Exclusive: NYC CIO Steven Meier Discusses Strategies and Sectors for AI Investments


Steven Meier is the chief investment officer for $278 billion New York City Retirement Systems, which includes New York City Employees' Retirement System (NYCERS), the Teachers' Retirement System for the City of New York (TRS), the New York City Police Pension Fund (POLICE), New York City Fire Pension Fund (FIRE), and the New York City Board of Education Retirement System (BERS). As fund managers and allocators gravitate towards investing in AI, Meier discussed the strategies and sectors he finds most attractive.

Watch the Interview

Muskan Arora: I wanted to understand the areas that you see promising or developing in the next five years in AI.


Steven Meier: Well, if you look at artificial intelligence, it's been around for a while. I'm told several decades, versus years, but I still consider it a somewhat nascent technology, which is rapidly developing. We look at this as being transformational technology, which will have broad implications for investors, the economy, and society.


In terms of what we think will be significant over the next 10 years in terms of, driving, gross domestic product growth, impacting inflation, impacting productivity, and ultimately impacting government spending and private investment. A number of those things include decarbonization demographic changes in the impact on the labor force and labor force dynamics, and debt sustainability in terms of sovereign debt levels that are out there that are very high at this point.


Also, I'd say, again, digitization of transformational technology as it applies to artificial intelligence, globalization, or nearshoring, or onshoring. And lastly, all this is wrapped in a geopolitical change dynamic... those are the themes that we think will drive the investment landscape over the next 10 years.


So, again, artificial intelligence, we believe that transformational technology will have broad implications, and it touches every aspect of our public and private holdings at this point, I'm happy to go into that in some detail.


Arora: Yeah. That sounds fascinating. Before we dive deeper into that, I wanted to also understand that while decision making, are there certain activities that AI would be best suited to help you with?


Meier: We look at artificial intelligence in terms of how it's impacting our portfolio with the opportunity sets, or both in public and private markets. We are also developing a use case in terms of how we can deploy artificial intelligence within our own offices to be more efficient, to use it for the support writing data analysis to improve our performance in terms of our quantitative system development.


It's a wonderful tool in terms of its pilot capabilities and in terms of helping with programming. So those are some of the low-hanging fruits that we see. But again, as I said earlier, Muskan, it’s still a nascent technology, and we think it has an enormous potential impact. There are a few different paths that the development, deployment, and use of artificial intelligence can take.


Those are broadly quite uncertain. But we are starting to develop the fact that there will be a significant impact on what we do, how we do it, and how our portfolios perform, frankly.


Arora: What approach are you taking when it comes to investing in AI? (Both investing in AI and using AI?)


Meier: When I think about investing in artificial intelligence, there are two aspects of it, directly investing in artificial intelligence-related companies, but perhaps more importantly, more broadly, the way it touches our portfolio. In the picks and shovels element, those activities, those asset classes, those sectors that are on the edge of artificial intelligence that are being significantly impacted.


We see several things that are being touched at this point in terms of the use of energy, and the impact on utilities, the impact on data center development, so it’s interesting: artificial intelligence uses about 10 times the amount of energy through its data centers to support artificial intelligence operation. So, we do think that's another area where we're investing through either real estate or through infrastructure in the private market sector. Strategies that can support the development of data centers, centers for various users as tenants, that we think will be very promising.


We also look at the fact that those data centers will need to use a lot of electricity. And certainly, that's a big part of, you know, leveraging the initiatives around clean energy. Also, midstream data in terms of midstream infrastructure, in terms of, say gas pipelines, and the like, that are considered not necessarily dirty technologies, but cleaner tech, fossil fuels, as we work through that transition to a clean economy to a clean energy source.

I'd say other things are software companies, both public and private, that are in our portfolios are dramatically impacting, are being impacted by the development of artificial intelligence and the deployment of artificial intelligence. So, it's really, it's  everything in the public sector. It's everything in the private sector as well, in different forms, direct and indirect. 


So, I did a quick query a little while ago, just in terms of our holding or aggregate holdings across the five plans in six companies, those being Google or Alphabet, both their AMC shares, Amazon, Apple, Meta platforms, Microsoft and Nvidia. And our current holdings, as of last night, were a little over $15 billion in assets just across those six stocks. And again, that's off a $278 billion asset base. We’ve seen some nice growth there... also quite substantial in terms of their size and impact on the market.


In the US, I'd say as a general matter, we're more of a passive investor, though, we do have about 25, or 30% of our assets in active strategies, we're a little bit more active in the developed market X-US and emerging markets. There we see significant opportunities to invest in companies also on an active or passive basis at the touch of artificial intelligence in different ways. I should also mention the data set and development is not just here in the United States, it's in Europe and Asia as well. And we're investing in those technologies through again, real estate and infrastructure development products.


Arora: Any regions that you could highlight in Europe and Asia, where you're looking at investments?

Meier: It's broad in terms of the demand, I mean, the demand for artificial intelligence technology, artificial intelligence-related data centers, and how that impacts the need for, you know, higher levels of energy, be it more fossil fuel related, or hopefully more clean energy or sustainable energy sources.


Arora: I think it's very similar market sentiment right now that everybody is looking to invest in data centers through real estate and infrastructure. I'd also like to know, why you think this is a good time to invest in AI considering other market variables.


Meier: So as a general matter, we're more strategic, we're not tactical, it's hard to be an institutional investor of our size and scale. We have a long-term strategic asset allocation. We're in these markets for years, we're active owners, even though we may have a passive investment through say, the S&P 500 index, we vote all our proxies in-house. So, we're consistently engaged. And again, we think broad exposure to those economic drivers over time will make a difference in terms of our portfolio returns, but if we're not tactical, we're not looking at valuations, which I believe are a little on the elevated side, particularly as it applies to artificial intelligence.


But again, the path of deployment is still somewhat uncertain in terms of how it will impact the underlying companies that use artificial intelligence to improve their productivity to reduce their cost structure. And as I said, all that is still playing out in the economy and the marketplace.


Arora: Which sector and sub-sectors are you focusing on in AI?


Meier: The sectors and sub-sectors are either passive index weights or through active strategies. We have significant investments in say, chip technology, software companies, and some of those other companies, such as Amazon, or Alphabet that leverages artificial intelligence for other reasons, to leverage their business and be more effective in delivering services to customers. So, again, it's direct and indirect. And as I said, you know, again, good old-fashioned picks and shovels, we invest in the infrastructure that supports artificial intelligence through again, a data center investment as well as clean energy.


Arora: What are the challenges you think that allocators will face when investing in AI?


Meier: Well, for us, it's straightforward as we're an allocator. But we're not a market timer, we believe that you get better results at our size and scale having time in the market as opposed to market timing. So again, we're consistent investors over time. And you know, again, we're not tactical, and we're not in and out of opportunities, by design.


Arora: Are there any emerging markets and emerging managers that you're focusing on with AI?


Meier: Yeah, not necessarily directly. So, we have emerging market equity positions, we invest in private equity, we invest in global infrastructure, we invest in global real estate opportunities that will have different elements that will touch either software or chip development, or data centers or sustainable energy infrastructure. But again, in our US domiciled fund, we have about 84% of our assets embedded in our investments and strategies. But again, 16% of our assets are abroad. Yeah, that's closer to $40+ billion in assets we have deployed elsewhere to take advantage of those opportunities.


Arora: With the geopolitical tensions in China and China being known for chipmaking. How, how are you handling that 16% bit of the portfolio?


Meier: Yeah, it's more Taiwan that's known for the chipmaking. There's a lot of infrastructure built up there. I know Nvidia has its chips made by a manufacturer that's physically located there. There has been, as I mentioned earlier, some significant movement to, let's say, shore up supply lines by bringing production closer to home. So those technologies mean even look at the states with the chipset, there's a lot of development of technology and infrastructure here in the United States that can ultimately support and feed the artificial intelligence initiatives, as well as a good and a stable source of energy, which I think is very, very important as well.


Arora: How are you assessing these opportunities in artificial intelligence, as you can perhaps, also talk about managers or consultants that are helping you out?


Meier: We have each of the five plans that we manage has a general consultant. So, we also have several specialty consultants, and specialized, typically and nontraditional alternative assets. Again, I mentioned private equity infrastructure, core non-core real estate, hedge funds, as well as private credit, we're big investors in private credit. And we do think there's a significant opportunity to invest in in-depth strategies that support the development of that critical infrastructure. So again, in terms of our use of consultants, it's not specifically, can help us invest in artificial intelligence. But a lot of the deals that are getting done now in software, somehow, in many ways, touch private equity. I believe this statistic is something like 98% of all software companies are private. We have those investments into private software companies, again, through a private equity initiative. And we do look at the performance drivers, the sector allocations, and the focus areas of those managers. And that's something that makes it attractive for us.


Arora: My next question was very similar about, how is AI interacting with private equity and private investments. So, could you expand a bit more on what you find attractive? And what do you think is going to happen in the next five years with these interactions, private equity, and AI?


Meier: it's a nascent technology; the paths of how it's going to be developed and deployed are uncertain. But I can say with humility and confidence that this is a transformational technology. It is changing the way we think, the way we function, and our opportunity set from an investment standpoint, it's picking up speed as well through the deployment, elements of artificial intelligence throughout our society and in the companies that we invest in. So, again, an uncertain outcome.


But its a technology, which I think has tremendous potential for positive impacts in terms of just looking at the impact of artificial intelligence on the broader global economy. We think it is going to contribute to GDP growth around the world, anywhere between 1% to 1.5% We think it'll have a deflationary 10 or disinflationary 10, meaning moving the level of inflation a little lower, and we think it's positive for productivity.

We also have to recognize that there will be some social dislocations at least early on, in terms of people that are displaced from their jobs, as associated with an increased use of artificial intelligence that lends itself to supporting an FTE producing more knowledge work. So, I do think this is more of will probably be more of a challenge for white-collar workers, to say that they can't be retooled to still be actively part of this, this expanding economy but in a different capacity.


Arora: Are there any sectors in private equity or any strategies in private equity that excite you the most about AI?


Meier: Buyout and growth, we certainly think that we'll see a consolidation in software companies, again, that occurs over time, as I mentioned, 90% of all software companies are private. There’s always a lot of activity in those areas. And you know, software is something that can typically have a significant amount of positive leverage, because the technology once it's developed, it's easier to deploy.


The margins are significant there. So again, I'd say just looking at our traditional private equity managers that look for strategic fit, to drive, process efficiencies, and those companies. But again, growth and buyouts seem to be the beneficiaries. I suspect you were also a very small investor in venture capital. But I suspect over time, that that will be a significant area of opportunity as well.


Arora: I am curious to know if you are keeping a separate amount just to allocate in AI. Essentially, how do you go about allocating to AI?


Meier: We're a traditional institutional investor, everything we do is professionally managed. Again, we don't time the market, we're a consistent investor, we adhere to a strategic asset allocation plan, and we deploy money, say, into the private markets, which we're in the process of increasing our exposure to over some time to make sure we get vintage year diversification.


We're not tactical, we're not in and out of the markets, and we're not specifically targeting any managers who are only somehow impacting artificial intelligence. But we have, as I said, significant public stock holdings as well as private equity holdings, in either chip companies or software companies that will benefit from or be involved in the development of technology.


All those adjacencies, those companies, and industries that are at the edge of artificial intelligence, somehow support the delivery of artificial intelligence benefits over time.


Arora: How are you accessing managers when it comes to investing in AI?


Meier: In every meeting, we have with external managers, I always ask about their approach to artificial intelligence, how is that impacting their business? How is it impacting their business in two ways? What's the impact on the opportunity set, as we move forward with the development deployment and use cases for artificial intelligence? And how are they using it internally, and some of that's a little self-serving, and that we're trying to come up with a strategic plan of how we can incorporate the use of artificial intelligence.


Again, protecting confidential information, making sure that nothing about our portfolios around in the general cloud is available for general access. We want to be careful about confidential information. We want to be respectful of intellectual property rights out there. We’re still going through that exercise so I can tell the investment professionals I have the privilege of working with here at the Bureau of Asset Management, using artificial intelligence in some way. A lot of it is simple chat GPT 4.0. In terms of how that works, a lot of that work is done on their private computers, just to make sure that we don't in any way jeopardize the confidentiality for information on confidential information that we have within our operating group.


Arora: Are there any red flags that you look for while accessing managers?


Meier: We don't use artificial intelligence as a specific screen, we do ask questions to see how the technology is being deployed and used. We're curious about the potential impact on our portfolio companies from an investment and return standpoint, and how this potentially impacts the risk associated with those operating companies.


We don't necessarily focus on specific use cases. But we ask everyone in terms of what they think the potential is for artificial intelligence, and what's the path over the next three to five years, I do think that the world will look different in five years than it does today. And I do think that the deployment of artificial intelligence will certainly surpass and exceed what we saw in terms of the use of the Internet.


And obviously, we had a little bit of a questionable track record as an industry. And as a marketplace back in 2000-2001, where things had gotten to such an excessive level in terms of just a bubble in internet-related companies. The difference today is that these companies pick Nvidia in particular, they are a tremendously profitable company generating just enormous levels of free cash flow that can be reinvested in the business or distributed over time. So those are real companies with real earnings. And the growth trajectory has been just incredible.


So, we're optimistic, also recognizing that the valuations have probably gotten to relatively high levels, I think perhaps they were a little higher last year than they are today. But certainly, something to watch. That's necessary to act on, we're not going to disinvest or divest from any artificial intelligence, or chip companies, because we think that the valuations are too high, but I think it's not unreasonable to expect some level of volatility over time.


I also think as a general matter, musket, we're going to see the level of right now, the real benefits of artificial intelligence, in terms of from an investment and stock perspective, on a direct basis are concentrated in a handful of companies. We talked about a few of them earlier, I do think you'll see a broadening of the positive impacts of artificial intelligence throughout the market over time. We haven't seen that yet. But I do think it's reasonable to expect. And that will come over time.


Arora: How do you think climate, technology, and AI will meet? And what do you think that it's going to happen in that sector?


Meier: I think that's a wonderful question. So well, maybe look at it this way, from a sustainability, climate sustainability issue. We need just enormous amounts of energy to be able to support the functioning of artificial intelligence at the data center. So, we think that's a positive impact that there'll be more hopefully more focus on, well, probably traditional, but also nontraditional sources of energy, which we think will be positive.

Chip generation requires a lot, you know, clean rooms, cool, cold rooms, a lot of water, as you said. And I do think the current administration, Joe Biden's administration has done a very good job in terms of the Chips Act, the Infrastructure Act, and the Inflation Reduction Act in terms of putting more money into our local economy, to develop those manufacturing industrial base that we think is important, and again, to mitigate the risks of supply chain disruptions down the road.


Perhaps there are some national security issues associated with those, those chat changes as well, which we view as positive. So that's the interaction now in terms of how we think we can use our artificial intelligence to improve our performance around sustainability issues. I had a meeting earlier today with an investment manager, a world-class investment manager who showed me a quick example in her life: how they were able to use artificial intelligence to develop a prototype of a hotel, located outside the United States that was energy efficient in terms of how many windows it needed, and what level of heating and ventilation was required.


Artificial intelligence for very little money can generate a complete analysis in terms of how you build sustainable infrastructure, what the cost is, the things you need in terms of watering, so it will automate that whole process and perhaps, speed the process of development. And this was just an example of how their architects use our artificial intelligence, again, to improve the delivery, of building properties out there, hotels, office space, multifamily housing, etc.


Arora: Before you leave, I’d love to ask you about that one thing that keeps you up at night.


Meier: I'd say my only concern would be the missed opportunity. There are probably two ways to answer that to make sure we're getting good relative value as we continue to deploy capital into those areas, again, not just as a public equity holder and a big investment and say, passive and active strategies. When we deploy money into private markets, are we taking advantage of the opportunity from, again, picks and shovels, infrastructure, and sustainable energy initiatives.


I don't lose a lot of sleep, because things I would lose sleep over are probably out of my control, as, as an investor as an institutional investor. But again,  I think we've done a wonderful job in terms of building all-weather portfolios that will, hopefully, perform consistently well, over a long period. One last point I would make is that we are long-term investors. I'm not looking at the next quarter, or the next year, or even the next five years, we're thinking about what the dynamics in the marketplace are. And what are the opportunities going to drive our investment performance in the next ten, fifteen, or twenty years? So really, to try to match the time horizon for investments to the liabilities. So, the promises we have made, collectively to the pension plan beneficiaries and participants.