Omnicom: cheap on normalized cash flow, but not cheap enough

Omnicom's operating cash flow for the first half of 2026 was -$932.4 million. Its own free cash flow measure, before working capital, was +$1,502.0 million. Working capital absorbed $2,434.4 million in between, and how much of that comes back is the central question of this analysis.

Omnicom is the largest traditional advertising holding company, and since November 2025 it owns its old rival IPG. Operations are working: organic growth was 6.1% in the second quarter and margins are rising as costs are cut. The cash flow is harder to read, because half a year of a merged company isn't enough to say what a normal year looks like.

At $76.15 on September 25, 2026, the shares trade 17.8% below my triangulated Base value of $92.69, an upside of 21.7%. The Bear value of $58.76 is 22.8% below the price and the Bull value of $121.91 is 60.1% above it. The framework's required margin of safety for a Confidence B score with this much integration risk is 30%, which puts the maximum buy price at $64.88. The shares would need to fall about 14.8% to get there. Until then this is a hold: cheap, but not cheap enough for a new position.

Route: SF-01 Standard Value, price to free cash flow, with SF-10 Dividend Income as a supplement. Economic classification: Standard, borderline Quality growth (6/12). Primary method: price to normalized owner free cash flow. Secondary method: price to normalized earnings. Values are present fair-value estimates, not 12-month price targets.

What Omnicom is

Omnicom owns agencies and networks that sell media planning and buying, advertising, precision marketing and data, public relations, healthcare communications, branding, commerce and experiential work. In the second quarter, 52.5% of core revenue came from Integrated Media, 15.7% from Advertising, 11.3% from Public Relations, 11.2% from Experiential and other, and 9.3% from Health. The United States was 59.0% of the total.

On November 26, 2025, Omnicom completed an all-stock merger with IPG. It converted 361.5 million IPG shares into 124,352,188 new Omnicom shares, at a ratio of 0.344 Omnicom shares for each IPG share, for total consideration of $8,893.5 million. After closing, legacy Omnicom holders owned about 60.6% of the combined company and former IPG holders about 39.4%, on a fully diluted basis. IPG's senior notes, about $2.95 billion, moved onto Omnicom's balance sheet.

That makes the company economically different from the one before 2025: more revenue, more debt, more shares, more goodwill and intangibles, and a large integration job. Old Omnicom history is a poor guide. The merger brings media scale, data capabilities and a cost-synergy opportunity, and it also brings leverage, execution risk and restructuring costs. That's why I flag this as a character change.

The latest quarter

EBITA means earnings before interest, taxes and amortization of acquired intangibles. Adjusted EBITA also removes merger and severance costs. Organic growth strips out currency, acquisitions and disposals, so it is the cleanest read on demand. At 6.1%, it is inconsistent with a business that is falling apart.

The cash flow needs more care. Omnicom buys media on behalf of clients, and the timing of paying suppliers and collecting from clients swings cash within the year. The first half of 2025 showed the same shape: operating cash flow of -$576.7 million on a $1,411.7 million working-capital outflow. The 2026 outflow is larger, which fits a bigger company but is also the number to watch. Only the second half will show how much reverses.

Which numbers matter for Omnicom

Annualizing the first-half operating cash flow gives a meaningless negative number. Using the company's own figure isn't much better: Omnicom calls the $1,502.0 million "Free Cash Flow," and it is free of exactly one thing, working capital, which is the item that took $2,434.4 million. The measure starts from net income and adds back depreciation, amortization, stock-based compensation (SBC, pay in shares) and severance charges. So the primary metric is normalized owner earnings per share: what a shareholder could take out in a normal year after capital spending, interest, tax and a normal working-capital cycle. I measure it after SBC, which is a real cost: $52.9 million in the first half, about $106 million a year or $0.38 per share.

My estimates are Bear $7.0, Base $8.3 and Bull $9.3 per share. They are estimates, pending a full year of combined results. A check against reported numbers:

My Base of $8.3 sits 10.3% below that run rate. The gap is deliberate: the run rate still has $99.5 million of integration costs in it that should fade, but it also adds back $51.1 million of severance charges, which are real costs, and it says nothing about working capital or a full year of combined trading. On this evidence the Base is not aggressive.

Shares, debt and buybacks

Diluted shares averaged 281.0 million in the second quarter and 290.2 million in the first half, against 196.0 million and 197.1 million a year earlier. The increase is the merger, not ordinary pay in shares. Omnicom has been buying back stock hard: $2,988.2 million in the first half, roughly twice its pre-working-capital free cash flow for the period. The company says $3 billion has been bought so far this year and plans to finish the whole $5 billion program by April 2027, which leaves about $2 billion to go.

That was funded from cash. Cash and equivalents fell by $3,544.9 million in the first half, and Omnicom also refinanced, borrowing $2,384.9 million and repaying $1,400.0 million of long-term debt. The debt position on June 30:

The company's own net debt figure is $6,617.0 million because it leaves out the bank loans. Its reported ratio of net debt to trailing twelve-month EBITDA is 4.1x, but that EBITDA of $1,613.8 million contains IPG only from late November and carries the merger costs. On the definition in its credit agreement, which uses pro forma adjusted EBITDA, net debt is 1.5x and total debt is 2.3x. The true level of leverage is somewhere between those two views, and it is materially higher than legacy Omnicom's 1.2x a year ago. Interest cost has already more than doubled.

Primary valuation: price to normalized owner free cash flow

Legacy Omnicom traded on higher multiples. I use 10x to 13x for the Base to Bull range because the company now has more debt, more shares and an integration to finish, and I would not pay the old multiple until the new structure has shown per-share cash flow growth. The primary Base is $91.30.

Sensitivity: owner earnings per share and multiple

The price is close to the value of Base cash flow at 9x ($74.70). Every cell at 11x or above on $8.3 or more is worth at least $91.

Secondary valuation: price to normalized earnings

Reported GAAP earnings are distorted by amortization of acquired intangibles and integration charges, so the earnings figures are also my estimates. For scale: first-half GAAP earnings of $3.41 per share plus $0.52 of one-off charges is $3.93, or $7.86 annualized; the second quarter alone, $2.08 plus $0.26, annualizes to $9.36. My Base of $8.5 sits between them. The company's own adjusted EPS, $4.53 for the half, also removes non-cash amortization and I do not use it. The secondary Base is $93.50.

Historical cross-check

Legacy Omnicom supported higher multiples than today's company deserves, and a clean history of the merged company does not exist. As a judgment range I use $90 to $105 per share, with a midpoint of $97.50. This is an estimated framework range, not a sourced trading statistic, and it carries only 10% of the weight. The check says the shares are discounted, and that the discount is partly justified.

Valuation triangulation

The weighted result is $92.6900, and I use it without rounding. The primary and secondary methods differ by 2.4%, well under the 20% warning threshold. Both rest on the same estimated cash flow and earnings, so their agreement says the multiples are consistent, not that the estimates are proven.

What is the market already pricing in?

At $76.15 the shares trade at 9.2x my Base owner earnings of $8.3. Turn it around: at the fair multiple of 11x, the price implies sustainable owner earnings of only $6.92 per share. That is 16.6% below my Base and slightly below my Bear estimate of $7.0.

So the market is pricing some mix of synergy disappointment, client losses, weak cash conversion, leverage that stays high, and lasting damage from AI. Those are reasonable worries, and one of them has just become concrete (see below). What the price does not need is anything heroic. Base cash flow at only 9x is worth about today's price.

Bear, base and bull scenarios

Bear and Bull blend the primary and secondary methods at their 55% and 35% weights, rescaled to 100%, because the historical range has no Bear or Bull. The Base is the full triangulation above. I assign no probabilities to the scenarios. The Bear value of $58.76 is below the 52-week low of $66.33, which is the point of a Bear case.

How the buy price is set

For a Confidence B score, the framework's base margin of safety is 25% to 35%. Integration and leverage push toward the top of that range, and recurring client economics pull back toward the middle. I select 30%. Applied to the $92.69 Base, the maximum buy price is $64.88. That is 2.2% below the 52-week low of $66.33, so it is a level the shares have not touched in a year. The preferred accumulation zone runs from about $58.76, the Bear value, to $64.88. Below the Bear value, first check that the decline is not a thesis break.

Dividend

The quarterly dividend is $0.80, or $3.20 a year, a yield of 4.2% at $76.15. It was raised from $0.70 a year earlier. Dividends paid in the first half were $481.5 million, covered 3.1x by company free cash flow before working capital. On my Base owner earnings the payout is 38.6% ($3.20 of $8.3), and on GAAP earnings 46.9% for the half ($1.60 of $3.41). The mechanical GAAP payout ratio is distorted by merger amortization and restructuring. Coverage from normalized cash flow is the test that matters, and it passes with room.

Reasons to own OMC

• Scale and clients. The largest traditional advertising holding company, with a global client base, agency brands, media buying scale and proprietary data.
• Organic growth. 3.9% in the first quarter and 6.1% in the second, with the margin on core operations up from 15.9% to 17.8%.
• Synergies. Cost savings from combining two large holding companies are only starting to show in the numbers.
• Buybacks. $3 billion repurchased this year at prices below my Base value adds to per-share value if cash flow holds.
• Income. A 4.2% dividend yield at the reference price.

What could go wrong?

• Integration. Persistent client or talent losses, or weak cash conversion, would break the thesis. Purchase accounting was still incomplete at June 30.
• A first concrete client loss. In early September, PepsiCo moved its global media account to Publicis after more than 25 years. Trade estimates put Omnicom's fee revenue from it near $100 million a year, about 0.4% of its run-rate revenue, on $1.7 billion to $1.9 billion of client media spend. The CFO called it a disappointment, said Omnicom is reviewing what went wrong and does not expect a significant effect on its 2027 outlook, and the shares fell about 5% on the news. Omnicom keeps PepsiCo work in public relations, creative and some sports marketing. It is small for the Base value, but it is exactly the kind of loss the thesis breaker describes.
• Balance sheet. Net debt of about $6.7 billion and interest expense that has more than doubled make capital allocation more sensitive than it was.
• Capital allocation. Roughly $2 billion of buybacks are still planned. If they come at the expense of paying down debt and cash flow disappoints, the balance sheet gets tighter when it should be loosening.
• AI disruption. AI is a margin opportunity and also a threat to parts of traditional agency work and fee models.

Management execution

Management closed a very large merger, delivered 6.1% organic growth in the second quarter, lifted core margins by nearly two points, bought back $3 billion of stock, and raised the dividend. Against that, first-half operating cash flow was negative, the PepsiCo loss is unexplained for now, and the case for synergies becoming lasting per-share cash flow is not yet proven while leverage is high. Assessment: 15/20, good. The unresolved test is converting synergies into durable cash flow per share without letting debt drift up.

Stock Analyza scorecard

The score sits at the edge of Quality growth, but merger-driven growth and leverage argue for a conservative standard valuation until per-share cash flow growth is demonstrated. It is not a compounder. Business quality is 74/100: moat 15/20, returns 14/20, balance sheet 8/15, cash quality 11/15, growth and runway 12/15, management 9/10, governance and alignment 5/5. The moat rests on global client relationships, agency brands, media purchasing scale, proprietary data and the cost of switching.

Confidence score

Company filings are current and detailed. Confidence is limited by the lack of a full combined cash-flow year, by the size of purchase-accounting and integration adjustments, and by reported numbers that must be normalized.

What I would watch from here

Review with the third-quarter report and the cash-flow update.

The three most important thesis breakers

1. Integration failure: persistent client or talent losses despite the merger.
2. Structural deterioration in cash conversion, meaning working capital that does not reverse.
3. Leverage rising despite the expected synergies.

Adversarial review

The first draft's core arithmetic reconciled: net debt, the dividend yield and the upside followed from its own inputs. But checking it against Omnicom's second-quarter release, investor presentation and 10-K, and against recent news, found rounding and definition problems, and one missing event.

Net effect: the Base value moved from about $93 to $92.69 and the maximum buy price from $65.10 to $64.88. The Hold and watchlist verdict is unchanged.

The strongest counter-thesis: Omnicom may look cheap because the market correctly expects client attrition, AI-driven fee pressure, integration friction and permanently worse cash conversion after IPG, and large buybacks could delay deleveraging. PepsiCo shows that clients can leave a merged holding company without a pitch. The response: organic growth of 6.1% is hard to square with an imminent collapse, the two valuation methods agree, and the reported first-half run rate is above my Base. Still, only a full combined cash-flow year can validate the normalized figures, which is why Confidence stays at B and the margin of safety at 30%.

Upward corrections considered: the stronger second-quarter growth, synergies, buybacks and a first-half run rate above the Base. Downward corrections considered: higher leverage, working-capital volatility, integration risk, the PepsiCo loss and only two full quarters of combined results. Audit flags: owner earnings, normalized EPS and the historical range are my estimates, not company data. The trailing EBITDA behind the reported leverage ratio is distorted by the merger. The PepsiCo revenue figure is a trade estimate, not a company number. No unresolved arithmetic error remains. The publication gate status is PASS WITH WARNING.

Final verdict: hold at $76.15, watchlist for new money below $64.88

Omnicom looks fundamentally inexpensive, and early performance after IPG is encouraging. But the price does not give the margin of safety that integration risk, higher leverage and an incomplete cash-flow history call for. It takes a 14.8% fall to reach the buy price, and you'd want a full year of combined results to know whether the Base cash flow is real. The first half showed what Omnicom can earn. The second half will show how much of it Omnicom gets to keep.

Verdict: WATCHLIST at $76.15. Triangulated Base value $92.69, upside 21.7%, Confidence B (19/25). Required margin of safety 30%, maximum buy price $64.88.

The case improves if normalized owner earnings hold at $8 or more while net debt falls, or if the price falls into the buy zone with the fundamentals intact.

Sources

Primary sources include Omnicom's second-quarter 2026 earnings release and investor presentation (Form 8-K, July 28, 2026), its Form 10-Q for the quarter ended June 30, 2026, its FY2025 Form 10-K, and its investor-relations dividend releases. The PepsiCo account loss is drawn from trade reporting by Ad Age, Digiday and Campaign. The market price comes from Morningstar and MacroTrends.

The reference market price is $76.15, the September 25, 2026 close. Balance sheet figures are as of June 30, 2026. Normalized owner earnings, normalized EPS and scenario values are analytical estimates, not company guidance.