By James Mackintosh

Are the big AI companies giving up their status as the highest caliber stocks in the market? The question is at the heart of a debate about "quality" companies that has left two popular ETFs with wildly different performance after one ditched Nvidia and most of the rest of Big Tech.

The $48 billion iShares MSCI USA Quality Factor (ticker: QUAL) and Invesco's $15 billion S&P 500 Quality $(SPHQ)$ specialize in what the investment industry has dubbed quality companies. The exchange-traded funds invest only in companies that show they are safe and steady on financial measures including high profitability and low leverage.

The benchmarks that these two ETFs use to measure quality have a crucial difference, which leads to one having much more exposure to the major artificial-intelligence stocks. QUAL, which bases its definition of quality on an MSCI index, has almost a third of its holdings in five of the leading eight Big Tech stocks. SPHQ, which is based on an S&P index, holds only one, Apple.

The result has been wild swings in performance. Up to June, when Invesco dropped Nvidia, SPHQ beat QUAL by the most in any six-month period. In the past six months SPHQ has lagged behind by the most, aside from the period immediately after QUAL got going in 2013.

This is much more than a semantic debate between ETFs about what quality means. It goes to the heart of a much bigger question in markets: Is the bet on AI being made by the country's biggest companies a vast potential profit pool or a money pit that should be sounding alarm bells?

In addition to Nvidia, SPHQ dropped Meta and Netflix in June, and Microsoft last December. The culprit, according to Nick Kalivas, head of factor and equity ETF strategy at Invesco, was the accounting concept of accruals.

Accruals are a way to gauge how much of reported earnings are based on sales that generate cash right away, rather than money owed by customers in the future or other noncash items. Cash today is obviously preferable to a promise that money will come through in the future. Things can go wrong. Customers go bust, economies change.

The S&P index behind the fund uses accruals as one of its three indicators, because high accruals are sometimes a sign of trouble ahead.

"You start to see a deterioration in the cash component of earnings, and that might indicate that your strength and the durability of your earnings might be coming under pressure," Kalivas said.

In Nvidia's case, there has been a jump in working capital as rapid sales growth has led to immediate costs, while payments from customers lag behind. In the latest quarter, money owed by customers, known as accounts receivable, leapt by $16 billion from the year before, to $33 billion, while money owed to suppliers, accounts payable, rose only $3 billion, to $8 billion. The gap between the two has to be funded by the company while it waits to be paid.

QUAL's index doesn't exclude stocks based on accruals, instead looking for steady earnings growth, which these Big Tech names have in spades. MSCI, which runs the index behind the fund, consulted customers about adding accruals last year, but decided against the move after finding it would lead to more churn in the stocks that qualify.

Rising accruals may be part of the problem that the market began worrying about a few weeks ago: Big Tech companies are pouring hundreds of billions of dollars into AI without a clear path to profit, eating into their cash on hand and increasingly needing borrowing to finance it.

Investing real money now and hoping for payments some time in the future is the core of investment. But big companies rarely spend so much on a technology where uncertainties are so high.

Wei Li, chief investment strategist at the BlackRock Investment Institute, thinks there is still more to go in the bull market. But she said the uncertainties are so big it isn't even worth trying to forecast how much productivity -- and so adoption and profit -- AI might lead to.

"We spend first, and we're hoping that at some point revenue will come," she said. "But that hasn't happened yet."

Quality stocks don't usually come with so much uncertainty about where they allocate their capital. But it's also undeniable that Nvidia and most other Big Tech stocks have been exceptionally profitable as the AI boom intensified.

The accruals happening here aren't the sort of accruals that make investors the most nervous. Forensic accountants worry about rising accruals as a sign of earnings manipulation or fraud, but that isn't the concern here.

Helen Jewell, chief investment officer for fundamental equities at BlackRock, uses an accruals-like measure -- cash conversion -- as part of stock selection and said the point is to identify companies able to get into a "virtuous cycle" of investment.

"If you have got strong earnings and cash flow, it allows you to continuously reinvest," she said. That investment makes more money, allowing more investment, and so on.

A separate academic finding used by neither ETF has long established that when companies splurge on capital spending, as the AI firms are now, their shares typically lag behind the rest of the market in the future. Investment strategists have been rediscovering the depressing historical link recently.

Yet, all this effort might be beside the point, said Mamdouh Medhat, research director at Dimensional Fund Advisors. He argues that there's no need to overcomplicate quality gauges: A strategy of just buying the most profitable stocks trading at lower valuations, on average beat combined measures with accruals, stability of earnings or other tricks.

Outside the smallest companies, such a strategy already captures any gains from avoiding high-capex stocks, he says. Put simply: Don't buy bad companies, and don't buy the most expensive companies.

For me, the point of any quality approach is to spot strong companies where management resists the temptation to throw money at the latest fad. AI is definitely in fashion, and management is throwing money at it.

Investing based on factors aims to deliver a small premium that adds up over time. The problem with today's market is that a handful of stocks are so big that including them, or not, can deliver, or miss out on, years of these gains in a few months. The result is that which ETF does better from here depends on investor appetite for AI at least as much as it does on quality.

Write to James Mackintosh at james.mackintosh@wsj.com

(END) Dow Jones Newswires

December 07, 2025 05:30 ET (10:30 GMT)

Copyright (c) 2025 Dow Jones & Company, Inc.