Fluence Energy, Inc. (FLNC)

Priced as of 2026-09-11 — market data goes stale within days; treat this as a starting point, not a current quote.

Fluence Energy trades at $9.98 (September 11, 2026). This Sum-of-the-Parts (SOTP — adding up a separate fair value for each business line instead of valuing the whole company at once) analysis puts central fair value at **$21.82 per share**, a gap of **+119%**. More importantly, the price sits **below the conservative end of the fair-value band ($11.20)** — meaning the stock is cheap even when every discretionary choice in this model is set to its most pessimistic defensible setting. Under the Decision Framework in `valuation_rules.md`, that is the strongest signal this methodology can produce, and it is stated prominently here for that reason. The reason is simple and checkable. Fluence trades at 0.73x its revenue (EV divided by trailing twelve-month revenue). Every named comparable company in this file trades higher: Array Technologies at 0.97x, Sungrow at 1.94x, Itron at 2.16x, Shoals at 2.46x, Stem at 2.55x. Fluence is the cheapest name in its own peer set on revenue — cheaper even than Array Technologies, a company whose revenue fell about 38% year-over-year (YoY) and which carries a distressed balance-sheet score. There is an obvious reason the market is doing this, and this valuation does not pretend otherwise: **Fluence has no profits.** Trailing twelve-month (TTM) Earnings Before Interest, Taxes, Depreciation, and Amortization (EBITDA) is −$77.97 million (stockanalysis.com, September 11, 2026), so EV/EBITDA is not computable. Every comparable company in the set is profitable. A revenue multiple applied to a loss-making business quietly assumes a profit-margin recovery that the comps already have and Fluence does not. That single issue is the biggest weakness in this analysis, it is why the harvest-scenario comp range could not be narrowed (see the Rule A cross-check below), and it is one of two reasons conviction is capped at Moderate rather than High. The other reason for the cap: the hardware segment is about 90.5% of Total Fair Enterprise Value, and it rests on both a Question Mark probability weighting and an unresolved Stage 4 market-share conflict. `valuation_rules.md` requires a Moderate cap in exactly that situation.

Fair-value band

Rather than a single point estimate, this valuation reports a range: every discretionary judgment call set to its defensible low end, its midpoint, and its high end.

Price sits below the conservative fair value — cheap even on pessimistic assumptions. This is the strongest signal this methodology produces.

Portfolio at a glance

Key assumptions & swing factors

• Question Mark probability weight is the 50/50 default. Deviation requires a quantified signal and the quantified signals conflict: Fluence's FY2025 shipment volumes were flat while the global BESS market grew ~50% by GWh (share loss), but Q3 FY2026 order intake was $1.44B versus $508.8M a year earlier and backlog hit a record $6.4B, up over 30% YoY (share gain in bookings). No net direction, so no licensed deviation.
• Rule B: the Stage 7 Job Audit's 'Blocked' verdict on the core hardware contract job was used ONCE — to select the harvest-scenario comp set (Array Technologies and Fluor, both explicitly sub-scale or low-margin). It was NOT also used to reduce the probability weight. The prior August 2026 valuation used it in both places, which is the single largest reason the answer changed.
• Rule A cross-check FAILED on the harvest comp range. Fluor at 0.27x EV/Revenue and Array at 0.97x differ 3.6x on revenue but sit within 1.4 turns of each other on EV/EBITDA (8.26x vs 9.65x) — the whole spread is a profit-margin artifact (3.25% vs 10.0%). The 0.62x midpoint is only coherent at an assumed ~6.9% steady-state EBITDA margin, which no source discloses for Fluence (TTM EBITDA margin is −3.0%). The range could not be narrowed on defensible grounds, so per Rule A step 3 the midpoint is used but labelled unresolved and conviction is capped. Moving this multiple alone spans $19.34 to $24.30 per share.
• Rule A cross-check PASSED on the invest-and-win range. The 2.20x midpoint of Sungrow 1.94x and Shoals 2.46x implies 12.1x–12.9x EV/EBITDA at those comps' 17–18% margins, inside the comp set's own 10.58x–14.39x range. It does assume Fluence eventually reaches a Sungrow-like margin, which it does not have today.
• Rule C: the base is TTM revenue of $2.63B (September 11, 2026), 16% above the FY2025 figure of $2.26B the prior run used. Fluence cut FY2026 guidance on August 6, 2026 to $2.9–3.1B revenue and −$30.0M to +$10.0M adjusted EBITDA; that cut is reflected, and the $3.0B guidance midpoint is used only in the generous case because the comp multiples are trailing.
• Market-implied probability, solved backward from the price: at $9.98 the market implies roughly a 0% chance of the invest-and-win outcome on this file's central scenario values (about 20% on the conservative scenario values). Using 50/50 is an explicit claim that the market is wrong, argued on the $6.4B contracted backlog and near-tripled order intake versus a delivery-timing miss the company attributes to factory ramp, not demand.
• The SOTP rests entirely on EV/Revenue because Fluence has no EBITDA (TTM −$77.97M) while all four hardware comps are profitable. This is the most likely single explanation for the 54% gap between Total Fair EV ($4.08B) and market EV ($1.90B), and this method does not resolve it.
• Stage 4 borderline call inherited and handled in the band, not re-resolved: ~4% global share by GWh shipped and rank #8 (Benchmark Minerals, 2026) versus a top-3 rank on installed-and-contracted capacity (S&P Global Commodity Insights via Fluence IR, 2026) whose denominator cannot be verified. The conservative case ($11.20) corresponds to the #8-by-GWh reading; the generous case ($36.01) approximates the top-3 reading, under which this SBU would be a Star rather than a Question Mark.
• Digital (Fluence IQ) is an explicit placeholder at 9.5% of Total Fair Enterprise Value — a two-comp sector midpoint applied to ARR of $164.0M (the Rule A midpoint of the $148.0M FY2025 actual and the $180.0M FY2026 target reaffirmed August 6, 2026). ARR is not GAAP revenue; Digital's GAAP revenue is only about $6–8M. It carries a wide $0.32B–$0.46B range into the band by design.
• Conviction is capped at Moderate for two independent reasons: the largest SOTP component (hardware, 90.5% of Total Fair EV) rests on both a probability weighting and an unresolved Stage 4 borderline call; and Rule A step 3 applies because the harvest comp range could not be narrowed.
• Source conflict flagged rather than silently resolved: stockanalysis.com reports a +47.19% twelve-month price return while a separate aggregator reports −22.71%. The former is used as primary because it shares a source and date with every other market figure here and is consistent with a 52-week low of $6.60.
• Cross-file inconsistency flagged: Stage 8's synthesis JSON labels Digital a 'Question Mark' while Stage 4's more detailed file leaves it unclassified. This valuation follows Stage 4.

Portfolio-level tensions

• Fluence OS (embedded control software bundled into Hardware's sale) and the standalone Digital ARR business run on the same Fluence IQ technology base; leaning harder into bundling to prop up Hardware's Blocked position risks slowing standalone Digital sales to owners of competing hardware (Tesla, Sungrow, Powin) — the exact independence Digital's Live-bet case depends on.
• Hardware is nominally expected to fund Digital's build-out, but Hardware itself is margin-squeezed, not a stable Cash Cow — a fragile, reversed funding arrangement flagged in Stage 6.
• Both SBUs list the same 37-patent, 17-family intellectual-property base as a Key Resource without either canvas noting it is shared, risking overstated resource exclusivity in each SBU's coherence assessment.