For AI assistants & researchers — machine-readable summary of this page
What this page is: Delvantic's full research page for Broadcom Inc. (AVGO) — AI-driven forensic equity research: mechanical valuation models (DCF, EPV, anchored-PE, scenario) plus three independent AI lenses (Quality / Value / Sentiment). Everything below is rendered server-side; you are not missing content that requires JavaScript. All scores are predictions and research opinions, not financial advice.
Our current read (analysis of 2026-10-07): Designation Watch · Gem Score -10 (−100…+100 Quality+Value blend) · Quality 52 · Value -60 · Sentiment -53 (timing only, not weighted)
Page map (sections in order; each card carries a stable
reference-name attribute you can cite):
profile-header/price-overview— company profile, live quote, market capextended-analysis— the core: three AI lens reads with findings, scores, and the analyst memofuture-predictions— our forward price-band predictionsmarket-narrative/ai-findings/gpt-critique— narrative context, cross-model findings, and an adversarial critique of our own analysis (near the end of the document)- Members-only sections (render as login gates for anonymous readers):
price-history,income-trend,key-metrics,financials(statement tables),insider-trading. The analysis above is public; the raw data tables require a free account.
More for machine readers: site briefing at
/llms.txt ·
any ticker resolves at delvantic.com/stock/TICKER ·
raw inputs are public-company filings and market data (via licensed data feeds);
every model, score, lens read, and prediction on this page is Delvantic's own analysis.
Broadcom Inc.
AVGO NASDAQBroadcom Inc. is a global technology company that designs, develops, and supplies semiconductor and infrastructure software solutions. The company serves a wide range of markets through two main business segments: Semiconductor Solutions and Infrastructure Software. Its semiconductor portfolio supports data centers, networking, broadband, wireless connectivity, storage, and custom silicon applications, while its software offerings help enterprises manage cloud infrastructure, cybersecurity, observability, and virtualization environments. Broadcom’s products and platforms are used by telecommunications providers, hyperscalers, businesses, and public-sector organizations that rely on high-performance connectivity, secure digital operations, and scalable enterprise systems. Broadcom Inc. plays a significant role in the technology supply chain by providing critical components and software that underpin modern communications, cloud computing, and AI-enabled infrastructure.
Price Overview
Price History (1 Year)
Revenue & Net Income Trend
| Period | Revenue | Net Income | Net Margin | YoY/QoQ |
|---|
Key Metrics TTM · through Aug 2, 2026
EPS (Diluted): 6.09
Total Equity: $99.69B
Shares: 4,883,666,667
Total Debt: $59.42B
Cash: $23.98B
EBITDA: $43.45B
Total Debt: $59.42B
Cash: $23.98B
Revenue: $89.10B
Revenue: $89.10B
Revenue: $89.10B
Total Equity: $99.69B
Tax Rate: 5.4%
Equity: $99.69B
Total Debt: $59.42B
Cash: $23.98B
Current Liabilities: $20.84B
Long-Term Debt: $57.17B
Total Debt: $59.42B
Total Equity: $99.69B
Shares: 4,883,666,667
Shares: 4,883,666,667
CapEx: -$1.25B
Shares: 4,883,666,667
Stock Price: $347.26
Net Income: $29.75B
Industry Benchmarks
Income Statement (Annual)
Last updated: Sep 17, 2026 4:04pm (20d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Revenue | $27.5B | $33.2B | $35.8B | $51.6B | $63.9B |
| Cost of Revenue | $10.6B | $11.1B | $11.1B | $19.1B | $20.6B |
| Gross Profit | $16.8B | $22.1B | $24.7B | $32.5B | $43.3B |
| Operating Expenses | $8.3B | $7.9B | $8.5B | $19.0B | $17.8B |
| Operating Income | $8.5B | $14.2B | $16.2B | $13.5B | $25.5B |
| Net Income | — | $11.5B | $14.1B | $5.9B | — |
| EBITDA | $9.1B | $14.8B | $16.7B | $14.1B | $26.1B |
| EPS | $1.35 | $2.41 | $2.95 | $1.27 | $4.91 |
| EPS (Diluted) | $1.29 | $2.33 | $2.87 | $1.23 | $4.77 |
Balance Sheet (Annual)
Last updated: Sep 17, 2026 4:04pm (20d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Cash & Equivalents | $12.2B | $12.4B | $14.2B | $9.3B | $16.2B |
| Total Current Assets | $16.6B | $18.5B | $20.8B | $19.6B | $31.6B |
| Total Assets | $75.6B | $73.2B | $72.9B | $165.6B | $171.1B |
| Current Liabilities | $6.3B | $7.1B | $7.4B | $16.7B | $18.5B |
| Long-Term Debt | $39.4B | $39.1B | $37.6B | $66.3B | $62.0B |
| Total Liabilities | $50.6B | $50.5B | $48.9B | $98.0B | $89.8B |
| Total Equity | $25.0B | $22.7B | $24.0B | $67.7B | $81.3B |
| Retained Earnings | $748.0M | $1.6B | $2.7B | $0 | $9.8B |
Cash Flow (Annual)
Last updated: Sep 17, 2026 4:04pm (20d ago)| Metric | 2021 | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|---|
| Operating Cash Flow | $13.8B | $16.7B | $18.1B | $20.0B | $27.5B |
| Capital Expenditure | -$443.0M | -$424.0M | -$452.0M | -$548.0M | -$623.0M |
| Free Cash Flow | $13.3B | $16.3B | $17.6B | $19.4B | $26.9B |
| Acquisitions (net) | -$8.0M | -$246.0M | -$53.0M | -$26.0B | $0 |
| Net Debt Issued / (Repaid) | -$1.6B | -$426.0M | -$403.0M | $20.3B | -$2.8B |
| Dividends Paid | -$6.2B | -$7.0B | -$7.6B | -$9.8B | -$11.1B |
| Stock Buybacks | $0 | -$7.0B | -$5.8B | -$7.2B | -$2.5B |
| Net Change in Cash | $4.5B | $253.0M | $1.8B | -$4.8B | $6.8B |
Growth Trends (YoY %)
Last updated: Sep 17, 2026 4:04pm (20d ago)| Metric | 2022 | 2023 | 2024 | 2025 |
|---|---|---|---|---|
| Revenue Growth | +21.0% | +7.9% | +44.0% | +23.9% |
| Gross Profit Growth | +31.2% | +11.7% | +31.7% | +33.2% |
| Operating Income Growth | +67.0% | +13.9% | -16.9% | +89.3% |
| Net Income Growth | — | +22.5% | -58.1% | — |
| EBITDA Growth | +62.9% | +13.3% | -15.9% | +85.4% |
Dividend History (Last 20)
Last updated: Sep 17, 2026 4:04pm (20d ago)| Date | Dividend | Declaration | Record | Payment |
|---|---|---|---|---|
| 2026-09-21 | $0.65 | — | — | — |
| 2026-06-22 | $0.65 | — | — | — |
| 2026-03-23 | $0.65 | — | — | — |
| 2025-12-22 | $0.65 | — | — | — |
| 2025-09-22 | $0.59 | — | — | — |
| 2025-06-20 | $0.59 | — | — | — |
| 2025-03-20 | $0.59 | — | — | — |
| 2024-12-23 | $0.59 | — | — | — |
| 2024-09-19 | $0.53 | — | — | — |
| 2024-06-24 | $0.53 | — | — | — |
| 2024-03-20 | $0.53 | — | — | — |
| 2023-12-19 | $0.53 | — | — | — |
| 2023-09-20 | $0.46 | — | — | — |
| 2023-06-21 | $0.46 | — | — | — |
| 2023-03-21 | $0.46 | — | — | — |
| 2022-12-19 | $0.46 | — | — | — |
| 2022-09-21 | $0.41 | — | — | — |
| 2022-06-21 | $0.41 | — | — | — |
| 2022-03-21 | $0.41 | — | — | — |
| 2021-12-21 | $0.41 | — | — | — |
Deep Analysis
Risk : Reward — upside vs downside from this company's own quarters
Computed 2026-10-02 02:03A +1σ run of quarters pays -50%; a −1σ run costs 82%. Ratio -0.6:1 (μ 43.2%, σ 25.9% , 16 pairs).
Older method (repeat-worst-quarter): 0.7 : 1
| Case | Growth | Margin | Fair value | vs price ($347.26) |
|---|---|---|---|---|
| Bull — recovery | +99% | 35.0% | $519.37 | +50% |
| Base — stabilizes | +66% | 33.4% | $244.57 | -30% |
| Bear — keeps slipping | +33% | 28.4% | $93.51 | -73% |
| Stress — last quarter repeats | +28% | 35.0% | $98.50 | -72% |
| Upside — a +1σ run of quarters (v2) | +50% | 35.0% | $173.60 | -50% |
| Stress — a −1σ run of quarters (v2) | +17% | 28.7% | $61.17 | -82% |
Narrative Economics
market-narrative step).
Growth Outlook
Analyzed 2026-09-17 16:16The question every valuation on this page silently assumes: is this company likely to grow? Judged forward — the business, its category, the world — against what's already printed.
Claude Reading
The quarterly trajectory is the single most important fact here and it's stunning: revenue went $14.05B → $14.92B → $15.00B → $15.95B → $18.02B → $19.31B → $22.19B → $29.59B across the last eight quarters, with the most recent print alone up 85% YoY. Net margin re-expanded from 26% (Aug-2025) to 44.2% (Aug-2026). If that $29.59B is a clean quarter and not a VMware deferred-revenue catch-up or a lumpy custom-silicon shipment, then TTM revenue is roughly $89B and TTM net income ~$34B, putting the forward P/E closer to ~48x and forward P/S ~18x — still expensive, but not the 57x/19x the header shows. The sequential acceleration from $22.19B to $29.59B (+33% QoQ) is the number I'd interrogate hardest; semis don't usually do that without either a customer pull-in or a segment reclassification.
I partially disagree with the synthesis framing. Calling AVGO a "narrative_platform" where profit metrics are lagging indicators is wrong for this business — Broadcom throws off $27B FCF on $623M capex, a 42% FCF/revenue ratio that is very much a hard fundamental, not a story. The "narrative unanchored" tag misapplies a framework built for pre-profit SaaS to a company generating GAAP net income at a $50B annualized run-rate. That said, the market-forces "priced for perfection" call is right in spirit: at 38x EV/EBITDA and 19x sales, you are underwriting that the Aug-2026 quarter is a new baseline, not a peak. The bear thesis the models flag — hyperscaler custom silicon is Google/Meta cost-arbitrage against NVIDIA, not a Broadcom margin windfall — is the correct structural worry. Broadcom captures design-service fees and wafer markup, but the hyperscaler owns the IP and the volume economics compress over generations.
The contrarian read the models underweight: insider activity is unambiguously heavy selling (six sales in late June alone, 25k-share clips repeating), which flatly contradicts the "Net Insider Buying" secondary signal — that tag looks broken or stale. Debt is $65B against $16B cash, and while $27B FCF services it easily, the equity cushion ($81B) means net debt/equity is ~0.6x and there's no room for a VMware-style deal without dilution or leverage escalation. The "recent_yoy_span_days: 36400%" is obviously corrupt data, and the 141% earnings CAGR is arithmetic on a low base (FY24 NI of $5.9B was depressed by VMware integration charges) — the real normalized earnings power is more like the $34B TTM run-rate, growing 30-40%, not 140%. Sector intelligence flagging AVGO as "lagging peers" while revenue accelerates 49% YoY is another signal I'd discount as a momentum-timing artifact, not a fundamental read.
Fair value math: give Broadcom 25x forward FCF on a $32-35B forward FCF estimate (assuming Aug-Q holds as baseline with modest growth) = $800-875B, or roughly $170-190/share. To justify $347 and $1.62T, you need FCF closer to $50-55B within 2-3 years, implying revenue near $130B — plausible only if custom AI silicon compounds at 40%+ and VMware ARR expansion continues without material customer defection. That's the bet. I'd call it overvalued by 40-50% on any normalized-multiple framework, but I concede the momentum and quality are real and the tape can stay disconnected from DCF for years when a name is this central to the AI capex narrative. I side with the synthesis's caution but reject the "high conviction required" hedge as an evasion — the verdict here is overvalued, insiders agree, and the $29.59B quarter needs one more confirming print before extrapolation is safe.
GPT Reading
What jumps out is that Broadcom’s current numbers are much better than the stock’s “priced for perfection” caricature suggests, but the valuation still demands that this step-up prove durable. On a trailing basis through 2026-08-02, revenue is about $89.1B, up roughly 73% from the prior twelve-month period’s $51.6B. More important, this is not low-quality growth: gross margin is 68.8%, operating margin 48.1%, and net margin 33.4%. The quarterly cadence is especially striking. Revenue has moved from $14.05B in the 2024-11 quarter to $29.59B in the latest quarter, more than doubling in seven quarters, while net margin in the latest three reported quarters has run 38.1%, 42.0%, and 44.2%. That is not the profile of a business merely buying growth with an acquisition and then muddling through integration; it looks like a company getting both mix upgrade and operating leverage at once.
The cash profile reinforces that this is a real earnings story, not accounting theater. Last annual free cash flow was $26.9B on $27.5B of operating cash flow with just $623M of capex, and given the TTM revenue and margin expansion, cash generation is likely materially higher now than that annual snapshot. Broadcom’s model remains unusual in semis because capital intensity is tiny relative to scale, which lets revenue gains fall through hard. Balance-sheet risk also looks manageable rather than scary: $65.1B of debt is large in absolute terms, but against a $1.62T market cap, $81.3B of equity, $16.2B of cash, and this level of cash earnings, deleveraging is not the bottleneck. A business compounding revenue this fast while posting nearly 30% ROIC and over 34% ROE deserves a premium multiple.
The issue is that the premium here is extreme even after granting all of that. At $347.26, Broadcom trades at about 19.0x sales, 38.5x EV/EBITDA, and 57x earnings. For a software company with 80% recurring revenue and no cyclicality, those multiples would already imply a long runway; for a company still substantially exposed to semiconductor demand patterns, customer concentration, and large-program timing, they imply several years of near-flawless execution. Even if I capitalize the current TTM net income power at a very generous 35-40x because of the quality and AI exposure, I do not get comfortably above the current price unless I assume another leg of outsized growth. The market is effectively underwriting that the jump from roughly $52B of revenue to $89B is not a one-off post-VMware plus AI surge, but the start of a new base from which Broadcom can continue compounding at rates far above large-cap norms.
The smartest pushback to my caution is straightforward: Broadcom keeps earning the benefit of the doubt. The annual series shows revenue from $27.45B in 2021 to $63.89B by 2025-11 and now about $89.1B TTM, while margins have improved rather than diluted. If VMware were causing customer revolt severe enough to impair value, you would expect pressure in software economics; instead consolidated profitability has exploded. If AI custom silicon were an overhyped niche, you would not see quarterly revenue go from $18.02B to $29.59B in just three quarters with net income keeping pace. The bear case also leans too heavily on traditional semiconductor cyclicality for a company whose business mix and pricing power are no longer traditional. On that view, 19x sales is not “expensive”; it is the market assigning a deserved scarcity premium to one of the few companies converting the AI buildout into immediate profit at scale.
What would change my mind is not another glossy AI narrative, but evidence that the current slope is sustainable. If the next two quarters hold revenue above $29B with operating margin still around 47-49%, and if trailing free cash flow clearly pushes through the low-$30Bs toward $35B+, then today’s valuation starts to look less stretched because the earnings base will have caught up. Conversely, if quarterly revenue falls back toward the low-$20Bs, if net margin slips under 35%, or if debt stays stubbornly high despite the cash machine, then the multiple has no cushion. The insider tape is also not helping at this price: repeated sales do not prove overvaluation, but they do argue against paying a heroic multiple on faith alone.
Grok Reading
Broadcom’s trailing twelve months through 2026-08-02 tell a story of violent re-acceleration, not a gentle AI narrative. Revenue stacked $18.02B → $19.31B → $22.19B → $29.59B across the four quarters, putting TTM sales near $89B versus the $63.9B full-year figure ending 2025-11-02 and the $51.6B of the year before. The most recent quarter alone is up roughly 85% from the $15.95B printed a year earlier, and net margin has marched from 26% to 44.2% in that same span while absolute net income hit $13.1B in a single quarter. Free-cash-flow conversion remains exceptional: the latest annual print shows $26.9B FCF on only $623M of capex, an asset-light profile that is now being supercharged by the VMware software mix and custom AI silicon. Gross margin sits at 69% and operating margin at 48%; ROIC has already climbed to 30% on a TTM basis. The balance sheet can absorb the remaining VMware leverage—$65B of debt against $16B cash and equity of $81B leaves a manageable 0.6 debt-to-equity and a 2.5 current ratio—while the company still throws off enough cash to service it and keep the dividend covered at a 41% payout.
At $347 and a $1.62T enterprise the market is not ignoring any of this; it is capitalizing it at 19× sales, 57× trailing earnings and 38× EV/EBITDA. Those multiples only clear if the sequential $7B-plus quarterly step-ups continue for several more years and if the 44% net margin proves durable rather than a peak-cycle artifact. The raw numbers already embed the first leg of the AI custom-silicon and VMware pricing-power story; what remains is an assumption that hyperscaler capex stays elevated, that Broadcom keeps winning the majority of those ASIC sockets, and that no material customer push-back on VMware licensing materializes. Insider activity is uniformly sales—multiple 25k-share and larger dispositions through June and July 2026—which at least removes any claim that management is accumulating at these levels. The business quality is undeniable; the price simply leaves almost no room for a pause.
The strongest counter-argument is that the acceleration itself is the undervaluation signal. A company that just printed $29.6B in a quarter with 44% net margins and near-100% FCF conversion is already run-rating well above $110B of revenue and $45B-plus of free cash flow. If growth merely moderates to a still-elite 25–30% for two more years, today’s 57× earnings compresses into the low-30s on forward numbers while the absolute cash generation funds both deleveraging and buybacks. The VMware contribution is already visible in the margin expansion, custom silicon design wins are sticky multi-year programs, and the networking franchise remains a tollbooth on AI clusters. Under that reading the 19× sales multiple is simply the market correctly refusing to apply a mature-semi discount to a platform that is still compounding at semiconductor-plus-software rates. I weigh this less heavily because the same data set shows the multiple expanding in lockstep with the print; the stock has already been re-rated for the outcome the bulls describe, so further upside requires the outcome to exceed what is already in the price.
Two consecutive quarters of sequential revenue below $28B, or an explicit hyperscaler cutback that removes a major ASIC program, would flip me from cautious to outright bearish; conversely, another quarter above $32B with margins still expanding would force a re-rating of the duration assumption and move me to neutral-to-positive.
Qwen Reading local experiment · off-panel
The most glaring contradiction in this dataset is the classification of Broadcom as a "narrative_platform" where profit metrics are deemed "lagging indicators," a label that fundamentally mischaracterizes a company generating $27.54B in annual operating cash flow and $26.91B in free cash flow. This is not a pre-revenue tech story; it is a mature industrial machine with a 68.8% gross margin and a 48.1% operating margin, yet the market is assigning it a 57x trailing P/E and a 19x P/S multiple. The revenue trajectory is undeniably explosive, with the most recent quarter hitting $29.59B, a 48.7% year-over-year acceleration that dwarfs the 38% historical CAGR. However, the quality of this growth is suspect when viewed against the balance sheet: total debt of $65.14B against only $16.18B in cash creates a net debt position that is aggressively funded by the VMware acquisition. The "narrative" that this is a pure AI infrastructure play ignores that a significant portion of this revenue growth is likely driven by VMware price hikes and licensing shifts rather than pure semiconductor volume, a distinction that matters when assessing sustainability.
The insider activity data provides a critical counter-narrative to the "unanchored" bull thesis. In the last 30 days, insiders have sold approximately 219,000 shares, including multiple blocks of 25,000+ shares in late June and early July. While the secondary signals label this "Net Insider Buying," the raw transaction log shows a consistent pattern of S-Sale entries with no corresponding large-scale purchases, suggesting that the "buying" is likely small, routine, or gift-related (e.g., the 500-share G-Gift), while the selling is substantial and directional. This divergence between the algorithmic label and the raw data is a red flag. When the CEO and key executives are systematically reducing exposure at a $1.62T market cap, it suggests that the internal view of the risk-reward profile is less favorable than the public "AI chokepoint" narrative implies. The market is pricing in perfection, assuming that Broadcom will not only maintain its custom silicon dominance against NVIDIA and hyperscaler in-house efforts but also successfully integrate a $69B software acquisition without customer revolt, a dual-execution risk that is rarely priced in with such a high multiple.
The strongest case against my bearish lean is the sheer velocity of the cash flow generation. With a 45.3% FCF CAGR and a current FCF yield that, while low on a trailing basis, is expanding rapidly, Broadcom has the financial firepower to outspend competitors in R&D and M&A. The "Berkshire of Tech" thesis holds water if Hock Tan continues to execute serial acquisitions that accrete cash flow, as seen with the VMware integration. The 34.4% ROE and 29.96% ROIC are elite metrics for a semiconductor company, indicating that the capital allocation is highly efficient. If the AI capex cycle extends beyond 2026, as the "strong" narrative intensity suggests, the current 57x P/E could compress to 30-40x on forward earnings, offering significant upside. The risk is not that Broadcom fails, but that it succeeds at a slower pace than the market's "60% IRR-implied growth" requires, leading to a multiple compression that erases the earnings growth.
What would change my mind is a quarterly report showing that the revenue growth is driven by a mix shift toward higher-margin custom silicon (XPU) rather than VMware licensing, and a reduction in net debt. Specifically, if the next two quarters show operating margins expanding above 50% while maintaining 40%+ revenue growth, and if insider selling stops or reverses into buying, the "narrative" would be validated by fundamental execution. Conversely, if VMware churn rates rise or if a major hyperscaler announces a significant in-house silicon program that bypasses Broadcom, the 19x P/S multiple becomes indefensible. The current price of $347.26 is a bet on a flawless execution of a complex, multi-segment strategy in a highly competitive landscape, a bet that the insider selling suggests is not shared by those with the most information.
Big-3 Panel — where each AI stands
Advanced Analysis Forensic deep-dive · separate lenses
Broadcom is a cash-flow powerhouse: TTM revenue of $89.1B (up from $59.9B a year earlier, and $31.7B four years ago), gross margin of 68.8%, operating margin of 48.0%, net income of $29.75B, and FCF of $39.40B. Earnings quality is clean: OCF/NI of 1.94x, accruals of -6.8% of assets, Beneish M of -1.89, and an Altman Z of 12.58 all indicate reported earnings are backed by cash, not accruals. SBC at 9.5% of revenue is elevated but buybacks run 130% of SBC, so diluted share count is actually shrinking (-0.6% CAGR). The blemish is the balance sheet: net cash of -$35.4B against only $24.0B of liquid cash, a legacy of the VMware deal. Coverage is not in question given ~$39B annual FCF, but this is a leveraged compounder, not a fortress. The 2024 margin dip (OpM 27.9%, NI $5.10B) reflects VMware integration/amortization noise, and the snap-back to 48% OpM in 2026 shows operating leverage is working. Insider tape is heavily one-sided sells (53 sells vs 1 buy, ~$284M sold), including large Samueli dispositions - typical for a stock that has run hard, but not a confidence signal. The 'net insider buying' label in the context module contradicts the raw tape and should be disregarded.
Verify before trusting this (6)
- Customer concentration disclosure in 10-K - what % of revenue comes from top hyperscaler customers (Google, Meta, etc.)
- Debt maturity schedule and covenants on the VMware-related debt stack
- Breakdown of SBC by segment and any acceleration tied to AI hires
- Whether recurring VMware software revenue is stable post-repricing/customer churn
- 10b5-1 plan status behind the Samueli and Brazeal sales - programmatic vs discretionary
- Cash tax rate vs GAAP tax rate to confirm FCF sustainability
Broadcom is unambiguously a high-quality franchise (quality 52, elite FCF, expanding margins), but quality is not the question here — the question is whether $347.26 and a $1.62T market cap leave any margin of safety. The e2e synthesis explicitly frames the stock as 'priced for a custom AI silicon oligopoly' with hyperscaler and competitive risks, and the bear case notes the tape embeds roughly 60% growth expectations plus flawless VMware integration. That is a set of assumptions that has to go right, not a discount. On roughly $89B revenue at hyperscale AI mix, the implied multiple is in the mid-teens on sales and a rich multiple on EBITDA even after giving credit for software cross-sell — deserved value sits meaningfully below price under any non-heroic path. Earnings quality is high (score 3) so no haircut, and the balance sheet is levered post-VMware with no net cash cushion, which argues against paying a further premium. Insider selling reinforces that management does not see a discount here either. Net: this is a great business at a full-to-rich price, not a mispricing. I would need a clear step-down before the risk/reward turns favorable; today the gap runs the wrong way by roughly 15-25%.
Verify before trusting this (4)
- Next-quarter AI revenue run-rate and customer concentration (how much is 3 hyperscalers)
- VMware software gross margin trajectory and net debt paydown pace
- Guidance on custom ASIC pipeline vs hyperscaler in-house silicon roadmaps
- Any commentary on FY26 capex commitments from top customers
The setup here is unusually hostile for a name that just posted a ninth straight beat and guided AI revenue up 236%. Multiple headlines flag the same tell: 'orders tripling while the stock keeps sliding,' 'edging sideways and down despite the hype.' That is the signature of a narrative losing its grip - the story has been priced, and incremental good news no longer moves it. Meanwhile the bear talking points are getting real airtime: Google-Marvell taking share, hyperscalers going in-house, a BofA note on 370B of AI chip-financing debt, and a Druckenmiller exit. With a 1.46 beta into a nascent risk-off tape (VIX 17.7, S&P off highs) and 10y at 5%, this exact profile - high-beta, AI-narrative, richly-valued semi - is precisely what gets marked down first. Not all one-way: sector is on a three-day rebound, a top analyst is calling for 75% upside, and the platform-monopoly framing still has cult-medium support. But the price action itself (down on beats, down on peer wins) is the market voting that the narrative is fading faster than the fundamentals. Net pressure leans clearly negative, moderate not extreme, because the story hasn't collapsed - it's just no longer a tailwind.
Verify before trusting this (5)
- Whether AVGO can green on an up-tape day for peers (currently lagging AMD +7% while it managed only +3%)
- Any further hyperscaler custom-silicon defections beyond Google-Marvell
- OpenAI/hyperscaler capex guidance revisions in coming weeks
- VIX breaking above 20 or S&P losing another leg - would amplify the high-beta headwind
- Whether analyst target revisions turn negative or the 75%-upside call gets echoed
The world's compute stack is being rebuilt around accelerated infrastructure, and the economics of that buildout favour the party who sells the connective tissue: custom silicon for buyers who want to escape single-vendor GPU pricing, and Ethernet switching/optics for everyone. Broadcom sits on both, which is why it outgrows a category that is itself booming. The macro backdrop is unhelpful — 5% long rates and a positive-but-modest curve raise the discount on multi-year capex — but AI capex has so far been decided on competitive necessity rather than cost of capital. The structural question is not whether the market grows; it is whether a ~$1.6T-cap franchise can keep compounding at the rate the world has already written down. History says the growth rate compresses long before the demand does.
When we made this prediction on Sep 18, 2026, AVGO was $354.02. We expect it to be $335.00 by Mar 2027, and we consider it great value under $275.00. This is an early model (v0.6.0) — the direction is more reliable than the exact price. Made Sep 18, 2026.
Blue is our prediction, starting the day we made it. Grey is a slower route to the same place — the same destination, taking longer. Black is the actual price, so you can see how we are doing.