The "Great Rotation" Out of Artificial Intelligence (AI) Stocks Has Arrived. Here's What Smart Money Is Buying Instead.

The article argues that AI stocks may be in a bubble and highlights three defensive dividend options: Procter & Gamble, Realty Income, and Brookfield Renewable. It cites P&G’s 69-year dividend-increase streak and 2.9% yield, Realty Income’s 31-year streak and 5.1% yield, and Brookfield Renewable’s 3.6% corporate yield or 4.6% partnership yield. The piece is primarily a portfolio rotation argument toward stable income and away from volatile AI exposure.

Analysis

This is less a “dividend stocks are good” note than a rotation signal out of duration-sensitive growth and into cash-flow visibility. If the AI trade continues to de-rate, the first beneficiaries are not just the named income stocks, but the entire cohort of staples, regulated infrastructure, and contract-backed owners that can absorb slower growth without multiple compression. The second-order effect is that capital may migrate from speculative AI beneficiaries into yield-heavy proxies, temporarily supporting REITs and utilities even if rates remain sticky.

PG is the cleanest defensive beneficiary because its earnings quality is largely insulated from macro volatility, but the stock’s bigger implication is relative: when investors pay up for “certainty,” premium staples can re-rate even if operating growth is mediocre. O is more interesting tactically because it sits at the intersection of rates and risk appetite; a recessionary AI unwind could compress the long end, helping the spread economics, but a disorderly credit event would widen financing spreads and partially offset that benefit. BEP/BEPC are the most asymmetric: they are not pure defensive names, they are AI-adjacent power supply claims, so any data-center capex durability keeps demand intact even in a broader tech drawdown.

The consensus may be overestimating how much a pullback in AI leaders would hurt infrastructure demand. Data-center load is already being contracted years forward, so the more relevant risk is not AI equity sentiment but project financing and interconnection delays; those are slower-moving and create a 12-24 month lag before fundamentals inflect. Conversely, the market may be underpricing rate risk for O: if the disinflation story weakens, income stocks can outperform on earnings resilience but still suffer multiple compression.

Best setup is to separate business quality from sentiment beta. PG is the lowest-risk parking place, O is a rate-sensitive income trade with levered upside if growth fears pull yields lower, and BEP/BEPC offers a cleaner way to express “AI picks and shovels” without owning the expensive software winners.

Key Decisions for Investors

• Long PG vs short a basket of high-multiple AI names over 3-6 months: favors cash-flow durability if AI sentiment keeps de-rating; lower beta, lower drawdown, but expect muted absolute upside.
• Add O on any post-rate-spike pullback for a 6-12 month income trade; if the 10-year yield trends lower on growth fears, O can benefit from both cap-rate support and yield-seeking inflows. Risk: financing spreads widen if credit stress replaces rate relief.
• Use BEPC/BEP as a selective AI-infrastructure hedge for 12-24 months; upside comes from data-center power demand persisting even if AI equities compress. Risk/reward is best if entered on broad renewable selloffs rather than strength.
• Pair trade: long PG / short consumer discretionary ETF for a defensive slowdown hedge; if consumers weaken, staples should hold margin better than cyclicals, with relatively low factor overlap.
• If rate volatility spikes, prefer O over BEP/BEPC for tactical income exposure; if rates stabilize and power demand headlines stay hot, rotate from O into BEP/BEPC to capture higher thematic upside.