ABBV: The Contrarian Case for US Humira Trend

Print Friendly, PDF & Email


Richard Evans / Scott Hinds / Ryan Baum

203.901.1631 /.1632 / .1627

revans@ / shinds@ /


August 16, 2017

ABBV: The Contrarian Case for US Humira Trend

  • ABBV claims +/- 11.5 percent end-user unit demand growth for US Humira in 2Q17, as compared to Symphony Health Solutions’ (SHS) estimate of 5.4 percent; ABBV’s claims and third-party estimates have differed materially since 3Q16
  • If ABBV is right, both SHS and IMS Health are independently and simultaneously making very large projection errors; this is possible, but is to our minds less likely than the alternative
  • The alternative is that SHS is right and ABBV is shipping Humira into the US channel faster than it’s being pulled out. The three major drug wholesalers’ days inventory outstanding (DIO) metrics tend to grow in 1Q of each year, which (despite routine presence of inventory management agreements) strongly implies wholesalers are routinely loading prescription brands in ahead of traditional start-of-the-year pricing actions. Few if any brands have a better combination of total sales and rate of price inflation than Humira, making Humira a priority target for inventory loading
  • AMGN’s Enbrel, the second largest brand in Humira’s category, publicly gave up on US list price inflation in late 2016. AMGN’s reported unit growth for Enbrel matches SHS’ estimates; this may be because, as compared to Humira, wholesalers have far less motive to load Enbrel inventories ahead of pricing actions
  • Medicaid reports a census of claims for US brands; over time the change in end-user demand reflected in Medicaid claims has closely tracked the change in end-user demand across the entire US market. Medicaid data for Humira are available through 1Q17, and show a deceleration of end-user demand that is far more consistent with SHS’ estimate than with ABBV’s claim
  • ABBV’s claim of higher end-user demand growth is feasible, but is on balance less likely to be correct than SHS’ estimate of slower end-user demand. If the slower SHS’ projection is accurate, US Humira inventories would hold approximately $560M more Humira (about 18 days of demand) than ABBV has estimated

A growing discrepancy

Since 3Q16 ABBV has reported significantly greater US Humira unit volumes than have been reported by third party data vendors, in our case Symphony Health Solutions (SHS). Using the 1Q16 SHS count of total prescriptions (TRx’s) as a starting point, at ABBV’s claimed rate[1] of US unit growth TRx’s in 2Q17 would have been approximately 754,225, as compared to the 712,956 TRx’s reported by SHS (Exhibit 1)

What ABBV can and cannot see

ABBV plainly knows exactly how many units it ships to US wholesalers in any given period, and presumably ABBV’s agreements with the major wholesalers give the company some degree of insight into wholesale inventories. However, ABBV’s proprietary line of sight into total trade (wholesale + retail) inventories arguably ends there, as the company would not have the same line of sight – if any – into retail inventories. Like everyone else, ABBV relies on 3rd-party prescription data[2] to estimate end user demand, which then allows them to back into an estimate of retail inventory

The difference between ABBV’s and third-parties’ estimates of end-user demand can be explained only by: 1) expanding trade inventories; or, 2) third-party projection error

It’s worth noting that despite the prevalence of inventory management agreements (IMA’s) intended to reduce fluctuation in brand inventories, wholesalers’ inventories nevertheless show a clear tendency of swelling in the first quarter of each year (Exhibit 2). This is strong evidence that wholesalers are buying in ahead of traditional 1Q list pricing actions for brands, which account for +/- 80 percent of dollar sales. And, it is certainly reasonable to believe that wholesalers would focus their buy-ins on the brands that have the greatest potential to produce inventory pricing gains, i.e. those with the highest product of gross sales and rate of list price increase – which makes Humira among the most desirable, and perhaps the single most desirable, brand to load in at the start of each year. Because of this it’s entirely plausible that trade inventories on Humira are greater than ABBV has been led to believe[3]

It’s also plausible that third-party projections of Humira end-user demand are inaccurate. The third-party data vendors sample prescriptions, and from this sample project an estimate of total national prescription demand for each product. It is entirely feasible that – as ABBV has claimed – a single large supplier has stopped supplying its prescription data to IMS and SHS, thus creating a forecast error that has resulted in underestimates of end-user demand. What’s less feasible is that a single vendor would be sufficiently large and/or fast-growing to account for the observed difference (Exhibit 1, again). As of 2Q17 the gap between ABBV’s estimate of end-user demand and SHS’ estimate was 41,269 TRx’s, which is equal to 5.5 percent of the 754,25 TRx’s that ABBV’s end-user estimate implies. And, this gap grew quite rapidly, from 13,718 TRx’s in 3Q16 to 41,269 TRx’s in 2Q17. So, while it’s possible that loss of a single large data vendor would compromise SHS’ (and IMS’) projections, it’s challenging to accept that this would result in such a large and rapidly growing discrepancy. And, it’s further difficult to accept that both IMS and SHS would independently make similarly large and rapidly growing projection errors

We cannot prove one explanation (trade inventory build) over the other (3rd party projection error); however, we find it easier to believe that the trades would load Humira inventory to benefit from inventory pricing gains than to believe that two data vendors independently and simultaneously committed impractically large and fast-growing projection errors

What others in the category are, or aren’t, saying about 3rd party projections

Other than ABBV, among the other DMARD competitors only UCB has made specific mention of 3rd party data under-estimating US end-user demand. PFE reported a minor discrepancy for Xeljanz in 1Q17 (25 percent TRx growth for PFE v. 32 percent for SHS). Notably AMGN’s reported Enbrel unit growth has all but exactly matched SHS’ reported unit growth, except for a relatively small difference in 2Q17 (AMGN reported -5 percent v. SHS’ report of -9 percent). AMGN waived the white flag on pricing gains ahead of 1Q17, which would have all but eliminated wholesalers’ interest in carrying excess Enbrel inventory into the start of 2017 – so as compared to Humira, Enbrel would have been less likely to be affected by changes in trade inventory

An independent (of sampling error and trade inventories) proxy for end-user demand

For product categories with significant Medicaid volumes, Medicaid claims data can provide an alternative basis for estimating the unit trend. And, because Medicaid claims are a census, rather than a sample, of Medicaid prescriptions, the Medicaid trend data are unaffected by the types of vendor issues that sometimes affect 3rd party estimates of total market prescriptions. Exhibit 3 compares total Medicaid units by quarter (dashed line, right axis) with SHS’ reported TRx’s by quarter (solid line, left axis). For the most recent three quarters in the exhibit, the orange solid line indicates TRx demand as implied by ABBV’s estimates, as compared to the black solid line which indicates SHS’ TRx estimates. Note that despite a difference in these lines’ slopes, the inflections in demand for the two series are highly similar – in other words Medicaid demand tends to accelerate or decelerate in parallel with total market end-user demand as estimated by SHS TRx’s. Another way to look at this (Exhibit 4) is to compare the 4Q rolling log-difference between Medicaid Humira units (x-axis) and SHS- or ABBV-reported Humira TRx’s (y-axis)[4]. The blue dots in Exhibit 4 compare Medicaid and SHS since 1Q14, and have a fairly tight relationship as evidenced by an R-squared of 0.73. The orange dot in Exhibit 4 represents ABBV’s 1Q17 estimate of Humira units; the point here is that ABBV’s estimate of 1Q17 end-user demand can only be correct if the normal relationship between Medicaid demand and whole-market end-user demand is suddenly and seriously broken. 2Q17 Medicaid claims are not yet available, as such we can’t yet speak to the relationship between ABBV’s 2Q17 estimate of end-user demand and what’s implied by Medicaid claims. For the record we did verify that the 1Q17 Medicaid data included submissions by all 50 states, so we have no reason to believe the Medicaid figure will be revised upward

Best case, worst case …

The best case for ABBV is, of course, that their estimates of end-user unit demand are correct, and that trade inventories are in fact normal. Nevertheless even if we assume ABBV is correct, it’s worthwhile noting what this implies about Humira’s US price / volume trend

At ABBV’s claimed rate of end-user demand, we estimate that the average annual net cost of a Humira regimen is now $38,404, and that growth in net price is decelerating (Exhibit 5). And, again using ABBV’s estimate of end-user demand, Humira’s TRx share was static in the most recent quarter (Exhibit 6). ABBV has effectively pledged to limit list price increases on Humira; however this pledge has no effect on the competitive discounting behavior of other brands in the space, and several of these brands (particularly the anti-psoriatics) recently have grown their absolute dollar discounts rather dramatically (Exhibit 7). Because the absolute dollar value of discounts largely determines the gross profitability of a brand to formulary managers, ABBV is forced to stay in range of these absolute dollar concessions if it wants to continue defending its share. The result is that Humira’s US net pricing is being squeezed between a single-digit list price pledge and far more rapid growth in absolute dollar discounts. Unless competitors either stop discounting for share or ABBV violates its single-digit list pledge, either of two things will happen: US Humira net pricing will decline, or Humira will lose unit share

The worst-case scenario is of course that SHS is correct about end-user demand, which implies that trade inventories are carrying about $560M in excess Humira (equal to about 18 days’ demand, Exhibit 8), and that Humira’s TRx share has fallen (Exhibit 9). And, the worst-case scenario does nothing to alter the pricing squeeze outlined under the best-case scenario


  1. ABBV has given approximate rather than exact values for US Humira unit growth; for this analysis we assume the most rational point estimate implied by these approximations
  2. From the most recent 10-k, emphasis added: “… the company uses both internal and external data to estimate the level of inventory in the distribution channel … The company evaluates inventory data reported by wholesalers, available prescription volume information, product pricing ….”
  3. ABBV has consistently maintained that US wholesale inventories of Humira are at less than ½ month
  4. Very roughly, the y/y growth rates

©2017, SSR, LLC, 225 High Ridge Rd, 2nd Floor, Stamford, CT 06905. All rights reserved. The information contained in this report has been obtained from sources believed to be reliable, and its accuracy and completeness is not guaranteed. No representation or warranty, express or implied, is made as to the fairness, accuracy, completeness or correctness of the information and opinions contained herein. The views and other information provided are subject to change without notice. This report is issued without regard to the specific investment objectives, financial situation or particular needs of any specific recipient and is not construed as a solicitation or an offer to buy or sell any securities or related financial instruments. Past performance is not necessarily a guide to future results. In the past 12 months, through a wholly-owned subsidiary SSR Health LLC has provided paid advisory services to Pfizer Inc (PFE), Gilead Sciences (GILD), Bristol-Myers Squibb (BMY) and Sanofi (SNY) on both securities-related and non-securities-related topics. One or more of SSR Health’s analysts owns long positions in the following stocks: AAAP, ACOR, AERI, AMGN, ARRY, BMY, DVA, FLXN, GILD, HRTX, KITE, NSTG, ONCE, PTCT, PTLA, RARE, RHHBY (Roche), RIGL, SNPHY, TBPH, THERF, VRTX

Print Friendly, PDF & Email