Three shorthand stories about the economy that professionals are pushing back on

The consumer is less split than the shorthand

The first shorthand under challenge is the split consumer. Apollo's chief economist Torsten Slok argues that the US consumer is less K-shaped than commonly perceived. Nominal spending growth has been broadly similar across income groups since 2024, the commentary says, with all groups running at roughly 4-5% year over year. That breadth, rather than a split between strong higher-income spending and weak lower-income spending, is the commentary's central claim. m:…9469

The commentary extends the same point to balance sheets. Household net worth has risen substantially across the income distribution since 2019, it says, including by around 80% for the bottom 20% of the income distribution. The commentary does not give figures for the other income groups, so it does not show whether the gains were equal, only that growth was broad. The claim is about the shape of the recovery, not its size for any single group. m:…9469

Chip stocks fell; AI spending did not

The second shorthand is that a retreat in chip stocks means the AI build-out is fading. Vivek Arya, managing director at BofA Global Research, told TJ Thorne on a Bank of America podcast that AI spending remains robust even as chip stocks have retreated from their June highs. Arya said hyperscaler spending forecasts have been revised upward by approximately 20% intra-quarter. The argument is that stock prices and spending plans are telling different stories, and that the spending plans are the one to watch. m:…4401

The discussion gave two reasons for that reading. Demand for high-bandwidth memory currently outstrips supply by a factor of 3 to 4, Arya said, a shortage that points to demand running ahead of what suppliers can deliver rather than demand falling away. And AI is a critical driver of GDP growth, contributing approximately 70 basis points in 2026, according to the discussion. Arya also argued that the current semiconductor cycle is likely mid-cycle, with potential for at least two more years of growth, and that open-weight AI models do not reduce demand for underlying computing infrastructure. m:…4401

The typical millionaire is 62 and unglamorous

The third shorthand is the millionaire made in tech. Guests on Bloomberg Surveillance argued that most millionaires in America build wealth through private, mundane businesses rather than high-profile tech roles, and that the typical everywhere millionaire in the United States is 62 years old. The picture is of wealth accumulated gradually through ordinary businesses, not of a single high-profile career path. m:…4282

That claim sits alongside a warning in the same discussion: two out of five Americans are heading toward retirement with no real plan. The discussion does not connect that figure to the millionaire claim directly. The two points share a theme, that household finances look different from the headline version, but they are separate observations, not one finding. m:…4282

What the pushback does not settle

Each of these arguments is a professional reading of data or of a client conversation, not a measurement that settles the question. The Apollo commentary does not break out spending or net-worth growth group by group, so a reader cannot check from it whether any group is lagging. The BofA discussion's spending forecasts are themselves projections, not observed spending. The Bloomberg discussion does not say which kinds of private businesses produced its millionaires, or over what time period the wealth was built. m:…9469 m:…4401 m:…4282