Why this money manager is buying Microsoft, Meta and selling Disney

Money manager Allan Small says he is buying Microsoft, Meta Platforms, and Blackstone after each fell about 25% to more than 30% from recent highs, arguing the sell-offs are overdone and fundamentals remain strong. He is trimming Walt Disney and data-storage chip names such as Micron and Western Digital to lock in gains and generate cash. The piece is mainly portfolio-positioning commentary, with a focus on buying weakness in AI-related and high-quality growth stocks and selling into strength.

Analysis

The important signal here is not the specific names being added or trimmed, but the rotation logic: the manager is harvesting winners in memory/storage while reallocating into large-cap AI beneficiaries that have de-rated on spending fears rather than on deteriorating demand. That usually matters over a 3-6 month horizon, because the market tends to punish capex-heavy AI names first and reward them later once monetization evidence catches up. The second-order effect is that capital is moving from a semis subsector with more cyclical pricing risk into platform/software and alternatives, which can be a better way to express AI exposure when the market is skeptical of direct infrastructure plays.

Meta and Microsoft look attractive here because the selloff creates a better entry point into businesses with self-funding AI optionality, not speculative AI dependency. The key nuance is that both can absorb elevated AI spend without impairing returns if ad pricing, cloud mix, or enterprise attach rates remain intact; if those metrics stall, the market will quickly re-rate them as capex traps rather than AI compounds. For Microsoft specifically, the near-term catalyst is proof that AI features are not cannibalizing software margins, while for Meta the catalyst is ad ROI and incremental monetization from AI-assisted tools.

Blackstone is a different expression of the same theme: if public tech re-rates lower on AI uncertainty, private-market capital allocators with real asset exposure and fee streams become relatively more defensive. Data-center assets are the hidden beneficiary of AI infrastructure demand, but the stock can still re-rate violently if rate expectations rise or if fundraising slows. The contrarian miss is that trimming Micron and Western Digital may be early if memory tightness persists longer than expected; storage tends to lag in price realization, so gains can continue even after the first leg of the cycle.

Overall, the portfolio stance implied is to buy quality AI-linked franchises after pullbacks and sell cyclical semiconductor strength into euphoric pricing. The risk is that the AI spend cycle broadens faster than expected, causing the under-owned infrastructure names to keep outperforming for longer than valuation models suggest. That argues for keeping exposure selective and using strength in memory names to finance longs in higher-quality compounders rather than chasing the entire AI basket.

Key Decisions for Investors

• Add MSFT on a 3-6 month horizon into weakness; target a re-rating if AI margin concerns fade and enterprise adoption evidence improves. Risk/reward is attractive because the stock is being priced as if AI is a near-term margin headwind, while the more likely outcome is delayed but durable monetization.
• Add META on dips over the next 1-2 quarters; use it as a funded AI-adjacent compounder rather than a pure capex story. If ad pricing and user engagement stay resilient, multiple expansion can resume quickly; if spend growth outruns revenue growth, cut exposure aggressively.
• Initiate a pair trade: long MSFT / short MU for 1-3 months. Thesis: the market is overpaying for cyclical memory upside while underappreciating durable AI monetization franchises; the spread should widen if AI capex optimism cools.
• Use BX as a relative-value long versus regional banks over the next 6-12 months, especially if rate volatility remains elevated. BX offers fee-based AI infrastructure optionality plus dividend support, while banks face slower loan growth and more balance-sheet pressure.
• Trim or sell into strength in WDC and MU rather than chasing cycle continuation. Re-enter only on a 10-15% pullback or if pricing data confirms a fresh leg of supply tightness; otherwise the risk/reward skews toward giving back gains.