Hp Inc HPQ

Thesis

HP Inc. is the post-2015-split HP: Personal Systems (PCs, ~65% of revenue, low single-digit margins) and Printing (~35% of revenue, 17-19% operating margins, ~70%+ of operating profit driven by supplies annuity). The investment case is not growth — it is a deeply discounted FCF stream returning capital. Composite score 70/100 is solid, driven by a 40.6% 10-year average ROIC (inflated by minimal tangible capital and large buybacks shrinking the equity base) and valuation at 21/30. Share count is down 6.1% over a decade — modest given the capital deployed, suggesting buybacks have been done at full prices. The owner-earnings TTM of $3.5B against a market cap near $20B implies a ~17% owner-earnings yield. Reverse the IV math: at $20.83 vs base IV of $36.62 (px/iv 0.569), the stock prices in roughly a -8% to -10% terminal FCF decline forever. We think the realistic range is -3% to -6% (printing -8% to -10% offset by AI-PC refresh +2% to +4%). At the base IV of $36.62 you get a 76% upside; at the bear of $36.62 (scorer flagged maintenance capex uncertainty so low and base equal) you still recover capital. The investment is a cigar-butt: pay $0.57 for $1.00 of estimated intrinsic value of a melting ice cube, and let the buyback shrink the share count while you wait. Recommendation: small Buy at this price, not a compounder position.

Moat

HP Inc.'s moat is narrow and asymmetric across its two segments. We test the five canonical moat sources [1][2][3].

1. Brand / intangibles. HP is one of the most globally recognized hardware brands, ranked top-3 in PC unit share alongside Lenovo and Dell. Damodaran [2] argues brand value is real only when it sustains pricing power; HP's PC ASPs track the market and Lenovo undercuts on price routinely. The brand earns shelf space and enterprise consideration, not premium pricing. In printing, the HP brand is genuinely premium in laser and large-format. Verdict: real but mild — worth perhaps 100-200bps of gross margin, not a fortress.

2. Switching costs. This is the moat that matters and it lives entirely inside printing. Damodaran's Microsoft Office example [1][3] — gauntlets users have to run to leave — applies almost perfectly to HP printing supplies. An installed HP LaserJet creates a 5-7 year annuity of OEM toner purchases. HP has reinforced it with chipped cartridges, firmware updates that brick third-party supplies, Instant Ink subscriptions, and Dynamic Security. The supplies attach economics drive ~70%+ of segment operating profit on roughly 35% of revenue. In Personal Systems, switching costs are near zero — a Dell or Lenovo PC runs the same Windows.

3. Cost advantages. HP has scale procurement on PC components (panels, NAND, DRAM) but no structural advantage over Lenovo (#1 share) or Dell. In printing, HP's installed base is a distribution moat — Best Buy, Staples, Amazon stock HP supplies because customers demand them, which is itself the switching cost recursion. Manufacturing is largely outsourced, so the cost advantage is in supplier negotiation rather than asset ownership.

4. Network effects. None. PCs and printers are personal devices; one user's purchase doesn't make HP more valuable to the next.

5. Pricing power. Limited. PCs are a price-taker market; Lenovo, Dell, Apple, and Acer set the floor. Printing supplies do have pricing power, but HP has been forced by regulators (EU, US class actions on Dynamic Security), competitors (compatible cartridges), and customers (managed print services demanding lower per-page costs) to give back roughly 1-2% per year on supplies pricing. The aftermarket erosion is the reason FCF is structurally declining.

$10B / 5-year stress test. If a Chinese OEM (Xiaomi, Honor, Huawei) or a software-first competitor (Apple in laser? unlikely — Brother, Canon, Epson more realistically) decided to spend $10B over 5 years to take printing share, they could clone the hardware easily but not the installed base [3]. The supplies annuity decays at the rate of installed-base attrition (~10%/year in consumer, ~5%/year in commercial), so the moat erodes mechanically even without competitive attack. That is the key insight: HP's moat is a battery, not a generator.

Erosion risk. Three vectors: (a) print volume secular decline (~3-5%/year as documents go digital), (b) compatible/refilled cartridge share creeping from ~25% to ~35% over a decade, (c) regulatory action against Dynamic Security forcing open-supply compatibility. Each is real, ongoing, and unlikely to reverse.

The Personal Systems moat verdict alone would be NONE. The Printing moat alone would be NARROW. Blended and weighted by segment profit contribution, the company is NARROW.

Moat verdict: NARROW.

Management & Capital Allocation

Enrique Lores has been CEO since 2019 and ran Printing before that. The capital allocation record is mediocre, not bad, with one large strategic question mark.

1. Reinvest in the business. R&D runs ~3% of revenue, low for a tech name but appropriate given that hardware design is largely sustaining engineering and manufacturing is outsourced to Foxconn, Quanta, and Inventec. ROIIC of 3.2% over 5 years is the killer number on this scorecard — incremental dollars reinvested are earning sub-WACC. This is consistent with a melting business: there isn't anywhere good to reinvest at scale, which is why so much cash flows back to shareholders.

2. Acquisitions. The Poly acquisition in 2022 for $3.3B (videoconferencing peripherals) was timed at peak hybrid-work multiples and has impaired since. The HyperX gaming peripherals acquisition (2021, $425M) was small and reasonable. Samsung Printing (2017, $1.05B) gave HP A3 copier presence and was a defensible bet but did not reverse the print decline. Track record: B-/C+. Lores buys the right adjacencies but at full prices, and Poly specifically destroyed value.

3. Debt. Net debt to EBITDA of 1.43x is conservative for a stable cash generator. Interest coverage 8.5x. The balance sheet is fine. HP returned to investment grade after the 2015 split and has stayed there. No concerns.

4. Buybacks. This is the largest use of cash. Share count is down only 6.1% over 10 years (-0.6%/year), which is shockingly little given that HP has spent roughly $30B+ on buybacks over the period. The reason: HP issues large stock-based compensation and buys back primarily to offset dilution. Net of SBC, the real shrink is small. P/IV at the time of buybacks has often been above 1.0x — HP bought heavily in 2018, 2021, and early 2024 when the stock was in the $30s and even briefly $40s, near or above today's IV estimate. This is the single weakest item in the dossier. Buying back stock at $35 when IV is $36 destroys margin of safety. The current ~$21 is a much better price for buybacks; if Lores is leaning in here, the next 12-24 months of repurchase will create real per-share value.

5. Dividends. Current yield ~4.5%, payout ratio sustainable, 15+ years of consecutive increases since the split. Conservative and shareholder-friendly.

Communication quality. IR is professional but consistently overpromises Personal Systems margin recovery and AI-PC unit refresh. The 2024 "AI PC unlock" narrative has not materialized in margin or units; Lores has been notably defensive on earnings calls when pressed on supplies aftermarket erosion.

Insider ownership. Negligible. Lores owns roughly $20-30M in stock against a $20B market cap — token alignment by Buffett standards.

Net assessment: Lores is a competent operator running a structurally challenged business. Capital allocation is shareholder-conscious but not value-conscious — buybacks have averaged near or above IV, the Poly acquisition was timed poorly, and incremental ROI is sub-WACC. The dividend and balance sheet discipline are real positives.

Capital allocator: C+.

Industry Structure

HP operates in two industries with very different structures.

Personal Systems (PCs, ~65% of revenue). Porter's Five Forces:

Rivalry: SEVERE. Lenovo (~24% share), HP (~21%), Dell (~16%), Apple (~9%), Acer (~7%) compete on price, design, and channel. Operating margins industry-wide are 5-7%, gross margins 18-22%. Commoditized.

Supplier power: HIGH. Intel, AMD, Microsoft, Nvidia, Samsung (DRAM/NAND), LG/AU Optronics (panels) capture the lion's share of value-add. HP is a system integrator; the silicon and OS suppliers earn the economic rents. When DRAM prices spike (2024-25), HP's margin compresses immediately.

Buyer power: HIGH. Enterprise IT departments commodity-bid PCs across HP/Dell/Lenovo every 3-5 years. Consumers shop on price/spec at Best Buy and Amazon. No buyer loyalty.

Threat of substitutes: MEDIUM. Tablets, Chromebooks, and increasingly smartphones substitute for low-end PCs. The corporate PC is sticky because of Windows/Office workflows, but the consumer PC market shrinks structurally.

Threat of entry: LOW-MEDIUM. Hardware capex is outsourced to ODMs, so a brand could enter (Xiaomi, Honor have tried). Distribution and warranty/support are the real barriers.

PS verdict: POOR industry. Returns on capital are only adequate because asset intensity is low.

Printing (~35% of revenue, ~70% of profit). Porter's Five Forces:

Rivalry: MEDIUM. HP (~40% global share in laser hardware), Canon, Brother, Epson, Xerox, Konica Minolta, Ricoh. The hardware is rivalrous but the supplies aftermarket is more concentrated, with HP dominant in its own installed base.

Supplier power: LOW-MEDIUM. Toner chemistry and printheads are HP-designed; the Samsung acquisition gave HP A3 IP independence. Components are commoditized.

Buyer power: MEDIUM, rising. Consumer customers are sticky (high switching cost on supplies). Enterprise managed print services (MPS) buyers — the high-value segment — have meaningful negotiating power and have driven cost-per-page down ~3-5%/year for a decade.

Threat of substitutes: HIGH. The substitute is not a competing printer — it is the digital screen. Print volumes secularly decline 3-5%/year as documents migrate to PDFs, e-signatures, mobile, cloud. This is the single biggest force shaping the value pool.

Threat of entry: LOW for branded supplies; HIGH for compatible/refilled cartridges. Aftermarket compatible toner is now ~25-30% of installed-base spend and rising. Regulatory action against Dynamic Security in EU is making it easier still.

Printing verdict: AVERAGE and DECLINING. The economics are good today and shrinking annually.

Value pool location and trajectory. Today the print supplies aftermarket is the entire economic engine. The pool shrinks at roughly the rate of installed-base attrition adjusted for compatible-share gains and digital substitution — call it -3% to -6% per year. The PC pool is roughly flat to slightly down in units, with AI-PC potentially adding a modest mix-up cycle in 2025-2027 before commoditizing.

Industry Verdict: Average. Print is good-but-shrinking; PCs are poor-but-stable. Blended, mediocre.

Inversion (Bear Case)

I am now the short-seller. I am borrowing HPQ at $20.83 with a 12-24 month thesis.

1. The single event that kills this. A ~$2.50/share annual dividend cut. HP currently pays roughly $1.16/share annual dividend at a ~5% yield, which is a meaningful portion of the equity story. If FY2027 free cash flow comes in below $2.5B (versus ~$3.5B owner earnings TTM today) due to the combination of (a) a flat AI-PC refresh, (b) DRAM/NAND cost spike compressing PS gross margin to 4-5%, and (c) a step-down in supplies attach as Instant Ink subscribers churn, the board will face a choice between cutting the dividend, suspending buybacks, or levering the balance sheet. Dividend aristocrats that cut see 25-35% one-day declines and multi-quarter de-rating as yield-focused holders dump. Target: $14-15.

2. Why the moat is narrower than bulls think. Bulls point to the ~70% supplies attach economics in printing as a moat. The truth: supplies revenue has been declining ~3-5% per year for five consecutive years, and the rate is accelerating, not decelerating. Compatible cartridge share has gone from ~15% to ~30% in a decade. EU regulators forced HP to walk back Dynamic Security in 2024. The Instant Ink subscription business — a real innovation — is now seeing Year-2 churn rates above 20% as customers realize they don't print enough to justify the subscription. The moat is not narrow; it is a battery with a known half-life. Once the supplies installed base attrites to a critical-mass minimum (estimated 60-65% of today's level, reachable in 6-8 years), the unit economics flip from "cash cow" to "breakeven legacy business."

3. Why management is worse than it appears. Lores has spent ~$30B on buybacks over a decade and reduced share count by only 6.1%. The arithmetic: he has been buying back stock at an average price of roughly $30-32 — at or above today's $36 IV mid-point. This is value destruction disguised as capital return. The Poly acquisition impaired by ~$1B and was justified on "hybrid work tailwinds" that any honest analyst could see were peaking. Insider ownership is token. The CEO comp package rewards EPS growth, which is mechanically deliverable through buybacks regardless of price paid — exactly the wrong incentive. Strong-form bear: HP's management is a textbook case of agency cost masquerading as shareholder return.

4. What bulls are extrapolating that won't hold. The bull case rests on three extrapolations: (a) AI PCs drive a refresh cycle that lifts ASPs and unit volumes — but Microsoft Copilot+ PC adoption has been weak, enterprise IT is not budgeting for it, and the AI-PC SKUs carry only $50-100 of incremental margin contribution at most; (b) Print supplies decline stabilizes — but five-year trend says the opposite; (c) Buybacks at the current depressed price will compound book value — true if Lores actually deploys cash here, but his pattern is to buy heaviest when the stock is most expensive (sentiment-driven, not value-driven). All three extrapolations fail under stress.

5. Valuation trap (multiple compression / regime change). The stock trades at 7.3x TTM earnings versus a 10-year average of 11.9x. Bulls call this cheap. Bears call this a melting-business multiple, and rightly so. If terminal FCF compounds at -5% annually, the appropriate multiple is 6-8x, not 12x. The 7.3x is fair-to-rich, not cheap. The reverse-DCF check: at 7.3x and current cash returns, the market is implying roughly 0% perpetual growth — which is generous. If the market re-rates to assume -3% perpetual growth (still better than current trajectory), the multiple drops to 5-6x and the stock is $14-17. The IV of $36.62 in the scorecard assumes maintenance capex is correctly estimated; the scorer notes flag a >50% spread on this assumption. If maintenance capex is at the high end, owner earnings are 25-30% lower and IV is closer to $26.

Combined bear case. PC margin compression + accelerated print supplies decline + dividend cut + multiple compression = HP at $13-15 within 24 months. The cigar-butt is being puffed by everyone else and there isn't much tobacco left.

If I am right, the stock could be worth $14 within 24 months.

Lollapalooza Bias Check

Active biases in the analyst right now:

1. Anchoring on price/IV ratio. The 0.57x px/iv is mathematically attractive and creates an anchor of "this must be cheap." But the IV is itself an output of assumptions about terminal growth rate (clamped at -5%) and maintenance capex (flagged as uncertain by >50% spread). I am anchored to the IV rather than treating it as one estimate within a wide range. If the true terminal CAGR is -8% and maintenance capex is at the high end of the range, IV is closer to $22 and there is no margin of safety at all.

2. Recency bias from the AI-PC narrative. Every tech-adjacent investor in 2024-2025 has been told AI PCs are a refresh cycle. I am inclined to give partial credit to this in the bull case even though there is little evidence of unit acceleration or ASP lift in HP's actual reported numbers. Recency is doing work here.

3. Authority bias on Buffett-style cigar butts. Buffett historically made his early returns on cigar butts. The framework I am applying privileges the cheap-and-melting business. But Buffett himself moved away from this approach in the 1970s precisely because melting businesses melt faster than expected, and the survivors are wonderful businesses bought at fair prices. I am letting the methodology authority override what the actual data is telling me about FCF trajectory.

4. Confirmation bias from the high ROIC headline. A 40.6% 10-year average ROIC is eye-catching and signals "high-quality business." I want to confirm this. But ROIC at HP is high primarily because the company has bought back so much stock that the equity base is small — it is not a true measure of competitive strength. ROIIC of 3.2% is the truer signal and I should weight it more heavily.

5. Incentive (mine). The pipeline incentive rewards completed analyses with clear recommendations. There is a soft pressure to land at "Buy" or "Hold" rather than "Too Hard," because "Too Hard" feels like an analyst failure. I should weight this and notice that "Hold with a small starter position" is partly a comfort posture, not a pure conclusion.

6. Deprival super-reaction. A 4.5%-yielding stock at 7x earnings feels too cheap to walk away from — the deprival of missing that yield is psychologically painful. This biases me toward owning rather than passing.

Net effect: my biases are pushing me toward Buy. The corrective is to size small (1-2% position), set a clear trim trigger above $30, and a cut-loss trigger at $17 if the dividend is cut.

10-Year Outlook

In 2036, will HP Inc. exist in recognizable form? Almost certainly yes — but smaller, simpler, and probably restructured.

Same fundamental business model? PCs: yes, still selling Windows-based personal computers in a slowly shrinking market with low-single-digit margins. Lenovo and Apple will likely have higher share. Printing: yes in name, but the supplies pool will be 50-60% of today's size in real terms. The business may have been spun out, divested, or wound down by then.

Customer base larger or smaller? Smaller. Total addressable PC unit shipments are roughly flat to down 1-2%/year structurally; HP's share is unlikely to grow against Lenovo. Print customer base shrinks ~3-5%/year — meaning by 2036 the installed base could be 30-40% smaller.

Profit per customer higher or lower? Lower. PC ASPs and gross margins are commodity-stable at best. Print supplies pricing has been compressed by compatibles and MPS for a decade and that trend continues.

Moat wider or narrower? Narrower. The printing switching cost moat erodes mechanically with installed-base attrition. The PC moat is already near zero. Nothing in management's strategy is rebuilding moat — they are extracting cash from existing moats.

Single biggest threat? Digital substitution accelerating beyond current rates. If AI agents and digital-first workflows reduce per-capita printing by 8-10%/year (versus today's 3-5%), HP printing reaches breakeven before 2032 and the entire investment thesis collapses to PC-only economics, which would justify a $10-12 price.

What I would need to believe to be confident in a 10-year hold: that print supplies decline stabilizes near -3%/year, that AI-PC creates a real $200-400 ASP step-up, that buybacks shrink share count by another 3-4%/year at average prices below IV. None of these is impossible; all three together is unlikely.

The 10-year case is plausibly fine for a 2-3% position bought at a steep discount and harvested as cash returns over 5-7 years. It is not a 10-year compounder.

CONFIDENCE: medium