IBM Stock Gains on Quantum Optimism. This Analyst Isn’t Convinced.

IBM teamed up with Vanguard to explore how quantum computers could work with classical machines to construct optimized portfolios. (Johannes Eisele/AFP/Getty Images)

Key Points

• IBM’s quantum-powered bond-trading experiment with HSBC showed a 34% reduction in errors compared with classical methods.
• An analyst questioned the enthusiasm surrounding IBM’s quantum experiment, citing concerns about replicability and the classical baseline.
• IBM’s new research with Vanguard demonstrated quantum computers could optimize portfolios on par with classical methods.

International Business Machines caused shock waves last week after it shared the results of a quantum-powered bond-trading experiment conducted with HSBC, but one analyst is questioning whether the news was overplayed.

The company unveiled additional research on Monday showing how algorithms run on hybrid quantum-computing systems could be used to construct optimized portfolios. IBM described portfolio optimization as “one of the most computationally demanding problems in finance.”

The results, reached in collaboration with Vanguard, indicated that quantum computers could work with classical machines to deliver results on par with purely classical methods.

IBM shares barely reacted to the news in early trading Tuesday, a contrast to the jump that occurred after the HSBC experiment was unveiled on Thursday. J.P. Morgan analyst Brian Essex believes that last week’s reaction was unwarranted.

Essex noted that IBM stock gained more than 6% following the back of the bond-trading study, which found a 34% reduction in errors compared with purely classical methods. The stock was up 1.7% at $284.61 on Tuesday.

“The news highlights an exciting step forward for quantum research, but we think press outlets may have been a bit too enthusiastic with their characterization of the announcement, which likely contributed to the move in the stock price last week,” he wrote.

Essex argued there were no guarantees the results could be replicated outside a research setting. Moreover, he took issue with the classical baseline, arguing that it “may not fully reflect the capabilities of state-of-the-art classical methods.”

D-Wave Quantum attracted similar criticism when it said its Advantage2 quantum processor had achieved supremacy over classical methods earlier this year. One group of scientists argued that more powerful classical machines had emerged since the study was conducted, undercutting the claim that quantum could outperform traditional computers.

“The HSBC results have meaningful implications for quantum progress, but it is not clear that the results reflect an actual quantum advantage at this point,” Essex continued.

Notably, neither HSBC nor IBM themselves characterized the result as an example of quantum advantage, also known as supremacy: the completion of a practical task that would be impossible for a classical computer to achieve within a reasonable time.

In an accompanying press release, HSBC noted that IBM’s Heron quantum processor was “able to augment classical computing workflows” to parse through market data better than “standard, classical-only approaches in use by HSBC.”

Still, Essex sees promise in the technology. The analyst said IBM was well positioned to be one of the leaders in quantum computing, even as the study fell short of being a “Sputnik moment,” as some news outlets called it.

Wall Street may feel impatient for results in the context of the eye-watering moves of quantum stocks. D-Wave and Rigetti Computing are up 2,651% and 3,838% over the past 12 months, respectively. Critics say those gains are unwarranted, as the companies remain unprofitable and the technology currently has limited real-world applications.

As Essex pointed out, most scientific breakthroughs occur in a research setting, and there is limited evidence of how the machines will perform in the real world. Companies including IBM itself are targeting the end of the decade for truly fault-tolerant computers, which catch and correct errors as they arise.

The HSBC paper wasn’t peer-reviewed, which leaves room for scrutiny. While it doesn’t show that hybrid systems can trump purely classical methods all the time, every development helps legitimize the nascent technology in the eyes of those outside the scientific community.

That is worth something.

Write to Mackenzie Tatananni at mackenzie.tatananni@barrons.com