- More than 40% of US insurers surveyed by Moody’s Ratings at the end of last year said they intend to increase exposure to private debt. Moody’s estimates private credit already made up 35% of life insurers investments and 8% of property and casualty insurers investments at that point.
- Life insurers rated by Moody’s hold roughly $8 billion in data center-related assets and property and casualty insurers about $2 billion. Moody’s separately estimates the entire US insurance sector carries as much as $20 billion of data center exposure, roughly double what its rated universe accounts for.
- Energy projects tied to the AI buildout are the far larger position. Life insurers hold about $26 billion in related energy assets and property and casualty insurers about $1 billion, meaning life insurer exposure to AI-adjacent power is more than three times their data center exposure.
- The concentration sits against liabilities that must be paid on schedule regardless of market conditions. Private credit is illiquid and valued infrequently, so a third of the life insurance investment base now sits in assets that cannot be sold quickly or priced continuously.
What Happened?
Moody’s published survey results on Friday showing that the shift of insurance capital into private credit is set to continue rather than plateau. The survey covered both the scale of existing private debt holdings and, separately, exposure to assets tied to artificial intelligence, breaking those into data center holdings and related energy projects across life and property and casualty insurers.
Why It Matters?
The energy figure is the one that reframes the AI exposure question. Insurers have put roughly $26 billion into AI-related energy projects against $8 billion in data centers, so the industry is financing the power supply rather than the compute. That is a rational preference, since generation assets have long-lived contracted cash flows that match insurance liabilities, while data centers carry technology obsolescence risk. It also means insurance capital is concentrated in the part of the AI buildout that is hardest to relocate and slowest to repurpose, and a power plant contracted to a facility that never reaches full utilisation has limited alternative buyers. The two data center numbers do not reconcile cleanly either: $8 billion and $2 billion across Moody’s rated universe against an estimate of up to $20 billion for the sector overall implies roughly half the exposure sits with insurers Moody’s does not rate, which is the segment with the least visible disclosure. The broader concentration deserves direct treatment. Private credit at 35% of life insurer investments is a large allocation to instruments that are illiquid and infrequently marked, held by institutions with guaranteed obligations to policyholders. That combination works while credit performs and becomes difficult if defaults rise, because the assets cannot be sold quickly and their carrying values have not been tested by continuous pricing. More than 40% intending to add exposure, at a point when the Federal Reserve has resumed raising rates, means the industry is increasing that position into a tightening cycle rather than trimming it.
What Next?
Watch state insurance regulators, particularly the National Association of Insurance Commissioners, for movement on capital treatment and valuation standards for private credit holdings, since a 35% allocation is large enough to attract supervisory attention and any change to risk-based capital charges would force reallocation. Moody’s next survey will show whether the stated intention to increase exposure survived the rate move, and the gap between intention and execution is the useful measure. On the AI side, track whether energy project exposure keeps growing faster than data center exposure, because that ratio indicates whether insurers are becoming more or less confident in compute demand itself. The clearest early warning would be any impairment or restructuring in a large AI-linked energy financing, which would test valuations across the whole category at once. Disclosure of exposure at non-rated insurers is the visibility gap worth pressing on.
Source: The Wall Street Journal
