One of My Top AI Dividends Just Popped 43% (and It's Still a Bargain)

AI stocks sold off last week, with NVIDIA down 8.5% and Micron falling 16.6%, but the article frames the decline as a profit-taking opportunity rather than a fundamentals-driven break. Columbia Seligman Premium Technology Growth Fund (STK) now trades at a 4.6% discount to NAV versus a 2.3% average premium over the past decade, while offering a 3.6% yield. The piece also highlights four other AI-focused closed-end funds yielding about 10% on average, including one pharma fund yielding 13% and trading at a 9.7% discount.

Analysis

The tape is being driven less by a durable change in discount-rate assumptions and more by a crowded de-risking event: a strong macro print gave investors permission to trim extended AI exposure all at once. That matters because the undercurrents are still constructive for the franchise names inside these funds—enterprise AI spend is increasingly funded from operating budgets, not speculative capex, which makes demand stickier than the market is pricing after a one-week drawdown. The most important second-order effect is that the selloff likely improves forward returns for wrapped products by widening discounts faster than the underlying holdings can fundamentally deteriorate.

The opportunity is asymmetric in closed-end funds with leveraged exposure to AI leaders because you can get both alpha from the basket and mean reversion in the vehicle’s discount. If sentiment stabilizes over the next 2-6 weeks, these funds can re-rate faster than single stocks since income buyers are mechanically attracted to wider discounts and elevated distributions. The risk is that the market shifts from “rates may stay higher” to “growth is re-accelerating,” which would initially keep pressure on long-duration tech multiples even if fundamentals remain fine.

The contrarian miss is that investors are treating job creation from AI as a bearish macro narrative, when it is actually a diffusion signal: AI adoption is moving from lab/productivity story into organizational budget line items and labor reallocation. That should support semis, hyperscalers, and select infrastructure winners more than it hurts them, but the path will be choppy because positioning is still crowded. Near term, the setup favors buying weakness in high-quality AI income wrappers; over months, the bigger tailwind is a broader earnings revision cycle as AI-linked productivity starts showing up outside pure tech.

Key Decisions for Investors

• Initiate a starter long in STK on any further 1-2 point widening in the discount to NAV; target a 3-6 month mean reversion to a low-single-digit premium with downside limited by the underlying mega-cap quality of the basket.
• For higher yield exposure, accumulate the AI-linked CEFs trading at the widest discounts now; use a 3-8 week window and size for volatility, since the main catalyst is sentiment normalization rather than near-term fundamentals.
• Pair trade: long STK / short a broad tech ETF into strength if rates rhetoric intensifies; the spread should outperform if income buyers continue to support the CEF discount while the short leg absorbs duration pressure.
• Add upside convexity via call spreads on NVDA or MSFT with 2-4 month tenor; the risk/reward improves if the current pullback is just a positioning reset and not a true demand shock.
• Avoid chasing MU on the dip until there is evidence that AI memory demand is translating into pricing power; its beta to sentiment is high, but the fundamental follow-through is less certain than for the platform names.