Description

Masquerading as a low-return, low-margin business, Medco Health Solutions is really a high-return operation with respectable profit margins, double-digit growth prospects, and a conservative balance sheet, selling at an attractive price. As I’ll explain shortly, Medco has good reasons for portraying its financial results in a poor light, leading to some misconceptions, which, along with overreaction to the recent lawsuits and state investigations launched against Medco’s two competitors, Caremark and Express Scripts, have contributed to create an attractive investment opportunity -- the stock is selling at about $31, down from its 52-week high of $39.
A previous write-up of Medco is available on this website. Rather than cover the same ground again, it might be best for me to provide a link to that write-up at http://www.valueinvestorsclub.com/Members/view-thread.asp?id=1163&more=dtrue, and then focus the rest of this write-up only on updated and/or different information which I hope you find value-added.

HOW DO PBMs MAKE MONEY?
Officially, pharmacy benefit managers (PBMs) like Medco make money by administering drug benefit programs. PBMs achieve volume-based cost savings in the form of rebates from drug companies and discounts from third-party retail pharmacies, and then share these cost savings with plan sponsors. Superficially, it doesn’t seem like a recipe for consistent profitability.
IN REALITY, HOWEVER, the large PBMs can make good money, consistently, because they have an informational advantage over all counterparties -- plan sponsors, drug companies, and third-party retail pharmacies. See, PBMs negotiate separate contracts with each of these counterparties, so no single plan sponsor, drug company, or retail pharmacy can see a complete picture of the economic arrangements the PBM has in the aggregate with all the other parties, putting the PBM in a convenient position to skim a bit off the top, here and there, from each transaction processed on its network.
This dynamic is evident in the design of drug-benefit plans for clients, and particularly in the design and management of each plan’s formulary. The so-called “formulary” is a long list of specific drugs covered by the plan, organized along multiple categories, sorted in a particular order of preference, and tinkered with regularly to incorporate new drug and market developments; this list is displayed on the computer screen used by retail pharmacists to confirm eligibility and process claims immediately prior to filling a prescription at the point of sale.
When clients negotiate with the PBM to design their drug-benefit plans, they rely to a significant extent on the cost savings and clinical outcomes predicted by the PBM for each potential configuration. For example, the PBM might suggest that the plan sponsor consolidate certain drug categories in the formulary, or adopt a particular set of generic substitution policies for certain drugs, or implement a certain co-pay structure that gives preference to one brand of drugs over another, and so on, while providing the plan sponsor with the estimated financial and clinical impact of each such initiative. Evidently, many moving variables and preferences are involved in designing a plan, effectively limiting the client’s ability to make simple apples-to-apples comparisons between competing PBM offers.
Separately, each drug company negotiates product rebates with the PBM in exchange for achieving specific volume or market-share targets, or for inclusion/better positioning of its products across formularies. The drug company’s negotiating power against the PBM depends primarily on the uniqueness and sales volume of its drugs. For example, in the case of a blockbuster drug for which there are no substitutes, the drug company has the most negotiating power, whereas in the case of a drug for which several substitutes are available, the PBM has more power.
Meanwhile, third-party pharmacies, particularly the smaller chains and independents, have little choice but to sign up with the large PBMs; otherwise they wouldn’t be able to accept the drug discount cards used by many of their own customers. Not surprisingly, PBMs have been taking advantage of these counterparties’ limited negotiating power by imposing on them one-sided contracts, which, for example, prohibit the retail pharmacy from providing more than a 30-day supply of prescription drugs (while the PBM’s own pharmacy routinely provides a 90-day supply). Third-party pharmacists also have little choice but to rely on the PBM formularies for approving and processing claims at the point of sale. On each sale, the PBM charges the pharmacy a transaction-specific price-spread and/or a cut of the pharmacist’s professional dispensing fees.
Note that only the PBM knows precisely how much each of the above parties is willing to pay, or give up, for which product, under which circumstances, in each transaction. But it’s more than just that: it’s also that aggregate cost savings are a function of the collective actions of all parties, often making it difficult to attribute savings to the actions of any one particular party, therefore also making it easier for the PBM to extract profits.
This may seem too good to be true, so here’s a concrete example. Say Merck offers to pay a PBM the following market-share rebates for Zocor: no rebate if the PBM delivers a 50% market share; 5% off the wholesale price for a 55% share; 10% for a 60% share; 15% for a 65% share; and so on. Some clients decide to include Zocor in their formulary; but others decide to include not just Zocor, but also its archrival, Lipitor; and yet other clients decide to include both drugs but assign a higher co-payment to Lipitor; and then other clients require not just a higher co-payment but also “prior authorization” for Lipitor; and so on -- each plan favors Zocor versus Lipitor to a greater or lesser extent. What happens if Zocor’s share hits, say, 65%? How much of the aggregate 15% rebate should the PBM allocate to each client? Should some clients be allocated more on a proportional basis? How much more? How could the decision be explained to each client without disclosing key terms of other clients’ deals? These questions aren’t easy to answer, but a good guess would be that no matter how the rebate is ultimately allocated among clients, the PBM will keep a sizeable chunk for itself -- no one else will rightfully claim it.
Short of legal and regulatory threats, the only force truly constraining the PBM’s ability to extract profits from its network is competition from the two other PBMs. Sure enough, every now and then a client will get fed up with its current PBM, and upon contract expiration will switch to one of the other two PBMs -- only to get another taste of the same medicine, so to speak.

MEDCO’S TRUE ECONOMICS
The two biggest misconceptions about Medco have to do with its return on capital and profit margins, which at first glance appear to be lackluster. In fact, Medco has been generating consistent double-digit returns on invested capital for years, and its profit margins since the spinoff have been understated by a factor of about 2.6 times.
Return on invested capital appears to be lower than it really is because Medco’s balance sheet vastly overstates the amount of capital actually invested in the business. Until 1993, Medco was an independent publicly traded company; in that year, Merck acquired it at a significant premium to book value -- using purchase accounting for the acquisition; then, last year, Merck spun it off, in the process “pushing down” the excess of price over book paid in 1993 unto Medco’s balance sheet, where it now shows up as goodwill and intangible assets to the tune of more than $5.5 billion.
But this huge sum was never really invested in Medco’s business. In fact, it never even flowed through the company: it went straight from Merck’s pockets to Medco’s pre-Merck shareholders -- back in 1993. Thus the amount of money actually invested in the business is about $5.5 billion lower than reported. Page 75 of the latest 10-K contains a brief explanation of this accounting maneuver; for further reference, consider looking at pages F11-F12 of the S-1 filed with the spinoff, or even Merck’s 1993 10-K, which shows Merck’s intangible assets jumping from $153 million in 1992 to $6.6 billion in 1993 -- largely a result of the Medco acquisition.) It goes without saying that it was in Merck’s best interest to dump this balance-sheet albatross unto Medco.
Subtracting $5.5 billion from any measure of Medco’s invested capital gives a very different picture of the capital efficiency and return power of the business. For example, in fiscal year 2003, Medco generated $584 million in “free cash flow” (net income, plus depreciation & amortization, less capital expenditures); dividing this figure by $9.9 billion in total assets at the beginning of the period yields an after-tax return on assets of 6%; however, if the asset figure is adjusted downward by $5.5 billion, as it should be, a more accurate estimate of 14% is obtained. Note that this figure is after-tax and on total assets -- and that the $5.5 billion adjustment would have a similar impact on the results reported for all past years.
The true return on invested equity capital is more difficult to estimate, because the company paid Merck a one-time $1.5 billion dividend immediately prior to spinoff, effectively taking out all equity invested in the business as of that date. Indeed, it appears that more equity capital has been taken out of Medco that has been plowed into it -- a rare feat that few businesses, let alone your average steel mill, can even hope to achieve.
Profit margins are also significantly understated. Take the second quarter of 2004, for example. During the quarter, Medco reported revenues of nearly $8.8 billion and a net profit (adding back amortization of intangibles) of $172 million -- an apparently miserly net margin of about 1.9%. I say apparently, because by the same logic eBay would have to be considered a low-margin business: eBay’s net profit as a percentage of the gross value of merchandise sold on its websites is in the low single digits too. Of course, eBay doesn’t book as revenues the gross value of merchandise and services sold by third parties. But Medco does.
Indeed, Medco books as revenues the gross value of drugs sold by third-party pharmacies processing claims on its network. Medco can book such third-party revenues as its own thanks to a perfectly legal (but still quite creative) interpretation of accounting rules, as explained under “Critical Accounting Policies” in page 44 of the latest 10-K: “We have determined that our responsibilities under our client contracts to adjudicate member claims properly and control clients’ drug spend, our separate contractual pricing relationships and responsibilities to the retail pharmacies in our networks, and our interaction with members, among other indicators, qualify us as the principal under the indicators set forth in EITF 99-19, ‘Reporting Gross Revenue as a Principal vs. Net as an Agent,’ in most of our transactions with customers.” I give them an A for ingenuity and an F for clarity on this one.
Looking again at the second quarter of 2004, $5.4 billion of the reported revenues were from prescription drugs sold, not by Medco, but by third-party retail pharmacies. Medco didn’t buy the inventory, close the sales, or even dispense the drugs; it simply processed the claims (perhaps requiring, in some cases, that the third-party pharmacist dispense an alternative product, e.g., an equivalent from a different brand or a lower-cost generic). Thus a more accurate estimate of Medco’s true, honest-to-God revenues during the quarter would be obtained by subtracting the above $5.4 billion from the reported figure of $8.8 billion, leaving $3.4 billion as the more accurate estimate of quarterly revenues -- 2.6 times lower. Consequently, the reported gross, operating, and net profit margins understate the company’s true margins by a factor of about 2.6 times.

WHY IS PROFITABILITY UNDERSTATED?
Like its two competitors, Medco has good reasons for portraying itself as a low-return, low-margin business.
The first reason is historical. The PBM business originated out of the need of plan sponsors to reduce claims processing costs via economies of scale, and for many years Medco executives and employees correctly thought of their company as a straightforward “cost-center” operation that existed only to save money for clients. Today, this view of the business is far too simplistic, but many inside the company continue to think of the business in the same terms -- it keeps them focused on cost efficiency. Hence financial results continue to be reported, not in relation to true revenues, but in relation to the gross value of claims processed -- a measure of total cost-efficiency that might make profits seem relatively immaterial in the eyes of clients.
The second reason is political. With healthcare expenses spiraling out of control, government officials at all levels are actively looking for ways to take costs out of the system. As evidenced by recent state investigations, the PBMs haven’t escaped regulatory scrutiny, and will likely continue to be the target of more such scrutiny in the future. At issue in their case is a thorny question of whether they act as agents or as principals in each transaction -- a question that isn’t easily answered. This constant threat of regulatory scrutiny gives PBMs an incentive to report low margins -- why waste time going after the poor PBM making only “1.9%” on revenues?
The third reason is tactical. In short, it is in Medco’s best interest to avoid disclosing exactly how much of its profit comes from each of its counterparties -- all of whom, no doubt, would love to get their hands on such information. Medco keeps much of this information confidential by grossing up reported revenues and costs while simultaneously netting out all amounts charged to drug manufacturers and pharmacies, making it difficult for outsiders to quantify each individual source of gross profit. (If anyone here knows how much is contributed to profits by volume and market-share rebates, pharmacy price-spreads, cuts of pharmacist fees, and how much of each of these is retained versus passed on to clients, please let me know.)
The final reason is strategic. By lumping third-party revenues and costs together with its own, in less than a decade Medco has been able to build, somewhat stealthily, what today is perhaps the world’s largest and most profitable online pharmacy -- without attracting too much attention from traditional retail pharmacies, at least not until it was too late for them to do much about this unpleasant (for them) development.
STEALING THE BEST CUSTOMERS FROM TRADITIONAL RETAIL PHARMACIES
Medco calls its pharmacy a “mail-order pharmacy,” a rather un-sexy way to describe a highly automated online drug-retail operation that is accessible 24x7 via the web or automated voice-response system, and that in the second quarter of 2004 dispensed 22 million 90-day prescriptions, generating $3.3 billion in sales -- up 22% from the previous year. Only five years ago, most of those 22 million prescriptions would have been purchased at a traditional retail pharmacy; today, they are purchased from Medco.
Unlike a traditional pharmacy, a PBM mail-order pharmacy sells primarily to patients taking medications for prolonged periods of time, resulting in more predictable, annuity-like revenues from each patient; dispenses only prescriptions, significantly reducing the risk of buying excess inventory that might go unsold; fills prescriptions only in 90-day supplies, versus a 30-day supply for retail, reducing from three to one the number of professional dispensing fees paid for a 90-day supply of drugs; has no physical stores, eliminating the need to invest in and maintain a network of real-estate properties; and centralizes many functions (for example, it doesn’t require a pharmacist on each of many locations), resulting in greater economies of scale and operating leverage.
As a consequence, a mail-order pharmacy can deliver drugs directly to the consumer at much lower cost than the traditional retail channel. Not surprisingly, many plan sponsors, seeking to lower costs, increasingly require that patients on maintenance drugs fill prescriptions via mail. The result has been a massive migration of patients, especially those with chronic long-term drug needs, from traditional retail to mail-order. As shown below, in the second quarter of this year, Medco and the two other PBMs sold a combined 43 million 90-day prescriptions. Again, only five years ago, most of those drugs would have been purchased at a traditional pharmacy.

2nd Quarter of 2004 (Million) Express Scripts Caremark MEDCO
Retail prescriptions 95.7 133.9 103.4
Mail prescriptions 9.8 11.4 22.0

The impact is only now becoming visible. Rite Aid, for example, recently blamed lower guidance on revenues and profits on loses of pharmacy sales to PBM mail-order operations. "Payors are taking additional steps to slow health care spending and mandatory mail programs have increased," explained Rite Aid’s CEO, Mary Sammons, in a press release, immediately adding: "while we are disappointed with current sales trends, we believe this is a near-term situation...” blah, blah, etc. CVS and Walgreen, meanwhile, refuse to participate in PBM plans that require prescriptions to be filled through the mail, putting them at odds with plan sponsors who seek to reduce costs -- a stance that reminds me a bit of those third-world governments that refuse to devalue their currency until the last minute, when facing unpleasant but unavoidable economic forces. Anyhow, it’s the smaller chains and independent pharmacies that have suffered the most so far.
Medco doesn’t provide sufficient information to figure out with precision the cost structure and profitability of its in-house pharmacy (if anyone on this website has figured out a way to estimate them, please let me know); but the available facts suggest that mail-order prescriptions are more profitable to the company than prescriptions dispensed by a third-party retail pharmacy. For example, one analyst estimated that just reducing the number of professional dispensing fees from three to one can bring down the cost of dispensing a 90-day supply of drugs by 10%.

GROWTH PROSPECTS
Profits in the PBM sector are poised to grow rapidly, both in the short run and in the long run, primarily as a result of two factors.
First, in the near term, more than $38 billion in brand-name drugs are expected to lose patent protection within the next four years. PBMs, with their enormous purchasing power, get much better deals on generics than on brand-name drugs, so this wave of patent expirations represents a mouth-watering opportunity for PBM margin expansion. Mail-order patients, in particular, will be quickly switched over to the new generic alternatives; so Medco, having the largest mail-order operation of the three PBMs, will likely benefit the fastest and the most from this opportunity.
Second, over the next couple of decades, millions of baby boomers will be entering their elderly years, boosting demand not just for prescription drugs in general, but for maintenance medications in particular. Thus the number of patients for whom PBM mail-order pharmacies are best suited is poised to grow rapidly for a long time. (Think about that for a moment.) Since prescriptions dispensed via mail are typically more profitable to PBMs than those dispensed by retail pharmacies (see the table below, for example), this large-scale demographic trend represents an opportunity not just for long-term revenue growth, but for further margin expansion as well. Medco, having the largest mail-order operation of the three large PBMs, should benefit the most from this opportunity too.

2nd Quarter of 2004 Express Scripts Caremark MEDCO
Mail % of total volume 23.3% 20.4% 39.0%
Gross profit/prescription $1.78 $2.44 $2.52
EBITDA/prescription $1.14 $1.68 $1.89

VALUATION
At $31 per share, Medco has a market capitalization of approximately $8.5B, equivalent to roughly 11x its run-rate “free cash flow” (earnings plus depreciation & amortization less capital expenditures) of $779MM -- an attractive multiple for a business whose profits are poised to grow at double-digit rates for years to come. Here’s the calculation:

($MM) 1st Half 2004 Annualized
Earnings $230.8 $461.6
Depreciation 111.8 223.6
Amortization 90.0 180.0
Capital expenditures (43.1) (86.2)
FREE CASH FLOW 389.5 779.0

MARKET CAPITALIZATION 8,525.0
MULTIPLE OF FREE CASH FLOW 10.9x

Annualizing the results for the first half of 2004 is appropriate given that quarterly profits have been growing sequentially, and that the company estimates full-year earnings of approximately $465-$580MM, versus the $462MM used for the calculation above.
Earnings are likely to understate free cash flow for many years to come, for two reasons: first, Medco completed a multi-year capital-spending plan prior to spinoff, so capital expenditures should be lower than depreciation for the foreseeable future; and second, all those intangible assets pushed down by Merck last year are now being amortized over the next two decades, reducing reported earnings (but not free cash flow).

RISKS
The major risks with Medco are legal and regulatory. Both were discussed in some detail in the previous write-up, incorporated here by reference. My view on the matter is that the economic rationale for continued migration of millions of patients to mail-order is so powerful that it should be able to withstand lots of legal and regulatory threats over the years.

Catalyst

$38B in drug patent expirations in next four years
Continued growth in mail-order pharmacy
Some deleveraging
Management considering dividends and/or buybacks