High Administrative Costs and Low Enrollment Led To the Demise of Startup Insurer Crystal Run


Identifying the reasons for poor financial performance of health plans is important for startups and health plans expanding their markets and for state and federal regulators who aim to promote competitive markets, for health insurance. This is especially important during a time when 112 members of the House of Representatives have cosponsored HR 1384, the Medicare for All Act of 2019, which would eliminate health insurers from all U.S. markets. In addition, identifying the reasons for poor financial performance is important given the evidence that more competition in insurance markets leads to lower premiums (Trish 2015).

In 2014, there were three startup health plans in New York State: Oscar, Health Republic, and CareConnect, all based in New York City. In December 2018, this journal published a financial analysis of these startup plans. That analysis showed that high claims costs, driven by high-cost contracts with providers, was the most likely reason for plan closure, not risk adjustment or high administrative costs (Block 2018). In 2015, a fourth startup, Crystal Run Health Plan, entered the market in New York State in Orange and Sullivan Counties as part of the Hudson Valley rating region, a market contiguous to New York City. Orange and Sullivan counties are two exurban counties about 70 and 100 miles northwest of New York City with populations of approximately 373,000 and 78,000, respectively (NYS-DOH, 2010). On March 29, 2019, Crystal Run Health Plan announced it was ending its health plan at the end of 2019.

Both the New York City health insurance startups and Crystal Run cited problems with CMS’s risk-adjustment program. (Axelrod 2019). Crystal Run CEO Hal Teitelbaum said, “Unfortunately, like many new plans, the risk adjustment methodology in the Affordable Care Act created an impractical and unsustainable financial obligation for insurers” (Axelrod 2019). Local reporting cited that the concerns of this plan were similar to the risk-adjustment concerns of Health Republic and CareConnect, two plans that exited the New York City ACA exchange market and closed in 2015 and 2017, respectively.

However, financial reports submitted by Crystal Run to the New York State Department of Financial Services (NYS-DFS Exhibit 17) as part of the state’s rate-review process show that Crystal Run’s losses in the individual market were unrelated to risk adjustment.Individual market includes plans both on and off the ACA exchanges. Rather, its losses in the individual market in 2016 were driven by below-market premiums.

This analysis shows that Crystal Run’s losses in the small-group market were driven, in part, by risk adjustment. But another major factor was excessive administrative costs, which were three times as high as the administrative costs at established health plans.

These assessments of Crystal Run Health Plan differ from the financial analysis of CareConnect, Oscar, and Health Republic in the December 2018 issue of this journal, which showed the NYC plans had premiums that were similar to the rest of the market, but they had medical costs far in excess of other plans operating in New York City in 2016.


This analysis uses rate-review data filed with NYS-DFS that are publicly available on the NYS-DFS website. Rate-review data lag behind actual financial peformance; rates for plan year 2018 were submitted in June 2017 and contain claims data for plan year 2016. NYS-DFS requires submission of aggregate historical financial data for each ACA exchange product for the most recent year available in Exhibit 17 of the 2018 and 2019 rate submissions (NYS-DFS 2018 & 2019).

Individual market financial performance

Figure 1 shows the financial performance of the nine plans operating in New York City in 2016 and Crystal Run. In 2016, Crystal Run lost, on average, $227 per member per month (PMPM). On a PMPM basis, Crystal Run’s performance was worse than plans operating in New York City in 2016.

Figure 1. Adjusted and unadjusted net income PMPM in 2016

Crystal Run executives blamed risk adjustmentRisk adjustment: A federal program, required for all plans participating in the individual market (on- and off-­exchange), in which health plans with a lower-risk population relative to plans in the same market make a risk-adjustment payment, and plans with a higher-risk population receive a risk-adjustment payment. for its 2019 decision to close. But in the individual market, Crystal Run collected $68 PMPM in risk-adjustment payments in 2016 (Figure 2). It collected $348 in premiums PMPM on average and paid out $501 PMPM in medical costs, meaning for every $1 it collected in premiums, it spent $1.44 in medical costs before accounting for administrative expenses.

Figure 2. Revenue and expenditures for Crystal Run in 2016

Average administrative costs for Crystal Run in the individual market were $142 PMPM (Figure 2). These costs were very large compared with the average for plans in New York City ($65 PMPM) (Block 2018). In Figure 1, the adjusted net income metricAdjusted net income is calculated as (net income – [risk adjustment + administrative costs – average administrative costs]) and is defined in detail in Block 2018. shows Crystal Run losing $218 PMPM, slightly less than Crystal Run’s net loss of $227 PMPM. This small difference of $9 PMPM means that even if it had received no risk-adjustment payment and if its administrative costs were the $65 PMPM NYC administrative cost average, the Crystal Run ACA individual-market products would have operated at a loss. Crystal Run was spending $1.63 for every $1 it received in individual-market premiums in 2016. It was only in the small-group market where risk adjustment was a major issue because Crystal Run made risk-adjustment payments of $228 PMPM, which had the effect of driving the $240 PMPM net income loss in the market.


In 2016, Crystal Run’s individual-market silver plan premium of $370 per month was the lowest in the Hudson Valley rating region by 4% (Affinity was second lowest) and 33% below the median premium of $555 (UnitedHealthcare). Crystal Run responded to its 2016 underpricing by increasing its 2017 silver plan premium to $617—a 67% increase versus the average silver plan premium increase of 12% (NYS-DFS 2016). In 2016, Crystal Run had low enrollment in the individual market, enrolling on average only 403 (7.5%) of the 5,354 people in Orange and Sullivan counties who enrolled for coverage through the exchange. After the 67% silver plan premium hike, enrollment fell to just 113 (2%) of the 5,347 people who bought coverage through the exchange. This dynamic was different from the New York City startup plans, which were competitively priced within 10% of median premium but lost money because of medical costs in excess of premiums (Block 2018).

Figure 3 shows that in the individual market in 2016, Crystal Run incurred a $227 PMPM net loss because the average $348 premiumAverage premium of $348 aggregates plan purchases of bronze, silver, gold, and platinum plans, which have varying premiums. The actual silver plan premium in 2016 was $370. was $85 PMPM less than medical costs net of risk adjustment of $433 PMPM. However, in 2017, the average $593 premium was sufficient to both cover $403 in medical costs net of risk adjustment and $151 in administrative costs and still leave a $39 PMPM in income.

Figure 3. Crystal Run individual market financials, 2016 and 2017

Small-group market

Crystal Run and other startup plans, most notably Oscar and CareConnect, originally targeted the individual market. After initial years of large financial losses in the individual market, however, Oscar, CareConnect, and Crystal Run all added the small-group market as a core part of their businesses. For example, Crystal Run’s small-group enrollment grew from 15 in 2015 to 1,291 in 2017 (NYS-DFS 2016). While health plans are permitted to enter and exit the individual and small-group markets separately, startups with low membership may make participation decisions across both markets to ensure that administrative costs can be scaled, a factor that likely contributed to Care­Connect and Crystal Run closing their plans in 2017 and 2019 respectively. In the small-group market in 2016, Crystal Run took in premiums of $369 PMPM, made risk-adjustment payments of $228 PMPM, had medical expenses of $241 PMPM, and incurred administrative expenses of $140 PMPM (Figure 2). The result was a net loss of $240 PMPM in the small-group market. The risk adjustment payment was large, but even if it had been zero, the administrative costs of $140 PMPM would have led to a loss of $12 PMPM.


Although Crystal Run blamed the federal risk-adjustment program for its exit from the ACA exchange market, publicly available financial data on other aspects of its financial performance show that other factors, such as low premiums and high administrative costs, might have been of greater importance. Crystal Run Health Plan had poor financial performance in 2016, driven by premiums 33% below the median. However, when it increased premiums to the median level in 2017, enrollment declined. In the small-group market, Crystal Run’s poor financial performance was driven by risk adjustment and disproportionately high administrative costs due to plan membership an order of magnitude smaller than the next smallest plan’s membership. While risk-adjustment costs were about 30% of premiums in the small-group market, the tiny overall membership of the product being offered in two small New York counties resulted in unsustainably high administrative costs.


Axelrod D. Crystal Run to pull plug on its health plans. Times Herald Record. March 29, 2019.

Block AE. A financial analysis of New York City start-up health plans and reasons for their losses. Manag Care. 2018;27(12):34–36.

NYS-DFS (New York State Department of Financial Services). Exhibit 17–historical claims data by policy forms included in rate adjustment filing. NYS-DFS. 2018 & 2019.

NYS-DFS. 2017 Average approved premium rates. Updated 2016. www.crainsnewyork.com/assets/pdf/CN106720824.pdf. Accessed June 13, 2019.

NYS-DOH (New York State Department of Health). Table 2: Population, land area and population density by county, New York State–2010. https://www.health.ny.gov/statistics/vital_statistics/2010/table02.htm. Accessed June 13, 2019.

Trish EE, Herring BJ. How do health insurer market concentration and bargaining power with hospitals affect health insurance premiums? J Health Econ. 2015;42:104–114.