Navigating the six key issues affecting data centre investment and operation

Investment considerations

The investment thesis is asset specific

Data centres are a complex investment proposition. As with the discussion of funding models above, data centre investments may span a range of investment mandates depending on the structure of the transaction and the characterisation of the revenue profile and risks inherent in the chosen structure.

The investment thesis will vary with the lifecycle

Early-stage (or greenfield) developments typically appeal to sponsors and developers with a higher risk tolerance and sector expertise across a blend of traditional infrastructure and information and telecommunication developments. Legal risks at this stage often centre on land acquisition, environmental, planning and regulatory compliance, tenancy arrangements and customer contracts, construction and supply chain risks, as well as lender considerations and grid connection arrangements. Who these risks sit with will depend on the delivery model and whether these can be passed on to developers or to contractors.

By contract, operational assets (or established portfolios seeking to expand) tend to attract institutional investors, infrastructure funds and superannuation capital seeking stable cashflows and inelastic demand. These investors will usually seek stable input costs and revenue profiles via assets anchored by power purchase agreements, or captive generation and long term service agreements backed by quality credit. A premium for mature assets is currently being paid in the market to secure these relatively stable returns. Indeed, as we mention above at Section 1 (Energy considerations), data centre capacity demand is usually so stable that data centres are expected to be important for the stability of the grid until long-duration storage options mature.

Common investment considerations

Investors should be alert to the following high-level considerations:

  • Operational continuity, as well as a potential mismatch between the projected data consumption demand curve and the realised trajectory.
  • Ensuring that modelling assumptions are thoroughly tested, as well as ensuring there is a pass-through of rising energy costs and generation risk.
  • A rapidly deepening market and the speed of technological advancement means that identifying quality assets is becoming more difficult (eg assets are now at varied lifecycle stages and significant capex expenditure can be required to align them with market expectations).
  • Stranded capital risk:
    • the intangible nature of data means data centre investors face a relatively higher risk of customer demand relocating to another location or jurisdiction (as compared to other hard infrastructure assets like roads or electricity generation infrastructure)
    • the rapid pace of technological advancement can result in assets becoming obsolete or underutilised (for example, advancements in AI workloads or changes in data localisation laws or practices could render certain assets less competitive or non-compliant, leaving investors with sunk costs and limited recovery options).

Investors will place a premium on the resilience of a data centre's revenue model (ie one hyperscale client or multiple smaller clients), limiting financial exposure and maximising recovery options in the event of underutilisation.  

Looking ahead to overseas market trends, our colleagues in the UK are also seeing that lease lengths are extending, and capacity uptake has increased eightfold in the last five years, from 5MW in 2019 to 40MW in 2024.1 If this trend is replicated in Australia, this may somewhat mitigate the risks associated with demand mobility.

As the data centre sector matures, a trend has emerged of segregating operational assets from those under development. This allows investors and funders to take exposure to different stages of the data centre life cycle: from a high-risk, high-reward construction proposition to a stabilised, yield-producing product.

As we mention above at Section 3 (Funding models and trends), lenders operating in the sector are becoming more flexible and the finance package will be dictated by whether the investment proposition has typical development characteristics, or infrastructure-link inelasticity.

Considerations on disposal

The flexibility of data centres as an investment proposition is also evident in the key considerations for investors on disposal. These considerations are generally similar, whether the investor is exiting at the culmination of a real estate development, or is returning/recycling capital as part of an infrastructure fund.

Typically, we see the following common challenges and risks in the disposal phase of the investment cycle:

  • Identifying and attracting the buyer pool for a specialised asset class (although this risk seems somewhat alleviated by interest in the sector, for the time being).
  • The challenge of securing a renewed income stream (if required).
  • Regulatory/national security risk (FIRB, hosting approvals and security clearances to acquire critical infrastructure).
  • Significant capital expenditure risk and potentially short asset lifecycle, given rapidly evolving technology and demand for data supply.