Interpublic trades for 5x consensus 2012 EBITDA while peers trade for 8x.  This year Aegis was bought for 9.7x EBITDA.  Interpublic has net cash while investment grade peer Omnicom has net debt/EBITDA of 1.2x.  If Interpublic dividended enough cash ($1 billion) to get their net debt to .5x EBITDA, the post-dividend equity would trade for 7x my estimate of 2013 FCF.  Interpublic has reduced their shares outstanding by 13% over the last 12 months, and they are on track to buy a similar percentage over the next 12 months.  They will also lower their cost of debt from 7% to below 4% in July 2013.  I believe the fundamentals of ad agencies benefit from digital advertising.  At a valuation comparable to peers I believe Interpublic could triple in value in three years.

How the Business Works

Historically, ad agencies took a 15% cut of their clients’ ad budgets and for that clients got creative services and benefited from the agency’s media buying power, which is often 20x or more than even its largest client’s budget.  Realizing the agency dilemma inherent in such a compensation arrangement, powerful advertisers moved to individually-negotiated contracts usually based on hourly rates.  The hourly rates are usually explicitly designed to provide the agency with 13-15% EBITDA margins when SG&A is allocated. 

Agencies purchase media for clients.  Agencies pay production and media charges after they have received funds from clients.  Because payables are paid more slowly than receivables are collected, agencies maintain a float.  They act as the client’s agent rather than the primary obligor.  In some cases agencies agree with the media company that the agency will only be liable to pay the production and media costs after the client has paid Interpublic.  Unique to Interpublic, the finance team is incentivized to grow the float.  An agency requires very little PP&E and hence maintains both negative working capital and negative tangible equity.

Agency work encompasses a very broad array of services.  The traditional creative services and media buying are still a large part of their work and their value-add.  Services can also include market research, public relations, direct marketing, Web site design, and social media management.  As an industry, the agencies are a toll business on global advertising spending.  Agencies still attract the brightest creative talent.  Agencies’ share of media purchases has continued to grow from 56% in 2001 to 62% in 2010.

Interpublic’s Lost Decade

As recently as 2000, Interpublic was the largest advertising agency in the world.  Berkshire Hathaway owned 17% of Interpublic from 1974-1984 and made a ten-fold return.  Following a reckless acquisition binge in the 1990s (they acquired 96 companies in 2000) Interpublic entered 2001 levered 2x EBITDA and hemorrhaging clients due to unforeseen client conflicts.  For the next six years Interpublic’s organic revenue declined at -2.2% per year while Omnicom logged 6.2% annual organic growth.  The current CEO was hired in 2005 and eventually returned Interpublic to organic growth and margin expansion.  From 2006 to 2011 Interpublic’s organic growth has averaged 1.8% per year while Omnicom has averaged 2.6% organic growth per year. 

Beginning in 2010 Interpublic began to include this sentence at the top of their 10-K: “Our long-term financial goals include maintaining organic revenue growth at competitive levels while expanding our operating margin.”  In 2010 and 2011 Interpublic delivered on these goals, but they have not in 2012.  Because of their protracted period of poor execution I believe the market likens Interpublic to Buster Bluth and Omnicom to Michael Bluth.  The decision to hire one ad agency versus another has no rational basis, so is to some extent reflexive.  So Interpublic’s recent history of underperformance in account wins/losses relative to Omnicom contributes to a perception among advertisers that Interpublic is somewhat inferior to Omnicom.  However I don’t think the difference in perception is as stark as that between Buster and Michael, but maybe more like choosing a more attractive twin.  I could however easily imagine Interpublic continuing to grow more slowly than Omnicom for the foreseeable future.  My hope is that a 40% lower multiple is enough to compensate for a 1% lower growth rate. 

Why Digital Advertising is an Opportunity

Advertising supply is growing.  As the price of media declines in absolute terms or relative to advertising spending, I believe advertisers will buy a larger quantity of media.  Advertising budgets are now spread across a larger array of media.  Digital ad campaigns are generally broadcast across many more outlets than traditional ad campaigns.  The complexity of the ad spending decision is growing, and complexity strengthens the value-add of the agencies.  I believe the whole range of agency services are as valuable as ever to advertisers.  The supply of useful data about the effectiveness of digital advertising is also growing, opening an opportunity for agencies to help clients analyze the data.  Ad Age estimates that the agencies generated 30% of their revenue from digital services in 2011.  Interpublic states, “The share of client expenditure available to agencies as opposed to media owners is greater in the digital space.”  Agencies are capturing a higher share of digital advertising spending than of traditional advertising.

On an unrelated note, relative to the media owners, the agencies are considerably less encumbered from entering emerging markets.  Local media regulation often prohibits foreign media companies from entering.  Local agencies are also relatively less entrenched than local media owners, as in the case of Baidu.  Agencies generate a higher percentage of their revenue internationally than almost all of the media owners.

Management

CEO Michael Roth came to Interpublic in 2005 to pull the company out of its tailspin.  Today he owns $11 million of common stock and $61 million notional value of options, so he is well incentivized to execute.  At the Goldman conference three weeks ago Roth said, “We're in a mode right now where we're returning capital to our shareholders in the form of buybacks.  We've already spent over $500 million on share buyback.  We've reduced our outstanding shares by 13%.  We instituted a dividend.”  I think he realizes there’s a big gap between Omnicom’s multiple and Interpublic’s multiple.  After a rough 12 years I think the company has been forced to get religion and has adopted a focus on per share value that Omnicom has practiced with great success since 2003.  When your competitors’ stocks are up four times as much as yours in 12 years, it’s time to look at your life.

In mid 2010 Nick O’Brien was put in charge of McCann, Interpublic’s largest agency brand, making up an estimated 40% of Interpublic revenues.  McCann is the source of Interpublic’s recent account losses.  O’Brien came to Mediabrands, Interpublic’s media business, in 2008 and by 2010 pulled it out of ten years of no growth into the fastest-growing media business in the industry.  Many are optimistic about O’Brien’s leadership of McCann but so far results have been mixed.  I have no special insight into the likelihood of a successful turnaround of McCann, but I think it's a bigger challenge to turn around a creative organization than a media organization.

The Bad

One doesn’t have the warm fuzzy reassurance of past success when investing in Interpublic, as you would with Omnicom.  Even this year two large account losses will cause organic growth to trail expectations by about 2%.  Interpublic doesn’t have a long history of buying in shares, the buyback program was implemented in 2010.  Instead they have a rich history of issuing shares faster than Zimbabwe nickels.  To buy this investment thesis one has to believe marginal growth isn’t impaired and marginal capital allocation will be value maximizing.  Neither premise is a certainty.  My hope is that Interpublic is priced for failure and I have a free call option on just doing average.

Model

¹ For this model I’ve converted the convertible notes to equity for maximum dilutive effect.  In the worst case I do not convert them since they are anti-dilutive below $11.9952.

² For this model I’ve converted the preferred shares to equity for maximum dilutive effect.  In the worst case I do not convert them since they are anti-dilutive below $13.12.

³ I’ve added back option expense because Interpublic is a net canceller of options.  I’ve included the outstanding options in the fully diluted shares outstanding.  Including both expense and dilution would be double counting.

⁴ I refinance the 10% notes in July 2013 with the undrawn $1 billion credit facility.

I model 2013-2015 organic growth at the low end of long term expectations, and 2012 at a below-consensus 2%.  My 2013 and 2014 EBITDA and net income estimates are slightly below consensus.  However, my free cash flow per share estimates are substantially above consensus because the sell side doesn’t model much, if any, repurchase activity, so I think they are missing a huge source of potential EPS growth.  In this scenario I have Interpublic slowly raising their net leverage through buybacks to .5x EBITDA, which is in line with the aggregate industry leverage ratio and hardly irresponsible, even relative to their covenant of 2.75x.  It’s worth noting that although my modeled buybacks might seem aggressive, Interpublic has reduced their diluted shares outstanding by 13% over the past 12 months.

My worst case scenario mirrors exactly what happened in 2009, so I believe I’m being adequately punitive.  Interpublic is more heavily weighted toward North America than all of its peers, and less weighted toward continental Europe, so one could imagine a European/Asian meltdown might not be as bad for Interpublic as the 2008-2009 US recession.

Valuation

                             3 Yrs Out            Worst Case

FCF/Share          2.55                     1.26

Multiple              14                        8.5

Stock Price         35.70                   10.70

I think Interpublic has substantial upside potential with moderate downside risk.