Bet on These 5 Dividend Growth Stocks Amid Rising Treasury Yields

With the 30-year U.S. Treasury yield at its highest level since 2004, the article advocates dividend-growth equities as a lower-volatility alternative to high-multiple growth stocks. It highlights EnerSys, Amgen, Cheesecake Factory, Vertiv and Taiwan Semiconductor, with 2026/27 consensus revenue growth ranging from 3.7% to 36.9%; Vertiv and TSM lead at 36.9% and 36.5%, respectively. The five stocks were selected using dividend, sales and EPS growth, relative valuation and momentum screens, with Zacks ranks of Buy or Strong Buy where disclosed.

Analysis

The proposed basket is not a coherent duration hedge: VRT and TSM remain long-duration AI capex exposures whose equity multiples are materially more sensitive to real-yield changes than their token dividends imply. A continued rise in long-end yields would likely favor AMGN and ENS on relative terms, while pressuring VRT/TSM despite intact operating demand. The more relevant transmission channel is capex affordability: higher financing costs can slow marginal data-center builds, affecting VRT orders before it affects TSM's leading-edge wafer demand.

ENS is the underappreciated second-order beneficiary of data-center power redundancy and grid instability, with less crowded AI positioning and potential upside if hyperscalers expand battery-backup procurement. However, its industrial and telecom exposure makes it more cyclical than the screening framework suggests. CAKE is the weakest fit for a higher-yield regime: consumer discretionary traffic and restaurant-level margins are vulnerable to renewed energy inflation, wage pressure, and credit-sensitive spending; its dividend provides limited valuation support.

Over the next 1-3 months, this article itself is unlikely to move prices; it is a generic screen with no independently verified earnings revision or capital-allocation catalyst. Over 6-18 months, the key debate is whether AI infrastructure spend converts from capacity reservation into delivered revenue and free cash flow. Consensus may be underestimating the dispersion: quality dividend growth is preferable to high yield, but only where payout growth is backed by recurring cash flow rather than elevated capex and valuation assumptions.

Falsification points: reduce the VRT/TSM caution if 10-year real yields fall decisively while hyperscaler capex guidance is raised, or if VRT backlog conversion and gross-margin expansion exceed consensus for two quarters. Increase the CAKE short bias if same-store sales weaken or restaurant margins fail to offset food and labor inflation; abandon it if traffic accelerates without promotional intensity.

Key Decisions for Investors

• Pair trade, 3-6 months: long AMGN / short CAKE in equal dollar amounts. AMGN offers more defensive cash-flow and idiosyncratic pipeline optionality; CAKE has greater consumer and input-cost sensitivity. Target 8-12% relative return; stop if CAKE same-store sales outperform AMGN revenue growth by more than 300 bps for two consecutive reported quarters.
• Accumulate ENS on broad industrial or AI-infrastructure pullbacks rather than chase VRT. Use a 6-12 month horizon and require evidence of data-center power/back-up order growth in the next earnings release; without that disclosure, treat as a watch position rather than a full-size long.
• Maintain TSM as the preferred semiconductor AI exposure over VRT for 6-18 months, but hedge duration risk with a partial short in VRT or SMH during periods of rising real yields. TSM's strategic foundry position supports earnings durability, while VRT has greater multiple and execution risk from elevated infrastructure expectations.
• Do not treat VRT as a dividend-defensive holding. For existing longs, trim into strength ahead of hyperscaler capex updates and re-add only after confirmed backlog-to-revenue conversion; use a 10-15% position-level drawdown limit or a break in forward revenue guidance as the risk trigger.