The stock of one of the oldest technology companies in the world collapsed during Monday’s session. Investors sold shares fearing that artificial intelligence would replace an army of expensive consultants.
According to market data quoted by foreign economic websites, shares of International Business Machines (IBM) fell by over 13%. This sudden collapse means the evaporation of approximately USD 31 billion from the company’s capitalization in just one day. Analysts point out that it was the worst day for the giant’s quotations since October 2000, when the Internet bubble burst. Since the beginning of the year, the company’s valuation has already shrunk by a quarter.
However, the direct impulse for the sale was not the financial results of the company itself, but the announcement of the competition. Startup Anthropic, dealing with the development of artificial intelligence, presented the Claude Code tool. It is intended to solve one of the biggest and most costly problems of modern banking and administration: handling the outdated COBOL programming language.
Americans gave China “the best gift ever”
Table of Contents
Old world foundations
To understand the scale of the stock market reaction, you need to go back to the 1950s. This is when COBOL (Common Business-Oriented Language) was created. Although it is considered a relic in the world of modern technologies, it is still the lifeblood of global finance. It is estimated that systems based on this language support 95 percent. U.S. ATM transactions and key operations of airlines and government agencies.
The problem is that experts who can operate these systems are retiring in large numbers, and the technical documentation has often not kept up with decades of code modifications. Therefore, maintaining and modernizing this infrastructure previously required employing thousands of consultants who manually analyzed millions of lines of code. It is on this process – tedious, long-term and extremely expensive – that a significant part of the revenues of consulting and technology companies such as IBM, Accenture and Cognizant are based.
AI changes the rules of the game
The new solution presented by Anthropic hits the very heart of this model. The company claims that its artificial intelligence can automate the exploration and analysis phases of systems, which have previously consumed the most resources. Instead of an “army of humans” mapping workflows for years, AI is expected to do it in a fraction of the time.
In their announcement, the creators of Claude Code put forward a thesis that shocked investors. They found that modernizing legacy systems had stalled because “understanding the old code cost more than rewriting it.” Artificial intelligence is intended to reverse this equation, enabling not only analysis but also quick migration to modern languages such as Java or Python.
Experts: It’s an escape from the ecosystem
Dr. Łukasz Olejnik, an independent researcher and cybersecurity consultant, in his analysis quoted by the media indicates that the threat to IBM is real and multidimensional. In his opinion, it is not only about improving the work of programmers, but about fundamentally changing the relationship with the customer.
The expert explains that IBM profits from the fact that COBOL is difficult to use. The company keeps customers on its mainframe servers by selling them upgrade services but making sure the final product still runs on their hardware. Meanwhile, the competition’s proposal may enable customers to completely “escape” from this ecosystem to the cloud computing of any provider.
Olejnik notes that the statement that the costs of understanding the code exceed the costs of rewriting it is something that “should never be said publicly” from the perspective of IBM’s interests. Although no bank will decide to migrate key systems overnight based on a blog post, the signal sent to the management boards of financial institutions is clear: an alternative to expensive consulting is becoming real.
Fortress defense and cool heads
In response to the sharp decline in shares, IBM issued a statement in which it tries to tone down the mood. The company argues that the value of their mainframe computers does not come from the COBOL language itself, but from the unique architecture of the platform. According to the company, it provides security and transaction performance that other distributed environments cannot guarantee, regardless of the programming language used.
Some Wall Street analysts also suggest that the market reaction may be exaggerated. JPMorgan representatives in a note to clients called the logic behind the selloff “flawed,” arguing that implementing AI in critical sectors such as banking will take years, not days. In turn, Dan Ives from WedBush Securities described Monday’s session as “the most detached from reality trade” he has seen in his career. In his opinion, AI may even help companies such as IBM in the long run, making their services more effective.
Despite these voices of reassurance, market sentiment remains nervous. The declines affected not only IBM, but also other companies living on “technological debt”, such as the aforementioned Accenture and Cognizant. Investors have adopted a “sell first, ask later” strategy, fearing that the development of artificial intelligence is progressing faster than large corporations can adapt to it.
