Thunder of the 2026 “AI Bull Market Narrative”: America's Tech Giants' “Depreciation Trick”

Generally speaking, when an accounting issue becomes a hot topic on Wall Street for a period of time, it definitely doesn't bode well for the stock market. This is why recent popular discussions about the “depreciation period” arrangements and plans of the US tech giants for AI computing power infrastructure should keep investors who are deeply addicted to the AI investment boom to stay alert as they move towards 2026.

Generally speaking, matters in financial markets debated by those who prepare corporate financial statements are too obscure — and, frankly, too boring, so there was little interest among stock investors to actively explore until the beginning of this century.

However, since the beginning of this century, the issue of accounting fraud was revealed by popular companies such as Enron (Enron), WorldCom (WorldCom), and Adelphia Communications (Adelphia Communications), which had been popular in the capital market, using deceptive accounting practices. From time to time, the term “accounting fraud”, which had been accumulated in the capital market over a long period of time, made headlines, and often caused a wave of rapid decline in the stock market.

These companies quickly slipped from what seemed like glorious “everyone admires” success to bankruptcy, triggering a widespread market chain reaction. For a while, investors began to wonder if they could trust any company's comprehensive statement of its profitability and financial position.

This heinous accounting fraud — in fact, is also criminal fraud — has triggered major changes in the US stock market, the most famous of which is the Sarbanes-Oxley Act (Sarbanes-Oxley Act). It has largely made outright fraud in US listed companies a thing of the past.

However, that doesn't mean investors can let their guard down for a long time.

A somewhat misleading yet legitimate financial/accounting disclosure remains a significant risk to the market. If a company's financial reporting profits are revealed to be completely out of touch with the reality of its underlying economic fundamentals, the company's stock price may drop sharply over a period of time. And if the company's market capitalization is large enough (such as the current Magnificent Seven), it may even drag down the broader US stock market.

The “super bull market” where the S&P 500 index has accumulated a cumulative increase of about 30 trillion US dollars over the past three years is largely driven by the world's largest tech giants (such as the seven major US tech giants), and also by companies that promote large-scale investment in AI computing power infrastructure (such as Micron, TSMC, Broadcom, etc.) and power system suppliers (such as Constellation Energy).

The so-called “seven tech giants”, or “Magnificent Seven,” which account for the high weight of the S&P 500 index (about 35%), include Apple, Microsoft, Google, Tesla, Nvidia, Amazon, and Facebook's parent company Meta Platforms. They are the core driving force behind the S&P 500 Index's record highs, and are also regarded by top Wall Street investment institutions as the combination most capable of bringing huge returns to investors in the context of the biggest technological changes since the Internet era.

A major concern — a “big short” tweet sparks a buzz about “depreciation” in the market

In the macro context where the market capitalization weight of the S&P 500 index is increasingly concentrated on the “Big Seven Tech Giants” (about 35%-40%) and the “AI bubble argument” that has completely taken the market by storm, the “Big Short” published a tweet about the depreciation controversy of US tech giants, the current controversy surrounding the depreciation period schedule of these technology companies is worrying.

Not long ago, the “AI circular investment” led by OpenAI was increasingly exaggerated, and the unprecedented large-scale borrowing of AI data centers by giants such as Oracle jeopardized financial fundamentals, causing the market to increasingly worry that “the AI bubble is bursting.” As for 2026, as the market became more afraid of the AI bubble storm that blew up in November, they even began to doubt that this irrational AI bubble is forming and getting closer to bursting.

A number of major Wall Street banks, including Goldman Sachs, Bank of America, Yardeni Research, and Morgan Stanley, recently stated in their annual summary and outlook that the market's growing skepticism about high valuations of technology stocks and whether huge AI investments can bring significant returns is driving the market to pay more attention to traditional cyclical sectors such as industry and energy, rather than the “Big Seven US stocks” such as Nvidia and Amazon, which have high valuations and are at the “center of the AI bubble.” Ed Yardeni, founder of Yardeni Research, recently even suggested that investors “reduce” their holdings of the top seven tech giants compared to the rest of the S&P 500 index. This is the first time since 2010 that he has changed his position on increasing his holdings in the major US technology sector.

Take Nvidia (NVDA.US) as an example — the “AI chip hegemon” with a market capitalization of over 4 trillion US dollars and has the title of “Earth's Most Important Stock”. Every move in the company's stock price and market capitalization can be said to have a bearing on whether the “AI bull market narrative” logic works smoothly, and even whether the global stock market can continue to be bullish.

Michael Burry (Michael Burry), a well-known investment manager with the title of “Big Short” — his successful bet to short the US housing market on the eve of the 2008 global financial crisis was told to the world by screenwriters in the Hollywood film “The Big Short” (“The Big Short”).

This tweet posted by Michael Berry shows that he believes that US tech giants such as Meta and Oracle that have spent huge sums of money to lay out AI computing power infrastructure, and even Nvidia itself have overestimated their own profitability by amortizing their semiconductor procurement/production costs over many years. The longer the depreciation period, the lower the depreciation expense included each year, and thus the lesser the reduction in profits.

Burry believes that due to advances in technology, the value of chips will fall faster than the figures assumed by the company's depreciation schedule. Investors who are bullish on Nvidia, on the other hand, believe that the longer depreciation period only reflects the actual useful life of these chips in terms of wear and tear.

Nvidia isn't the only technology company whose “big short” profit statements have been reviewed due to depreciation practices. In his tweet, Bury also pointed the finger at cloud computing giant Oracle (ORCL.US). Furthermore, since 2020, as global technology companies' capital expenses (capex) soared, other “seven tech giants”, including Google parent company Alphabet (GOOGL.US), Amazon (AMZN.US), Facebook parent company Meta (META.US), and Microsoft (MSFT.US), have also been extending the “assumed useful life” of some major heavy assets.

Also, there is some evidence that the established US tech giant IBM (IBM.US), which has the title of “Blue Giant,” is also participating in this “accounting game.” According to the research institute Stock Analysis on Net (which mainly analyzes the financial statements of large US companies), IBM's total depreciation expenses fell from US$4.2 billion in 2020 to only US$2.2 billion in 2024, while revenue increased from US$52.3 billion to US$62.8 billion, which is clearly not in line with the fundamental growth of IBM's economy.

Of course, asset disposal may also contribute to changes in IBM's revenue data, and there may also be other adjustments to depreciation policies; however, it is undeniable that extending the assumed useful life of assets may also be one of the main reasons.

The “storm” hidden in tax forms

Regarding depreciation period arrangements for technology companies, one thing to keep in mind: they are completely irrelevant in terms of actually creating shareholder value.

Specifically, how to set a depreciation period in terms of financial reporting (depreciation plan) itself will not create or increase the real economic value of the company. Shareholder value (intrinsic value) is determined by future free cash flow, not by the “method of apportionment” of accounting expenses. DCF/FCF valuations consider the value of a company as the present value of future cash flows.

Depreciation is a non-cash expense in financial reports: it simply accounts for capital expenses that have already been incurred in the past (buying equipment/building a data center) to future periods. Extending the depreciation period will only “reduce the depreciation expenses accrued every year and increase the book profit,” but it will not increase the free cash flow in the cash flow statement.

It's also important to remember that these accounting techniques are important for tax purposes, but the financial statements these technology companies submit to the US Internal Revenue Service (IRS) are very different from the financial statements they provide to investors.

Therefore, the channel through which depreciation may actually have an impact on cash flow is mainly tax (depreciation deduction), but the company's tax calibration statement to the IRS and GAAP caliber financial reporting for investors are two systems; therefore, “changing the depreciation period for financial reporting” is usually not equivalent to “changing tax depreciation,” which necessarily results in a reduction in cash tax.

In a technology company's annual report, depreciation expenses are simply an accounting entry in one statement. Reducing depreciation expenses by changing assumptions does not increase free cash flow, and therefore does not enhance the real economic value of the enterprise. Therefore, it is not surprising that stock prices usually do not rise sharply or fall rapidly as a result of such “superficial” actions.

Therefore, this kind of “lengthening the depreciation period” is more about making earnings per share (EPS) better looking, making the P/E denominator larger, and making the valuation multiplier look less expensive. It is an “cosmetic modification” rather than an operational improvement method that creates cash flow.

So why might they write an article on the depreciation period arrangement?

This was easier to understand in the 1980s and 90s, when CEO pay was often tied to reported earnings per share (EPS). Today, however, most listed companies pay CEOs based on stock price/market capitalization rather than simple EPS data.

Perhaps these accounting frauds or reasonable retouching techniques are a way to reduce what appears to be “too high” price-earnings ratio, leading some investors to believe that the company is still undervalued even after experiencing a sharp rise in stock prices — over the past few years, tech executives have had to explain this over and over again.

According to agencies' statistics, as of December 18, based on Wall Street analysts' unanimous estimates, the current average price-earnings ratio (P/E) of information technology stocks (that is, IT stocks) is as high as 36x, while the S&P 500 index as a whole is only 25x.

By increasing the denominator, you can make the price-earnings ratio look less high, which may make life slightly better for CEOs, chief financial officers, and public relations managers of high-valued tech giants.

However, the spread of such practices does not help investors who are trying to determine whether the value of a stock is reasonably priced; on the contrary, they mistakenly believe that the company's P/E valuation is still appropriate.

To be clear, there is no indication that today's large technology companies have fraudulent/fraudulent accounting practices.

Overall, depreciation period adjustments are more like a “visual management/communication tool for profit statements” to serve valuation, manage market expectations, and the “profit quality” narrative logic (and some indicators) dominated by CEOs, rather than generate cash flow. For example, depreciation affects core profit indicators such as EBIT, and many credit indicators (such as interest coverage) use EBIT as a numerator to measure solvency; therefore, reducing depreciation and increasing EBIT may make these ratio figures better.

However, these technology stocks, which have huge weight in the S&P 500 index and where any change in market capitalization has a huge impact on global stock markets, are almost always priced by the market's “perfect profit expectations”; therefore, if the debate over depreciation heats up, investors may begin to re-evaluate the actual profit levels of these companies. In the stock market where valuations have reached the highest level in history, the “AI bull market narrative” almost single-handedly supports investors' bullish sentiment and market capitalization concentration is extremely high, they are already nervous about any factor that is perceived as “bad news”. Any negative controversy about depreciation may be enough to trigger a very painful “bear market edge” correction.