**(Excerpt from today's Focused Compounding article)**

*"My belief is that the market undervalues the ad agency business model. It doesn't understand the P/E premium over the market that an ad agency would need to trade at to equalize the likely future return of the ad agency with that of a "normal" business. Every year, an ad agency both grows organically (in line with growth in the ad budgets of its existing clients) and is able to payout 100% of its earnings. Normal businesses can't do both of these things at the same time. So, they can grow 5% a year, but they can only pay out say 50% of earnings. That means a normal business trading at a P/E of 15 would be priced to return 8.33%. Actual returns in the stock market have not been exactly 8.33%. But, they've been close and they've been close for the reason I just explained.*

*The typical stock grows 5% a year organically, it pays out half its earnings in dividends (or share buybacks), and it trades at a P/E of 15. Those conditions will - in the very, very long-run - give you a return of 8.3% a year.*

*A P/E of 15 is an earnings yield of 1/15 = 6.67%. Half of 6.67% is 3.33%. So, I am saying that to the extent stocks tend to trade at a P/E of 15 and retain 50% of earnings they will tend to have an annual payout (in either the form of dividends or reductions in shares outstanding) of 3.33% of their market price. When you add this "yield" to the growth rate, you get a return for the investor of 8.3%. In some periods, the Shiller P/E - or whatever normalized valuation measure you want to use - will expand and returns may reach 10% or 12% over a certain 15 years. But, in other times valuations may contract and there will be 15 year periods where returns are just 6% or even 4%. For a typical business: what's really underlying all this is about an 8% to 8.5% increase in intrinsic value (which is normally turned into about 5% sales growth and like 3% to 3.5% dividends and share buybacks).*

*Now, do the same math with an ad agency and you get a different number. Ad agencies retain no earnings as they grow 5% a year. So, if an ad agency stock trades at the same P/E ratio of 15 as a "normal" stock, it will have a 6.67% yield in terms of what it is going to use on dividends and buybacks. Add that to the growth rate and you get an 11.7% long-term expected return instead of an 8.3% long-term expected return. That's an inefficient pricing.*

*How would the market efficiently price ad agency stocks?*

*The market would need to put a P/E of 30 on ad agencies instead of a P/E of 15. Value investors don't like hearing this. But, it's true. If Company A can grow 5% a year and pay out 3.3% and Company B can grow 5% a year and pay out 3.3% - they're equalized in terms of future return expectations. Because ad agencies have a 100% earnings payout, they will only reach intrinsic value parity with the market when they trade at a P/E of 30 and the market trades at a P/E of 15. The earnings yield on an ad agency stock should be half the earnings yield on the overall market for it to be correctly valued.*

*In almost all years, ad agency stocks don't trade at double the P/E ratio of the market. And yet, in literally all years, they do have free cash flow that isn't needed to grow their business organically. Therefore, the best and simplest capital allocation policy for any publicly traded ad agency would be to pay no "regular" annual dividends and instead commit to "regular" share buybacks.*

*Historically, the stock that has come closest to this ideal capital allocation policy is Omnicom (OMC)."*