Re: Student Loan Data Could Shake Up the For Profit Education Industry

The Department of Education released preliminary data regarding its repayment rate calculation for its proposal on gainful employment. The initial data suggests that repayment rates are worse than we anticipated. Several companies, including Washington Post'sWPO Kaplan unit, could be materially affected if their students lose access to Title IV funds. While we don't plan on changing our fair value estimate for Washington Post at this time, we've long considered Kaplan a lower-quality institution relative to some of its peers, and have assigned the company a high uncertainty rating as a result. While the data released looked bleak, numerous companies have mentioned inconsistencies between their data and the numbers released by the Department of Education. For example, Strayer STRA held a conference call this morning to explain some of these inconsistencies. Strayer's management had previously stated it believed it would pass the 45% repayment rate measure, based on its internal data. However, the DoE's release suggested only a 25% repayment rate, well below the 45% standard for full eligibility. Additionally, DeVry DV noted in a press release that it understands that the DoE intends to rectify an oversight in regards to the methodology used to calculate medical school repayment rates, so the data could change. DeVry's Ross University School of Medicine had a repayment rate of only 16%, but Harvard's medical school repayment rate was also very low at only 24%. This suggests that the repayment rate numbers released today may not reflect the quality of the programs, which could cause the government to rethink its new proposals for determining which programs are eligible for Title IV funds. Only three of the 12 companies that Morningstar covers were above the 45% rate: Universal Technical Institute UTI (49%-64%); Bridgepoint BPI (45%-52%); and Grand Canyon LOPE (52%); Apollo's APOL University of Phoenix was close at 44%. Programs with repayment rates below 45% can still remain fully eligible if they pass a debt-to-income ratio test. However, the data surrounding this metric is even more obscure. Given the thresholds (8% debt-to-income or 20% debt-to-discretionary income) required to remain fully eligible, many of these schools could fall into a restricted category, which could limit enrollment growth substantially. If a school falls below a 35% repayment rate and fails to meet a 12% debt-to-income and a 30% debt-to-discretionary income measure, it could lose access to financial aid support for its students. This would have a material impact on any school, as a significant portion of revenue usually comes from financial aid sources. According to the data released, Kaplan's various programs had repayment rates between 13% and 51%, with a 28% average. One of the key issues in the calculation, according to Strayer, has to do with how consolidation loans are calculated. When Strayer noted that it would likely pass the repayment measure, management had assumed repayment rates on consolidation loans would be similar to its other loans. However, the DoE counts consolidation loans against a school if the student isn't paying any principal. With many consolidation loans offering interest only and graduated income-repayment options, this could hurt many schools. Also, there are $130 million of loans missing from the DoE's calculation, compared to Strayer's calculation. The company had no insight into this difference and is working with the DoE to rectify it. We believe that counting consolidation loans against schools doesn't make a lot of sense given that consolidating loans are often in a student's best interest, and schools do not control the terms lenders might offer. We believe that there are many flaws in the DoE's proposal and that its data may not be accurate. There is no guarantee that the DoE will change its stance on its methodology surrounding consolidation loans. However, we believe once more accurate data is available, the DoE may decide to revisit its proposal. Overall, we think the murkiness of the data the DoE released and the inconsistencies with the companies' own internal data confirm our belief that the current gainful employment proposal could be difficult to actually implement. Also, we think the lack of perfect correlation between the reported repayment rates and the quality of the schools suggests that the proposal is somewhat ineffective in spotlighting the difference between valid programs and the bad actors in the industry. However, if the DoE decides to move forward with the proposal anyway, lower-quality institutions would certainly be affected. Todd Young also contributed to this note.