Generating cash is one of the clearest signs that a business model is working. For investors, however, the next question is equally important: how is that cash being used?
A recent debate around two of the world’s largest semiconductor manufacturers provides a useful example. On 6 August, Reuters reported growing investor pressure on Samsung Electronics and SK Hynix to increase shareholder distributions as booming demand for AI memory drives strong profits and rapidly expanding cash reserves.
According to LSEG data and Reuters calculations, the two companies are projected to hold a combined $263 billion in net cash by the end of 2026. That would be more than twice Nvidia’s estimated $102 billion and greater than the combined cash position of the other six companies commonly grouped within the ‘Magnificent Seven’.
The numbers are striking, although the underlying issue is relevant far beyond technology. Once a business has funded its operations and generated surplus cash, management has to decide where that capital can create the most value. Reinvestment may be the right answer when attractive projects are available. Debt reduction may take priority when leverage is high. Acquisitions can make sense when they strengthen the core business at a reasonable price. Dividends and other distributions become increasingly important when the business produces more cash than it can productively deploy.
For investors focused on cash generating projects, understanding this decision process is central to assessing long term returns.
Revenue and accounting profit can provide useful information about business performance, although neither tells investors exactly how much capital is available for distribution. Free cash flow provides a closer view because it considers the cash generated by operations after investment in assets. Even then, investors need to understand what sits behind the figure.
A capital intensive business may produce substantial operating cash flow while requiring equally substantial investment to maintain equipment and capacity. Another company may require comparatively little reinvestment and therefore convert a larger proportion of operating profit into cash available to owners. This is why the ability to pay dividends should be considered alongside the capital requirements of the underlying business.
The semiconductor sector illustrates the trade off particularly clearly. Samsung has historically maintained large cash reserves because memory manufacturing is highly cyclical and requires considerable investment. Reuters reported that Samsung and SK Hynix have also committed substantial capital to meeting future AI demand. At the same time, analysts and investors argue that current cash generation may allow both companies to increase distributions while maintaining sufficient resources for investment.
Both companies already have formal frameworks connecting shareholder returns to free cash flow.
Samsung’s current policy covers 2024 to 2026 and provides for an annual regular dividend of KRW 9.8 trillion. The company targets total shareholder returns equivalent to 50% of free cash flow over the three year period and can consider additional returns when there is a significant surplus.
SK Hynix introduced its current programme for 2025 to 2027 with a similar principle. The company intends to allocate half of accumulated free cash flow towards shareholder returns and increased its annual fixed dividend by 25% to KRW 1,500 per share. It also linked the programme explicitly to capital expenditure discipline and financial stability.
These policies illustrate an important feature of sustainable dividend models: distributions are connected to the economics of the business rather than treated as an isolated target.
A company that promises an aggressive payout while ignoring maintenance requirements, debt obligations or changes in demand can eventually weaken its own ability to generate cash. Conversely, continuously accumulating cash without a convincing reinvestment case can lower capital efficiency and leave investors questioning how management intends to create value from the surplus. The most durable approach sits between those extremes.
The debate is taking place as global dividend payments continue to grow.
Janus Henderson reported that global dividends reached $424.5 billion during the first quarter of 2026, an increase of 10.1% from the same period a year earlier. Dividend growth was broad based across North America, Europe, Japan and the UK.
The firm now forecasts global dividend growth of 8.3% for the full year, compared with 6.8% in 2025. Buybacks moved in the opposite direction during the first quarter, declining 3.1% year on year to $425.7 billion.
That divergence is useful because dividends and buybacks communicate different things about capital allocation. Buybacks can be increased or reduced relatively quickly as market conditions change. Regular dividends usually require greater confidence in the durability of future cash generation because companies are generally reluctant to establish a payout they may later need to cut.
For an investor evaluating a dividend project, predictability therefore matters alongside headline yield.
An attractive distribution policy does not require a business to distribute every available unit of cash.
Healthy businesses need reserves. Assets require maintenance. Opportunities to grow can arise unexpectedly. Economic conditions change, customers can delay payments and financing markets can become less accommodating.
The relevant question is how much cash can be distributed after accounting for those requirements.
That requires looking at the business from the bottom up. How recurring is its revenue? How volatile are its margins? What level of maintenance expenditure is required? How much debt has to be serviced? How concentrated are its customers? How much working capital does growth consume? Which expansion projects can realistically produce returns above the company’s cost of capital?
Only after these questions are answered does a distribution yield become meaningful. A lower payout supported comfortably by recurring free cash flow can be more valuable over time than a high payout that depends on favourable conditions continuing indefinitely.
The debate around Samsung and SK Hynix also highlights another factor that matters across public and private businesses: capital allocation provides a practical test of management. Generating a surplus creates choices, and those choices reveal priorities.
Management can reinvest because a project offers an attractive economic return. It can retain cash because the operating environment requires a larger safety margin. It can reduce expensive debt. It can make an acquisition when there is a clear strategic and financial case. When none of these alternatives offers a compelling return, distributing surplus capital can be the more disciplined decision.
Reuters reported that investors have increasingly focused on exactly this question at Samsung and SK Hynix. Both companies have said they believe stronger shareholder returns can be balanced with future investment and financial stability.
For private cash generating projects, the principle can be even more relevant. Investors do not necessarily have the same liquidity available in public markets, which places greater importance on the underlying project’s ability to produce cash and on having a clear framework for how that cash will eventually reach investors.
A strong operating business and an attractive dividend project are closely related, though they are not automatically the same thing.
The business first needs to generate cash consistently. That cash generation needs to remain resilient after realistic maintenance expenditure and working capital requirements. The balance sheet needs enough capacity to absorb weaker periods. Management then needs a clear policy for dividing surplus cash between reinvestment and distributions.
When these elements work together, investors gain something valuable: a direct connection between the performance of the underlying business and the return on their capital.
Recent developments in the semiconductor industry offer an unusually large scale example, but the underlying principle applies across sectors. The most interesting cash generating businesses are often those that can continue investing enough to preserve their competitive position while regularly returning part of the value they create.
For investors, the question therefore extends beyond whether a company generates cash. The more revealing question is what management does with it once it arrives.
☑️ Reuters, 6 August 2026: Samsung, SK Hynix shareholders call for bigger payouts from AI cash mountain
☑️ Samsung Electronics: Shareholder Return Policy
☑️ SK Hynix: Shareholder Return Program and Value Up Plan
☑️ Janus Henderson, 21 July 2026: Global Dividends Rise 10.1% in Q1 2026
