NRG NRG Energy
A cyclical merchant power and retail business on a tiny book; residual income can't capture its buyback-inflated ROE.
The story
NRG is not a regulated utility. It is a competitive generator and retailer (Reliant, Direct Energy, Vivint Smart Home) that has added LS Power's gas fleet to capture rising data-center power demand in ERCOT and PJM. Book equity is only about $1.03B because years of aggressive buybacks and acquisition accounting have hollowed it out, so the trailing 82% ROE mostly reflects a shrunken denominator, not a durable franchise return. Earnings are tied to merchant power prices and capacity auctions, which are currently near a cyclical peak, and leverage has risen after the acquisition.
I cut ROE from the 82% trailing figure to 40%. That assumes power spreads and capacity prices fall back from peak while the buyback-depleted book stays small, so measured ROE stays structurally high. Book growth is modest because most free cash flow goes to buybacks rather than retained equity. Beta is 1.05, well above the 0.55 regulated-utility anchor, because NRG carries merchant commodity and retail-margin risk.
Value drivers
| Return on equity (normalized) | 25.0% |
| Book-value growth (Y1) | 8.0% |
| Terminal book growth | 3.0% |
| Beta | 1.05 |
| Failure probability | 3.0% |
| Cost of equity | 9.8% |
Valuation bridge
| PV of excess returns | 782.33M |
| PV of terminal excess | 303.69M |
| Equity value | 2.05B |
| ÷ shares → per share | $9.77 |
Projected excess returns on equity
| Yr | Book equity | ROE | Excess return | PV |
|---|---|---|---|---|
| 1 | 1.03B | 25.0% | 156.57M | 142.58M |
| 2 | 1.11B | 23.6% | 154.02M | 127.72M |
| 3 | 1.20B | 22.3% | 149.28M | 112.73M |
| 4 | 1.28B | 20.9% | 142.25M | 97.82M |
| 5 | 1.36B | 19.6% | 132.85M | 83.19M |
| 6 | 1.44B | 18.2% | 121.05M | 69.03M |
| 7 | 1.51B | 16.9% | 106.88M | 55.50M |
| 8 | 1.58B | 15.5% | 90.42M | 42.76M |
| 9 | 1.65B | 14.2% | 71.81M | 30.92M |
| 10 | 1.71B | 12.8% | 51.24M | 20.09M |
Key risks
- Power prices and PJM/ERCOT capacity prices fall from current highs, compressing generation margins
- High leverage after the LS Power deal combined with a thin equity cushion; a Winter Storm Uri-type event could wipe out book
- Data-center load growth arrives slower than priced in, or regulators intervene in capacity markets
- Residual income on a buyback-depleted book understates economic value, so model output is unreliable in either direction
Catalysts
- Signed long-term data-center power contracts that lock in premium pricing
- PJM capacity auction results and ERCOT summer scarcity pricing
- Deleveraging toward the target ratio and continued buybacks
⚠ Extreme gap to market price — large, well-covered names are rarely mispriced this much; likely embeds risk the model underweights. A flag, not a verdict.