Amazon (NASDAQ:AMZN | AMZN Price Prediction) is a stock worth owning for decades because its three highest-margin engines (AWS, advertising, and subscriptions) are now compounding faster than its low-margin retail business, and that mix shift is structural.
For a retirement-focused investor who has watched fads come and go, the appeal here is the unusual combination of a consumer utility, a cloud utility, and an advertising platform under one corporate roof, all funded by an operating cash flow machine that produced $139.5 billion in 2025, up from $38.5 billion in 2019.
Pillar One: Durability That Compounds in the Background
While the mainstream story still fixates on shipping costs and Prime Day promotions, the forever case lives in the high-margin engines. AWS generated $37.587 billion in Q1 2026 revenue at a 37.7% operating margin, growing 28% year over year, its fastest pace in 15 quarters on a roughly $150 billion annualized base. Advertising, almost invisible a decade ago, now runs at over $70 billion in trailing revenue. The custom silicon business, including Trainium and Graviton, sits at a $20 billion+ annual run rate, with CEO Andy Jassy disclosing “over $225 billion in revenue commitments for Trainium”. These are utility-like revenue streams locked into multi-year customer contracts.
Pillar Two: Compounding Without a Dividend
Amazon does not pay a dividend, and that is the point. Capital that would otherwise be distributed is being redeployed into infrastructure with measurable returns: return on equity of ~22.3%, return on assets of ~10.8%, and a gross margin of ~50.3%. Over the past decade, AMZN shares are up 620.18%. Retirees who want income can manufacture it by trimming a sliver of a position that has historically generated its own returns. Income strategies layered on top, like covered-call ETFs, have significantly lagged Amazon’s direct stock performance, which is itself an argument for owning the equity directly.
Pillar Three: Surviving the Cycles
Amazon has already survived a dot-com collapse, the 2008 financial crisis, the 2020 demand shock, and the 2022 cost reset that briefly produced a $2.7 billion net loss. Each time, operating cash flow returned and then expanded. The balance sheet today carries $101.816 billion in cash against $119.1 billion in long-term debt, with interest coverage of ~35.2x. That is a fortress balance sheet.
The Scenario Where It Underperforms
Heavy capex periods like this one compress free cash flow. Trailing free cash flow has fallen to $1.2 billion, a 95% decline, as 2026 capital spending heads toward roughly $200 billion to build AI infrastructure. In a recession, that combination can produce flat or negative price action for a year or longer. The forever thesis does not require avoiding those years. As Jassy put it, AWS assets carry “many-year useful lives, 30-plus years for data centers”, with revenue arriving six to 24 months after the cash goes out. A retiree holding for two decades is on the right side of that math.
Long-term holders who can tolerate capex-heavy years have historically been rewarded for sitting through them.