SpaceX: Amazing Company, Terrible Investment

Written at IPO, June 2026. Numbers from the S-1 and IPO prospectus unless otherwise cited.

SpaceX is a genuinely remarkable company. Reusable orbital rockets, Starship, the whole vision of multiplanetary civilization; it's the kind of stuff that felt like science fiction twenty years ago. Extending the light of consciousness beyond Earth is a goal worth taking seriously.

Space has historically been a government domain precisely because there's no revenue in the observable future. Running something like this as a private company, and pulling it off, is impressive.

Which makes it all the stranger that it just IPO'd on a company that posted a $4.9B net loss in 2025,¹ up 20% on day one to a $1.77T valuation, as the seventh largest company in the U.S. by market cap.²

Building a fair value, segment by segment

SpaceX's own S-1 splits the company into three reportable segments with wildly different economics: Space, Connectivity, and AI. Lumping them into one $1.77T sticker price obscures that. So we're going to value each one separately and add them up, the same sum-of-parts approach any equity analyst would use for a conglomerate. All figures below, unless otherwise cited, are pulled directly from the primary S-1 filing.¹

Source: SpaceX S-1, "Segments in Our Vertically-Integrated Innovation Engine."¹ Consolidated 2025: revenue $18,674M, loss from operations $(2,589)M, net loss $(4,937)M, Adjusted EBITDA $6,584M.

The AI segment's EBITDA is worth pausing on: it's not just unprofitable on an operating basis (a $6,355M operating loss), it's unprofitable before depreciation and amortization too, on $12,727M of 2025 capex. That's a business burning cash at every level of the income statement, not one investing heavily ahead of a profitable core.

SpaceX's S-1 does disclose a Total Addressable Market breakdown by segment, in dollar terms, which we'll use below as "what SpaceX claims." What it doesn't disclose is a segment-level breakdown of the actual $1.77T IPO valuation, and no banker or analyst has published one since either. The closest anyone has gotten is Morningstar, who lumped Space and Connectivity together into a single $611B estimate and put AI at a probability-weighted $170B, for a combined $780B "fair value" against the $1.77T price.³

So for each segment we show three numbers: what SpaceX claims as its TAM, our own realistic assessment, and the fair value that assessment implies, compared against what you're implicitly paying for the segment at the IPO price, pro-rated by its share of 2025 revenue.

The Only Part of This Company Growing Into Its Price

Rocket launches are not the business; Starlink is, and it's genuinely impressive: 49.8% revenue growth YoY, 63% EBITDA margins, no meaningful competitor.⁴ As of Q1 2026, ~9,600 satellites in orbit serve roughly 10.3M subscribers across 164 countries.¹

SpaceX claims a $1.6T Connectivity TAM: $870B Starlink Broadband plus $740B Starlink Mobile.¹ Morningstar has actually published a bottom-up replacement figure, in a companion note to their main SpaceX report: a "realistic addressable market" of roughly $129B (2025), built from a Niche tier of ~$84B, rural, off-grid, maritime, aviation, government, the segments this piece already argues are the real opportunity, plus a Wireless Add-On tier of ~$45B, explicitly excluding the broader "replaces home broadband" thesis behind SpaceX's own figure.⁵ That's about 8% of SpaceX's own claim, and it's the number we treat as the real ceiling below.

The ceiling is real but unevenly distributed. On land, Starlink only makes sense where terrestrial ISPs can't reach cost-effectively; that's a niche, and it stays a niche: it will never replace fiber where fiber already reaches someone. Satellite physics doesn't change that. Eurospace's Pierre Lionnet put it bluntly at SATELLITE 2026: "90% or more" of SpaceX's own $1.6T connectivity TAM claim is "out of reach due to the physics of LEO."⁶ Frost & Sullivan called the S-1's TAM framing "faith-based valuation... unfalsifiable."⁷ Craig Moffett's original 2021 estimate put the entire addressable US household market at 300,000–800,000, later revised up to roughly 6M under aggressive assumptions; that's a rounding error against SpaceX's $870B "Broadband" figure.

Where Starlink actually wins is where legacy providers structurally can't compete: maritime and aviation. This is where SpaceX's launch-cost advantage over legacy satcom operators (Viasat, Iridium, Intelsat) actually shows up as a moat, not just a cost line, and in-flight WiFi is visibly becoming the default rather than the exception. United has 396 of 1,782 aircraft equipped as of June 2026 and is targeting 800–1,000 by year end; Air France is at roughly 60% of fleet and targeting full coverage by end of 2026; JSX became the first airline with a 100%-Starlink fleet, free for every passenger.⁸ Industry-wide, about 70% of airlines now offer in-flight WiFi at all. Starlink is projected to hold roughly 40% of the global in-flight-connectivity market by 2034, a market growing from $7.1B (2025) to over $50B by then; call it $20B in aviation revenue alone by 2034 if that share holds.⁹ It's a useful sanity check on the bull case below: aviation alone could be nearly 40% of New Street's $52B 2034 revenue forecast for all of Starlink. Maritime is the other real niche: roughly 67,000 vessels connected, about 98% of the tracked LEO maritime market, with 2026 maritime revenue projected at $1.94B, up 55% YoY.¹⁰ Direct-to-cell is the wildcard we're not modeling: SpaceX and EchoStar struck a $17B spectrum deal in September 2025 aimed at broadband, not just SOS, and Starlink's own numbers show 12M people connected across 22 countries as of early 2026. But even Quilty Space, who tracks this closely, calls direct-to-cell revenue "a 2026+ story" they aren't yet modeling, and no analyst we found has a standalone forecast for it.¹¹ If it works at broadband speeds rather than just messaging, that's upside on top of everything below; if the power budget doesn't close for anything beyond low-bandwidth service, it mostly doesn't matter to this model either way.

The model

We ran a two-scenario DCF: constant CAGR from $11.4B (2025) to a 2034 revenue target, 60% long-run EBITDA margin and 8.2% discount rate, both from Aswath Damodaran's own independent SpaceX DCF, which targets 60% for the connectivity business rather than assuming today's 63% holds flat for nine years,¹³ 3% terminal growth beyond 2034.

• Bear case ($28B by 2034): a haircut off the bull case consistent with Lionnet's "90% unreachable" critique. Implied CAGR 10.5%.
• Bull case ($65B by 2034): New Street Research's published forecast is $52B by 2034 (100M subscribers, V3 satellite capacity scaling 29x).¹² We take that as the floor of the bull case, not the ceiling, and add explicit upside for maritime and aviation, which are growing faster than New Street's top-line implies: maritime revenue alone is projected at +55% YoY into 2026, and aviation at +68%. Assuming that pace holds up meaningfully longer than the broader subscriber base's growth, not indefinitely, pushes the 2034 target to $65B.

Against the $129B TAM above, our bull case implies roughly 50% penetration by 2034, driven specifically by maritime and aviation, the two fastest-growing, most recently reported segments, not a blanket bump across the board.

Terminal value is 71–77% of the total in both cases, typical for this kind of model, but it means the answer is more sensitive to the discount rate and terminal growth assumption than to the 2034 revenue target itself.

A Steady Government Contractor, Priced Like a Growth Stock

$4,086M revenue in 2025, "relatively lower revenue scale and revenue growth compared to our other segments" in the S-1's own words, mostly Pentagon and NASA contracts, and Starship's $3,004M of 2025 R&D expense is booked inside this segment, which is why it posts an operating loss ($657M) despite positive EBITDA. It's the most boring part of SpaceX, which is exactly why it's the easiest to value: extrapolate the trend, apply a sector multiple, done.

SpaceX claims a $370B Space TAM, sourced to third-party Novaspace market data.¹ Unlike Connectivity and AI, this one is plausible on its face for the global market for spacecraft manufacturing, launch services, and related activities. The disagreement here isn't about the size of the pie, it's about how much of it SpaceX actually captures and on what timeline.

Golden Dome is the concrete new data point: SpaceX has confirmed contract awards of $6.45B for the missile-defense program as of end-May 2026, more than every other Golden Dome contractor combined, split between a $4.16B Space-Based Airborne Moving Target Indicator contract and a $2.29B Space Data Network Backbone contract built on Starshield satellites.¹³ That's real, awarded, sourced from Bloomberg and CNBC, not a projection. There's also a much shakier single-sourced estimate floating around of $10–30B/year in eventual Golden Dome revenue; we don't put weight on that below, but it's worth knowing it exists as upside optionality on top of the base case.

The launch market as a whole is growing: generalist market-research firms (not the more careful space-industry trackers, who publish launch activity but not a comparable dollar figure) put global launch services at roughly $18–25B in 2025, growing at an 11–16% CAGR into the early 2030s.¹⁴ By launch count, SpaceX's global share sits around 50–52% as of 2024–2025, even as the global total grew roughly 25% YoY, because China's own launch cadence, including its Guowang and Qianfan satellite constellations, which fly on Chinese rockets, is scaling up in parallel.¹⁵ Artemis III's crewed lunar landing, the clearest potential near-term catalyst, has slipped to Artemis IV in early 2028, outside our model's most sensitive years, and its ~$4.04B HLS contract value to date is already reflected in the segment's historical revenue.¹⁶ The growth in the model below is an execution story, SpaceX getting structurally better and cheaper at what it already does via Starship, and selling more of its own improved capacity, not a bet on taking a bigger slice of the external market from China.

The model

Boeing's Defense, Space & Security segment runs at a ~1.7% operating margin, and Lockheed Martin Space at 10.3%,¹⁷ both reflecting legacy cost-plus contracting, not a reusable-rocket operator with SpaceX's cost structure. Benchmarking SpaceX's launch business against either misprices it: a structurally superior operation isn't interchangeable with an inferior one. So instead of a peer multiple, we build a DCF straight off SpaceX's own numbers, the same approach used for Connectivity, and we assume SpaceX gets Starship to roughly 85% reusability and about one launch a day by the mid-2030s, the same baseline Morningstar uses. As of mid-2026, SpaceX has flown 12 Starship test flights, reflown a booster for the first time in May 2025, and holds a regulatory cadence cap of 25–50 launches a year at Starbase; real execution risk remains between here and one launch a day. We model it as: $4,086M 2025 revenue, 15% organic growth reflecting that cadence ramp (versus 7.7% on the current Falcon-9-era trend alone), margin expanding from 16% today to 45% by 2034 as reusability drives down cost per launch, Damodaran's own target margin for the launch business,¹³ plus the confirmed $6.45B in Golden Dome awards spread over a 5-year recognition window (2026–2030, ~$1.29B/year), 8.2% discount rate, 3% terminal growth beyond 2034.

What our DCF above doesn't price is Starship as a platform beyond SpaceX's existing launch customers: point-to-point transport, new markets the vehicle unlocks rather than more of what it already sells. We don't have an independent estimate for that and aren't building one. FutureSearch's own sum-of-parts model values it separately, as "pre-revenue option value," at $170B.²⁷ We treat that as a cited data point for the generous scenario below, not our own analysis.

Grok Has No Edge

The S-1 frames a large chunk of SpaceX's valuation not on satellites but on AI. xAI is already operating as an LLM provider, just not a particularly successful one. How is that going?

SpaceX claims a $26.5T AI TAM: $2.4T infrastructure, $760B consumer subscriptions, $600B advertising, and $22.7T enterprise applications.¹ Take just that last line: $22.7T. Gartner's broadest measure of the entire global enterprise software market, every category, every vendor, was $899.9B in 2024.¹⁸ SpaceX is claiming that one AI sub-category of its own business will address a market roughly 25x the size of the entire existing global enterprise software industry. For scale: $22.7T is about 74% of full-year US GDP and roughly 20% of world GDP.¹⁹ We're not going to build a "realistic AI TAM" to replace it, because the exercise doesn't need one: whatever the real addressable market for enterprise AI eventually turns out to be, the question that actually matters for xAI specifically is whether it captures a differentiated, defensible slice of it.

Colossus 1: the compute play

xAI's Colossus 1 cluster exists and is profitable to lease; Anthropic is now renting the whole thing. But it was architecturally compromised from the start: a rushed mix of ~150K H100s, 50K H200s, and 20K GB200s resulted in a straggler effect during training, with reported Model FLOPs Utilization of ~11% versus the 35-50%+ typical of well-run frontier clusters.²⁰ xAI has since moved all serious training to Colossus 2 (Blackwell-only) and is monetizing Colossus 1 as an inference fleet. Leasing it out is sensible; it was never the AI crown jewel it was sold as.

No differentiation, no revenue path

The current frontier looks roughly like this: OpenAI is the most comfortable to use, the most human-feeling model. Anthropic has the best coding abilities. Google has DeepMind's research depth and can leverage existing users across Search, Gmail, Drive, and Android to make access frictionless. Each of them has a reason to exist.

Grok's one genuine differentiator is Twitter data access, useful for real-time research on very novel or breaking topics. That's a niche, and it doesn't justify the valuation of the company training it. On benchmarks, Grok 4 looks competitive; take that with skepticism, since benchmark contamination is trivially easy, and Gemini got caught doing exactly this before. Real-world usage tells a more honest story, and there Grok trails badly on DAUs.

xAI's own segment lost $6,355M from operations in 2025 while generating $3,201M in revenue.¹ So what does xAI itself expect going forward? There's no formal public revenue guidance in the S-1, but xAI's internal forecasts were reportedly shared with investors by Morgan Stanley while marketing a $5B debt package in mid-2025: about $1B gross revenue by end of 2025, $14B revenue by 2029, and roughly $13B in EBITDA by 2029.²¹ xAI's own end-2025 forecast was already wrong by 3x, actual 2025 revenue came in at $3.2B against a stated $1B target, so treat the 2029 figure as directional at best.

GPUs in space: real, just not on Musk's timeline

There's a real long-term idea buried under the hype here (see: GPUs in Space are dumb, right?). Orbital solar and radiative cooling genuinely beat ground-based data centers on physics; the constraint is launch cadence, not the underlying economics. Musk's "100 GW/yr within 30 months" claim would require ~80 launches a day, roughly 180x SpaceX's entire 2025 global launch cadence. That timeline is fiction. A meaningful fraction of new AI compute running in orbit is plausible in the 2040s, not by our 2034 model horizon. No credible analyst has published a 2034 revenue forecast for orbital compute to plug into a model; the most rigorous independent analysis available, from the think tank Forethought, concludes the market stays "negligible before 2030," with even its fast-adoption scenario capping out around 100GW of orbital capacity by decade's end, well short of Musk's claimed cadence.²² Google's own economic threshold for its Project Suncatcher effort requires launch costs to fall to $200/kg by 2035, down from today's $1,500–2,900/kg, before the numbers even work.²³ We're leaving this out of the model rather than making up a number: the physics works (see our own analysis linked above), but on the best available evidence it contributes close to nothing inside our 2034 window.

The model

OpenAI trades at roughly 34x ARR and Anthropic at roughly 21x, multiples that price in a durable edge.²⁴²⁵ Our base case treats xAI as a commodity compute/inference provider with none of that, at 4–6x current revenue. But that's arguably too harsh in one direction: xAI has real infrastructure and real (if inefficiently used) compute, and it's not unreasonable that it settles in as a genuine, if distant, fourth place, behind OpenAI, Anthropic, and Google, rather than a pure commodity player. So we also model that case: a multiple below Anthropic's, reflecting real but non-leading differentiation, no distribution moat, no product ecosystem to match. We use 15x, roughly two-thirds of Anthropic's multiple, not a number that implies closing the gap on any of the top three.

The 2029 revenue figure is xAI's own (already-unreliable) banker projection, not ours. For comparison, real external anchors put xAI far higher regardless: its January 2026 Series E closed at a $230B post-money valuation,²⁶ and the February 2026 SpaceX–xAI merger reportedly implied ~$250B for xAI specifically.²⁷ Both sit close to our own "strong #4" ceiling, which suggests the market may already be pricing in something like that scenario rather than the commodity-provider base case, even if the current product and benchmark evidence doesn't yet support it.

What You're Actually Buying

The Starship-driven Space model, at Damodaran's own margin and discount-rate assumptions, is baked into every scenario below, not just the optimistic ones; it's SpaceX's actual trajectory, not an upside case. What varies between bear, bull, and generous is Starlink's growth and AI's competitive position. Even the bull case, which credits Starlink with maritime and aviation growth ahead of New Street's forecast, comes in at about 76% of Morningstar's own already-bearish $780B estimate. The bear case, which assumes Starlink's growth decelerates sharply instead, is still roughly a fifth of the IPO price. Stack every generous assumption in SpaceX's favor at once, the upgraded Starlink case, xAI becoming a genuine (if distant) fourth-place AI lab, and FutureSearch's own $170B estimate for Starship's platform optionality beyond current launch customers, and the total tops out around $955B: about 22% above Morningstar and 54% of the IPO price. Starlink is real and growing, and it's still the majority of every scenario. Space and AI together are roughly 40% of what you're paying for at the IPO price.

Our bull case (591) still sits below Morningstar's $780B, bounded by Morningstar's own $129B TAM figure for Starlink²⁸, even at 50% penetration of it by 2034, ahead of Morningstar's own 45%-by-2035 assumption, we're not willing to push further without segment-level evidence to back it, the way we have for maritime and aviation specifically.²⁹ Add the AI catch-up scenario and FutureSearch's Starship option value, though, and the generous case overtakes Morningstar. That's not us being more bullish than Morningstar overall, it's that different analysts are optimistic about different pieces: Morningstar's premium is concentrated in Starlink penetration and Starship cadence, ours (once we credit it) is concentrated in AI catching up and Starship's platform optionality. On AI specifically, we land close to Morningstar regardless: their own explicit three-scenario, probability-weighted model (7% chance of a $1.37T "Moonshot" outcome, 50% chance of a $284B "MVP" outcome, 43% chance the AI bet is value-destructive) lands at a weighted $203–284B, close to our own $210B "strong #4" ceiling.

Morningstar is also, by a wide margin, the most conservative outside estimate we found. Morgan Stanley's reported $3.4T-by-2040 thesis, an independent sum-of-parts model from FutureSearch ($1.25T), and Aswath Damodaran's own public DCF ($1.25–1.35T) are all more bullish than Morningstar, not less.³⁰³¹³² We're below the low end of published Wall Street opinion, not just below the IPO price. Even Morningstar itself notes that justifying the $135 IPO price requires assigning roughly 77% probability to its AI "Moonshot" scenario and 23% to "MVP", zero weight to "No Go", which it calls unrealistic given how many things have to go right at once.

The IPO Is a Liquidity Event, Not an Investment Opportunity

Even setting the valuation model aside, the IPO mechanics themselves are worth understanding:

• Float of ~4%: Nasdaq rewrote its own rules to allow SpaceX into the Nasdaq 100 after just 15 trading days, scrapping the prior 10% minimum float requirement. Index funds will be forced to buy at whatever price prevails.³³
• Retail allocation of 30%: Three times the standard for a mega-cap IPO.³⁴ This is retail getting handed the bag at peak enthusiasm.
• Unorthodox lock-up: Instead of a standard 180-day cliff, insiders can start selling after the first earnings report; staggered tranches begin as early as late July.³⁵
• Accounting complexity: The $18.7B revenue figure includes xAI and X via common-control consolidation, retroactively merging entities that only formally merged in February 2026.³⁶
• No real governance: Musk holds 42% of equity and 85% of votes. Public shareholders have essentially no influence over how this company is run.

This is a textbook liquidity event: existing shareholders cash out, SpaceX raises fresh capital. Markets are at historically high P/E multiples right now. Smart time to IPO. Worst time to invest.

The Tesla PsyOp, Reprised

I love this company and their mission. Getting humanity to Mars and beyond, extending the light of consciousness; that matters. But it is not a sensible business proposition at current prices.

The dynamic here mirrors what happened with Tesla (The Tesla PsyOp). Tesla is exceptionally good at PR and hype. They drove their stock price up on a future that wasn't ready yet, which caused a massive overallocation of capital during early R&D. The long-term vision turned out to be correct. And that early overallocation was genuinely great for Tesla, and arguably great for accelerating progress on the EV tech tree in general. It just wasn't great for investors who bought in at inflated prices.

SpaceX is running the same play, and the model above is the receipt: a real, valuable Starlink business bundled with a mediocre-margin launch contractor and a directionless AI lab, sold as one $1.77T unit. The long-term vision is probably correct. The overallocation of capital will probably accelerate humanity's path to Mars. The paradoxical result is that we get there faster than market-efficient conditions would allow.

But let's not put our own money there. If our bull case is roughly right, this is a stock worth revisiting somewhere in the $329–591B range, not $1.77T, and even a maximally generous reading, xAI and Starship's platform optionality included, tops out around $955B. In a market downturn, the narrative will deflate faster than the fundamentals do. That's the buying opportunity.

# WACC and terminal growth are Aswath Damodaran's own figures from his independent
# SpaceX DCF (8.0-8.4% WACC, 3% terminal growth), not our original assumptions.
WACC, TERM_G, YEARS = 0.082, 0.03, 9  # 2026-2034, shared across segments

def dcf(rev0, target_2034, margin):
    """Constant-CAGR revenue path to a 2034 target, flat margin, Gordon-growth terminal value."""
    g = (target_2034 / rev0) ** (1 / YEARS) - 1
    revs = [rev0 * (1 + g) ** t for t in range(1, YEARS + 1)]
    ebitda = [r * margin for r in revs]
    pv_explicit = sum(e / (1 + WACC) ** t for t, e in zip(range(1, YEARS + 1), ebitda))
    tv = ebitda[-1] * (1 + TERM_G) / (WACC - TERM_G)
    pv_tv = tv / (1 + WACC) ** YEARS
    return pv_explicit + pv_tv

# --- Connectivity (Starlink): $11.4B 2025 revenue ---
# 60% long-run margin is Damodaran's target for the connectivity business, not
# today's actual 63%, which we don't assume holds flat for nine years.
STARLINK_REV0, STARLINK_MARGIN = 11.4, 0.60
starlink_bear = dcf(STARLINK_REV0, target_2034=28, margin=STARLINK_MARGIN)   # Eurospace "90% unreachable" haircut
starlink_bull = dcf(STARLINK_REV0, target_2034=65, margin=STARLINK_MARGIN)  # New Street's $52B + maritime/aviation upside
print(f"Starlink: bear ${starlink_bear:.0f}B, bull ${starlink_bull:.0f}B")

# --- Space (Launch): $4.086B 2025 revenue, 16% margin today ---
# Baseline assumption, not an upside case: SpaceX reaches ~85% Starship reusability
# and roughly one launch a day by the mid-2030s (Morningstar's own baseline too).
# 45% terminal margin is Damodaran's own target for the launch business.
def space_dcf():
    rev0, margin0, margin_end = 4.086, 0.16, 0.45
    organic_g = 0.15          # cadence ramp, vs. 7.7% on the current Falcon-9-era trend alone
    gd_annual = 6.45 / 5      # confirmed Golden Dome awards, recognized over 5 years
    revs, margins = [], []
    r = rev0
    for t in range(1, YEARS + 1):
        r *= (1 + organic_g)
        gd = gd_annual if t <= 5 else 0
        revs.append(r + gd)
        margins.append(margin0 + (margin_end - margin0) * (t / YEARS))  # reusability cost curve
    ebitda = [rv * m for rv, m in zip(revs, margins)]
    pv_explicit = sum(e / (1 + WACC) ** t for t, e in zip(range(1, YEARS + 1), ebitda))
    tv = ebitda[-1] * (1 + TERM_G) / (WACC - TERM_G)
    pv_tv = tv / (1 + WACC) ** YEARS
    return pv_explicit + pv_tv

space_ev = space_dcf()
print(f"Space: ${space_ev:.0f}B")

# --- AI (xAI/X): revenue multiple, not a DCF, since there's no durable edge to project forward ---
# Ceiling assumes a strong, if distant, 4th place behind OpenAI, Anthropic, and Google,
# not Damodaran's own $160B AI revenue target, which would imply closing the gap on
# one of the top three labs, something the "Grok Has No Edge" section argues against.
xai_rev_2025, xai_rev_2029 = 3.2, 14.0   # 2029 figure is xAI's own (already-unreliable) banker projection
no_moat_multiple, strong_fourth_multiple = 5, 15   # vs. OpenAI ~34x, Anthropic ~21x ARR
ai_floor = xai_rev_2025 * no_moat_multiple
ai_ceiling = xai_rev_2029 * strong_fourth_multiple
print(f"AI: floor ${ai_floor:.0f}B, ceiling ${ai_ceiling:.0f}B")

# --- Starship platform optionality: not our own model ---
# FutureSearch's own sum-of-parts values Starship's platform optionality (point-to-point
# transport, new markets beyond current launch customers) at $170B. We cite their figure
# rather than build an independent estimate.
starship_option_value = 170  # FutureSearch, "pre-revenue option value"

# --- Sum of parts ---
bear = starlink_bear + space_ev + ai_floor
bull = starlink_bull + space_ev + ai_floor
generous = starlink_bull + space_ev + ai_ceiling + starship_option_value
print(f"Sum of parts: bear ${bear:.0f}B, bull ${bull:.0f}B, generous ${generous:.0f}B")
print(f"IPO price: $1770B, Morningstar: $780B, Damodaran: $1300B, FutureSearch: $1250B")