Palistar explains the capital challenges in digital

This article is sponsored by Palistar Capital

What is the current opportunity set in digital infrastructure?

It is one of those strange moments where almost every one of the fundamentals in the sector feels favourable. Data is growing at accelerated rates, demand for artificial intelligence is increasing and new technologies are connecting dots in a way not seen before.

Over the past five years, two very large events have changed the dynamics and accelerated demand exponentially. One was covid, which accelerated the digitalisation and virtualisation of work and life in general, and the second was AI becoming a more mainstream technology. That has unleashed people’s imagination of what can be done and is triggering a significant spend phase in compute, connectivity and ultimately digital infrastructure.

The investment windows are moving fast. Last year, we spent a lot of time in debt and structured investments because of where interest rates were and how fast they got there. Now some of those opportunities have closed up as companies start to adjust to a higher interest rate reality, but these sectors need a lot of capital fast, so there is a huge equity opportunity and a limited number of providers that can deliver it.

But it is important not to get caught up in the hype, as with high growth come high value expectations and a new set of challenges. Now is the time to be tactical and careful to deliver outsized uncorrelated returns.

The market is adjusting but people still have a hard time lowering their valuation expectations to account for more expensive and less available capital, particularly when they look at highly positive market projections. So, investors need to come up with creative structures to keep the seller’s valuation dream alive but protect their downside. Having said that, just like in any market, high-performing companies will still get high values.

Why has the focus in some investments shifted to the point where getting clients is key? How can managers navigate that?

Infrastructure is a world where historically you did not have to worry about clients, because if you built an electricity network, you knew clients were just there. We had those build-it-and-they-will-come models because clients had no other option. That worked for those businesses, but as the infrastructure definition has expanded, business models have had to be adjusted, particularly in areas where people have other options.

That strategy can work but you need to be focused on delivering for the client. You need to make sure the other options cannot compete. Transferring a client to a new broadband connection is significant; you have to be on the front foot to convince them to move out of their comfort zone.

In a market with cheap and easy capital, you can afford to wait; when capital gets more expensive, you can’t. Because of that, we have seen what were originally wholesale models having to be creative, pivot, and literally knock on doors to get people to sign up.

Managers must never underestimate people’s comfort zones and the competition to get clients. Never underestimate how hard that is in competitive markets.

How do you see the deployment opportunity in data centres evolving over time?

We think about the deployment opportunities in terms of phases. The first phase, which we are in right now, is taking advantage of capacity in the system.

Availability of power is one of the biggest threats to the growth of AI, and we are living on a power system that was built for something else. We are currently taking advantage of whatever spare capacity there is in the system but that will soon run out.

In phase two, once that capacity runs out, we will need to think about how we adjust data centres for the power systems we have available. For example, right now, data centres run 99.9 percent of the time, but we can’t add electricity generation capacity at the speed we want to add data centres.

Some people may consider structures where you do not need to run 99.9 percent of the time – at times of peak energy consumption, when everyone is demanding power, you might turn off the data centre or prioritise those data centres you really need. Years ago, we used to do that with aluminium smelters and other large power consumers, and that may wind up being necessary with data centres, too.

Then, when we reach phase three, power generation will be built specifically for data centres. Some of those contracts are starting to get signed, but they will take time to come online. So, until then, we will need to get smarter on usage and build to get to sustainable data centres.

It is not enough to say “I will sign a contract to buy power from a wind farm” to make a data centre “green”; the sun does not shine all the time and the wind does not always blow. You will need a mix of fuels/sources or storage to compensate for that. Data centre investors are going to be the ones backing these investments.

Digital infrastructure is not about investing in high-risk gold mines but in the picks and shovels, as it were, and power knowledge is becoming a key part of that tool kit. For the first time we are seeing far more co-operation between the teams that cover data centres, those that cover power and utilities, commodities and even industrial equipment that is unlocking some interesting opportunities. Understanding all these pieces to the puzzle is crucial.

With the rise of mega-deals, are we seeing assets getting too big to sell? How might that play out?

The reality is yes. You are seeing a whole new type of exit strategy that was not relevant before. Five or six years ago, you could find a lot of $1 billion to $4 billion companies that were growing and going on to become $15 billion to $20 billion listed companies. Most of those have now been taken private by the large funds.

If they do their job right, even before putting additional capital to work, those investors could potentially double their valuation in eight to 10 years. So, you will end up with a lot of $9 billion or $10 billion companies that need to exit.

The problem is that the strategics that were supposed to buy those companies have also been taken private, so they are not there and the assets are too big.

You are left with less traditional options: the IPO option is a slow partial exit, which is not always ideal, particularly when the market knows you need to sell down; there are continuation vehicles, but there you have to convince existing investors, and find some new investors, while answering difficult questions about price and liquidity; you can flip from one fund to the next, but again you need a third party to come in and put a price on the asset; or, lastly, you can chop up some of these assets into smaller more digestible parts or different types of risk (yieldcos and devcos as an example).

The traditional sale of an asset to others is now becoming the exception for the bigger deals. As a response to this, business models for these companies are also evolving.

Five or six years ago, investment in companies had the objective of holding assets for 20 years. Now, some of those models have migrated to investing in companies that recycle capital, so, for the company to have enough money and not become too big, they need to sell assets as quickly as they build them and deploy the capital in new ones.

As an asset manager, you have to consider a whole different set of risks, because you are still a long-term holder of the platform, but the platform may be a shorter-term holder of the assets and more subject to market risks.

Right now, the biggest challenge for some is that the assets are infrastructure, but the business models have become flip models. We saw that in renewables a decade ago, with the fragmentation of business models creating different risk profiles, and we are watching it happen in digital infrastructure now. Being mindful of short-term assets hiding in long-term vehicles should be top of mind.

With access to capital proving challenging, what do the most appropriate business models look like?

The most appropriate business models have not changed, but the timeline for delivery has. We are still looking at downside-protected business models with unique moats and hard assets that are predictable and – ideally – offer some upside via growth.

What has changed is that, before, people were buying into the “wave”, not looking at profitability or cashflow in the short/medium term; they just needed a light at the end of the tunnel and did not care how long the tunnel was.

We stayed away from those opportunities because that did not fit with our investment principles, and now see those companies coming back saying they need more capital to get them through the tunnel/break even, but that is a difficult proposition if you haven’t focused on it so far. The UK is a perfect example of that, as are other fibre markets.

So, the market felt a little bit toppy for a while and we saw people focused on deploy, deploy, deploy. That has now changed, particularly as capital has become scarce. Investors and banks need to see businesses that are or can quickly get to profitability and that generate money, otherwise they are not going to fund them. We see that as a fundamental shift in the market environment.