What Howard Marks Is Asking About AI

What Howard Marks Is Asking About AI

A summary of "A Layman Thinks About AI," the November 11, 2026 memo by Howard Marks, Co-Chairman of Oaktree Capital Management

Howard Marks is one of the most widely read thinkers in investing. His new memo takes on a question many of you have asked: how much of a portfolio's fate now rides on artificial intelligence? It is worth your time, and we have summarized the main ideas below. Every idea in this note is Mr. Marks'. We have restated them in our own words.

Highlights

• Many things must go right at once. An AI thesis depends on the technology, demand for computing power, competition, who profits, financing and timing. Even good odds on each shrink quickly when multiplied together.
• A great technology is not automatically a great investment. If AI models become interchangeable, the benefits may flow to the users rather than the vendors.
• Leadership can change fast. Today's top AI labs were little known a few years ago, so he asks how durable any lead really is.
• The market is concentrated. AI accounts for an outsized share of economic growth, earnings growth and stock market gains. He describes this as a fault line running through portfolios.
• Credit markets are exposed too. He points to a surge in AI-related borrowing and to "circular" deals among suppliers, customers and lenders, and urges lenders to be exceptionally vigilant.
• His conclusion: neither paralysis nor blind enthusiasm. Size positions with humility, know which risks you are being paid to take, and ask how many things must go right for an investment to work.

Not all AI risk is equal

Mr. Marks asked an AI model (Claude, from Anthropic) to score how predictable each part of the industry is, from 1 (unknowable) to 10 (fully predictable). These are the model's judgment calls as reported in his memo, not precise measurements, but they show how much the risk varies.

AI segment: Predictability (1 to 10)
Regulated power utilities serving AI load: 6
Data centers leased long-term to big tech: 5-6
Hyperscalers: 5
Chips: 4
Speculative "neocloud" data centers: 2-3
Frontier AI labs: 2

Overall, the model rated AI about a 3: the direction is predictable, but how fast it happens, how big the profits are and who keeps them are much less so.

Five questions he suggests asking about any investment

• How many things must go right for me to earn an acceptable return?
• If my central assumptions are wrong, how much could I lose?
• Which risks am I being paid to take, and which am I simply assuming?
• How much optimism is already built into the price?
• Does this add to the risk my portfolio already carries, or diversify it?

Why it matters to us

Our approach centers on protecting your lifestyle from large drawdowns, especially around retirement, when losses are hardest to recover from. Mr. Marks' concern about heavy reliance on a single theme is a diversification question that we review in every portfolio. If you would like to talk about how AI-related concentration may show up in your own accounts, please reply to this message or call us.

Read the original

The ideas above belong to Howard Marks and Oaktree Capital Management. His full memo is copyrighted by Oaktree and is available at oaktreecapital.com/insights/memos. We encourage you to read it in his own words.

Warm regards,

Rick Shrier
Investment Advisor Representative

Registered Investment Advisor
619 Devenwood Way | Clinton, MA 01510
5 Farleigh St. Suite 305 | Asheville, NC 28803
tel 508.898.9500  fax 508.464.0064

http://www.shrierassociates.com/

See how we plan for the next downturn. Visit our redesigned website:
www.shrierassociates.com

Source: Howard Marks, "A Layman Thinks About AI," Oaktree Capital Management, L.P., November 11, 2026. © 2026 Oaktree Capital Management, L.P. The memo expresses the author's views as of its date, which are subject to change. Figures and data referenced by Mr. Marks are as reported in the memo and have not been independently verified by Shrier Associates.

Disclosure: This message is for general educational purposes only. It is not personalized investment advice, an offer or a solicitation, or a recommendation to buy or sell any security. Shrier Associates, LLC is not affiliated with Oaktree Capital Management or Howard Marks, and his views do not necessarily reflect those of our firm. Investing involves risk, including loss of principal. Past performance does not guarantee future results.