My Forever Portfolio: 3 Undervalued Stocks to Buy Now and Hold Forever

The article argues for buying/adding Netflix, IBM, and Duolingo despite weakness: Netflix is down 32% YoY and trades at 21.6x forward earnings (FCF +61% to $11.1B; revenue +45% to $48.8B), IBM is yielding 2.9% (17.8x forward earnings) with dividend-focused total-return support, and Duolingo is down 55% YoY while reporting Q2 2026 results of 58.7M daily active users, $299M revenue, and $78.6M free cash flow. Overall tone is valuation- and cash-flow-driven optimism offset by stated stock underperformance, with impact likely limited to individual names rather than the broader market.

Analysis

The market is likely mispricing these names on different clocks. NFLX is now a cash-flow compounder rather than a hyper-growth story, so the next rerating will depend less on subscriber adds and more on whether adjacent bets like games/experiential can widen ARPU without lifting churn; if those initiatives stay capital-light, the stock can grind higher, but any step-up in content or real-world capex would cap the multiple. WBD looks like the cleaner loser from the current setup: without a premium bid embedded, it reverts to a slower monetization asset with a weaker balance sheet and less strategic optionality, which should pressure supplier economics across Hollywood talent and independent studios.

IBM is the quiet beneficiary of a rate-sensitive market. A high-dividend, low-beta cash generator tends to outperform when investors pay up for visible capital return and downplay narrative risk; the second-order effect is that AI exposure in IBM may be underappreciated because the market is still anchoring to legacy IT services. The real falsifier is not sentiment but execution: if AI and software mix fail to offset services drag in the next 2-3 quarters, the multiple will stay trapped despite the payout.

DUOL is the most interesting asymmetry: user engagement can keep expanding while the stock de-rates if monetization stalls, so this is more a 6-18 month operating leverage story than a quick rerate. The contrarian view is that the selloff may still be incomplete if growth is being subsidized by product novelty rather than durable pricing power; alternatively, if management keeps converting DAU growth into FCF, the market may be underestimating a software-like margin profile. Watch upcoming guidance, paid-user conversion, and FCF margin as the deciding data.

Key Decisions for Investors

• Long IBM / short WBD as a 3-6 month quality-vs-overhang pair: IBM offers visible cash return and lower balance-sheet risk, while WBD still lacks a credible catalyst without deal premium; exit if WBD announces a strategic transaction or IBM misses FCF guidance.
• Accumulate NFLX only on weakness, not on valuation alone: treat 21x forward earnings as fair, not cheap; upside depends on incremental monetization from games/experiences, while a content-spend reset or margin compression would negate the thesis.
• Watch DUOL for a 1-2 quarter entry window after the next earnings print: if FCF margin expands and paid conversion holds, the stock can rerate meaningfully; if DAU growth decelerates or monetization softens, avoid catching the knife.
• For income-oriented exposure, IBM is the cleaner long than most software names: it can work as a defensive dividend compounder if rates stay sticky; the risk/reward turns less attractive only if the company raises capex faster than operating income.