Why Consider AI Stocks, What to Be Aware of, per Morningstar

Risks and opportunities exist in machine learning for investors who want exposure.



With artificial intelligence playing an increasing role in many businesses, Morningstar Inc. has laid out opportunities and risks for allocators who want exposure to AI stocks.

“Artificial intelligence, or AI, has become far more sophisticated in recent years, creating both disruption and opportunity,” stated Morningstar’s 2024 wealth outlook report. “It brings risk, but there will be winners. We seek to support investors in their understanding of this space.”

According to Morningstar data, AI semiconductor companies generated $47 billion in revenue, and AI software companies generated $68 billion in revenue in 2023. Morningstar expects these numbers to grow to $143 billion and $137 billion, respectively, by 2027.

AI Ideas for 2024

Never miss a story — sign up for CIO newsletters to stay up-to-date on the latest institutional investment industry news.

According to Morningstar, there is opportunity in the “next rung” of AI adopters, even without focusing on the big AI players. The report recommended investors avoid high valuations, which are common for many of the top tech and AI companies at the moment. Instead, Morningstar pointed to smaller firms that can strengthen their products with AI and do not have such valuation risk. “As competition rises, we could see disappointments,” Morningstar analysts wrote, noting that overvalued AI stocks may not perform as well as up and coming competitors. 

According to the report, the 10 tech and AI stocks most commonly held in big-data exchange-traded funds and mutual funds globally are the so-called Magnificent Seven stocks, plus a few more:

  • Nvidia Corp.; 
  • Microsoft Corp.; 
  • Alphabet Inc.;
  • Amazon.com Inc.; 
  • Advanced Micro Devices Inc.;
  • Tesla Inc.;
  • ServiceNow Inc.; 
  • Meta Platforms Inc.;
  • Taiwan Semiconductor Manufacturing Co.; and 
  • Snowflake Inc.

Risks in AI

Morningstar’s report also listed several potential risks when investing in AI.

Regulation and Safety – AI companies are under scrutiny from governments and regulators around the world. These regulators have concerns about data, privacy and copyright issues. Regulations will likely be regional, and big policy changes are likely to be implemented in 2025.
Valuation – “Even fast-growing businesses can be poor investments if investors overpay for shares,” Morningstar’s analysts wrote. “It’s also important to keep in mind that the largest portion of a growth company’s value is derived from cash flows generated many years in the future. Companies that develop durable competitive advantages are more likely to sustain long-term free cash flow growth and could warrant richer valuations. …. A lot has to go right for the primary AI stocks to continue to deliver, which could happen, but the risk-to-reward can deteriorate if investors overpay.”

Concentration – Investors should acknowledge the uncertainty AI brings when considering position sizes.

Obsolescence – Some companies’ products and services could become obsolete or less relevant because of AI. Some AI companies could fall behind, as not every firm has the resources to build large foundation models, as OpenAI and others are doing.

Related Stories:

Will AI Compromise Security for Institutional Investors?

Asset Managers Ponder Investments in AI as Security Risks Compound

How Asset Managers Can Harness AI to Boost Profits, per EY

Tags: , ,

Markowitz Redux: Updating Modern Portfolio Theory

Two academics create a new measure to divine a security’s price after a certain holding period.


Seventy years ago, an economist named Harry Markowitz received his Ph.D. from the University of Chicago, based on his doctoral thesis about the proper allocation of investments. Now known as Modern Portfolio Theory, it laid out a framework for how to get the best return on stocks in light of the risk involved.

A new upgrade to MPT, seeking to refine how to assess risk for various types of securities, comes in a working paper, “Equivalent Expectation Measures for Risk and Return Analysis of Contingent Claim Portfolios,” by two young economists.

Markowitz, who won a Nobel prize for MPT, died last June at age 95. He invented a mathematical concept to measure the risk on a collection of assets in terms of how they move up and down together. Before his research, scholars focused on individual securities, not on how they might offset one another, or on the market writ large. Other refinements to MPT have been launched since, notably the work of another Nobel winner, William Sharpe.

The two academics who penned the new paper—Sanjay Nawalkha, from the University of Massachusetts, Amherst, and Xiaoyang Zhuo, at the Beijing Institute of Technology—created a new class of probability gauges, called  “equivalent expectation measures,” which produce formulas to find the future prices of virtually any financial securities over a given holding period. Their concept encompasses Treasurys, corporate bonds, mortgage-backed securities and derivatives: options, futures and swaps. These securities are collectively called “contingent claims.” 

Want the latest institutional investment industry
news and insights? Sign up for CIO newsletters.

The professors’ approach allows measurement of the risk (known in economist-speak as the “variance”) on, for example, a three-month call option for the S&P 500 for a one-month holding period. The EEM concept also permits investors to measure the “covariance” (i.e., how two different securities move in relation to one another, useful when constructing a portfolio). Such as that of a three-month call option on Tesla stock and a six-month call option on Apple stock during a two-month holding period for both.

A spin-off version of this paper is already accepted by the Journal of Investment Management. This journal is also organizing a conference in March at the University of California at San Diego, to honor Markowitz.

In an interview, Nawalkha termed it “serendipitous” that the new concept “extends applications to Markowitz’s work to all contingent claims [and] coincides with the conference.” 

Previously, Nawalkha and Zuo authored a paper refining another storied financial model, Black-Scholes, to calculate the expected risks and returns of derivatives over a finite period, say over three or six months.

Related Stories:

Black-Scholes 2.0: Classic Valuation Method Expanded to Risks and Returns

‘Smart Money’ Flows Into Options Markets, Not Equities

Goldman: Due to High Volatility, Now’s the Time to Bet on Earnings Reports

Tags: , , , , , , , ,

«