Version	Trigger	What changed	Target
v1	Initial draft	Sum-of-the-parts on consensus segment estimates, opioid-adjusted peer multiples	$935
v2	RBC sub-industry initiation, 19–20 Aug [16]	Corrected a factual error — RBC initiated, did not downgrade. Added IRA Part B headwind effective Jan 2028, previously missed entirely. Resized RxTS variance	$913
v3	Hermes multi-reviewer panel + directed primary-source verification [17][18]	FCF normalisation method revised, which inverts the peer ranking against us. Oncology multiple trimmed to 14.0x. Bear probability cut to 25%. Apollo instrument identified as redeemable convertible preferred — minority interest deduction understated by $1,846M, worth $16.3 per share. Rating firmed to Hold	$889
v3.1	Buyback & Bear dynamics refinement	Added §6.7 (buyback-float funding linkage, ~$73/share SOP impact, 3.8 pts of FY28 EPS growth) and correlated tri-factor bear dynamics in §8. Valuation targets and Hold rating unchanged	$889

Three revision rounds moved the target down 4.9%. None of them challenged the assessment of the business. All of them bit on the price. That pattern is itself informative and is discussed in Section 15.

Hold / Market Weight on a 12-month view. Conviction: Medium-High.

We rate McKesson the highest-quality asset among the three US drug wholesalers and we would own it in preference to Cencora or Cardinal Health. That relative judgment survived three rounds of independent challenge. The absolute entry point at $848.85 did not.

Our sum-of-the-parts produces a base-case 12-month target of $889 (+4.7%) and a probability-weighted target of $873 (+2.8%, or +3.2% including the dividend). The Street average sits at $978 (+15.2%) [1]. At an expected total return below the risk-free rate, this is dead money for twelve months and we say so plainly.

We have deliberately removed the "Neutral 12M / Buy 3Y" formulation of earlier drafts. A dual rating across two horizons is a hedge, not a call. The rating is Hold. The three-year case is set out in Section 8 as a portfolio note, not as a second recommendation.

Entry discipline. The accumulation zone that means something is below $800, where a buyer acquires the shares beneath our FY28-only sum-of-the-parts of $835 and receives the roll to FY29 for nothing. We reject the reviewer suggestion of accumulating at $830–845; that sits 0.5–2% below spot and is noise, not discipline.

The two dated headwinds, neither visible in consensus estimates:

1. The CVS Caremark mail and retail contract, served jointly by McKesson and Cardinal Health, expires in mid-2027 [2]. CVS represents 24% of consolidated revenue and 21% of trade receivables [3]. RBC frames the same exposure as pharmacy benefit channel steerage risk [16].
2. IRA price negotiations on Part B drugs take effect in January 2028, sized by RBC at roughly 70 basis points of revenue headwind [16]. Because branded compensation is indexed to a percentage of wholesale acquisition cost, a negotiated reduction in the price base cuts the fee with no offsetting cost relief and no volume compensation. Part B is physician-administered medicine, so this lands on the Oncology & Multispecialty buy-and-bill economics carrying 20% of segment profit.

Recommended action: hold existing exposure. Do not initiate at spot. Revisit below $800, or on renewal visibility for the CVS Caremark contract during 1H CY2027, whichever comes first.

---

Why this matters before the numbers. McKesson is not a logistics company that happens to be profitable. It is a credit and regulatory intermediary whose economic capital is close to zero, and whose profit is migrating out of distribution and into two adjacent businesses with fundamentally better economics. The market understands this. The question is not whether the transformation is real — it is — but whether an investor is being paid to underwrite a customer contract cliff that arrives before the transformation completes.

Four pillars:

A three-firm oligopoly with regulatory, not physical, barriers. McKesson, Cencora and Cardinal Health move more than 90% of US prescription volume. Property, plant and equipment total $2,668M against $403,430M of revenue — 0.66% [4]. Adding right-of-use lease assets of $2,058M brings the total physical footprint to 1.2% of revenue. The moat is not the warehouses, which are financeable by anyone; it is DEA licensing across fifty states, DSCSA serialisation, suspicious-order monitoring liability, and the ability to extend credit to tens of thousands of dispensing points. Manufacturers outsource this deliberately.

Supplier-funded working capital removes the need for equity. Accounts payable of $59,973M exceed receivables plus inventory of $52,192M by $7,781M [4]. The cash conversion cycle is negative 8.3 days. The consequence is a stockholders' deficit of $2,172M built entirely on treasury stock, not on losses — retained earnings stand at $22,291M against treasury shares carried at $32,005M. This makes ROE, P/B and gearing meaningless for this name and, critically, will cause automated screens to flag or reject it.

Profit mix has already shifted. North American Pharmaceutical contributed 49% of FY26 segment adjusted operating profit on 83% of revenue. Prescription Technology Solutions contributed 16% of profit on 1.4% of revenue, at a 19.45% operating margin — twenty times the margin of distribution [4][5].

The portfolio has been actively simplified while peers have added complexity. Full European exit completed with the Norway disposal in January 2026; Canadian retail largely divested. The Medical-Surgical separation, rebranded Wellverse, has an IPO targeted for 2H CY2027, with Apollo already holding roughly 13% at an implied valuation near $13B [6][7]. Cencora, by contrast, added $11.2B of debt-funded MSO acquisitions; Cardinal added roughly $5.85B and has already impaired part of it.

---

3.1 Reported structure (FY26, USD millions)

Segment	Revenue	% of rev	Adj. operating profit	% of segment profit	Margin
North American Pharmaceutical	336,652	83.4%	3,452	49.0%	1.03%
Oncology & Multispecialty	48,423	12.0%	1,435	20.4%	2.96%
Prescription Technology Solutions	5,805	1.4%	1,129	16.0%	19.45%
Medical-Surgical (Wellverse)	11,507	2.9%	1,029	14.6%	8.94%
Corporate	—	—	(654)	—	—
Total	403,430		6,465		1.60%

Source: Company FY26 results and Bloomberg segment consensus file [4][5].

3.2 What each segment actually earns money from

North American Pharmaceutical. Two distinct economies sit inside one line. On branded drugs, compensation is contractual and indexed to a percentage of wholesale acquisition cost, meaning revenue rises with manufacturer list price inflation even at flat volume. On generics, margin is captured on the buy side through ClarusONE, the sourcing joint venture with Walmart [8]. Generics are roughly a tenth of revenue and a disproportionate share of profit. GLP-1 distribution revenue reached $15B in Q1 FY27, up 24% year on year — high volume, dilutive to margin, and management says so explicitly [9].

Oncology & Multispecialty. A hybrid: specialty drug distribution plus practice management. The US Oncology Network now covers roughly 3,300 providers following the June 2025 acquisition of a 70% controlling interest in Core Ventures for approximately $2.49B, the vehicle established by Florida Cancer Specialists [3][10]. PRISM Vision, 80% acquired in April 2025 for approximately $850M, added ophthalmology and retina [10]. The structural logic is buy-and-bill: oncology and retina practices purchase the drug and bill the payer, so owning the management services organisation locks in distribution of the highest-value molecules in medicine.

Prescription Technology Solutions. The segment that justifies any premium McKesson carries. Three components: CoverMyMeds, the electronic prior-authorisation network sitting inside the prescriber's workflow; patient access and affordability solutions, where manufacturers pay to eliminate abandonment at the pharmacy counter; and third-party logistics, a fee-for-service model where the manufacturer retains title. Consensus models operating margin expanding from 19.45% to 21.09% by FY28 [5]. We think that is the most vulnerable assumption in the consensus model — see Section 7.

Medical-Surgical / Wellverse. Non-acute distribution to physician offices, surgery centres, long-term care and home health. Formal name transition begins January 2027, IPO targeted 2H CY2027 [6]. Revenue growth has averaged roughly 1% for three years; this is the asset being separated because it does not fit the growth profile of the remainder.

3.3 Concentration

Top ten customers: 73% of FY26 revenue. CVS alone: 24% [3]. Top ten suppliers: 72% of FY25 purchases [3]. This is the highest-quality customer concentration of the three wholesalers — CVS is a considerably stronger counterparty than Walgreens under Sycamore ownership, which represents 25% of Cencora's revenue — but it is concentration nonetheless, and the contract is up for renewal.

---

4.1 The repricing cycle is the sector's defining risk, and it has a calendar

Deutsche Bank's 12 August note is the most important piece of sell-side work in this file, and it is not a McKesson note at all [2]. Its argument, in sequence:

1. Cencora filed an 8-K on 12 August confirming that certain Walgreens volume served outside the prime vendor agreement began moving away from 1 July 2026. The market concluded McKesson is the recipient, most likely the specialty business.
2. This follows McKesson taking OptumRx from Cardinal, and Cardinal taking Publix from Cencora.
3. The CVS Caremark mail and retail contract, served by Cardinal and McKesson, expires mid-2027.
4. Drug wholesaler shares fell 2–4% against a flat S&P on the day.

The historical analogue matters. In the 2014–2016 cycle, a price war for the independent pharmacy channel collided with double-digit generic deflation. On 28 October 2016, McKesson cut FY17 guidance and the stock fell 22.7% in a single session, erasing close to $9B of market value; Cardinal fell 14.2% and AmerisourceBergen 13.0% [2]. Deutsche Bank adds an observation worth weighing: few of the senior executives at the Big Three were in post during that cycle.

Our reading. Large retail and PBM contracts are the lowest-margin business the wholesalers write, so the direct earnings exposure to a CVS renewal is smaller than the 24% revenue share implies. The risk is not the contract itself. The risk is that competitive repricing on headline accounts spreads into the independent and small-chain channels, which is where the sector actually earns its money — precisely the sequence of 2014–2016. That is a multiple event before it is an earnings event.

4.2 Structural drivers, unchanged

CMS projects retail prescription drug expenditure growth averaging roughly 6.0% from 2023 to 2032 [3]. Specialty and oncology mix continues to shift favourably. Biosimilar conversion is an emerging tailwind. Branded-to-generic conversion supported Q1 FY27 gross profit and management called out earlier-than-expected conversion [9][11].

4.3 The IRA Part B mechanism, and why it bites harder than it looks

Inflation Reduction Act price negotiations extend to Part B drugs with effect from January 2028. RBC sizes the consequence at approximately 70 basis points of revenue headwind for McKesson [16]. On FY29 consensus revenue of $496B that is roughly $3.5B of revenue.

The profit consequence is disproportionate to the revenue number, for a reason specific to this business model. Distributor compensation on branded pharmaceuticals is contractual and expressed as a percentage of wholesale acquisition cost. When a negotiated price reduces the WAC base, the fee falls proportionally while the physical cost of handling the same molecule is unchanged. There is no volume offset — the same patients receive the same treatments. At McKesson's 3.6% gross margin, a $3.5B revenue reduction of this type translates to roughly $100–125M of gross profit, most of which drops through to operating income. That is approximately 1.3% of FY29 consensus EBIT, or about $0.75 per share.

Part B covers physician-administered drugs — infused oncology, ophthalmic anti-VEGF. This is precisely the buy-and-bill economics underpinning the Oncology & Multispecialty segment and the entire MSO strategy across all three distributors. We reflect this in our Oncology & Multispecialty estimates in Section 7 rather than treating it as a generic sector caveat.

4.4 Regulatory overhang

Opioid settlement: $7.4B payable by McKesson over 18 years from April 2022, plus $518M to Washington and $250M to Oklahoma [3]. Long-term litigation provision stands at $5,091M at 31 March 2026, declining to a modelled $4,654M by FY28 [4][5]. Senator Warren urged the FTC to examine the McKesson/Core Ventures transaction in October 2024 on community oncology concentration grounds [10]. Site-of-care and Medicare Part B reimbursement policy now matters to McKesson in a way it did not three years ago, because the company owns physician practice economics.

---

COVERAGE NOTE — this section is not optional housekeeping. Three separate definitional traps materially distort McKesson's headline multiples.

Trap 1 — three different "adjusted EPS". Company non-GAAP FY26 adjusted EPS is $39.11, up 18%. GAAP diluted EPS is $38.38. The S&P Global Market Intelligence normalisation carried in Bloomberg's FA screen shows $33.64, down 3.65% [12]. A P/E computed on the third number gives 27.2x rather than 18.4x. We use company non-GAAP as reported and reconciled in EEO throughout.

Trap 2 — three different EBITDA. FY26 EBITDA appears as $7,194M in EEO, $6,811M under BEst, and $6,763M as "EBITDA, Adj" in FA [5][12]. We lock EEO as the single reference and every EV multiple in this report is computed on that basis.

Trap 3 — opioid liabilities and enterprise value. Bloomberg's standard enterprise value of $108,434M at 30 June 2026 excludes the opioid provision [12]. Wells Fargo includes it, adding $4,094M at their 2Q28 balance sheet date [11]. This is not a rounding issue — it is 4.5% of EV. Critically, the same exclusion applies to Cencora and Cardinal Health in the peer screen, so applying their unadjusted multiples to McKesson's segment profit and then deducting McKesson's opioid liability would penalise McKesson twice. Our SOP grosses up the peer multiples to an opioid-inclusive basis before applying them. See Section 6.2.

Lease accounting. All nine peer companies report under US GAAP and ASC 842. No IFRS 16 adjustment is required anywhere in the comparison [13].

Quality of earnings — favourable, but qualified. On FY26 the gap between GAAP and non-GAAP EPS is 1.9%. Cardinal Health's is 56%, of which roughly $1.16 per share is recurring cash compensation to physicians at acquired practices, excluded from adjusted earnings. Cencora's GAAP earnings exceed its adjusted figure.

We qualify the earlier version of this claim. FY27 consensus shows GAAP EPS of $41.06 against adjusted EPS of $44.65 — an 8.7% gap, against 2.4% in FY28 [5]. The cause is the Apollo instrument. Apollo's $1.25B investment in Medical-Surgical is redeemable convertible preferred equity, not common stock, so McKesson must recognise changes in its redemption value through GAAP earnings [17]. Q1 FY27 GAAP EPS absorbed a $293M redemption-value adjustment on Medical-Surgical redeemable non-controlling interests, worth $2.51 per share of the $3.59 FY27 gap [18].

This is structurally the same criticism we level at Cardinal: a real economic charge to the common shareholder, excluded from the adjusted figure. The order of magnitude remains incomparable — 8.7% against 56%, and terminating at separation against permanent — but the claim of a categorically clean adjusted number does not survive FY27. $293M of accretion in a single quarter on a $1.25B instrument is a rate that, if repeated, is a financing cost rather than an adjustment. Monitor the Q2 print on 5 November.

---

6.1 Why sum-of-the-parts, and why not a DCF

A discounted cash flow model on this business is actively misleading. Free cash flow is materially driven by expansion of the negative working capital position, which is a balance-sheet phenomenon and not a recurring earnings stream. Projecting it into a terminal value capitalises a treadmill. The consensus cash flow build makes this explicit:

Trade working capital contribution to OCF (USD m)	FY26A	FY27E	FY28E
Accounts payable	+4,317	+4,928	+3,743
Receivables	(1,999)	(2,409)	(2,168)
Inventory	(1,082)	(2,234)	(1,524)
Net contribution	+1,236	+285	+50
Reported free cash flow	5,410	4,840	5,925
Normalised FCF (ex-WC)	4,174	4,555	5,875

Source: Bloomberg consensus cash flow statement [5][12].

The FY27 free cash flow decline from $5,410M to $4,840M is entirely explained by the working capital contribution collapsing from $1,236M to $285M. This is not operational deterioration. It is the growth deceleration in the payables balance, exactly as we would expect with revenue growth stepping down from 12.4% to 7.1%. Consensus has already modelled it; CFRA independently arrives at approximately $4.7B [3]. Any committee reading the headline FCF decline as a warning sign will be reading it wrong.

Consolidated valuation must therefore rest on operating profit, and operating profit spans four businesses earning between 1.03% and 19.45% margins. A single blended P/E has no economic meaning here.

6.2 Peer multiples, opioid-adjusted

Calendarised consensus, CY2027, Bloomberg FPT=BA [13]:

Ticker	Company	EV/EBIT (reported)	Opioid liability	EV/EBIT (adjusted)	P/E
COR US	Cencora	12.8x	~$3.8B	13.5x	15.7x
CAH US	Cardinal Health	13.7x	~$4.3B	14.7x	17.4x
IQV US	IQVIA	18.1x	n/a	18.1x	18.0x
DOCS US	Doximity	17.2x	n/a	17.2x	17.4x
CERT US	Certara	12.5x	n/a	12.5x	22.2x
VEEV US	Veeva	26.9x	n/a	26.9x	25.4x
HSIC US	Henry Schein	10.9x	n/a	10.9x	14.8x

COVERAGE NOTE — Owens & Minor (OMI US): Bloomberg returned no live price, market cap or enterprise value as of 20 August 2026. Net debt of $1.77B and ND/EBITDA of 17.6x are available from the June 2026 filing period. OMI may be subject to a trading halt, delisting process or entitlement gap. It is excluded from the Wellverse comparable set and Henry Schein carries that anchor alone [13].

COVERAGE NOTE — Phreesia (PHR US): CY2028 EV multiples of 1.4x EV/EBITDA and 3.5x EV/EBIT are anomalous, most likely a calendarisation artefact on a small-cap January year-end with a thin consensus sample. Excluded from the RxTS comparable set [13].

COVERAGE NOTE — MSO transaction multiples: No EV/EBITDA multiple was disclosed in any of the nine physician practice or MSO transactions completed by the three distributors since January 2024, including Core Ventures, PRISM Vision, Retina Consultants of America, GI Alliance, Solaris Health and OneOncology [10]. Physician minority rollover structures make clean multiples underivable from public disclosure. The Oncology & Multispecialty multiple in our SOP is therefore a judgment anchored on Cencora's adjusted consolidated multiple plus a growth premium, not on transaction evidence.

6.3 Sum-of-the-parts, FY28E (year ending 31 March 2028)

Segment	Adj. op. profit FY28E	Multiple	Basis	EV
North American Pharmaceutical	3,999	13.5x	Cencora adjusted, in line — see note below	53,987
Oncology & Multispecialty	1,889	14.0x	Cencora + growth premium, less FTC/MSO execution discount	26,446
Prescription Technology Solutions	1,321	16.0x	Below IQVIA/Doximity; GLP-1 PA headwind	21,136
Medical-Surgical (Wellverse)	1,071	12.1x	Apollo transaction mark, confirmed enterprise value	13,000
Corporate	(619)	14.0x	Blended	(8,666)
Gross enterprise value				105,903
Less: net debt (ex-leases), FY28E consensus				(1,462)
Less: capitalised lease liabilities				(2,110)
Less: opioid litigation provision, FY28E				(4,654)
Less: minority and redeemable minority interests				(3,246)
Equity value				94,431
Diluted shares FY28E (m)				113.1
Value per share				$835

The minority interest line was understated by $1,846M in earlier versions — worth $16.3 per share. Directed primary-source verification established that Apollo subscribed redeemable convertible preferred equity, not common stock [17]. The corrected build:

Minority interests to deduct, FY28E
Apollo redeemable convertible preferred in Wellverse ($1,250M plus accretion)	1,600
Physician put options, Core Ventures and PRISM Vision	1,100
Conventional non-controlling interests	546
Total	3,246

Note on Wellverse debt and why we do not deduct it separately. Wellverse carries $4.25B of committed separation financing: a $1.0B Term Loan A funded 1 April 2026, a $1.0B undrawn revolver maturing 2031, and a $2.25B senior secured Term Loan B added 9 June 2026 [17]. These are additive, not alternatives, and our earlier reading of $1B was wrong. We nonetheless make no separate deduction, because the $13B Apollo mark is an enterprise valuation and therefore already reflects Wellverse's own leverage. Deducting it again at group level would double count. Note separately that $993M of net Term Loan A proceeds were used by the Medical-Surgical borrower to repay an intercompany loan to McKesson in April 2026 — not labelled a separation dividend, but cash extracted from NewCo to the parent ahead of the IPO [17].

Segment estimates from Bloomberg consensus [5], adjusted for our variances at Oncology & Multispecialty (−$25M) and Prescription Technology Solutions (−$85M), derived in Section 7.

Cross-check on the Apollo mark, which reconciles to within 1%: $1,250M for approximately 13% implies equity of $9,615M; adding drawn debt of $3,250M gives $12,865M against a stated enterprise valuation of approximately $13,000M [17]. Three independently disclosed figures close. This is the strongest single data point in the valuation.

Caveat on that mark, working against us. Redeemable convertible preferred carries downside protection and therefore prices above common. The $13B enterprise valuation probably overstates what Wellverse common equity is worth in an IPO. Henry Schein trades at 10.9x CY2027 EV/EBIT; Apollo's mark implies 12.1x on a protected instrument [13]. We retain 12.1x in the base case and use Henry Schein's 10.9x in the bear case.

Note on the North American Pharmaceutical multiple. Two reviewers argued for a 0.5–1.0x premium over Cencora on counterparty credit quality, CVS being investment grade against Walgreens under Sycamore leverage [17][18]. We decline, and the reason is a distinction worth stating: credit quality determines whether a customer pays; commercial power determines what terms it can extract. CVS is superior on the first and probably worse on the second — a larger buyer, PBM-affiliated, with its contract in renewal. Awarding a premium for credit quality at the precise moment commercial power is being exercised counts in the wrong direction.

Implied consolidated cross-check: $105,903M / $7,661M = 13.8x EV/EBIT, against a current blended forward EV/EBIT of 14.5x and an EEO FY28 multiple of 14.19x [5]. Our parts are valued, in aggregate, below where the market values the whole. The sum-of-the-parts uncovers no hidden value at $848.85. That is itself the finding: the market has already priced the mix shift.

6.4 Rolling to FY29E and the twelve-month target

The FY28 anchor understates a twelve-month target because a share purchased in August 2026 is valued in August 2027 against a forward window that is roughly 60% FY28 and 40% FY29. Applying identical multiples to our FY29E segment estimates:

	FY28E basis	FY29E basis
Gross EV	105,903	114,567
Net bridge items	(11,472)	(8,834)
Equity value	94,431	105,733
Diluted shares (m)	113.1	109.0
Value per share	$835	$970

FY29 segment splits are our own extrapolation at FY28 growth rates, less the full-year IRA Part B effect at Oncology & Multispecialty (−$100M) and the GLP-1 prior-authorisation headwind at RxTS (−$95M). Sanity check before those adjustments: our FY29 segment total of $8,470M sits within 0.8% of EEO consensus FY29 EBIT of $8,402M [5].

Twelve-month base target: (0.60 × $835) + (0.40 × $970) = $889.

Wells Fargo reaches $933 applying 19x to NTM+1 EPS [11]. RBC, initiating on 19 August, sets $845 [16]. Our $889 sits between two houses that arrived there independently and by different methods. One reviewer objected that anchoring on the base rather than the probability-weighted target is itself a hedge; we agree, and Section 8 gives both, with the weighted figure carrying the rating.

6.5 Sensitivity

Value per share, varying the two multiples that carry the most weight:

NA Pharma ↓ / RxTS →	14.0x	16.0x	18.0x
12.5x	$801	$824	$847
13.5x	$866	$889	$912
14.5x	$931	$954	$977

Each turn on North American Pharmaceutical moves the target by roughly $65. Each turn on RxTS moves it by roughly $12. The entire valuation debate is a debate about the multiple on commodity distribution, which is precisely the variable the CVS renewal threatens. Note that the top-right cell — 14.5x on distribution and 18.0x on RxTS — reproduces the Street average target almost exactly. That is what the consensus is implicitly assuming, and it requires being right on both variables simultaneously.

6.6 Normalised free cash flow cross-check — and a finding that cuts against us

METHOD REVISED IN v3, AND IT WORKS AGAINST OUR CONCLUSION. Earlier versions stripped 100% of the trade working capital contribution out of reported free cash flow. Reviewers objected that this is theoretically too punitive, and they are right. In a negative-working-capital oligopoly, the portion of payables expansion linked to structural industry growth of roughly 6% per annum is permanent funding, not a timing difference. Only the excess above trend should be removed. We now attribute the contribution pro rata to a 6.5% baseline growth rate.

FY26	Reported FCF	WC contribution	of which recurring	Normalised FCF	Yield on market cap
McKesson	5,410	+1,236	648	4,822	4.9%
Cardinal Health	4,525	+2,567	1,192	3,150	5.9%
Cencora	1,569	(1,505)	—	3,074	5.1%

Under the corrected method Cardinal Health screens best on normalised cash flow yield, not worst. The earlier version of this report used the punitive method and reached the opposite ranking, which we used as one of three arguments against Cardinal. That argument is withdrawn.

The two remaining arguments stand on their own and neither depends on the normalisation method: 56% of Cardinal's adjusted EPS consists of adjustments, roughly $1.16 per share of which is recurring cash compensation to physicians; and Cardinal itself guides FY27 free cash flow down 20–30% from $4,971M to $3,500–4,000M, which is the company's own acknowledgement that the FY26 contribution was not recurring. But a reader who weights cash flow yield above earnings quality now has a defensible case for Cardinal over McKesson, and RBC makes exactly that case [16].

On reported free cash flow yield Cardinal screens at 8.4% [13]. The gap between 8.4% reported and 5.9% normalised is the measure of how much of Cardinal's cash generation is balance sheet rather than earnings.

6.7 The buyback is partly funded by the float, and that matters to the bear case

Share count falls from 124.1M in FY26 to 113.1M in FY28 and 109.0M in FY29 — down 12% over three years [5]. This is incorporated in our valuation through the denominator, not added separately as a return component, which would double count. It is worth roughly $73 per share of our FY28 target. At a frozen FY26 share count our sum-of-the-parts would fall from $835 to $762.

It also drives a third of the earnings growth. FY28 adjusted net income rises 9.4% while adjusted EPS rises 13.2%; the 3.8-point gap is the buyback [5].

The point that is not obvious. FY26 repurchases of $4,750M were funded from free cash flow of $5,410M, of which $1,236M came from working capital expansion. Roughly a quarter of the buyback was funded by supplier float rather than by earnings. Consensus then models $5,222M of FY27 repurchases against free cash flow of $4,840M — more than the company generates — while the working capital contribution collapses to $285M. The shortfall is met from cash and new borrowing, which is visible in the modelled long-term debt line rising from $5,259M to $8,874M [5][12].

This compounds the bear case in a way the reviewer panel had exactly backwards. In a scenario where CVS is lost, three things happen at once: earnings fall, working capital consumes roughly $1.9B instead of producing cash (Section 10), and the buyback must slow at the precise moment it would be most useful in defending EPS. The reviewers assumed a working capital release would accelerate repurchases and neutralise the earnings drag. The mechanism runs the other way on all three legs.

---

Consensus is largely right, and we say so. Adjusted EPS growth of 13–14% per annum through FY29 is consistent with the company's stated 13–16% long-term target and with three consecutive years of delivery [5][14]. We diverge in two specific places, in opposite directions, and they substantially offset at the EPS line.

7.1 Above consensus: North American Pharmaceutical, FY27

Consensus models FY27 North American Pharmaceutical adjusted operating profit of $3,766M, up 9.1% [5]. Q1 FY27 delivered $894M, up 19.4%, accelerating from 10.7% in Q4 [9][11]. Wells Fargo models roughly 10% and states explicitly that they see room for upside [11].

Three concrete drivers support a higher number:

• Earlier-than-expected generic conversion. Management attributed Q1 performance to stable utilisation, specialty growth and earlier generic conversion [11]. Generic conversion is where distribution margin is actually made.
• The Walgreens specialty volume transfer. Effective 1 July 2026, volume served by Cencora outside its prime vendor agreement began moving, and the market consensus is that McKesson is the recipient [2]. This lands inside FY27 and was not contemplated in the original guidance.
• Guidance was raised only for the Q1 beat. Management lifted FY27 adjusted EPS from $43.80–44.60 to $44.20–45.00 — a $0.40 midpoint increase against a Q1 beat of approximately $0.38 versus consensus [9][14]. Nothing was added for the balance of the year.

Our estimate: North American Pharmaceutical FY27 adjusted operating profit of approximately $3,900M, roughly 3.5% above consensus.

The counter-argument, which we accept as partially valid: Wells Fargo notes that earlier generics pull forward benefit from later quarters [11]. Some of the Q1 outperformance is timing rather than run-rate. We have sized our variance accordingly rather than extrapolating the Q1 growth rate.

7.2 Below consensus: Prescription Technology Solutions margin

Consensus models RxTS operating margin expanding from 19.45% in FY26 to 20.36% in FY27 and 21.09% in FY28 — 164 basis points in two years [5]. Company guidance for FY27 is revenue growth of 2.5–6.5% against adjusted operating profit growth of 11–15%, which does imply operating leverage [15]. But the segment's economics rest on an equilibrium that is under active attack from two directions:

• Copay accumulators and maximisers. Payers have deployed these specifically to prevent manufacturer copay assistance from counting toward patient deductibles, which attacks the value of the affordability programmes that drive the segment's growth.
• GLP-1 access expansion. Wells Fargo names this as an explicit downside risk: broader GLP-1 coverage reduces demand for prior-authorisation services [11]. Prior authorisation is CoverMyMeds. The same GLP-1 volumes that inflate distribution revenue erode the case for the technology layer sitting above them.

Neither is speculative. Both are visible in current payer behaviour, and RBC independently quantifies the second in its 19 August initiation: GLP-1 prior authorisations pressured by coverage reductions, sized at a 5–10% operating income growth headwind for the segment [16].

Our estimate: RxTS FY28 operating profit of approximately $1,321M, roughly $85M or 6% below consensus.

The sizing deserves an explanation because it moved twice. Our bottom-up margin analysis gave $60M. RBC's top-down framing gives $70–140M. We initially adopted the upper end of RBC's range, then revised down to $85M — the midpoint of two methods that only partially overlap. Adopting the high bound of a single sentence in a news summary over our own build was not defensible. This remains the largest earnings variance in our model.

7.3 Below consensus: Oncology & Multispecialty, IRA Part B

Consensus segment estimates were compiled before the Part B negotiation timeline was widely modelled and we can find no evidence of it in the FY28 or FY29 segment build [5]. Applying the mechanism described in Section 4.3 to the physician-administered share of the segment: approximately $25M of operating profit in FY28, one quarter of effect, and approximately $100M in FY29.

We hold the Oncology & Multispecialty multiple at 14.5x rather than reducing it. The IRA effect is a level adjustment to fee income, not a change in the segment's competitive position or growth algorithm, and the volume trajectory in specialty and oncology is unaffected.

7.4 Net effect and valuation bridge

Our variances no longer cancel. The upgrade at North American Pharmaceutical is worth roughly $134M in FY27; the two downgrades are worth roughly $125M in FY28 and $210M in FY29. We are therefore modestly below consensus on FY29 — approximately 2.5% at the EBIT line — having previously been in line.

That said, the bulk of our difference with the Street remains a multiple call, and we state the bridge explicitly:

	Our SOP	RBC	Wells Fargo	Street average
Target	$889	$845	$933	$978
FY28E adjusted EPS	~$49.6	n/d	$49.94	$50.44
Implied P/E	17.9x	~16.9x	18.7x	19.4x
vs. MCK 5Y average BF P/E (15.8x)	+13%	+7%	+18%	+23%
vs. MCK +1 SD (18.55x)	below	below	at	above

The Street average target requires McKesson to trade above one standard deviation of its own five-year forward P/E range, sustained, through a period containing both the CVS Caremark renewal and the Part B negotiation effective date. We are not willing to underwrite that. Our target now sits below the +1 SD boundary rather than at it — a consequence of the v3 minority interest correction rather than of any further change of view on the multiple.

---

Probabilities: Bear 25% / Base 55% / Bull 20%. Revised in v3 from 30/50/20. Reviewers argued that the 2014–2016 analogue is overweighted because the independent pharmacy channel has been structurally weakened by PBM reimbursement pressure and can no longer transmit contagion as it did then [17][18].

We accept the magnitude argument and reject the mechanism argument. The 2014–2016 mechanism was not independent pharmacy bargaining power — it was wholesalers competing for GPO contracts. That mechanism requires wholesalers willing to buy share, not strong independents, and Deutsche Bank documents exactly that behaviour now: OptumRx to McKesson, Publix to Cardinal, Walgreens specialty to McKesson [2]. Volume displacement is the leading indicator; channel weakness is not a defence. A three-firm oligopoly is the natural terrain for a price war, not protection against one.

But if the independent channel is smaller than in 2014, the damage from contagion is capped. That is a legitimate reason to reduce the probability, not to discard the scenario. Hence 25%, not 20%.

Bear — 25% — "The repricing cycle returns"

CVS Caremark renewal is contested and competitive pricing spreads from headline accounts into the independent and small-chain channels, repeating the 2014–2016 sequence. Generic deflation compounds it. RxTS margin compresses as accumulators bite and GLP-1 prior-authorisation volume declines. Wellverse prices below the Apollo mark, at Henry Schein's 10.9x.

	12 months	3 years
Price target	$700	$700
Total return	−17.1%	−16.1%
FY28E / FY30E adj. EPS	$46.50	~$53
NA Pharma multiple	11.5x EV/EBIT	13.5x P/E (group)
NA Pharma operating margin	~0.95%	~0.90%

Historical calibration: McKesson fell 22.7% in a single session on 28 October 2016 when the last cycle was confirmed [2]. A 17% twelve-month decline is not a tail scenario for this sector.

Three effects compound here, and they are correlated rather than independent. Earnings fall; working capital consumes roughly $1.9B rather than producing cash (Section 10); and the buyback — which supplies close to a third of EPS growth and was already running ahead of free cash flow in FY27 — must slow at the moment it would be most useful in defending EPS (Section 6.7). Any scenario model treating these as separate draws understates the bear.

Base — 55% — "Excellent business, fully priced"

CVS renews on broadly acceptable terms. Generic conversion tailwind moderates but persists. Oncology compounds near the middle of the 13.5–17.5% guidance range before the Part B fee reduction from January 2028. RxTS grows profit at 8–10%, below consensus, as GLP-1 prior-authorisation volume declines. Wellverse IPOs near the Apollo mark. The multiple holds below the +1 SD boundary and does not expand.

	12 months	3 years
Price target	$889	$1,090
Total return	+5.1%	+29.8% (9.1% p.a.)
FY28E / FY30E adj. EPS	$49.6	$62.3
Blended EV/EBIT	13.8x	13.8x
Group adjusted operating margin	1.65%	1.72%

Bull — 20% — "Transformation re-rates the whole"

CVS renews without margin concession, removing the overhang. Walgreens specialty volume proves larger than modelled. Biosimilar conversion accelerates. Wellverse IPOs above the Apollo mark and the remaining company is re-rated on a cleaner, higher-margin profile with RxTS finally valued as the software-like asset it is. We raise the bull RxTS multiple to 18.5x on reviewer argument, but decline the suggested 20x: Doximity trades at 17.2x on a materially higher margin, and 20x would require Veeva-style recognition, which is a bet rather than a bull case [13][17].

	12 months	3 years
Price target	$1,045	$1,390
Total return	+23.5%	+65.2% (18.2% p.a.)
FY28E / FY30E adj. EPS	$52.0	~$72
RxTS multiple	18.5x EV/EBIT	—
Group adjusted operating margin	1.75%	1.85%

Probability-weighted

12 months: $873 (+2.8%, or +3.2% total return including the dividend). 3 years: $1,053 (+24.1%, or ~25.5% total return, 7.9% p.a.).

This is the arithmetic behind the Hold. An expected twelve-month total return of 3.2% is below the risk-free rate. The three-year figure is adequate without being compelling. The asymmetry does not justify initiating at spot.

---

Micro

Catalyst	Horizon	Direction
Q2 FY27 results — Wellverse trajectory clarity after the Q1 one-off admin expense; laps $51M USON equity gain	5 Nov 2026	Neutral / negative
CVS Caremark contract renewal	Q4 CY2026 – Q3 CY2027	Binary, high impact
Q3 and Q4 FY27 — verification season is the seasonal peak for RxTS	Feb / May 2027	Positive
Wellverse name transition begins	January 2027	Neutral
Wellverse IPO — the mark versus Apollo's ~$13B is the read-through	2H CY2027	Positive if above
H2 FY27 investment step-up in growth and AI capability; management guides payback from 2028	H2 FY27	Negative near term
Core Ventures and PRISM Vision integration ramp	Through FY28	Positive
Further oncology or ophthalmology MSO acquisitions	Ongoing	Neutral

Macro

Catalyst	Horizon	Direction
Generic price deflation rate — the single most important sector variable	Continuous	Two-sided
Branded WAC adjustments, typically January	Jan 2027	Two-sided
GLP-1 volume and coverage expansion	Continuous	Positive on volume, negative on RxTS
IRA Part B price negotiations effective — ~70bps revenue headwind, disproportionate profit effect on fee-on-WAC economics [16]	January 2028	Negative, dated, quantified
Medicare Part B / ASP reimbursement reform	2027	Negative for MSO economics
Biosimilar launch cadence	2027–2029	Positive
Tariff policy — management sees no single critical country and describes Mexico and Canada as immaterial suppliers	Continuous	Low
FTC scrutiny of distributor MSO concentration	Ongoing	Negative

---

Customer concentration and contract renewal. CVS at 24% of revenue and 21% of receivables, with the Caremark contract expiring mid-2027 [2][3]. The primary risk in this file.

Sector repricing spreading to the profitable channels. Independent and small-chain pharmacy remains the highest-margin customer category. A price war that reaches it compresses earnings and multiple simultaneously [2].

RxTS structural erosion. Copay accumulators and maximisers, Medicare exclusion of copay programmes under the Anti-Kickback Statute, and GLP-1 access expansion reducing prior-authorisation demand [11]. This segment carries 16% of profit at 19.45% margin and a disproportionate share of the valuation premium.

MSO execution and reimbursement. McKesson has moved into physician practice economics with $3.4B of acquisitions in FY26 alone [5]. Goodwill and intangibles now total $15,395M, 19% of total assets [4]. Cardinal Health impaired $184M against its Navista and ION oncology unit eighteen months after acquisition — the same strategy, further along. Redeemable non-controlling interests of $943M represent physician put options and dilution not yet crystallised [4].

Working capital reversal — and the most likely misreading of this file. Any material revenue deceleration converts the negative working capital position from a cash source to a cash use. Q1 FY27 already showed negative free cash flow of $372M on operating cash outflow of $220M, in a quarter with 20% adjusted EPS growth [9]. Seasonal, but instructive.

THE COUNTER-INTUITIVE POINT. All three independent reviewers of this report got this backwards, one of them quantifying a "$3bn+ working capital release" on a CVS exit [17][18]. The intuition is correct for a positive-working-capital business and inverted for this one. Losing 24% of revenue at an unchanged −8.3 day cycle:

	USD m
Receivables released (24% of 27,985)	+6,716
Inventory released (24% of 24,207)	+5,810
Payables to be settled (24% of 59,973)	(14,394)
Net cash effect	(1,868)

Losing CVS consumes roughly $1.9B of cash; it does not release any. The net position is negative $7.8B, and shrinking it by 24% is an outflow. Two qualifications: a repricing of CVS at unchanged volume barely touches working capital at all — only a loss triggers the effect, and it is negative on either reading — and the outflow spreads across the transition period rather than landing at once. If three competent reviewers with the balance sheet in front of them inverted the sign, a committee member will too. Expect this question and have the table ready.

Wellverse separation financing and the Apollo instrument. Wellverse carries $4.25B of committed facilities, of which $3.25B is drawn, including a $2.25B Term Loan B added 9 June 2026 that we had not identified in earlier versions [17]. Apollo holds redeemable convertible preferred, not common equity, which ranks ahead of McKesson's residual interest and generates GAAP redemption-value adjustments — $293M in Q1 FY27 alone [18]. FY27 income attributable to non-controlling interests is guided at $295–325M [17]. Wells Fargo states directly that they remain sceptical McKesson can create value through the separation [11]. A disappointing IPO removes an assumed catalyst and leaves a smaller parent carrying unchanged corporate overhead against a preferred instrument that must still be satisfied.

Opioid and litigation. $4,654M provision projected at FY28, payments running to 2038 [5].

Screening artefact risk — operational, not fundamental. Negative stockholders' equity of $2,172M, book value per share of $(17.50), and debt-to-capital of 76% will cause automated quality or leverage filters to reject or flag this name [3][4]. Retained earnings are $22,291M positive; the deficit is entirely a treasury stock artefact. Flag this to risk and compliance before committee.

---

Metric	Value	Reading
TBI Multi (composite)	91	Strong
TBI Risk	82	Strong
TBI Quality	75	Solid
TBI Growth	72	Solid
TBI Value	71	Solid
TBI Momentum	39	Weak — the outlier
S&P Global Quality Ranking	B	Average
Relative Strength Rank	83	Strong
Beta	0.30	Defensive

Source: internal TBI dataset as of 10 August 2026, and CFRA [3].

The composite of 91 is strong and the profile is internally coherent with our fundamental work — high quality, low risk, reasonable value — with one exception. Momentum at 39 is the only discordant signal in the entire file, and it sits alongside two rating downgrades in two weeks (RBC to Sector Perform on 19 August at $845, CFRA to Hold on 6 August at $945) against a consensus that remains 80% Buy [1][3]. We treat the momentum score as corroborating evidence for the Neutral twelve-month stance rather than as an anomaly to be explained away.

---

Metric	Value
Covering analysts	20 rated
Buy / Hold / Sell	16 / 4 / 0 (80% / 20% / 0%)
Consensus rating	4.55 / 5.00
Average 12M target	$978.11
Target range	$845 (RBC) – $1,080 (UBS)
Implied return	+15.2%
LTM return	+21.3%
Estimated long-term growth	12.45%
Past surprise, adjusted EPS	Beat 8 of 8

Source: Bloomberg ANR and EEO as of 20 August 2026 [1][5].

The dispersion is 28% and it is not about earnings. Wells Fargo's estimates sit 0.1% below consensus on FY27 and 1.0% below on FY28 [11]. UBS at $1,080 and Wells Fargo at $933 are not disagreeing about what McKesson will earn. They are disagreeing about what multiple it deserves. The entire sell-side range is a multiple debate, which means the consensus target of $978 is a statement about sentiment, not about fundamentals.

CORRECTION — RBC did not downgrade. An earlier version of this report characterised RBC's 19 August Sector Perform as a downgrade and paired it with CFRA's 6 August move to Hold as evidence of deteriorating sentiment. That reading was wrong. RBC initiated coverage of the sub-industry on 19 August, rating Cardinal Health Outperform at $276, and McKesson and Cencora both Sector Perform at $845 and $330 respectively [16]. The observed softening of the Bloomberg consensus rating from 4.74 to 4.55 is compositional — new analysts joining the panel — not a wave of downgrades. Only CFRA genuinely downgraded, on 6 August, and its stated rationale is valuation after significant gains rather than deterioration; CFRA remains constructive on the operating environment and low tariff risk [3].

What the RBC initiation actually tells us, which is more useful than a downgrade would have been. A house with no legacy position on these names, building models from scratch, arrived at three conclusions that bear directly on this report:

• It quantifies our RxTS thesis at 5–10% of segment operating income growth, more severe than our original estimate, which we have since revised upward to match [16].
• It identifies the IRA Part B headwind effective January 2028 at roughly 70bps of revenue, which we had missed entirely and have now built into Sections 4.3, 7.3 and 8.
• It names pharmacy benefit channel steerage risk, corroborating our CVS concern from an independent angle.

And it disagrees with our ranking. RBC prefers Cardinal Health, expecting its Other segment to reach approximately 25% of total operating income by 2028, and describes Cencora as lacking the differentiated high-growth, high-margin businesses that make the competitors interesting [16]. That second point is our RxTS argument almost verbatim. The first is a direct challenge to us and is addressed in Section 13.

Beat 8 of 8 on adjusted EPS. Management guides conservatively and delivers. This is a genuine argument that consensus is too low and cuts against our Neutral stance. We weight it into the bull case at 20% rather than dismissing it.

Where we sit versus the Street. Our base target of $889 is 9.1% below consensus and 5.2% above RBC. We are not contrarian on the company. We are contrarian on the price, and specifically on whether a multiple above the historical +1 SD band survives a contract renewal of this size alongside a dated regulatory fee reduction.

---

Three review rounds were conducted. Two were open-ended; one was directed at specific open items with a requirement to cite primary documents and to state "not disclosed" rather than estimate. The contrast between them is discussed in Section 15.

Round	Reviewers	Outcome
1	Gemini 3.1 Pro; Claude Sonnet via JC-flash	Four objections. One accepted in full, one in part, two rejected [17]
2	Perplexity Thinking (Claude Sonnet)	No new information. Two of three points reproduced the draft's own figures back to it [18]
3	Perplexity, GPT-5.6 Terra, directed at Section 14 open items	Closed two open items. Surfaced a $16.3 per share error nobody had found [17]

Accepted

Free cash flow normalisation was too punitive. The single best objection of the exercise. Stripping 100% of the working capital contribution assumes the float never grows. In an oligopoly whose underlying market compounds at roughly 6%, the growth-linked portion is permanent funding. Revised in Section 6.6, and the revision inverts the peer ranking against our own conclusion.

Oncology & Multispecialty multiple trimmed from 14.5x to 14.0x for FTC scrutiny and MSO execution risk, empirically evidenced by Cardinal's $184M Navista/ION impairment eighteen months post-acquisition. We declined the suggested 12.5x, because reviewers were also asking us to reflect Part B in the multiple when we had already removed it from the numerator — that would be double counting.

Bear probability reduced to 25%, on the magnitude argument only. See Section 8.

RxTS bull multiple raised to 18.5x. We declined 20x, which would require Veeva-style recognition.

The dual rating was a hedge. "Neutral 12M / Buy 3Y" with staged entry covered both sides. The rating is now Hold. We note that the same reviewer then asked us to anchor on the base target rather than the probability-weighted one, which is the identical hedge they had just criticised, and that the proposed $830–845 accumulation zone sits within 2% of spot and is noise rather than discipline.

Rejected

The working capital release on a CVS exit. All three reviewers had the sign backwards. See the boxed calculation in Section 10. This is the most valuable output of the review process, and it came from the reviewers being wrong rather than from what they said.

The 2014–2016 mechanism argument. See Section 8.

A counterparty credit premium on North American Pharmaceutical. Two reviewers argued for 14.0x. See the note in Section 6.3 on the distinction between credit quality and commercial power.

Contested — RBC prefers Cardinal Health

RBC initiates Cardinal at Outperform with 19.6% implied upside, against Sector Perform on McKesson, on the thesis that Cardinal's Other segment reaching roughly 25% of operating income by 2028 creates a flywheel across core distribution [16]. We agree with the observation and reject the conclusion — but with less conviction than in v2, because the free cash flow revision above removed one of our three counter-arguments. The two that remain: 56% of Cardinal's adjusted EPS consists of adjustments, roughly $1.16 per share being recurring physician compensation; and Cardinal guides FY27 free cash flow down 20–30%, its own acknowledgement that FY26 was not recurring. Note also that RBC describes Cencora as lacking the differentiated high-growth, high-margin businesses that make competitors interesting — which is our RxTS argument verbatim, applied by them against Cencora and by us in favour of McKesson. Two analysts, one criterion, two different winners. That belongs on the table at committee rather than discovered there.

Errors made and disclosed

1. RBC's 19 August action was an initiation, not a downgrade. We inferred deteriorating sentiment from a ratings snapshot without the underlying note. The Bloomberg feed had flagged the 4.74-to-4.55 move as compositional and we did not act on it.
2. Minority interests understated by $1,846M, because we did not verify the Apollo instrument. It is redeemable convertible preferred, not common equity. Worth $16.3 per share, and no reviewer found it until the directed round.
3. Wellverse separation debt stated as $1B; it is $4.25B committed, $3.25B drawn. No effect on the sum-of-the-parts, since the $13B mark is an enterprise valuation, but the earlier statement was wrong.
4. RxTS variance oversized in v2 by adopting the top of RBC's range over our own bottom-up build.

Appendix to Section 13 — self-challenges raised before external review

These were logged in v1 before the external panel convened. Two were subsequently confirmed by reviewers.

Challenge 1 — is the CVS risk overweighted? Large PBM and retail contracts are the lowest-margin business the wholesalers write. Direct earnings exposure to a renegotiation is far smaller than 24% of revenue implies. Cardinal lost OptumRx, an even larger contract at 17% of revenue, and its earnings recovered fully within two years while its multiple expanded 30% above its five-year average. Response, partially conceded: the direct earnings hit is manageable. Our concern is contagion into the independent channel, which is what happened in 2014–2016. External reviewers subsequently made the same argument on magnitude grounds and we reduced the bear probability from 30% to 25% in v3, while rejecting their mechanism argument — see Section 8.

Challenge 2 — does a sum-of-the-parts systematically undervalue a compounding business? Applying static peer multiples to a company shrinking its share count 3–4% annually and compounding EPS at 13–14% will structurally understate value against a market that discounts several years forward. Response, conceded. This is precisely why the FY28 anchor gives $875 and the FY29 anchor gives $1,025. We addressed it with the 60/40 blend, but the choice of blend weight is a judgment worth roughly $60 per share and should be visible to the reader.

Challenge 3 — the RxTS bear case may be stale. Copay accumulators have existed since roughly 2018 and RxTS has grown profit at a double-digit rate throughout. Management stated on the Q1 call that growth has diversified across access and affordability solutions and that the segment began supporting the Medicare Bridge programme in July [11]. Response, conceded in part. The Medicare Bridge involvement is genuinely new and directly addresses the Medicare exclusion we cite as a structural ceiling. We have not sized it because we lack disclosure to do so. Our 90 basis point margin variance may prove too conservative.

---

#	Item	Why it matters	Status
1	RBC note, 19 Aug 2026, Ryan Halsted, Sector Perform $845	Only target below spot	CLOSED — initiation, not downgrade. Incorporated: RxTS variance resized, IRA Part B added, ranking challenge logged in §13
1b	RBC full initiation report, as opposed to the Bloomberg news summary	Our IRA Part B sizing rests on a single sentence in a news wrap; the underlying methodology is unseen	Outstanding
2	Wellverse $13B — enterprise or equity value?	~$8 per MCK share	CLOSED — enterprise value, stated verbatim in McKesson press releases of 20 Apr and 2 Jun 2026 [17]
3	NewCo debt quantum	Affects the Wellverse bridge	CLOSED — $4.25B committed, $3.25B drawn; the facilities are additive, not alternatives [17]
3b	Apollo preferred accretion rate beyond Q1 FY27	$293M in one quarter on a $1.25B instrument. If repeated, it is a financing cost, not an adjustment	Outstanding — check Q2 print, 5 Nov
3c	Wellverse post-separation accounting and McKesson's retained stake	Not disclosed; determines whether a residual holding is equity-method, fair value, or nil [18]	Outstanding
3d	Whether the $2.25B Term Loan B is drawn	The 8-K discloses no funded principal [17]	Outstanding
4	Size of the Walgreens specialty volume transfer	Central to our above-consensus NA Pharma call; DB reports market inference, not confirmation [2]	Outstanding
5	OMI live pricing	Second Wellverse comparable; Bloomberg returned nothing [13]	Outstanding
6	Whether broker models capitalise MCK non-GAAP as reported	A $1.16/share treatment difference at Cardinal; check consistency at MCK	Outstanding
7	Segment consensus for FY29	Our FY29 splits are extrapolated, not consensus	Outstanding
8	CVS Caremark renewal timing confirmation from a second source	Anchors our bear case; single-sourced to DB [2]	Outstanding
9	Multi-LLM review per Step 5	Three rounds completed	CLOSED — see Section 13

---

Three review rounds moved the target from $935 to $889, a decline of 4.9%. Not one of them challenged the assessment of the business. Every one of them bit on the price. When independent challenge repeatedly attacks valuation and never the franchise, that is corroboration of the franchise and a warning on the entry point. It is the clearest signal in the file and it points to Hold.

The two open review rounds produced one good objection out of four, and three identical sign errors on working capital. The one directed round — specific questions, primary sources required, estimation prohibited — closed two open items and found a $16.3 per share error that nobody else had seen.

For subsequent deep dives we recommend splitting Step 5 into two distinct passes: an open pass challenging the reasoning, and a directed pass against the "Verify Before Publishing" checklist, with mandatory primary-document citation and an explicit instruction to answer "not disclosed" rather than infer. The second pass is where the value is.

One further observation, worth recording as a control weakness. In round 2, two of three reviewer points reproduced the draft's own figures — 20.2% margin, $60M impact — back to us as independent findings. A reviewer holding the draft cannot corroborate the draft. Reviewers checking arithmetic and internal consistency should see the document; reviewers asked to challenge conclusions or supply external facts should not.

---

1. Bloomberg ANR, McKesson Corp, 20 August 2026 — ratings, targets, analyst records.
2. Deutsche Bank Research, 10 Years Gone: Drug Wholesaler Switches Drive Repricing Fears, George Hill et al., 12 August 2026.
3. CFRA MarketScope Advisor, McKesson Corporation stock report, Daniel Rich, 6 August 2026.
4. McKesson Corporation, FY2026 fourth quarter and full year results, Form 8-K Exhibit 99.1, 7 May 2026.
5. Bloomberg EEO consensus overview and Company Financials segment file, MCK US Equity, 20 August 2026.
6. Business Wire, Introducing Wellverse: Reducing Complexity So Care Comes First, 6 August 2026.
7. Bloomberg News, McKesson's McKesson Medical-Surgical Sees IPO in 2H of 2027, 6 August 2026.
8. Competitive landscape and generic sourcing joint ventures — CFRA industry commentary, per [3].
9. McKesson Corporation, Q1 FY2027 results release, 5 August 2026.
10. Bloomberg M&A data and news, US physician practice and MSO transactions, January 2024 – August 2026.
11. Wells Fargo Securities, McKesson Corporation: Updating Model Following 1Q27, Stephen Baxter CFA, 10 August 2026.
12. Bloomberg FA Financial Analysis, MCK US Equity, 20 August 2026.
13. Bloomberg calendarised peer valuation screen (FPT=BA), nine-company comparable set, 20 August 2026.
14. McKesson Q1 FY2027 earnings call transcript, 5 August 2026.
15. McKesson Q1 FY2027 segment guidance, per [9] and [11].
16. Bloomberg News, McKesson, Cencora Newly Rated as RBC Initiates Drug Distributors, Angel Adegbesan, 20 August 2026 — RBC Capital Markets initiation, Ryan Halsted: Cardinal Health Outperform $276, McKesson Sector Perform $845, Cencora Sector Perform $330.
17. Primary-document verification round, 21 August 2026, covering: McKesson press releases of 20 April 2026 (definitive Apollo agreement) and 2 June 2026 (completion) — $1.25B convertible preferred for approximately 13%, "approximately $13 billion total enterprise valuation"; McKesson FY2026 Form 10-K filed May 2026 — Term Loan A-1 and A-2 of $750M and $250M funded 1 April 2026, $993M net proceeds applied to repay an intercompany loan, plus a $1.0B revolver maturing 1 April 2031; Form 8-K dated 9 June 2026, accession 0000927653-26-000167 — $2.25B senior secured Term Loan B due 2032.
18. Hermes multi-reviewer panel, 20–21 August 2026 — Gemini 3.1 Pro, Claude Sonnet via JC-flash, Perplexity Thinking, Perplexity GPT-5.6 Terra. McKesson Q1 FY2027 results press release, 5 August 2026 — $293M redemption-value adjustment on Medical-Surgical redeemable non-controlling interests; FY2027 guidance for income attributable to non-controlling interests of $295–325M.

---

Prepared for internal use at Hyposwiss Private Bank Genève SA. Not investment advice. Not for external distribution.