Where are the opportunities beyond AI?
Digitization, particularly the rise of AI, has emerged as a central theme across infrastructure and real assets. While opportunities persist in this domain, valuations remain elevated, with digital assets exhibiting a median EV/EBITDA multiple of 11.7x, compared to 10.2x for the broader infrastructure universe.17 We believe data assets developed without pre-existing tenant contracts face significant adoption risk.
After an era of flat power demand, we now see growing demand for energy solutions to power AI and digitization initiatives. In many parts of the world, the current power grid baseload capacity is insufficient to support projected demand growth, and assets themselves are aging. In the US, assets in the power grid infrastructure average 40 years old—making for a structural mismatch to the rapidly-advancing technologies underpinning AI.18 In Europe, potential data center power demand makes up around 90% of the EU28’s power demand; even 20% conversion of the potential to actual demand would represent a material change from the declining energy consumption rates in Europe of the past 15 years.19 Estimates indicate $12 trillion in capital demand for energy transition by 2030.20 Demand is expected to continue growing even if power usage per AI server falls over time due to improved efficiencies.21
We believe this demand will need to be addressed through a wide variety of energy sources, both traditional and sustainable, with a focus on distributed generation solutions, reliable and dispatchable supply, and demand flexibility. In an environment of aging infrastructure and more instances of extreme weather events, investments in the power grids will be required to improve the grid’s resiliency; assets that improve energy efficiency, reducing consumption intensity will become a priority. The physical nature of these assets and the costs of bringing energy solutions online mean significant demand for capital solutions. In some cases, the capital required makes these assets the purview of value-add/opportunistic, rather than core strategies.
Beyond AI and data centers, we see compelling opportunities in various themes that may currently be overlooked and that may have more attractive supply/demand characteristics. The circular economy, encompassing waste, water, and recycling, involves contracted, essential services largely insulated from macroeconomic fluctuations. These opportunities are often found in the relatively undercapitalized middle market, contrasting with the mega-cap space where most investor capital is concentrated. Transportation and logistics are undergoing transformation as supply chains reorient. Countries globally are realigning trade routes to prioritize supply chain resilience and geopolitical alignment over mere cost efficiency. In the US, we have observed a drive for onshoring and increased domestic production that has led to a resurgence in rail demand and evolving requirements across seaports, airports, and storage. This trend is expected to continue, as we anticipate domestic manufacturing to accelerate over the next decade.
In Europe, the opportunity landscape is evolving, supported by broad policy initiatives and an intensified focus on energy independence. However, we believe investors must acknowledge significant differences in regulations, permitting, and underlying growth across economies. Some of Europe’s larger economies, such as France, are experiencing slower growth, while others, particularly in Southern Europe like Spain, are expanding more rapidly. Broad policy changes across Europe, including the Clean Industrial Deal signed in February, are expected to be beneficial. Nevertheless, country-specific policies are also crucial; Germany, for instance, is likely to offer numerous opportunities following a substantial infrastructure spending bill passed in September. Broadly, large-cap assets trade at a 13.1x EV/EBITDA multiple, whereas middle-market assets trade at 11.2x,22 making for—in our view—a more attractive opportunity set in the middle market.
1 MSCI Private Assets. As of 2Q 2025.
2 EY Prive Equity Pulse Survey conducted by AlphaSights. Data as of June 30, 2025.
3 Goldman Sachs Global Investment Research. As of October 8, 2025.
4 Goldman Sachs Global Banking and Markets. As of October 24, 2025.
5 Goldman Sachs Asset Management. This is based on an updated 2023 value creation framework considering current valuations and cost of capital. Assumes targeted returns of 2.5x gross TVPI (Total Value to Paid-In), or roughly 2.0x net. Historically, margin expansion contributed just under 10% of the average value-add for private equity-backed companies, implying a required revenue growth closer to 9–10% to achieve typical return levels targeted by investors private equity managers; this target surpasses historical experience (7–9%).EBITA refers to earnings before interest, taxes, depreciation, and amortization.
6 MSCI Private Assets, as of 2Q 2025.
7 Pitchbook, as of June 30, 2025
8 Silicon Valley Bank, State Of The Market, 1H 2025
9 S&P Global, 5-year average trailing recovery rates. As of September 2025
10 Lincoln International as of 4Q 2024.
11 Cliffwater Direct Lending Index, PitchBook LCD, data through March 25, 2025. Data include repricings and extensions done via amendment process only.
12 Goldman Sachs Global Investment Research. As of October 2025.
13 MSCI RCA. As of 3Q 2025.
14 MSCI RCA. As of October 2025.
15 MSCI RCA. As of 3Q 2025.
16 MSCI RCA. As of October 2025.
17 EDHEC. As of October 2025. Enterprise Value to Earnings Before Interest, Taxes, Depreciation, and Amortization.
18 Goldman Sachs, “Powering the AI Era.” July 2025.
19 Goldman Sachs Global Investment Research: ‘Powering Up Europe: AI, data centers and energy needs: at Europe’s inflection.” October 2025.
20 Goldman Sachs, “Powering the AI Era.” July 2025. Estimates based on Goldman Sachs Global Investment Research reports: “Carbonomics: the GS net zero carbon scenarios – a reality check”; “Generational growth: AI, data centers and the coming US power demand surge”; Power & utilities figure globalized based on a US estimate of ~$1.4T.
21 Goldman Sachs Global Investment Research: AI/Data Center Power Demand: The 6 Ps driving growth and constraints. October 2025.
22 EDHEC. As of October 2025.
