Why the Market Assumes too Much Margin Pressure on Insurers, too Little on Innovators

Print Friendly
Share on LinkedIn0Tweet about this on Twitter0Share on Facebook0

Richard Evans / Scott Hinds

212.531.6101 / .6102

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

April 19, 2010

Why the Market Assumes too Much Margin Pressure on Insurers, too Little on Innovators

  • Narrow (below corporate level) application of MLR limits, and/or limiting premium growth to less than the rate of medical cost growth, work against regulators’ and the public’s interests. Thus neither practice is likely to be common; and insurers’ share-price reaction to these concerns appears over-done
  • If MLR limits are too narrowly applied, winning contracts cannot offset losing contracts, and insurers must price all contracts so that none lose. Gross premiums rise, reducing insurance uptake rates; and, beneficiaries’ excess paid-in capital is held prisoner for at least the duration of the contract period. Incurring these two costs produces no gains – after rebates, MLRs under narrow limits net to the same figure as would have been achieved under broad application (corporate level) of MLR limits
  • If MLR limits are narrowly applied AND premiums cannot be raised in response, private capital exits geographies with such restrictions, reducing both competition and consumer choice. Private insurers’ ROIC’s already are below SP500 averages
  • State regulators are unlikely to carry through on isolated threats to hold premium growth below rates of underlying cost growth. Obvious effects include simultaneously reducing reserves and raising underwriters’ risk; less obvious, states that rely heavily on non-profits (e.g. Massachusetts) face the likely outcome of their non-profits having to seek capital – which almost surely means converting to for-profit
  • We provide further explanation and analysis of the pricing effects of the Independent Medicare Advisory Board (IMAB), showing that all of the IMAB’s 2015 – 2019 cost savings must come from roughly one-quarter of the Medicare cost base, and that innovators (brand pharma, biotech, devices, +/- capital equipment) are the most likely targets
  • We project that IMAB policies will reduce the Medicare sales of innovators by 11 to 19 percent between 2015 and 2019, and show why we believe this estimate is conservative

Why Aggressive MLR and / or Premium Limits are Unlikely

Concerns have risen around insurers’ future pricing freedoms, both as a result of questions surrounding how federal reform-related MLR limits will be applied, and as a result of recent state regulatory actions to limit premium growth. Our view is that MLR and/or premium growth limits, if applied aggressively, create results that run counter to state regulators’(and public) interests; accordingly we expect few new constraints on insurers’ pricing.

Beginning with MLR, the new federal law sets limits of 80 percent for individual and small group coverage, and 85 percent for large group coverage. We again emphasize that the large underwriters meet the current limits at the corporate level, particularly as the new law’s MLR calculation results in a higher MLR than insurers report in routine financial disclosures[1].

The point has been raised that more narrow application of MLR limits has the effect of eliminating profitable contracts (or groups of contracts) without adjusting losing contracts, the net result being a substantial drop in total corporate MLR for a given underwriter. Mathematically this is plainly correct; and, the more narrowly MLR limits are applied, the more dramatic the effect. However as a practical matter, we view narrow application of MLR limits as self-defeating to both regulatory and public interests.

The law’s language calls for MLR limits to apply at the ‘issuer’ level[2]. For reference, using WellPoint data as reported to NAIC, and as found in a recent report by Senate Commerce Committee staff, we show the effect of applying MLR minimums, at the issuer level, across WLP’s corporate portfolio. Note that in calculating MLRs we reduce revenues by WLP’s disclosed tax rate – after eliminating one-time tax effects; we make no attempt to add to medical costs expenses that have been disclosed as operating costs, but may meet the law’s more liberal definition of medical costs. If we were to apply the law’s MLR limits to WLP in aggregate, we would raise the reported MLR by 50bp (wholly the result of WLP’s individual plans having an MLR below the limit). If we apply the MLR limits to each issuer within WLP, the aggregate MLR rises by an additional 110bp (Exhibit 1). Thus the effect of imposing the MLR limit is 50bp, and the effect of applying it narrowly is an additional 110bp, for a total increase in MLR of 160bp.

We note that insurers’ first line of defense against narrow application of MLR limits is a higher level of aggregation of their wholly-owned issuers, though we recognize that state regulations complicate this defense. Importantly, even assuming insurers cannot further aggregate their wholly-owned issuers, we still see practical limits on the narrow application of MLR limits. In short, narrow application of MLR limits lessens the socially desirable efficiency gains associated with effective risk-pooling.

The very point of insurance is to create efficiency through risk-sharing. Reducing the base across which risks are shared reduces these efficiency gains. Insurers naturally rely on very profitable issuers to offset the cost of money-losing issuers, the objective being an acceptable aggregate level of profitability across the entire portfolio. If more profitable issuers are disallowed – which is precisely the effect of narrowly applying MLR limits to individual issuers – then underwriters have no means of offsetting money-losing issuers, and so must avoid money-losing issuers entirely. The only means of doing so are to raise premiums to the degree that money-losing issuers happen only very infrequently; or, where such premium increases are not allowed, to exit the geography that issuer serves. On the assumption that premiums are raised, note that ultimately we arrive at the same average MLR as would have been achieved by applying the MLR limit at the corporate level, though the narrow application of MLR would have created 2 very large inefficiencies. First, average gross (before rebate) premiums would be much higher, which very likely reduces rates of participation in insurance. Second, the cash that ultimately is returned in the form of rebates will have been unavailable for beneficiaries’ use between the time it was paid as premium and the time it was received in rebate.

If premium increases are not allowed as a response to narrowly applied MLR limits, fewer issuers will offer coverage in geographies with faster-growing and/or less predictable claims costs, reducing both competition and breadth of consumers’ choice. We defend this view by noting that health insurers’ ROIC — under current underwriting rules – already runs below the average non-financial ROIC in the SP500. Reducing margins and adding risks through narrowly applied MLR restrictions inevitably reduces ROIC further which, given the modest starting point, argues that insurers’ capital would be re-deployed – i.e. that insurers would exit restrictive geographies.

This takes us to the second – and highly related – matter of regulating health insurers’ premium increases, an issue that has been forced into view by recent state regulatory actions, particularly in Maine and Massachusetts. Using Massachusetts as an example, state regulators are refusing to allow requested levels of premium increases. To our minds this ultimately is self-defeating, for three simple reasons. First, the primary engine of Massachusetts’ health cost growth is Massachusetts hospitals[3]. Beneficiaries in Massachusetts choose plans in large part according to which of the state’s several ivory-tower hospitals is included in the benefit package – all of which means plans are relatively powerless to offer coverage without one or more such hospitals in tow, which in turn means insurers have little or no pricing leverage over the hospitals. Thus unless the state regulator can reach behind the plans to reduce hospital price growth – which the regulator cannot do – then reducing premium growth simply means that the state’s underwriters are going to lose gross margin if they have to write 2010 business at or near 2009 rates. Second, there’s not a lot of gross margin to lose — in 2008 almost all (99.2% of premiums) of Massachusetts’ health insurance was underwritten by non-profit insurers, at comparatively high MLR’s (88.8%)[4]. Third, if the Massachusetts regulator disconnects premium growth from underlying cost growth, the capital reserves of the state’s non-profit underwriters will be reduced. In that claims costs are somewhat unpredictable, this courts the risk of putting the state’s underwriters below minimum capital levels – and this in turn creates risk that at least some of the states’ predominantly non-profit insurers are converted to for-profits. Our logic here is that the original capital behind non-profits tends to have entered the health insurance business some time ago for mission-related purposes. And, over the years, non-profits’ capital reserves accumulated as a normal consequence of freely relating premiums to claims costs. Eliminating, or at least severely reducing, the non-profits’ capital reserves means new capital would have to be found – and we see few if any sources of ‘de novo’ non-profit capital, but plenty of sources of for-profit capital. Simply put, state regulators that run their non-profit underwriters’ capital below critical minima are very likely to force their non-profit underwriters into for-profit hands. Whatever one might believe about for-profits v. non-profits and public interests, we plainly believe this outcome runs counter to state regulators’ objectives, and so believe state regulators will not carry through on what Massachusetts is threatening to do: hold the line on premium growth despite underlying medical cost growth.

Summarizing, MLR limits only reduce the profitability of insurers’ portfolios if we assume no price response (or willingness to exit unprofitable markets) on the part of insurers. Either response from insurers – raising premiums or exiting certain markets — runs counter to public (and regulators’) interests. Accordingly we do not expect sustained application of MLR limits at narrow levels within insurers’ portfolios. Likewise, holding rates of premium growth below rates of medical cost growth presumes insurers have a level of pricing authority over medical providers / suppliers that they simply do not have. These types of premium restrictions only serve to de-stabilize underwriters, which runs counter to both state regulators’ and public interest, whether the underwriters are for-profit or non-profit. In states with large non-profit underwriters, such as Massachusetts, we see further disincentive to aggressive regulation of premiums – namely the risk of converting the state’s non-profit underwriters to for-profit.

The Independent Medicare Advisory Board and Innovators’ Margins

We continue to believe that one of the more impactful and under-appreciated elements of the reform law is the Independent Medicare Advisory Board (IMAB). The IMAB is tasked with keeping per-beneficiary Medicare cost growth within certain ranges of CPI and GDP, depending on the year; and, also depending on the year, the IMAB’s savings target is capped at certain percentages of Medicare spending. Exhibit 2 summarizes the legislative text relevant to the IMAB’s savings targets.

Critically, the law places limits on where the IMAB can find the savings necessary to meet its statutory savings goals. Through a series of indirect references, from 2015 to 2019 the law precludes the IMAB from reducing payment rates to certain producers and providers whose revenues equal nearly half of Medicare spending. Exhibit 3 provides the relevant legislative text. In effect, the text refers both to the Social Security Act (SSA) and to language in the reform law itself to identify producers and providers that both meet particular SSA definitions, and who are subject to certain pricing provisions[5] of the reform law. The relevant passages serve to exclude inpatient hospitals, long-term care hospitals, inpatient rehabilitation facilities, psychiatric hospitals, hospice care, outpatient hospitals and clinical laboratories from IMAB-directed cuts in spending (Exhibit 4). Repeating, these producers / providers account for nearly half (in 2008, 47.6%) of Medicare spending.

This plainly means the IMAB’s savings goals – at least during the 2015 – 2019 period – must be met by reducing costs associated with producers and providers that do not meet the exclusionary criteria. We term these producers and providers ‘fair game.’ Fair-game producers / providers are listed in Exhibit 5, along with their 2008 Medicare sales, and the CMS Actuary’s forecast of per-beneficiary spending growth for the 2009 – 2019 period. In fact we expect the savings burden to be carried by an even narrower sub-set of the fair-game categories; consistent with history we believe it is unlikely that physicians will be given substantial reductions in payment rates, and so expect savings to be borne predominantly by the non-clinician subset of fair-game producers / providers, who account for roughly 25 percent of Medicare spending (also Exhibit 5). In summary, we expect that all of the IMAB’s savings must be found from only one-quarter of the Medicare cost base.

Taking the argument a step further, we would divide the fair-game, non-clinician producers and providers into ‘commodity’ and ‘innovative’ categories. Keeping in mind that the IMAB cannot make changes that reduce beneficiaries’ access to care, in theory the IMAB cannot reduce prices that have already been competitively reduced to levels approaching cost of capital. This leaves only producers / providers whose margins reflect any of several benefits – excessively generous legacy payment rates, dominant selling positions as a result of geographic concentration; or, in the case of innovators — patents. Of these, innovators’ margins strike us as the IMAB’s low-hanging fruit. Reducing payment rates to producers / providers whose service offerings are made predominantly to seniors runs the considerable risk of reducing the availability of these services – which the IMAB is restricted from doing. In contrast, reducing the prices of innovative products sold by firms that sell not only to seniors, but to non-seniors as well, appears far less likely to reduce the availability of these products to seniors. Because innovators’ marginal costs of production typically are very low, the IMAB could reasonably presume that innovators would continue to supply products to Medicare at substantially reduced prices. In effect, the IMAB can force innovators’ Medicare v. non-Medicare sales into the same ‘balance’ as export and US sales – wherein one category of sales (non-Medicare) produces adequate margin to cover fixed costs and return both costs of capital and the majority of profits, and the other category of sales (Medicare) is priced at or near marginal production costs.

Exhibit 6 steps through the relevant math. Lines ‘a’ and ‘b’ extrapolate per-beneficiary spending growth according to either of two assumptions: ‘a’ that incremental productivity-based adjustments to payment rates actually have a reducing effect on Medicare per-beneficiary cost growth; and ‘b’ that the productivity-based adjustments have no effect on historic cost growth[6]. From 2015 to 2017 excess growth is defined as per-beneficiary cost growth less the 5-year rolling average of CPI-U and CPI-M (line ‘c’); in 2018 to 2019 excess growth is defined as per-beneficiary cost growth less the 5-year rolling average of GDP, plus 1 percent (‘d’). Lines ‘e’ and ‘f’ show excess spending under each scenario; line ‘g’ applies the maximum cost savings target to produce the IMAB’s dollar savings goal (‘h’ or ‘i’) depending on scenario. The IMAB’s goal is further expressed as a percent of Medicare spending (‘j’,‘k’), as a percent of ‘fair-game’ spending (‘l’,’m’); and, as a percent of non-clinician fair-game spending (‘n’,’o’). Note that the IMAB must find the absolute dollar savings each year; thus, during the period of analysis, we estimate the dollar cost savings goal will fall between a cumulative total of $27B and $50B, and the percent reduction in sales from this category will be between 11 and 19 percent (Exhibit 6, again).

As we’ve argued, we believe these reductions would fall predominantly on innovators – namely branded pharmaceuticals, biotech, innovative medical devices, and perhaps even capital equipment. That said, for the moment we struggle to calculate a precise percent of sales at risk from these suppliers. The reason is two-fold. First, the NHE categories that we rely on in this call are not comprehensive – for example the prescription drug sales identified in Exhibit 5 represent only direct sales of prescription drugs to beneficiaries – the category does not include, for example, sales of prescription drugs to hospitals who ultimately administer these drugs to Medicare beneficiaries. Thus actual sales of prescription drugs to Medicare beneficiaries are much higher than what is shown, and the same is true for other innovators’ sales. Second, the financial disclosures of the innovators similarly fail to estimate indirect sales to Medicare beneficiaries. We expect to be able to produce a more precise estimate through other sources, though these sources are not immediately available to us.

For our immediate purposes, we can show that the IMAB’s savings goals are substantial, and that they are likely to be focused on innovators. Our estimate of an 11 to 19 percent reduction in ‘fair-game non-clinician’ sales is clearly a low-end effect of the estimate on innovators’ Medicare sales, as innovators are a subset of the fair-game non-clinician total. And, among the affected innovators, we recognize that earnings impact will also be a function of the percent of any individual innovator’s sales that go to Medicare – obviously innovators with higher percentages will be at greater risk.

  1. Federal and state taxes are deducted from revenues, as are additions to reserves and regulatory fees, reducing the denominator; and, the portion of operating expenses incurred for ‘activities that improve healthcare quality’ is added to the numerator.
  2. A health insurance issuer offering group or individual health insurance coverage shall, with respect to each plan year, provide an annual rebate to each enrollee under such coverage, on a pro rata basis, in an amount that is equal to the amount by which premium revenue expended by the issuer …’
  3. Please see: “The Biggest Health Reform Worry That No One Seems to be Watching; Labor’s Interests and Reform Prospects; and, Lessons from Massachusetts” Sector & Sovereign LLC, December 2, 2009
  4. These MLR’s look even higher when we consider the relatively small average size of Massachusetts contracts
  5. Namely, productivity reductions to inflationary updates in excess of nationwide non-farm changes in productivity
  6. Note that our growth assumptions exceed those of the CMS’ Actuary by a considerable margin – almost entirely the result of the Actuary having to forecast on the assumption that current law is unchanged. Current law calls for a substantial reduction in physician payment rates, which almost no one – us included – believes will ever come about. We do not include the reduction in physician payment rates in our estimates.
Print Friendly