Phase Space AI

Valuation

Stryker [SYK]

Stryker Corporation [SYK] — Valuation Analysis (Task 3)

investment-memo v1.6.0 · analysis dated 2026-07-28, migrated onto the Criteria framework 2026-07-29 Spot: $346.57 (scan) / $346.60 (Alpaca close used in the body) · Model of record: SYK_Financial_Model.xlsx

Archetype COMPOUNDER · Implied path: 15.1% required vs 10.8% demonstrated → margin −4.3pp · 12-month target $390 base (+12.5%)

This document no longer states a position verdict. Under v1.6.0 the memo scores every Criteria, blocks on none, and emits an analysis. Whether that analysis justifies a position is a question about a particular book.

---

0. Valuation under v1.6.0 — two horizons, both mandatory

Inputs are taken from reports/scan/SYK_analysis.json (as_of 2026-07-28) rather than recomputed, per the update spec. Reporting only one of the two outputs below is a defect: they answer different questions over different horizons.

0.1 The implied-path test (reverse DCF) — the Valuation Criteria · FAIL, margin −4.3pp

Sensitivity — over the exit multiple, never over scenario probabilities:

The margin turns positive at a 27.5x exit multiple — i.e. SYK's demonstrated 10.8% justifies today's EV if the multiple compresses only from 29.3x to 27.5x over five years, rather than to the growth-matched 22.7x. That is a 6% compression rather than 22%. Unlike EW and ISRG, SYK's breakeven sits inside the plausible range, and that is the substantive difference between them.

DISCLOSED DEFECT IN THE PRE-COMPUTED INPUT (2026-07-29). The update spec describes a "derived terminal margin" in the scan file. There is none — SYK_analysis.json has no terminal_margin key. The published required_cagr_pct inverts exactly to a flat 20.0% terminal EBIT margin applied to every name in the scan universe, the hardcoded 0.20 that coverage_scan.py's own comments record as a defect fixed after these files were written. SYK's own operating margin is 19.5%; substituting it gives required CAGR 15.7% and margin −4.9pp — slightly worse, not better. Stated rather than silently applied.

0.2 The 12-month target · $390 base, +12.5% to spot

Built from near-term consensus and named product-cycle events, with the multiple anchored on SYK's own trading history and percentile — never a peer median projected years forward.

Step 1 — near-term revenue. Alpha Vantage EARNINGS_ESTIMATES, cached 2026-07-28, 26 analysts:

Step 2 — named product-cycle events inside the window, each dated in the Catalyst Criteria calendar: Q2-2026 results (30 July 2026); the Mako Spine / Shoulder commercialisation ramp against the six 510(k) clearances already granted; recovery of the ~$440m of cyber-deferred Q1-2026 revenue; FY2027 guidance (~Jan-2027). No undated catalyst is asserted here.

Step 3 — the multiple, on SYK's own history. Price-to-sales on a constant current share count (383.4m) against an internally-consistent EDGAR-derived TTM revenue series, daily, 2021-01-04 → 2026-07-28 (n = 1,397).

SYK's multiple sits at the 9th percentile of five and a half years of its own history — corroborating, on an independent measure, the NTM-P/E observation in §4 below (20.73x vs a 24.03x five-year mean, −13.7%).

The base case contains no multiple opinion — it is consensus revenue growth at an unchanged multiple. The band is generated entirely by the multiple.

Sanity band vs the external reference. Street average target $389.24 (+12.3%) is within $1 of the flat-multiple base. That is a check on the output, never a calibration target.

CORRECTION — the old target was $376.58 (+8.6%), from probability-weighted scenarios on an asserted exit P/E (18x / 22x / 26x) with the sensitivity run over the probabilities. The new base is $390 (+12.5%), with the sensitivity run over the multiple and the multiple anchored on SYK's own observed distribution. The change is an instrument change, not a change of view.

0.3 Data hygiene and corrections to this document

Nothing below this line has been deleted. §§1–10 are the original body, retained because the evidence is unchanged. Where they state a conclusion this framework no longer draws, a superseding note marks it.

---

Every figure below was read back from the live Excel model after independent recalculation (the formulas engine, a separate implementation from Excel — a stronger check than an Excel readback, since it re-parses and re-evaluates every formula from scratch). Excel itself was unavailable: a concurrent agent held the application, and closing its workbook would have destroyed another memo's work.

---

1. Tie-out to the filed statements — done explicitly, per v1.4.2

"A zero balance check verifies internal consistency, NOT input accuracy. Every actual-year line must be tied back to the filed statement."

Balance check FY2026E–FY2030E: 0.0 / 0.0 / −0.0 / 0.0 / 0.0. The sheet balances and ties.

One formula error was found and fixed during verification: a mismatched parenthesis in the Valuation Criteria flip-point cell (Scenarios & Comps!C28), which the recalculation engine caught by refusing to compile the workbook.

---

2. Consensus vs the house view (Alpha Vantage EARNINGS_ESTIMATES, pulled 2026-07-28, cached)

Revision history (Alpha Vantage, trailing 30 days): FY2026 EPS 2 up / 1 down; FY2027 EPS 1 up / 2 down. 90-day drift is flat to marginally negative ($16.744 → $16.718 on FY2027). Estimate revisions are a mild headwind, not a tailwind.

Where the house view sits inside consensus dispersion (Exhibit 15)

• FY2026 revenue $27,555m sits at the 91st percentile of the $27,070–27,601m range — inside it.
• FY2027 revenue $30,172m sits above the $29,991m high, by 0.6%.
• FY2026 adj EPS $15.073 sits at the 97th percentile — inside.
• FY2027 adj EPS $17.083 sits above the $16.96 high, by 0.7%.

On three of four lines the house view is inside the range of published estimates. On the two where it is outside, the margin is 0.6–0.7%. That is not a variant view; it is agreement with a rounding difference.

Consensus rating and price target — and the full bridge (required)

Sourced by web search against public aggregators, per references/consensus-bridge.md:

• Consensus rating: Buy. 21 buy / 7 hold / 0 sell (28 analysts, S&P Global).
• Average price target: $389.24 (+12.3% from spot). Range $315 – $465.
• The house probability-weighted 12-month target is $376.58 (+8.6%).

The bridge — the house and the Street disagree by 3.3%, and it is entirely the multiple, not the numbers. Decompose on a shared FY2027E adjusted EPS base:

The house revenue and EPS forecasts are 1–2% ABOVE the Street's. The house target is 3.3% BELOW the Street's. The entire difference is a 1.3-turn multiple disagreement. This is the same decomposition that has now recurred across ISRG, NTRA, TXG and TWST (calibration item B8) — and see Section 7, where SYK breaks the pattern in an important way.

What the risk to the house view is, if the Street is right: if the FY2026 cyber recovery is complete rather than half, and the multiple reverts to SYK's own 5-year mean of 24.0x rather than the 22.0x used, the target is $17.08 × 24.0 = $410 — 18% upside, and declining to own it would be a real opportunity cost. That is the honest statement of what is being given up.

---

3. Discounted cash flow

Result (Base): $330.80 per share, −4.6% versus spot. Bear $278.48 · Bull $381.14. Terminal value is 82.6% of enterprise value.

The reverse DCF is the number that matters

Because 82.6% of the DCF is terminal value, the point estimate is not independent evidence — it is a restatement of the 3.0% perpetuity assumption. Reversing it:

At $346.60, the market implies 3.22% perpetual free-cash-flow growth, against a WACC of 7.53%.

That is roughly nominal GDP, for a business that has compounded organic revenue at 10.4% for five years and owns a switching-cost installed base. The market is not pricing Stryker for growth. It is also, therefore, not pricing in a disaster — this is a fair-value reading, not a mispricing.

---

4. Comparables (prices Alpaca SIP 2026-07-28; peer NTM EPS are estimates, explicitly flagged — no paid feed)

¹ Alpha Vantage consensus, 26 analysts — sourced. ᵉ estimate, unsourced.

SYK trades at a 9.3% premium to the peer median — for the fastest organic grower in the group, with the highest gross profitability and the only structural installed-base moat. That is a modest premium.

SYK versus its own history — the single most important valuation fact

NTM P/E, computed as the 28-July closing price (Alpaca SIP) divided by that year's eventual adjusted EPS:

Only July-2022 was cheaper (19.83x), and the stock returned +30.7% over the following twelve months.

This is a genuine, sourced observation — and it is NOT a thesis. The founding correction of this framework (NET, 2026-07-26) was that a rich multiple is not a short. The mirror is exactly as binding: a cheap multiple is not a long. It is scored as evidence weight in Valuation Criteria, and nowhere else.

---

5. Factor & Anomaly Scorecard (required) — scored for a long

Synthesis — the factors are genuinely split, and split along a specific fault line

The accounting-quality factors support a long (accruals −3.96%, GP/A 0.336, 155% cash conversion). The price and capital-allocation factors fight it (12-1 momentum −16.9%, asset growth +11.3%, F-score 6/9, no insider buys, revisions drifting down).

The split is not random. Every supportive factor measures how well the existing business converts sales to cash — and Stryker is excellent at that. Every hostile factor measures what the market has done to the stock and how the company deploys capital — and on both counts the record over the last year is poor.

This is exactly the configuration where the fundamental narrative is most seductive and least decisive. The scorecard does not resolve the name; it constrains how much weight the narrative may carry, and the answer is "less than it wants."

---

6. Serial-acquirer analysis — the most important analytical point in the name

The brief asked for this to be run with care, because the Cooper/Gulen/Schill asset-growth anomaly is precisely the flag for roll-ups. It is run here, and the conclusion is more nuanced than either "Stryker is a roll-up" or "Stryker is a compounder."

The capital record

• $14.9bn on M&A against $4.3bn of capex — 3.4x. M&A is Stryker's primary capital-allocation channel.
• $1.46 of M&A cash per $1 of revenue added over 2019-2025 ($14.9bn spent, $10.2bn of revenue added).
• Goodwill + intangibles = $24,972m = 52.2% of total assets. Shareholders' equity is $22,420m, so tangible book value is −$2,552m.
• FY2025 alone: $4,960m of acquisitions against $4,283m of free cash flow — the year's M&A exceeded the year's entire free cash flow, funded by a $2.67bn increase in long-term debt.

Organic versus acquired — the honest decomposition

The disclosed first-year acquisition contributions are small at the consolidated level: FY2022 +1.3%, FY2023 +0.1%, FY2024 +0.5%, FY2025 +0.4%. Compounding the disclosed organic-growth series from FY2020 to FY2025 gives +59.4% against reported growth of +68.7% — implying roughly 86% of the revenue added was "organic" on the company's own definition.

That number is real but it overstates the case, for two reasons, and both must be stated.

1. The disclosure convention is generous. Acquired revenue graduates into "organic" after twelve months. A business bought in year one is compounding inside the organic line from year two onward. "Organic" therefore measures growth of a base that was repeatedly purchased.
2. The consolidated line nets an inbound against an outbound. In FY2025, Inari added +4.7pp to MedSurg & Neurotechnology while the US spine divestiture removed −5.7pp from Orthopaedics, netting to +0.4pp at the total-company level (Exhibit 5). The consolidated figure is not small because M&A is small; it is small because two large opposite effects cancelled. A reader looking only at "acquisitions/divestitures +0.4%" would conclude Stryker barely acquires anything. It spent $4.96bn that year.

The verdict, and why it stops short of a short thesis

Stryker acquires like a roll-up and converts cash like a compounder. The two most damning roll-up signatures — accruals and cash conversion — are absent, and they are the ones with the strongest empirical support. What is present is a balance sheet that has been financialised (negative tangible book), a non-GAAP gap that is widening again, and disclosure that is getting less granular every year.

That combination is a quality discount, not a short thesis. It is reflected in the 22.0x Base exit multiple — below SYK's own 24.0x five-year mean — rather than in a directional call.

---

7. Base-rate check on the scenario growth assumptions (required)

Reference class: US-listed companies above $20bn of revenue sustaining 8%+ organic growth for five consecutive years. This is a top-decile outcome, and the modal modelling error is extrapolating the recent rate (Chan, Karceski & Lakonishok 2003: growth persistence beyond chance is nearly nonexistent).

Stryker has already produced the top-decile outcome once: organic growth of 12.6% / 9.7% / 11.5% / 10.2% / 10.3% in FY2021-25, with sixteen consecutive quarters between 9.2% and 13.6%. That is evidence the reference class is the wrong one for this specific company — but a company that has beaten the base rate for five years is not thereby exempt from it for the next five. The Base case splits the difference by fading.

Margin base rate. The Base case adds 280bp of adjusted operating margin over five years (26.3% → 29.2%), against the ~100bp/year Stryker actually delivered in FY2024-25. That is a deceleration of the recent rate, deliberately, because margin expansion of that vintage is usually mix- and scale-driven and both fade.

---

8. Scenarios (all read back from the live model) — retained as evidence; the Valuation Criteria itself is §0.1

Probability-weighted target: $376.58. Expected return, net of 15bp round-trip slippage: +9.55%.

Exit-multiple justification (this assumption carries the whole result — stated openly)

• Base 22.0x is below SYK's own five-year mean (24.03x), below the four-year post-2022 mean (23.25x), and above today's 20.73x and the DCF-implied 19.4x. It assumes partial, not full, reversion.
• Bear 18.0x is below the July-2022 trough of 19.83x — a genuine bear multiple, not a token one.
• Bull 26.0x is roughly the July-2025 level (26.48x) and below the July-2021 peak (27.31x).

The break-even exit multiple, computed live in the model: 20.03x against today's 20.73x.

SUPERSEDED (2026-07-29). The original sentence read "above that, Gate 4 passes … which still clears the hurdle". Both the hurdle and the pass/fail-on-E[R] construction are retired. The break-even multiple itself is the useful residue, and the v1.6.0 equivalent is in §0.1: the implied path's breakeven exit multiple is 27.5x EV/EBIT against 29.3x today. Note these are different multiples on different bases (NTM P/E vs EV/EBIT) over different horizons (12 months vs 5 years) and must not be compared to each other.

---

9. Valuation Criteria result — RETIRED IN FULL AND REPLACED BY §0.1

SUPERSEDED (2026-07-29). This section computed an expected return against a cash hurdle and ranged the result over scenario probabilities. Every component is retired:

• The cash hurdle is retired — replaced by slot competition inside the strategy, not an arithmetic test inside the memo.
• The E[R]-versus-hurdle test is retired — the Valuation Criteria is now the implied-path test in §0.1: required 15.1% vs demonstrated 10.8%, margin −4.3pp, exit multiple 22.7x (GROWTH_MATCHED, n=31), implied compression −22.5%.
• The probability range and the "flip point at P(Bear) = 38.2%" are retired as a sensitivity. Ranging over probability weights while point-estimating the exit multiple runs the analysis on the parameter that cannot change the answer — the defect the reverse DCF exists to remove. The sensitivity now runs over the exit multiple (§0.1), where the breakeven is 27.5x.

The old figures, for the record only: E[R] +9.55% net; cash hurdle 4.70%; excess +4.85pp; range +4.01% to +13.49%; flip point P(Bear) = 38.2%.

One observation from this section survives and is worth carrying forward, because it is about the structure of the old result rather than its arithmetic: the entire excess over cash was carried by the 25%-weighted Bull tail (+9.64pp of a +9.55pp net), with Bear and Base very nearly cancelling (+0.06pp combined). A result that depends on one tail is fragile regardless of which instrument computes it.

---

10. Calibration note — SYK breaks the B8 pattern, and the break matters

CALIBRATION_WATCH.md item B8 (escalated 2026-07-27) records four consecutive healthcare names — ISRG, NTRA, TXG, TWST — where the house forecast was at or above the Street and the name was rejected on the exit multiple. The stated concern is systematically conservative exit multiples.

Stryker is the fifth healthcare name and it is the first with the opposite configuration:

This is counter-evidence to B8 and is logged as such. When the multiple is a tailwind rather than a headwind, Valuation Criteria passes comfortably. The framework is not rejecting names because its exit multiples are structurally too low — SYK's exit multiple was set below the company's own five-year mean and the Criteria still passed by 4.85pp. What binds here is the dissolved variant-vs-consensus test (old Gate 2), and the dissolved variant-vs-consensus test (old Gate 2) binds for a completely different reason: there is no variant view.

That distinction is worth preserving. B8's hypothesis is "we reject on the multiple." SYK says: when the multiple cooperates, we do not. What we still cannot do is manufacture a variant against 28 analysts on the most-modelled story in medtech — and that is a different, and probably more honest, limitation.