A Reverse Valuation Analysis on FRMI, and the reason I am buying after selling in June
For context, I bought mid May, at around $5.9 a share, and sold $9 a share, because I believed it had entered a reasonable zone. Recent news have encouraged me to revisit the stock, and I have decided to buy back in.
The following analysis focuses on breaking down the stocks price, using 4 specific scenarios, forecasting their cashflows, and using a normal distribution to figure out what probabilities the market is assigning each situation.
My goal is to help us understand what outcomes exist and the odds at different price levels, not to really predict a perfect valuation.
1.TensorWave Breaks Down (Worse case scenario)
FRMI fails to convert the initial TensorWave agreement into a functioning operating campus due to delays, renegotiation, financing constraints, or customer failure. Existing land, generation assets, infrastructure, permits, and equipment retain some value, but the ~$1.55B of PP&E faces substantial impairment/restructuring risk. Long-term FCF settles around $0–25M.
Valuation: $0.75/share
2.Small Developer / Limited Survival
TensorWave partially materializes, but FRMI cannot scale Matador meaningfully. The company retains some infrastructure and power assets but faces heavy financing requirements and potential dilution, ultimately operating as a small specialized infrastructure developer. Normalized FCF of ~$50–75M/year, valued at roughly 10–12× FCF after accounting for capital structure/dilution.
Valuation: $2.50/share
3.222 MW Works, Then Stagnates
FRMI successfully delivers the initial 222 MW TensorWave campus, realizing the economics of the ~$6.5B / 15-year contract, but TensorWave does not expand materially and no second major customer emerges. The ~$433M/year headline contracted revenue includes turnkey development/construction economics, so normalized FCF is modeled at ~$250–300M/year after operating costs, corporate costs and recurring capex.
Valuation: $5.00/share
4.TensorWave Reaches ~650 MW
All three TensorWave phases are successfully developed, bringing the relationship toward ~650 MW. FRMI establishes a legitimate AI infrastructure campus but does not meaningfully expand beyond TensorWave. This implies roughly $1.2–1.3B of annualized contractual revenue equivalent, with normalized FCF of approximately $450–500M after depreciation, maintenance capex and financing costs.
Valuation: $8.00/share
5.Matador Reaches ~1–2 GW
TensorWave reaches ~650 MW and FRMI secures at least one additional large customer, bringing total contracted/operating capacity to roughly 1.2–1.6 GW. FRMI becomes a scaled AI infrastructure operator rather than a single-customer project. Normalized FCF reaches approximately $700M–$1.0B, after accounting for additional capital requirements and financing/dilution.
Valuation: $14.00/share
- Matador Becomes a 2–3+ GW Platform
FRMI successfully executes multiple large customer contracts and scales Matador to roughly 2–3+ GW of contracted/operating capacity. Customer validation and successful execution improve financing access, allowing more development to occur at the project level and reducing dependence on corporate equity. Normalized FCF reaches approximately $1.5–2.0B.
Valuation: $28.00/share
7.FRMI Becomes a Scaled AI Infrastructure Platform (Blue Sky situation)
Matador scales to roughly 5–7 GW with multiple hyperscaler/AI customers, while FRMI proves the model is repeatable beyond the initial TensorWave relationship. Project-level financing becomes increasingly available, allowing FRMI to develop substantially more capacity without proportional corporate dilution. FCF ramps from -$600M in 2027, -$400M in 2028, +$500M in 2029, +$2.0B in 2030, to ~$3.5B in 2031 and beyond. Long-term growth is modeled at ~3%.
Valuation: $55.00/share
Market-Implied Distribution at $6.40/share, normalized centrally
TensorWave breaks down → $0.75 → 25.1%
Small developer / limited survival -> $2.50 -> 22.6%
222 MW works, then stagnates -> $5.00 -> 19.4%
TensorWave reaches ~650 MW -> $8.00 -> 16.1%
Matador reaches ~1–2 GW -> $14.00 -> 11.2%
Matador becomes a 2–3+ GW platform -> $28.00 -> 4.7%
FRMI becomes a scaled AI infrastructure platform -> $55.00 -> 0.9%
Value of FRMI depends on what you think these probabilities fall out to. Personally, I believe the market is being far overly pessimistic at this level, and I am back in.