Healthcare Demand is (Cyclically) Improving Ahead of Estimates and Share Prices; Something Has to Give

eytan
Print Friendly
Share on LinkedIn0Tweet about this on Twitter0Share on Facebook0

Richard Evans / Scott Hinds / Ryan Baum

203.901.1631 /.1632 / .1627

richard@ / hinds@ / baum@sector-sovereign.com

twitter.jpg @SecSovHealth

April 16, 2012

Healthcare Demand is (Cyclically) Improving Ahead of Estimates and Share Prices; Something Has to Give

  • Estimates for healthcare stocks appear too low in light of improving healthcare demand; this is especially true for healthcare sub-sectors that are highly geared to changes in patient volumes (our favorites being Non-Rx Consumables, Hospitals, and HMOs)
  • Over the past decade the volume-sensitive healthcare sub-sectors’ revenue patterns have been reasonably well correlated with concurrent indicators of both economic activity (e.g. national employment) and demand for healthcare (e.g. aggregate hours worked in healthcare settings). Employment and healthcare hours worked both are growing off of their recent cyclical troughs, nevertheless revenue estimates imply decelerating growth rates for the remainder of 2012, and a continuation of 2011’s near-historically low rate of revenue growth through 2014
  • Estimates for the volume-sensitive healthcare groups have remained static since 4Q11 results, even though national employment and aggregate healthcare hours have since improved considerably. Valuations for these sub-sectors also have remained static, i.e. neither estimates nor share prices appear to reflect robust evidence of cyclical strengthening in healthcare demand
  • Accordingly we believe the frequency and magnitude of positive surprises may be higher than normal in 1Q12 earnings results for volume-sensitive healthcare sub-sectors; and, we anticipate steady and substantial positive revisions to full year 2012 and out-year estimates as expectations align more realistically with cyclically improving demand

Revenue estimates for US-listed healthcare names imply decelerating sales growth through the 3 remaining quarters of 2012, and annual growth rates (’12 – ’14) mirror 2011’s near-historic lows in per-capita healthcare demand. This contrasts with robust evidence that trailing US healthcare demand has been weak for cyclical reasons[1] and that healthcare demand is strengthening off of a cyclical low in sync with underlying economic improvement

We use the national employment rate as a proxy for underlying economic activity, and aggregate hours worked in healthcare settings as a proxy for current levels of healthcare demand[2]. Changes to national employment and healthcare hours worked correlate reasonably well with changes to healthcare revenue and/or EBITDA, and this is especially true for healthcare sub-sectors whose business models are more reactive to changes in patient volumes (Exhibit 1)

Nevertheless revenue growth expectations for the remainder of 2012 – even for volume-sensitive subsectors – are falling slightly despite growth in healthcare hours worked, i.e. despite robust evidence that healthcare demand is rebounding off of its recent cyclical trough (Exhibit 2). The discrepancy is repeated in annual estimates; revenue and EBITDA growth expectations for 2012 through 2014 imply 2011 (near cyclical trough) rates of unit demand growth, despite growth in aggregate healthcare hours (Exhibit 3), and also despite growth in employment and consensus expectations for employment (Exhibit 4)

Since 4Q11 results, positive revisions to volume-sensitive names have been larger than to non-volume-sensitive names (Exhibits 5, 6)[3]; however to be clear these revisions have only lessened the amount of deceleration called for in 2Q12 through 4Q12, and brought annual (2012 – 2014) growth expectations back to 2011’s near cyclical trough levels

Not only do estimates fail to reflect mounting evidence of a cyclical upturn in healthcare demand, share prices also reflect little or no expectation of cyclical improvement. Price to sales ratios have risen marginally among non-volume-sensitive sub-sectors since 4Q11 results became known, but not at all among volume-sensitive sub-sectors (Exhibit 7)

All else equal, we would expect positive revenue and earnings revisions as 2012 earnings unfold, and more than the usual frequency and magnitude of positive earnings surprises as 1Q12 earnings are reported in the coming weeks. This should be especially true for volume-sensitive names, which are demonstrably more reactive to underlying growth in patient volumes

Hospitals rank highly on our pro-cyclical list because of the anticipated effect of rising volumes on earnings. Hospitals are a high fixed cost / low operating margin business model with inherently high operating leverage, thus we expect cyclical improvement in patient volumes to have an inordinately positive earnings effect. Also bear in mind that Exhibit 1 defines cyclical reactivity in revenue terms, for the practical reason that as a cyclical measure change in revenue is not nearly as contaminated by accruals and other accounting effects as other measures. And, Hospitals’ revenue trend is a well known prisoner of reimbursement policy changes, which presumably reduces the correlation of hospitals’ revenue with cyclical changes in the broader economy. Hospitals’ pro-cyclical utility further benefits from these businesses being 100% US-focused, particularly because cyclical improvements expected in the US are considerably stronger than those expected in other US-listed healthcare multinationals’ export markets. And, finally, Hospitals’ current valuations appear to reflect a greater level of concern with the extent of pending Medicare price cuts, and whether these can be offset with commercial pricing gains, than we believe is warranted[4]

HMOs also rank highly among our pro-cyclical recommendations, despite a relatively low level of correlation between revenue and either employment or aggregate healthcare hours worked. Bear in mind that unlike most volume-sensitive sub-sectors (with the exception of Hospitals), HMOs’ revenue growth incorporates underlying rates of medical inflation – which are non-cyclical. And, the positive earnings effects of rising employment are considerable: not only does rising employment raise enrollment, driving both revenue growth and beneficial operating leverage, rising employment reduces underwriting risks, because marginal enrollees are on average considerably healthier than existing employees

On a more nuanced level, our specific ‘pro-cyclical’ preferences (Hospitals, Non-Rx Consumables, HMOs) differ somewhat from the purely quantitative pro-cyclical rank ordering of sub-sectors provided in Exhibit 1. In particular we have not recommended any of Diagnostic Laboratories, Dental, Drug Retail, or Drug Distribution, all of which appear on the basis of historic data to be highly cyclical. Because Diagnostic Laboratories’ Medicare prices are higher than their commercial prices, and further because it appears commercial prices may have been lowered in an effort to gain greater Medicare volumes at full Medicare prices, we have an overwhelming strategic / structural concern that the sub-sector and its pricing structure face growing risks from an enforcement action. That said we’re unable to offer any sense of when such an enforcement action might take place, and accept that upward revisions are highly likely to precede any change in perceived or real Medicare pricing risks. We have not recommended Dental sub-sectors simply because valuations among these names offer less room for upside as compared with other pro-cyclical alternatives. Our concerns with Drug Retail and Drug Distribution are longer-term / structural, and have a great deal to do with the sustainability of both generic selling margins and brand buying margins. Because we believe the average wholesale price (AWP) price index is fundamental to generic selling margins, and further believe the states will uniformly move away from the AWP standard this summer – clearing the way for commercial buyers ultimately to abandon the standard as well – we’ve been hesitant to recommend Drug Retail and/or Drug Distribution in front of these events. More importantly, because we’re just at the beginning of a surge in new generics, we sense near and mid-term valuations are more geared to changes in selling margin than to underlying volumes – and by extension that because expectations for positive generic effects are high, that cyclical improvement in unit demand would have less of an effect on share prices. That said, we’ll admit these sub-sectors are even more pro-cyclical (i.e. volume-sensitive) than we’ve realized, and that the transition of the states off of AWP is at best a ‘soft’ headline risk; accordingly we do see a short-term pro-cyclical trade in Drug Retail and Drug Distribution

  1. To be clear, there are both cyclical and secular headwinds to healthcare demand growth. However at the moment, cyclical headwinds appear far more relevant. We estimate cyclical headwinds have reduced trailing rates of annual per-capita unit demand for healthcare by 1.6 percent as compared to the ’00 – ’08 period preceding the economic downturn. In contrast, we believe secular headwinds (e.g. the transition from lower to higher deductible plans) are reducing annual per-capita unit demand by roughly 20bp annually. Please see US Healthcare Demand Slow for Cyclical (i.e. Temporary) Reasons…”, January 12, 2012, SSR
  2. Hospitals and physicians’ offices are by far the dominant settings in which care is delivered; however these settings are under-sampled in reported earnings. Because hospitals and physicians’ offices do not create inventory, aggregate hours worked are a useful co-incident indicator of both supply and demand. Aggregate healthcare work hours data are robust (approximately 23 percent of the relevant settings are sampled monthly), and are available during each quarter, before quarterly earnings are reported. Please see Accelerating Growth in Hospitals’, Physicians’ Offices and Other Care Settings’ Labor Hours…”, March 12, 2012, SSR
  3. N.B.: Exhibit 5, and to a lesser degree Exhibit 6, show growth expectations that are higher for volume-sensitive than non-volume-sensitive names. This is precisely the picture one might expect to see if estimates are anticipating a cyclical upturn; however in this case the picture is in place for unrelated reasons. Specifically, trailing growth rates for the non-volume-sensitive names were unusually high, thus lowering current quarterly growth rates by simple virtue of difficult comparisons. Also in 2012 patent losses, for reasons unrelated to the economic cycle, have a very large effect on the non-volume-sensitive names’ revenue growth rates
  4. Please see Hospitals’ Stable to Improving Net Pricing Power, January 24, 2012, SSR
Print Friendly