, expect the ETF to underperform the operators meaningfully over the next 1-3 months; if rates continue lower, cover quickly because the entire complex can re-rate higher in a reflexive squeeze.

Key Decisions for Investors

• Go long EQIX and DLR versus short DTCR in a 1-3 month pair trade; thesis is that high-quality operators outperform the ETF wrapper as allocators rotate into scarce capacity and away from broad basket exposure.
• Add APLD only on pullbacks over the next 2-6 weeks; higher beta gives better upside if AI infrastructure flows re-accelerate, but size smaller because it is more sensitive to financing conditions.
• Use a rate-sensitive hedge: short a broad REIT ETF or buy puts against DTCR if 10-year real yields rise; target a 2:1 payoff if the market reprices duration exposure.
• If already long the group, take partial profits on any further 5-8% rally in DTCR over the next month; the flow signal suggests marginal buyers are becoming less price-insensitive.
• For more convex exposure, buy 3-6 month calls on EQIX rather than the ETF; better risk/reward if the next leg higher comes from operator scarcity rather than broad sector beta.