5 Value Stocks to Buy as AI Volatility Reshapes the Market

A screen requiring earnings yield above 10%, projected EPS growth at least in line with the S&P 500, and adequate liquidity identified Cenovus Energy, Micron, Sandisk, Hewlett Packard Enterprise and Honda among 34 qualifying stocks. The article cites projected fiscal-year EPS growth of 168% for Cenovus in 2026, 135% for Micron in fiscal 2027, 204% for Sandisk in fiscal 2027, 97% for HPE in fiscal 2026 and 209% for Honda in fiscal 2027. It presents the stocks as value ideas amid AI-related market volatility, persistent inflation and elevated long-term yields; this is screening commentary, not a report of company results or share-price moves.

Analysis

Treat this screen as a cyclicality and estimate-revision signal, not evidence that five stocks are independently cheap. A high earnings yield can be a peak-earnings artifact: if commodity or memory pricing rolls over, the denominator in the P/E calculation can fall faster than the share price. Verify whether the screen uses trailing or forward EPS and stress-test earnings at mid-cycle prices before sizing.

The apparent AI diversification is also overstated. MU, SNDK and HPE all have exposure to hyperscaler investment, so a pause in AI infrastructure budgets could hit several positions at once. Memory suppliers may gain pricing power during tight supply, but that reverses quickly if capacity arrives or customers digest inventory; HPE could face demand without equivalent profit conversion if server competition or component costs pressure economics. Conversely, a capex pause may benefit buyers of memory and equipment over time, even as near-term supplier estimates compress.

CVE offers a different driver and may diversify a tech-heavy book, but is not a clean inflation hedge: oil, refining conditions, production execution and policy can offset one another. HMC adds FX, auto-cycle and trade-policy risks. Near term, estimate momentum may support the AI-linked names; over 1–3 months, earnings guidance and hyperscaler capex commentary matter more than screen rankings. Over 6–18 months, memory supply response and realized returns on AI spending are the key regime tests. Contrarian read: the consensus may be underweight correlation across the three AI-linked picks while over-crediting headline EPS growth that could reflect unusually favorable pricing or comparison periods.

Key Decisions for Investors

• Do not buy the five-name screen as a basket. Before adding MU, SNDK or HPE, check forward versus trailing EPS methodology, earnings sensitivity to memory prices, inventory trends, and cash conversion; treat estimate upgrades as provisional until guidance confirms them.
• Potential diversifier: build a modest, staged CVE position against an existing AI-heavy equity book, rather than treating it as a standalone value call. Reassess on oil-price reversal, production/capex guidance, or evidence that refining and upstream performance are diverging.
• Keep MU and SNDK exposure sized as one correlated memory-cycle risk, not two independent AI beneficiaries. Trim or hedge if supplier commentary points to customer inventory digestion, weaker pricing, or supply additions outpacing demand.
• For HPE and HMC, wait for evidence that projected EPS gains convert into operating cash flow and resilient guidance. Falsifiers include weaker server profitability or order trends at HPE, and adverse FX, trade-policy changes, or deteriorating auto/motorcycle demand at HMC.