Phase Space AI

Valuation

AST SpaceMobile [ASTS]

AST SpaceMobile, Inc. [ASTS] — Valuation

as of 2026-07-29 · spot $53.03 (SIP close) · two outputs, both required

---

1. Inputs — and the $2.8bn asset that does not exist

The primary correction on this name is not the valuation method, it is the balance sheet.

1.1 The tie-out that proves the $2.8bn is a double count

The 10-Q's liquidity section states the total independently:

"As of March 31, 2026, we had approximately $3,458.9 million of cash and cash equivalents and restricted cash on hand, including $429.3 million of restricted cash."

  3,029,591  cash and cash equivalents          (balance sheet)
        873  restricted cash, current           (balance sheet)
    428,400  restricted cash, non-current       (balance sheet)
  ---------
  3,458,864  =  the MD&A's "$3,458.9 million"      TIES EXACTLY

There is no room in that reconciliation for another $2.8bn. The $2.8bn is the portion of the $3,458.9m that is held in money-market funds and therefore classified as cash equivalents — a disclosure about the composition of the cash, tagged to ShortTermInvestments. The scanner added a description of the cash to the cash.

Root cause, for the calibration file: the net-cash builder adds any ShortTermInvestments tag without verifying it is a distinct balance-sheet line. On OUST and ONDS that tag is a distinct line and the treatment was correct; on ASTS it is not, and the same code produced a $2.8bn error. The fix is a cross-foot: gross cash and investments must tie to the cash-flow statement's ending cash plus separately-stated investments, or to the MD&A's stated liquidity total. Both were available here and both refute the number.

1.2 Net debt now, after two 8-Ks the screen never saw

  2,723,000  cash and cash equivalents and restricted cash at 2026-06-30
             (PRELIMINARY, unaudited, 8-K 2026-07-15 -- the company's own caveat is quoted below)
   (3,024,121) debt face at 2026-03-31
  -----------
    (301,121) net debt at 2026-06-30

    +983,600  net proceeds, $1.0bn of 1.625% Convertible Senior Notes due 2034 (8-K 2026-07-20)
     (96,900) paid for capped call transactions
  (1,000,000) new debt face
  -----------
    (414,421) NET DEBT, approximately, at the memo date

"As of June 30, 2026, total cash and cash equivalents and restricted cash was approximately $2,723 million. The Company's financial results as of and for the quarter ended June 30, 2026 are not yet complete… You should not place undue reliance on these preliminary estimates." — 8-K, 2026-07-15

Cash and restricted cash fell from $3,458.9m to ~$2,723.0m in one quarter — approximately $736m of consumption, against Q1-2026 capex of $261.6m and operating expenses of $164.1m. That is the number that matters more than any multiple on this page.

EV used throughout: 53.03 × 388.124572 + 414.4 = 20,584.3 + 414.4 = $20,998.7m.

---

2. Terminal margin and exit multiple are both UNIDENTIFIED

ASTS has never reported an operating profit, and it has never recognised one dollar of revenue from the product being valued. The company states it directly: "To date, the Company has not recognized any revenues from its SpaceMobile Service." Derived operating margin for Q1-2026 is −1,014%.

The screen assigned a 14.1% terminal margin as an "industry median of mature profitable peers (pre-profit subject)" and a 28.5x exit multiple described as GROWTH_MATCHED from 63 peers. Both are rejected.

• The terminal margin is the ADMA (+51pp) / INOD (+32pp) construction. Borrowing a satellite-industry or telecom median for a company with no product revenue asserts the answer.
• The exit multiple cannot be growth-matched, by construction. valuation.md requires a comparator set "whose growth brackets the subject's growth at the exit year". The subject's growth is undefined — its product revenue is zero, so no comparator's growth can bracket it. A 63-name "growth-matched" set for a company with no revenue in its intended business is definitionally impossible. Declared UNIDENTIFIED.

No exit anchor is shared with OUST or ONDS. Industrial lidar, Israeli counter-UAS and LEO direct-to-device satellite have nothing in common on the exit-multiple axis, and ASTS in particular has no margin history to anchor at. Each of the three names in this cluster is gridded independently.

Declared: terminal EBIT margin UNIDENTIFIED and exit multiple UNIDENTIFIED. What follows is a grid, and over much of it there is no solution at all.

---

3. The implied-path test — and the no-solution finding

Method as assets/reverse_dcf.py. Solved for the revenue CAGR, holding revenue at the verified TTM of $84.935m, the horizon at 5 years, the discount rate at 13.0%, and the EV at $20,998.7m. Solver bounds on the CAGR are the standard [−50%, +200%]; bounds on the multiple are [0.1x, 200x].

3.1 Required five-year revenue CAGR

Nine of the thirty cells return no solution. "NO SOLUTION" here means precisely what assets/reverse_dcf.py returns when the bisection interval does not bracket the target:

"NO SOLUTION — today's price cannot be justified anywhere in the plausible range of this parameter, holding the others fixed."

In those cells, even a 200% five-year revenue CAGR — revenue compounding to 243x the current level, $20.7bn in year five — does not produce enough terminal EBIT to support today's enterprise value at the stated margin and multiple.

3.2 Solved the other way — for the exit multiple

Holding the CAGR and solving for the exit multiple makes the boundary sharper:

At the screen's own assumed 14.1% terminal margin, the price requires an exit multiple of roughly 210x EBIT on the screen's own 72.5% demonstrated CAGR — outside the solver's bounds. The screen's record carries the note "reverse DCF unsolvable in range" and then populated terminal_margin: 0.141 and exit_multiple: 28.5 anyway. Those two numbers are mutually incompatible with the price by roughly an order of magnitude, and the record presents them side by side without the contradiction being surfaced.

3.3 The same result expressed as required revenue, which is the intuitive form

At an 18x exit multiple, the year-five revenue the price requires:

At a 10% terminal EBIT margin the price requires $21.5 billion of revenue in five years. For scale, that is larger than the annual revenue of most listed satellite operators combined, and it is required from a company whose product has never generated a dollar. The 5% row requires $43.0 billion, which is why it returns no solution at any multiple: the implied CAGR of 247.4% exceeds the solver's 200% ceiling.

3.4 The required parameter, with everything held fixed named

Holding revenue at the verified TTM of $84.935m, the horizon at 5 years, the discount rate at 13.0%, the enterprise value at $20,998.7m (388,124,572 shares across all three classes at $53.03, plus $414.4m of net debt after the July 2026 convertible issue), and solving for the five-year revenue CAGR:

• At a terminal EBIT margin of 5%: there is NO SOLUTION at any exit multiple between 0.1x and 200x. The price cannot be justified anywhere in the plausible parameter range.
• At a terminal EBIT margin of 10%: NO SOLUTION below an 18x exit multiple. At 22x it requires 190.5%; at 30x, 173.1%.
• At a terminal EBIT margin of 15% and an 18x exit multiple, the price requires a five-year revenue CAGR of 178.9%.
• At a terminal EBIT margin of 30% and a 30x exit multiple — both heroic — the price still requires 119.2%.

Demonstrated: not meaningfully measurable, and zero on the product being valued. The screen's 72.5% is a CAGR from a year in which the company had $4.4m of revenue; the product itself has never generated revenue, so there is no demonstrated rate for it. The margin demonstrated − required is therefore not computable, and the reportable finding is the no-solution region itself.

This is the correct output and it is not a failure of the instrument. Per the brief's own precedent — SMR requiring a 124.2% CAGR against 15.1% demonstrated with a flip point at 631x EV/EBIT, outside the solver's bounds — the right answer is to say so plainly rather than to force a number. Said plainly: across the lower half of the plausible terminal-margin range, today's ASTS price cannot be justified by any combination of revenue growth and exit multiple inside the model's bounds.

3.5 Implied compression, stated as a number

Today: 247.2x TTM revenue. The 15%/18x cell implies year-five revenue of $84.935m × 1.789^5 = $14,332m, year-five EBIT of $2,150m, exit EV of $38,700m, discounted at 13% to $20,999m. Year-five EV/sales = 2.7x.

Compression: 247.2x → 2.7x, a factor of 91.6x, or −98.9%. The buyer at $53.03 is paying 247x revenue for an asset that must grow revenue 169-fold and then still be worth only 2.7x sales.

3.6 Verdict

FAIL, with the no-solution region as the primary finding. There is no PASS WITH ARGUMENT available: that route requires the price to exceed demonstrated performance for "a specific, evidenced reason — a named product cycle, mix shift, pricing action", and while ASTS unquestionably has a named product cycle (Block 2 BlueBird launches, §3 of ASTS_Research.md, with genuine dated technical milestones including the first VoLTE call over satellite on AT&T spectrum), no product cycle can bridge a 179% required CAGR, and in the 5%-margin case no product cycle can bridge an infinite gap.

The honest characterisation of ASTS's price is that it is not a discounted-cash-flow valuation at all. It is an option on a spectrum-plus-constellation position, and the reverse DCF's job here is to say so rather than to manufacture a growth rate. It has said so.

---

4. Sensitivity — on the exit multiple, and on the two screen errors

Per criteria.md, terminal value is 100% of modelled EV (ASTS generates no interim free cash flow — it consumed ~$736m of cash in the last quarter), so the reverse DCF is mandatory as the primary output and the sensitivity must run on the exit multiple, never on scenario probabilities.

On the exit multiple (15% terminal margin): required CAGR moves from no solution at 12x to 151.8% at 30x. The boundary between "no solution" and "solvable" sits between 12x and 15x at a 15% terminal margin, and between 18x and 22x at a 10% terminal margin. Locating that boundary is more informative than any point estimate, because it says exactly how heroic the multiple must be before the arithmetic even closes.

On the screen's two errors — this is the sensitivity that matters most in practice:

The two errors together move the required CAGR by 23.3pp and the multiple by 87.4x. They do not change the verdict — every one of those rows fails — but they change it from "requires 156%" to "requires 179%", and in the 10%-terminal-margin column they change it from a number to no solution at all. That last cell is the sharp one: on the screen's inputs the price is justifiable at a 10% terminal margin and 18x; on the correct inputs it is not justifiable there at any multiple below 22x. The errors did not merely mis-size the answer, they moved the name across the solvability boundary.

Flip point

There is none inside any defensible parameter set, and the arithmetic of how far outside it lies is worth stating. Solving for the required CAGR at a 30x exit multiple:

Even at a 70% terminal EBIT margin — a level no telecom or satellite operator has ever sustained — and a 30x exit multiple, the price still requires 85.0% compound revenue growth for five years. Bringing the required CAGR down to 50% at 30x would need a terminal EBIT margin of approximately 200%, which is arithmetically impossible.

Stated as the brief requires: the price cannot be justified anywhere in the plausible parameter range, holding revenue at $84.935m, the horizon at 5 years and the discount rate at 13.0%. There is no flip point.

---

5. What the price is actually buying

The DCF framing above is the wrong instrument for this asset, and saying so is more useful than pretending otherwise. What $20,999m of enterprise value buys:

Book value of the tangible programme is roughly $1.9bn. The market is paying $21.0bn. The $19.1bn difference is the option value on spectrum plus subscriber revenue-share — which may well be the right way to think about it, and is not something a five-year reverse DCF on $84.9m of gateway-equipment revenue can price. The instrument's verdict is FAIL; the instrument's limitation is stated here so a book is not misled about what has and has not been measured.

---

6. The 12-month target — UNIDENTIFIED

6.1 Why

Own EV/TTM-revenue history, from the filed balance sheets and the SIP close at each date. Share counts are Class A plus the 89.4m of Class B and C; the pre-2025-12-31 Class A figures are weighted averages, so those rows are approximations and are marked as such.

Median 1,011.6x · min 372.7x · max 3,073.9x · today 247.2x — below every observation, the 0th percentile.

This is not a multiple range; it is the trace of a company acquiring a revenue line. For four of the six observations TTM revenue was between $4.4m and $18.7m, so the "multiple" is an artefact of a near-zero denominator. Across the series the share count rose 59%, net cash swung from +$417m to −$111m to +$560m to −$414m, and revenue rose 19-fold. Per valuation.md, the history is too short and spans a regime change: the multiple is declared UNIDENTIFIED. No peer median is substituted.

And a note on what "0th percentile" does not mean. On a naive read, ASTS at 247x against a 1,012x median looks cheap. It is not a signal — it is the denominator growing. A falling multiple driven by a denominator arriving from zero carries no information about value. This is worth recording because it is precisely the kind of number a screen would rank on.

6.2 What is offered instead, with its basis stated

A reference band, explicitly not multiple-anchored and not to be used as a target. All on 388.1m shares.

Reference band $3.84 – $142.98 — a 37-fold range. That spread is not analytical failure; it is the honest width of the distribution on a pre-commercial asset, and any point target inside it would be a fabrication. The mid cases sit above spot, which is worth stating explicitly against calibration item B16: this memo's reference range for ASTS is not uniformly bearish, and the instrument that says FAIL (the five-year implied path) and the instrument that says +52% (its own lowest historical multiple) are measuring different things over different horizons. valuation.md exists to keep them apart.

6.3 The number to watch instead of a target

Cash and restricted cash fell ~$736m in the June quarter, and the company raised $983.6m net three weeks after quarter-end. At that consumption rate the July raise funds roughly four quarters. The Sound Point facility's availability expires 2026-10-05. The next financing event, not the next revenue print, is what will set the price over the coming twelve months — and it is dateable, unlike the revenue.