Tuesday, November 15, 2016

US Drug Wholesalers


Part 1: The Good

Drug wholesalers handle the logistics of buying, warehousing and transporting drugs from manufacturers to retailers.  Three companies: McKesson (MCK), Cardinal Health (CAH) and AmerisourceBergen (ABC), form an oligopoly with an estimated 90% of the US market.

Its a low margin business where you make money on turnover.  Operating margins average around 2%.  All three companies are profitable, generate free cashflow, and have low debt1.  As expected in a business like this, their main costs are inventory (COGS) and SGA.

Most of the growth in this industry is by acquisition, which consumes part of the free cashflow.  This industry was previously fragmented, and over time consolidated into the oligopoly it is today.  Even in today's concentrated industry, all three companies make acquisitions every year to expand their offerings and move into new fields, such as specialty drugs and oncology.

The long term trends look good.  The population is getting older and weaker.  Drug usage is expected to grow 5% pa.  As the middleman, these companies take their small cut off the drugs passing through their pipelines.  When analysing them, we don't have to worry about patent expiry, competition from other generic suppliers, or the intricacies of insurance payments in the US health system.  We just sit back, let the drugs flow through, and collect our money.  Whats not to like?

Part 2: The Bad

When we look at the details of this industry, there are some issues.

First: retailer consolidation.  Distributors are most profitable when they work in "barbell" model.  That is: a small number of distributors in between a large number of suppliers and a large number of customers.  Giving pricing power to the distributor.  But one side of the barbell is shrinking.  The pharma retail industry has consolidated, with more and more sales going to the largest players (Walgreens and CVS) - see point 2 here.  This gives two problems:

  • Distributors margins for their small customers, such as independent pharmacies, are a lot higher that for their large ones.  Distributor operating margins average around 2%, but this is estimated to drop to 0.5% for their large customers.  Even these large customers make up 20-30% of a distributor's revenue, most profits come from their many smaller customers.
  • When retailers are taken over, they usually switch their drug distributor to the acquirer's one.  This makes it harder to project future revenue and profits.  For example, Walgreen's first indicated it will takeover Rite Aid in the middle of the year. Rite Aid's distribution contract with McKesson expires in March 2019, but will probably be terminated before then.  McKesson still does not know when, except to say that they will probably service Rite Aid at least till March next year.

To overcome this, all three companies have started franchises to help independent pharmacies start up.  They provide a broad range of services to help the viability and profitability of their smaller customers.  They have also been acquiring the smaller distributors that supply the independent pharmacies.

Second: pricing is complicated.  Distributors do not earn a fixed fee, like a postage rate.  Instead they buy drugs at some discount to the manufacturer's list price ("buy side profit") and sell at a profit ("sell side profit"), which may still be less than the list price.  See here for a good diagram of payments.  The distributors also benefit from price inflation when the price of items in their inventory increases, though less than before2.  

This leads to three more points:

  • The industry is hard to analyse.  Contract terms are not disclosed, so analysts have to estimate the revenue and profitability of each customer.  When analysing profitability, we end up making estimates based on other peoples estimates.  Its not modelling, just guessing.
  • In the past few years, wholesaler profits have been artificially boosted by price inflation in generic drugs.  This "generics bubble" was caused by few generics being approved for use - this pipeline is now clearing.  So those excess profits have gone down.
  • Theres a risk that the whole pricing model will have to change, so its no longer based around the manufacturer's list price.  Or medicaid may eventually be extended to everyone as a "single payer" system, like in a civilised country.  Even though drug wholesalers are performing an essential task in the system, with sustainable (that is, low) margins, any uncertainty associated with an industry overhaul could play havoc with healthcare stock prices.

Part 3: The Ugly

In October, McKesson said in their conference call that AmerisourceBergen has started a price war in the independent segment.  The stock fell 22% overnight.

I think the price war is temporary,and won't last in a market with 3 players.

Still, the level of stock volatility is sickening.  I may hold a smaller amount.

Comparing the companies

The companies are pretty similar in terms of cashflows generated, debt and the industry they serve. There's no point comparing their profit margins, since they vary so much between the large and small customers.  The main thing I want to look at is valuation.  How much would I be willing to pay for them?

I'm leaning towards McKesson or Cardinal.  Among independent pharmacies - the most profitable segment - AmerisourceBergen's market share has been falling, and they have a lower estimated margin3.  And they are the ones starting the price war.


All three companies report both GAAP and non-GAAP earnings.  Non-GAAP is supposed to exclude one-off items.  Guidance is only provided for non-GAAP.

The differences are:

  • Changes to LIFO Reserves: Stock is sold on a FIFO basis, as with any perishable good.  But the accounting is done on a LIFO basis.  When prices are rising, this lowers profits and taxes.  In other words, the profits on paper are lower than the cashflows.  The difference is added back to their non-GAAP earnings.  I think this is makes sense, and the amount is objective and easily calculated.
  • "Acquisition costs" and "Amortisation of acquisition related intangibles"4.  This is the largest part of the difference.  Since all three companies are serial acquirers, then do we ignore this by considering it a one-off cost that should be excluded from earnings?  Not sure, but I can probably accept it - its better than leaving the costs on the balance sheet as goodwill.
  • Litigation costs.  McKesson has such large and frequent litigation costs which makes me wary.  I'm wondering if they can conduct their business for another 5 years without getting their asses sued off.  Cardinal's litigation costs are small enough to ignore.

Earnings Estimates

For MCK, after accounting for:
  • the loss of OptumRx, Omnicare, Target, Rite Aid, partially offset by Rite Aid and Safeway
  • less generics price inflation than in FY2015.
  • a price war for supplying independent pharmacies, assuming operating margins are halved.
I get an EPS reduction of $3.84. This would give an non-GAAP EPS of $8.51.  At 12X earnings, thats a target price of $102.12.  At 15X, its $127.65.

If we exclude the price war, its $112.24 (12X earnings) and $140.31 (15X earnings).

For CAH, after accounting for:
  • the loss of Safeway, plus the potential loss of Pharmerica
  • less generics price inflation than in FY2015.
  • a price war for supplying independent pharmacies, assuming operating margins are halved.
I get an EPS reduction of $1.12. This would give an non-GAAP EPS of $4.24.  At 12X earnings, thats a target price of $51.03.  At 15X, its $63.78.

If we exclude the price war, its  $58.39 (12X earnings) and $82.87 (15X earnings).


1 Less than 2X EBIDTA
2 McKesson2Q17 Conference call: "90% of our [branded] income is fixed, 10% is variable."  And Cardinal 1Q17 Conference call: "about 15% of our branded margins are based on a contingent basis, which means that generic or a branded inflation is a piece of the driver of the value that we receive."
3 From the DrugChannels 2016-17 Wholesalers and Specialty Distributors Report.  All my estimates are calculated from the estimates in the report. 
4 Defined in Supplemental non-GAAP information here.

No comments: