Digital Infrastructure May Last Decades While The Economics Are Re-negotiated Much Sooner

Photo by Scott Rodgerson on Unsplash

Digital infrastructure can remain operational for decades while its underlying economics change at a much faster pace. Customer contracts get renewed, power gets repriced, technology requirements shift, and new capital injection is needed to keep the facility competitive, all while the physical asset keeps running.

A mobile network can carry far more traffic without corresponding revenue growth. An IRU can remain legally enforceable long after its original use case has weakened. A power agreement can expire just as prevailing rates spike. Each of these moments is a fresh negotiation, and each one carries its own execution and market risk. The building survives the reset but the value doesn't necessarily stay with the owner. It can move to customers, suppliers, competitors, or the next buyer.

That is the paradox of long-lived digital infrastructure. The longer the asset stays in service, the more times its economics have to be won again.

Physical Durability Does Not Preserve Bargaining Power

Digital infrastructure investing leans heavily on physical longevity, recurring demand, and high barriers to entry. But the durability of the asset doesn't necessarily mean that its owner has a sustained claim on the value it creates.

Telecommunications is a good example. Mobile networks have become more essential as data consumption has grown, yet more traffic hasn't translated into proportionately higher operator revenue. Each generation of network technology demands fresh investment, yet customers rarely pay materially more for the upgrade.

While networks continue to enable value creation, bargaining power may have shifted, and the value can be captured elsewhere by device manufacturers, cloud providers, applications and digital platforms.

The distinction is important because utilisation only proves the infrastructure is still relevant. Whether it translates to stronger economics for its owner is debatable.

Technological Change Can Reallocate Value Without Stranding the Asset

Data centres illustrate a different version of the same reset.

The sector was already growing through cloud adoption before the current AI cycle. AI has accelerated demand while changing the underlying design of the data centre. Power availability, rack density, cooling capacity and room to expand now matter more than they used to. A conventional data centre can stay fully leased and profitable while new facilities, built specificially for these requirements, pull ahead on pricing premiums and capital access. Older ones need to either invest heavily to keep up, settle for less demanding workloads, or accept that their role in the market has changed.

There is also a layer of uncertainty ahead. The current AI infrastructure buildout is frontloaded against demand projections that haven't yet been proven. There is real current demand. But whether it persists at renewal depends on the actual returns those workloads generate, and whether customers are still willing to pay on the same terms once that becomes clear.

Contracts Reduce Risk. They Don't Eliminate It

Long-term contracts are an important feature of infrastructure investing. They can improve revenue visibility, mitigate key risks, and support financing. But they can only mitigate foreseeable risks for defined periods, and can be a double-edged sword.

An IRU gives a customer a long-duration right to use specified fibre or network capacity. That right can remain valid even after the route becomes less important, or the utilisation did not scale as projected. In this case, the IRU seller may have already captured most of the value through the original deal, while the buyer carried the risk of revenue non-materialisation.

A PPA works the same way by offering price or supply certainty for its term. However, that term does not match the operating life of the data centre it powers. Once the contract lapses, power has to be secured again at prevailing market rates. If costs have risen and there is no avenue for the operator to pass the increase to customers, margins get compressed. If the increase gets passed through, the competitiveness and customer economics will shift. It boils down to bargaining power, alternative supply, and demand at the point of renewal, rather than the conditions of the preceding contract.

Selling Early Does Not Eliminate Risk

Holding periods add a further layer of timing risk and cost of capital. An owner can plan an exit before a major customer renewal, power repricing or technology upgrade. However, that only shifts the risk but doesn't eliminate it. The next buyer would have already priced in the reset.

If the remaining contractual protection is short, more capital is clearly needed, or the asset's competitive position looks uncertain, it shows up in the exit valuation before it ever reaches the accounting books.

A shorter holding period doesn't eliminate exposure. It just moves it from operating performance into terminal value. The seller doesn't experience the reset directly, but someone else's willingness to underwrite it is baked into the price they offer.

The Asset Does Not Own Its Economics

Digital infrastructure may remain useful for decades, but the value it creates is repeatedly redistributed through contract renewals, market pricing, technology cycles, reinvestment and bargaining power.

The owner may retain that value through scarce capacity, strong customer relationships and successful reinvestment. It may also shift to customers through lower prices, suppliers through higher input costs, competitors with better-suited assets, or to the next owner through the price paid at exit.

None of this is unfamiliar to investment banks and funds that are underwriting these investments. Forecast periods, terminal values and sensitivity analysis exist precisely to manage this kind of reset risks. What they don't fully solve is the confidence placed in the assumptions on either side of that reset, particularly when a clean forecast horizon quietly depends on what the next buyer is expected to believe about everything beyond it.

Some investments will still command a premium for reasons extending beyond standalone cash flows, including portfolio exposure, strategic access, platform value or the opportunities they enable elsewhere. These considerations introduce additional dimensions to what the asset is worth to a particular buyer.

The physical asset may last decades but its economics will be contested and re-established many times. The party that captures the value it creates is dependent on prevailing conditions at each reset.