AI companies are growing rapidly, investment in data centers keeps rising, and mega IPOs are drawing close market attention. But for investors, how much return can all this growth and spending ultimately generate?
On October 6, 2026, during the Greenwich Economic Forum, the author of "North America Outpost" spoke with Rich Nuzum, head of OCIO at Franklin Templeton, about how institutional investors are responding to the current interest rate environment and how they view investment opportunities in data centers, robotics, and AI companies. His judgment always revolves around cash flow.
He said the core indicators are whether a company can continue generating cash in the future, whether money lent out can be recovered, and whether the current investment price is reasonable.
SpaceX completed its listing this year, and potential large IPOs from companies such as Anthropic have also attracted market attention. Will these offerings attract new capital, prompt investors to adjust their holdings, or change the market's valuation of the entire sector?
Nuzum cautioned that a company's overall valuation and the amount of capital an IPO needs to absorb from the market are two different things. "A large IPO usually initially puts only a portion of its shares into public trading; not the entire valuation needs to be taken up by new capital."
A listing may also allow founders, employees, and existing private investors to sell part of their shares to public market investors. But a company's transition from private to public does not automatically change its business. Investors still need to look at the business itself and how much cash it can generate in the future.
Nuzum is currently head of OCIO at Franklin Templeton, mainly working with institutional investors, family offices, and others to help them formulate investment strategies, improve investment decision-making processes, and select investment managers, with a focus on long-term investment returns.
Who ultimately gets the gains brought by AI? Nuzum believes that if an AI company is already profitable and has positive free cash flow, further judgment is needed. Can these cash flows continue to grow? How much growth expectation has already been priced into the current valuation?
If a company is still burning cash, investors need to judge when and how it can move toward profitability and generate positive free cash flow.
But there is an even bigger question: the fact that AI benefits the entire economy does not mean that all companies investing in AI can deliver good returns to shareholders. This also means investors need to be especially cautious when entering AI company investments.
Nuzum used the historical construction of railways as an example. Railways made transportation cheaper and faster, and exporters, importers, and consumers all benefited, but many railway operators still went bankrupt, and buying railway stocks did not necessarily make money.
The buildout of the internet had a similar pattern. Companies that built bandwidth did not all achieve ideal returns, but the internet brought great benefits to other businesses and consumers.
Nuzum said that today, efficiency gains brought by AI are already appearing in daily work. But whether these benefits ultimately turn more into stock returns for technology suppliers or into profit growth for companies using AI still needs to be observed.
He also pointed out that this round of AI investment differs from the past. Some large technology companies already have substantial positive free cash flow. Therefore, investors need to analyze company by company, and can also invest in both companies providing AI technology and companies using AI to diversify.
When asked whether robotics has already entered institutional investment portfolios and what commercialization progress investors need to see, Nuzum focused the discussion on the practical uses of robots.
He is bullish on the combination of robotics and AI, believing that improvements in autonomous operating capability can help people complete more work. But he pointed out that robots themselves are not new. In areas such as automobile factories, automated ports, and forestry machinery, related applications have long existed.
For example, in the past, cutting trees and moving them required human labor. Now operators can sit in the cab of large machinery and let machines complete the work faster and more effectively. For companies, what matters is whether they can complete more work with the same number of people.
At the same time, he is reserved about whether robots must be humanoid. "I am not convinced that the human form is the best design for robots." He gave examples: if a robot needs to see farther, sensors can be mounted on drones or tall pillars. If it needs to move, wheels can also be used. The same chassis can also be fitted with different tools to complete different tasks.
"Do we need something that looks like a person, or a tool that can complete practical tasks?" Nuzum did not rule out the possibility that humanoid robots could attract investment. He is more concerned with whether this design is suitable for the work to be done, whether it can improve efficiency, and how much economic value it creates as a result.
When asked whether infrastructure investment places more emphasis on cash flow stability, inflation protection, or growth potential, Nuzum said that his institutional clients and wealth management clients are currently first considering diversification and making their portfolios more resilient to market changes.
They want to avoid the situation where they appear to have bought different assets but in fact are all betting on the same direction of AI and digitalization. "They may invest in private credit financing for data centers, or invest more broadly in data center infrastructure. But if they increase this kind of exposure, they may also reduce related exposure in other parts of the portfolio, avoiding excessive concentration on the same theme across the entire capital structure and value chain."
For new capital, Nuzum said the first consideration now is diversification and portfolio resilience, followed by inflation hedging, and then improving returns. When participating in large financings, institutional investors also spread funds across different regions, technologies, and project stages.
As for data center financing, Nuzum stressed that in some of the financing arrangements he described, investors need to focus on whether the companies that have already signed contracts to purchase computing power have the ability to perform and can support repayment of the related debt. Changes in interest rates will also affect this judgment. A company may originally be able to pay interest, but if the cost of the next round of financing rises significantly, the original financial arrangement may be difficult to sustain. Therefore, investors should not only watch whether the Federal Reserve raises or cuts rates next, but also look at how medium- and long-term rates affect corporate financing and whether companies have the ability to bear new interest costs.
Regarding investment opportunities in large IPOs such as Anthropic, Nuzum believes it is not possible to judge all candidate companies together; investors also need to look at different businesses within a company. Some businesses are already profitable and can generate cash, while others are still spending money continuously, and whether their technology will succeed is more uncertain.
"The latter may have a more attractive future growth story, thereby driving the overall valuation. Therefore, analysts need to look beyond the company's overall label and analyze the fundamentals of each business."
He pointed out that another benefit of these AI companies going public is that the shares held by employees and existing shareholders have a public market price and are easier to trade. Companies can also use such shares as an employee incentive tool.
As for which indicators investors should ultimately focus on after listing, Nuzum's answer is still cash flow. "Whether it is equity, debt, real estate, infrastructure, or private credit, investing involves paying money today and expecting to receive a return in the future. We can discuss dividends and capital structure, but in the end it still comes down to cash flow."
Comments