For investors trying to determine whether corporate earnings are genuinely improving, the latest reporting season offers an uncomfortable lesson: reported earnings can be accurate under accounting rules and still be a poor measure of the earnings power of a business.
The issue is becoming particularly important among the largest technology companies, where strategic equity investments in other businesses can produce enormous gains that flow through the income statement. Those gains may be legitimate under U.S. GAAP, but they are not the same thing as revenue growth, margin expansion or stronger operating cash flow.
That distinction matters enormously when investors are paying premium valuations for future earnings.
According to data compiled by S&P Global Market Intelligence, S&P 500 companies generated approximately $2.64 trillion in combined net income over the most recent four quarters. But a meaningful portion of that figure includes paper gains from companies marking up equity investments they own.
The most striking examples come from Amazon and Alphabet.
Last quarter, the two companies reported roughly $121 billion combined in after-tax “other income,” almost all of it attributable to investment gains. At Alphabet, other income represented approximately 71% of quarterly profits. At Amazon, it represented about 66%.
Alphabet benefited from revaluations of equity holdings that include SpaceX and Anthropic, while Amazon’s gains primarily reflected its stake in Anthropic.
For an accountant, these gains are real enough to appear in GAAP net income. For an investor attempting to determine what a company is worth, however, the analytical treatment should be very different.
The problem isn’t accounting. It’s interpretation.
This distinction is important for advisors because it is tempting to describe these numbers as an accounting problem. They are not.
GAAP is doing what GAAP is designed to do. When certain equity investments are required to be marked to market, changes in their estimated value can flow into reported earnings.
The problem arises when investors—or the earnings machinery surrounding Wall Street—treat those gains as though they represent recurring economic profits generated by the underlying business.
They don’t.
If Alphabet’s investment in Anthropic rises sharply in value, Alphabet has not suddenly sold more Google advertising. Amazon hasn’t shipped more packages because Anthropic became more valuable. Neither company necessarily generated additional operating cash from the revaluation.
The gain is largely a reflection of what somebody believes an investment is worth today compared with yesterday.
That can be economically meaningful. It simply isn’t the same thing as operating performance.
For wealth advisors, this distinction should become part of the standard earnings-review process. A client who sees a company reporting a huge earnings beat may reasonably assume that the business is becoming more profitable. The advisor’s job is to determine whether the improvement came from the business—or from something sitting outside the core business model.
Why valuation makes this especially important
The issue becomes more consequential when markets are expensive.
The S&P 500 is trading at roughly 27 times trailing earnings, compared with a historical average closer to 16 times. At that valuation, investors are paying a substantial premium for each dollar of reported profit.
That makes the quality of those earnings increasingly important.
Suppose a company reports $10 billion of earnings, but $2 billion comes from an unrealized increase in the value of an investment. A conventional earnings multiple treats the company as though it produced the full $10 billion through its business activities.
An investor might therefore conclude that a stock trading at 27 times earnings is reasonably valued because the company is generating strong profits.
But if the $2 billion investment gain is unlikely to recur, the more economically relevant earnings figure is closer to $8 billion. The effective valuation on recurring earnings is substantially higher.
This is the hidden danger of using headline P/E ratios without examining the composition of earnings.
At elevated market valuations, relatively small differences in the definition of “earnings” can produce very large differences in what investors are actually paying.
Wall Street’s earnings framework makes the problem worse
There is another wrinkle that advisors should watch closely: Wall Street doesn’t use one consistent definition of earnings.
Analysts routinely adjust reported results to exclude expenses they consider unusual or less relevant to underlying performance. Stock-based compensation is a familiar example. Restructuring charges, acquisition costs and other supposedly nonrecurring items may also be removed.
There can be legitimate reasons for making those adjustments.
But the process becomes problematic when recurring costs are excluded while convenient gains are allowed to remain.
That creates an asymmetric definition of “adjusted earnings”: expenses are treated as distractions, while gains are treated as evidence of stronger profitability.
Investors should be skeptical of any earnings presentation that consistently improves the company’s story regardless of whether the underlying economic event is positive or negative.
The appropriate question isn’t simply, “What did the company earn?”
It is:
“What did the company’s core business earn, and how much of that earnings figure is likely to be available again next year?”
That is a much more useful question for portfolio construction.
Advisors should separate three types of earnings
A practical way to handle the issue is to divide reported earnings into three buckets.
First, operating earnings. These come from the company’s core business: selling products, providing services and controlling costs. This is generally the most important category when assessing the sustainability of profits.
Second, recurring non-operating income. Interest income, for example, may be outside the core business but can still be economically meaningful and reasonably predictable.
Third, valuation-driven gains and losses. These include mark-to-market changes in equity investments and other assets. They can be economically significant, but they should generally receive a much lower weight when estimating sustainable earnings.
The third category is where investors can get into trouble.
An investment gain can eventually become cash if the asset is sold. It can also prove to be a very valuable strategic investment. But until the economics are realized, it should not automatically be capitalized at the same multiple as recurring operating profits.
This is particularly relevant for technology companies because many of them increasingly resemble investment portfolios in miniature. Large cash balances, strategic stakes in private AI companies and venture-style investments can create a second source of reported earnings that has little connection to the company’s primary operating engine.
What advisors should do differently
Advisors don’t need to become forensic accountants every quarter. But they should add a few questions to their earnings review process.
Start with the income statement, not the earnings headline. Look for unusually large changes in “other income,” investment gains, gains on securities or similar line items.
Reconcile reported earnings with operating performance. Revenue growth, operating margins, free cash flow and cash conversion tell a different story than headline net income. When those measures diverge significantly, investigate why.
Calculate a normalized earnings figure. If a large portion of quarterly earnings comes from an investment revaluation, consider what the company’s earnings multiple looks like after removing that gain.
Read the footnotes. This is especially important for companies with large strategic investment portfolios. The accounting treatment and source of the gains may be buried well below the headline numbers.
Be cautious with earnings beats. A company can beat analysts’ estimates and still fail to improve its underlying economics. The size and source of the beat matter.
For client conversations, the language can be straightforward: “The company reported higher earnings, but a significant portion came from increases in the value of investments it owns. That’s positive, but we don’t want to value a one-time paper gain the same way we value recurring profits from the business.”
That explanation is both more accurate and more useful than simply telling clients that earnings were strong.
The bigger lesson: earnings quality matters more when valuations are high
The market’s problem isn’t that companies are reporting misleading numbers. The problem is that investors can too easily attach a misleading interpretation to perfectly legitimate numbers.
This distinction will become increasingly important as technology companies accumulate larger stakes in other businesses, particularly in artificial intelligence. Private-company valuations can change dramatically between funding rounds, creating potentially enormous swings in the reported value of strategic investments.
That can make quarterly earnings appear far more volatile—or far stronger—than the underlying operating business actually is.
For advisors, the implication is not to dismiss these gains. Some investments will eventually become enormously valuable. Instead, the task is to place them in the right analytical bucket.
A $10 billion increase in the value of an investment is wealth creation. It is not necessarily $10 billion of recurring earnings power.
That difference may sound subtle. At a 27-times market multiple, it is anything but subtle.
As markets become increasingly concentrated in companies with substantial investment portfolios and strategic stakes in emerging technologies, earnings quality deserves as much attention as earnings growth. Advisors who focus only on whether a company beat expectations risk missing the more important question: How much of those earnings can the business realistically produce again?
For investors, that is ultimately the number that deserves the multiple.
