Description

The for-profit education companies have been among the worst performing stocks over the past 18 months due to increased regulatory scrutiny on Wall Street. While it may be tempting to view them as once-in-a-lifetime bargain, a closer analysis of the impact on proposed regulatory changes reveals that several of the more vulnerable stocks may continue to decline. Based on my analysis, the stock that remains the most exposed to further weakness is Education Management (EDMC). Below is an analysis of four for-profit companies, which demonstrates that all the companies analyzed are affected but EDMC fares by far the worst. 

The industry's fortunes over the next 12-24 months hinge almost entirely on a single 'binary event.' The binary event is the Department of Education's recently proposed 'gainful employment' initiative, which ties a program's tuition to a debt-to-income ratio of 8% based on likely post-graduate employment. The Department of Education will decide in the coming weeks whether to: 1.) proceed with its proposal in the current form; 2.) modify the proposal to less onerous terms; or 3.) drop the gainful employment proposal and instead simply tighten the definition of and guidelines for gainful employment. Despite the significance of this event, Wall Street sell-side analysts have offered little fundamental analysis to determine what the actual impact of gainful employment would be on the profitability of the companies and the valuations of the stocks.

The twin factors one must examine with the for-profit education companies and gainful employment are: probability and outcome. This analysis makes no effort to 'handicap' the probability of gainful employment. It focuses solely on the potential outcome the rule would have on the revenues and profits of four companies: Apollo Group (APOL), ITT Educational Services (ESI), Education Management (EDMC) and Corinthian Colleges (COCO). I will go into great detail to 'show the work' of the analysis, and I openly invite Club members to offer recommendations on refining the work. It uses the best available data on Bureau of Labor Statistics on employment classifications and disclosure/estimates on APOL, EDMC, COCO and ESI's student population mix. Predicting outcomes with incomplete data sets is inherently imprecise. The outcome, based on this analyst, is a considerably negative one for all four companies: the industry's profitability would be significantly hindered by the gainful employment proposal, particularly for high-tuition providers such as ESI and EDMC.

Before going into the full explanation and findings of the analysis, I will summarize the findings:

• The Street must fundamentally alter its perception of this industry amid the specter of gainful employment. This industry is loved on Wall Street for its robust operation margins and consistent growth. As the analysis shows, the operating leverage the works in the industry's favor in a growth environment reverses violently when tuitions decline.
• Schools with high-cost Bachelor's and Associate's programs are most profoundly hit. While APOL suffers significant erosion in profitability, it fares better than ESI and EDMC because its programs are lower cost. EDMC, among the highest-tuition institutions in the industry, fails to generate a profit under any scenario examined in the analysis.
• Part of the industry's survival/adaptation post-gainful employment would most likely be a reduction in advertising spending. While this would likely choke off avenues for enrollment growth, maintaining marketing spending of 15%-plus is simply uneconomical in a post-gainful employment environment.
• If gainful employment takes effect, these stocks are severely overvalued. While the companies can take measures to maintain profitability, the profits and growth characteristics of these businesses are curtailed meaningfully.
• The analysis contradicts the notion that the effect of gainful employment would be to 'kill' the for-profit industry. While it is clear there would be severe losers, other companies with lower tuitions should survive - and may ultimately thrive because the lower-cost programs would likely ensure better outcomes for students.
• Masters, Ph.D and diploma programs are less affected by gainful employment. Due to the relatively short time horizon and relatively modest tuition, many certificate/diploma and Associate's programs priced below $25,000 manage to pass muster in a post-gainful employment landscape. At the other end of the spectrum, many Master's and Doctoral programs manage to pass the 8% debt-to-income test because the salary data from the Bureau of Labor Statistics are considerably higher on average than middle-ground programs such as Bachelor's and Associates.

Below is a summary of the estimated impact on gainful employment as currently proposed on the four companies' previous full fiscal year results - to use forward-year earnings estimates would add an additional layer of speculation to the analysis. The analysis uses two different scenarios:

• 8% debt to income with no changes in operations. The first scenario assumes the 8% debt-to-income ratio takes effect and the companies have no other changes to their operations. Of the four companies analyzed, only APOL earns a profit.
• 8% debt to income with 20% cut in marketing costs. The second scenario assumes an 8% debt-to-income ratio. This scenario also assumes the companies cut advertising/marketing spending by 20%, as well as an 85% reduction in doubtful accounts and a reduction in debt-collection expenses. The latter two assumptions are based on the notion that a reduction in tuition would result in fewer delinquencies and debt-collection issues for the companies. Under this second scenario, three of the four companies are profitable: APOL, ESI and COCO, albeit at lower levels pre-gainful employment. The fourth, EDMC, remains money-losing due to heavy concentration in high-cost 'liberal arts'-type programs for entry-level positions.

 The analysis also examined the potential impact of gainful employment if the Department of Education modified the proposal to a 10% debt-to-income ratio. Some concession from the Department of Education is a distinct possibility, including a relaxation of the debt-to-income standard employed in determining the calculation. Again, the analysis uses two scenarios:

• 10% debt-to-income ratio with no changes in operations. This scenario assumes a 10% debt-to-income ratio takes effect with no operational changes by the companies. Under this scenario, APOL and ESI manage to deliver a profit, albeit at lower levels than pre-gainful employment. COCO posts a modest loss, while EDMC remains deeply unprofitable.
• 10% debt-to-income ratio with 20% reduction in marketing costs. The second scenario assumes a 10% debt-to-income ratio. This scenario also assumes the companies cut advertising/marketing spending by 20%, as well as an 85% reduction in doubtful accounts and a reduction in debt-collection expenses. Under this scenario, APOL is actually slightly more profitable than it was pre-gainful employment - although its growth trajectory may be hamstrung by the reductions in selling and promotion. ESI and COCO manage reasonable levels of profitability - although their valuations appear stretched. EDMC remains unprofitable.

 The conclusions of the analysis are evident, but the reader may require greater clarity on how this analysis was conducted. Before a company-specific explanation, I will explain the overall inputs required for the analysis. The research comprises a two-part analysis. First, it analyzes for-profit companies' tuitions in relation to the debt-to-income ratio required by the Department's proposed change and determines how tuition levels need to change to adhere to the proposed new standard. Second, it estimates what the impact of these changes would be to the companies' income statement in the most recent reported fiscal year.

 Gainful Employment Impact on Tuition

The first portion comprises essentially six components: tuition, salary, debt-servicing requirements, debt levels, programs and enrollment.

• Salary: The analysis refers to the BLS data on occupations, using the 25^th percentage, per the Department of Education's recommendation. (http://www.bls.gov/oes/current/oes_stru.htm#23-0000)
• Debt servicing: For interest rates on student debt, the analysis uses the 6.8% applied by Stafford Loans. Many for-profit colleges rely in part on private loans that carry a higher interest rate, but the analysis assumes all debt at the 6.8% level.
• Debt levels:  Most companies disclose the overall percentage of tuition revenue that comes from Title IV financing, but it is an average across many programs and degree levels. The analysis uses public-company information when available, and estimates from Wall Street analysts when specific data are unavailable. The analysis reduces the reported or estimated figure by 10%. The reason: some for-profit companies say the average student has completed roughly 10% of the credit requirements before enrolling; the analysis assumes this 10% was paid for entirely in cash.
• Tuition: The analysis uses information made available by the companies on tuition levels when available. When not available, it uses estimates from Wall Street analysts.
• Programs: Some companies offer more than 100 programs, many with employment outcomes that cannot be easily defined by BLS classifications. The analysis includes as many programs as possible where tuition and salary levels are available. When a direct comparison is not available, the analysis takes the highest salary among the next-closest matches.
• Enrollment: The estimates on program enrollment in the programs analyzed come from percentages provided by the companies in degree levels: diploma/certificate, associates, bachelors, masters and PhD. The analysis uses program percentages from companies when available, and estimates from Stifel Nicolaus on percentages of students enrolled in particular programs of study.

Company Specific Analysis

Education Management (EDMC) Assumptions and Analysis: Education Management, which reported 136,500 students for the year ended December 2009, fares the worst in this analysis. In this analysis, EDMC loses money in every scenario. Before detailing the analysis, a few assumptions merit explanation:

• Percentage of debt financed is lower than EDMC's disclosure due to assumption of 10% credit hour transfers. Some for-profit education companies indicated that the average student enters the program with 10% of the necessary credits for completion. In EDMC's case, the assumed percentage of tuition financed by debt is 72%, 10% less than EDMC's reported 81.5% tuition funded by Title IV. Private loans, 13.5% of EDMC's total tuition revenues in 2009, charge a considerably higher rate than the Stafford Loan rate used in the analysis, so the debt-servicing levels would likely be higher than the analysis' assessment. That debt is not included in the analysis.
• Programs are compressed when necessary by field of study. The EDMC analysis uses typical EDMC programs in the media, design, health, culinary, fashion, business, legal, education, IT and behavioral science segments to approximate the closest possible tuition and salary levels.
• Bachelor program tuition assumption is $85,140. EDMC's tuition runs higher in locations such as Hollywood, where programs run as high as $96,576. This analysis uses $85,140 for Bachelor's tuitions across the board -- $473 per credit hour with 180 hours required for graduation. This was the lowest Bachelor's price found during my research.
• Bachelor's program analysis employs generous assumptions on salary. In its recruiting material on its college websites, EDMC routinely notes that its Bachelor's degree programs typically result in entry-level positions. Further, to its credit, EDMC lists average salaries for its recent graduates. However, the analysis uses BLS statistics and often uses more tenured professional classifications - for example, the Associate's degree in Fashion Design analysis uses an assumed salary of $42,150 compared with the $31,878 reported by the Art Institutes website. An analysis with 'entry-level' positions for Bachelor's considers worsens EDMC's profitability beyond the current analysis.

Earnings Impact from Gainful Employment Scenarios: The table below shows the estimated impact of gainful employment on EDMC's 2009 results. The tables show four scenarios: an 8% debt-to-income ratio with no operational changes by EDMC; a 10% debt-to-income ratio with no operational changes; an 8% debt-to-income ratio and a 20% reduction in selling/promotion costs and other operational changes discussed; and a 10% debt-to-income ratio and a 20% reduction in selling/promotion costs and other operational changes discussed.

EDMC Program Examples: Below are three examples of the analysis I conducted on Education Management's programs. I analyzed the revenue implications on 34 programs.

Apollo Group (APOL) Analysis and Assumption: Apollo Group, which reported 443,000 students for the fiscal year ended in August 2009, registers a decline in profitability under most assumptions in the analysis. There are several assumptions to discuss:

• Tuition levels for non-online Bachelor's degrees assume on-ground credit hour costs rather than online costs. APOL's website indicates it charges $530 for most Bachelor's level online credit hours and a mix at on-ground Bachelor's programs: $345/credit hour for lower-level courses and $475 for upper-level courses. In the analysis, all non-online Bachelor's programs analyzed use the assumed on-ground costs rather than online tuition costs. This computes to $49,200 per Bachelor's program (60 credits at $345 plus 60 credits at $475) rather than the apparent online cost of $63,600.
• Percentage of debt financed is lower than APOL's disclosure due to assumption of 10% credit hour transfers. Some for-profit education companies indicated that the average student enters the program with 10% of the necessary credits for completion. The analysis for APOL and all other companies assumes this 10% credit hour completion. In APOL's case, the assumed percentage of tuition financed by debt is 77%, 10% less than the APOL's reported 86% tuition funded by Title IV sources.
• Student percentages within programs are estimated due to lack of disclosure by APOL. APOL provides the breakdown of its student population by degree, but not by program: it offers degrees in business, criminal justice, education, arts & sciences, human services, healthcare and technology. The analysis assumes a breakdown based on conversations with Wall Street analysts, a review of courses offered and the length of time each program was available. While imprecise due to the lack of disclosure, the assumptions are: 30% business; 21% healthcare; 21% education; 18% technology; 5% criminal justice; and 5% in arts & sciences/human services. In analyzing each program by degree mix, the analysis assumes a degree mix proportionate with APOL's total student population mix.
• Programs are compressed when necessary by field of study and degree mix. APOL offers at least 97 programs, many of which do not have readily applicable Bureau of Labor Statistics SOC Codes. The analysis examines 29 APOL programs as an approximate barometer of the tuition/salary mix across the various APOL programs. The analysis attempts to choose programs with directly-corresponding SOC codes, and selects the closest match when no direct SOC code exists (a limitation noted by the Department in its Issue Paper on Gainful Employment). The analysis uses an approximation of the students enrolled in each program by dividing the assumed enrollment mix by program and degree by the number of similar programs analysts (example: four business Bachelor's programs are analyzed, each with an estimated enrollment of 12,270 - 49,080 assumed business Bachelor's students in APOL divided by four).
• Master's and Doctoral programs do not assume accumulated undergraduate debt, resulting in all programs meeting the Department of Education's proposed standards. The analysis assumes Master's and Ph.D program students have no previous debt from undergraduate programs. This results in all Master's and Doctoral (16% and 2% of total enrollment, respectively) tuition costs meeting the 8% and 10% debt-to-income requirement, because the assumed salaries for these advanced degrees are higher.
• Apollo Global programs are not analyzed. APOL derives less than 2% of its revenues from its global education programs and was not yet profitable in FY09. Because of the non-material impact of these programs, they are not factored in to the analysis.

Earnings Impact from Gainful Employment Scenarios: The table below shows the estimated impact of gainful employment on APOL's FY09 results. The tables show four scenarios: an 8% debt-to-income ratio with no operational changes by APOL; a 10% debt-to-income ratio with no operational changes; an 8% debt-to-income ratio and a 20% reduction in selling/promotion costs and other operational changes discussed; and a 10% debt-to-income ratio and a 20% reduction in selling/promotion costs and other operational changes discussed.

APOL Program Examples: Below are three examples of the analysis I conducted on Apollo's programs. I analyzed the revenue implications on 29 programs.

ITT Educational Services (ESI) Assumptions and Analysis: ITT Educational Services, which reported 80,766 students for the year ended December 2009, suffers a sharp decline in its earnings under the various scenarios in the analysis. In this analysis, ESI loses money under the worst scenario, and in the other three scenarios would sport a P/E multiple between 24x and 75x. Before detailing the analysis, a few assumptions merit explanation:

• Percentage of debt financed is lower than ESI's disclosure due to assumption of 10% credit hour transfers. Some for-profit education companies indicated that the average student enters the program with 10% of the necessary credits for completion. In ESI's case, the assumed percentage of tuition financed by debt is 68%, 10% less than ESI's reported 75% tuition funded by Title IV and private lending sources. Private loans, 5% of ESI's total tuition revenues in 2009, charge a considerably higher rate than the Stafford Loan rate used in the analysis, so the debt-servicing levels would likely be higher than the analysis' assessment.
• Programs are compressed when necessary by field of study. As in the case with APOL, the ESI analysis uses typical ESI programs in the technology, criminal justice, business and healthcare segments to approximate the closest possible tuition and salary levels. 
• Bachelor's program analysis employs generous assumptions on salary. In its recruiting material on its website, ESI routinely notes that its Bachelor's degree programs can 'help graduates begin careers in a variety of entry-level positions in various fields.' However, the analysis uses more tenured professional classifications - such as the designation of 'software engineer, applications' in the analysis of its Bachelor's program 'software engineering technology.' The 25% level for a software engineer is $67,790 versus an entry-level-type position in IT such as 'computer support specialist,' which has a $33,680 salary at the 25% level. An analysis with 'entry-level' positions for Bachelor's considers worsens ESI's profitability beyond the current analysis.
• ESI's Master's programs are approximately less than 1%, and immaterial. The analysis omits ESI's estimated 808 graduated students because the programs are relatively new and have little impact on the analysis.

 Earnings Impact from Gainful Employment Scenarios: The table below shows the estimated impact of gainful employment on ESI's 2009 results. The tables show four scenarios: an 8% debt-to-income ratio with no operational changes by ESI; a 10% debt-to-income ratio with no operational changes; an 8% debt-to-income ratio and a 20% reduction in selling/promotion costs and other operational changes discussed; and a 10% debt-to-income ratio and a 20% reduction in selling/promotion costs and other operational changes discussed.

ESI Program Examples: Below are three examples of the analysis I conducted on ITT Education Services' programs. I analyzed the revenue implications on 14 programs.

 Corinthian Colleges (COCO) Assumptions and Analysis: Corinthian Colleges, which reported 86,088 students for the fiscal year ended June 2009, loses money in the 8% and 10% debt-to-income ratio scenarios without operational changes, but is profitable when it curtails its advertising spending. Before detailing the analysis, a few assumptions merit explanation:

• Percentage of debt financed is lower than COCO's disclosure due to assumption of 10% credit hour transfers. Some for-profit education companies indicated that the average student enters the program with 10% of the necessary credits for completion. In COCO's case, the assumed percentage of tuition financed by debt is 77%, 10% less than COCO's reported 86% tuition funded by Title IV. Private loans, not included in this analysis, charge a considerably higher rate than the Stafford Loan rate used in the analysis, so debt-servicing levels would likely be higher than the analysis' assessment.
• Programs are compressed when necessary by field of study. The COCO analysis uses typical COCO programs in the health, criminal justice, business and technology segments to approximate the closest possible tuition and salary levels.
• Diploma/certificate programs pass muster due to lower cost. 64% of COCO's students are in diploma/certificate programs that are shorter in duration and cost less than Associate's and Bachelor's programs. In this analysis, COCO doesn't require tuition cuts for most of these programs because the BLS statistics indicate that the average student can subsidize the debt.

Earnings Impact from Gainful Employment Scenarios: The table below shows the estimated impact of gainful employment on COCO's 2009 results. The tables show four scenarios: an 8% debt-to-income ratio with no operational changes by COCO; a 10% debt-to-income ratio with no operational changes; an 8% debt-to-income ratio and a 20% reduction in selling/promotion costs and other operational changes discussed; and a 10% debt-to-income ratio and a 20% reduction in selling/promotion costs and other operational changes discussed.

COCO Program Examples: Below are three examples of the analysis I conducted on Education Management's programs. I analyzed the revenue implications on 34 programs.

Additional Thoughts and Conclusions: The analysis, using the best available data on BLS employment classifications and disclosure/estimates on APOL, EDMC, COCO and ESI's student population mix, indicates that the industry's profitability would be significantly hindered by the Gainful Employment proposal, particularly for high-tuition providers such as ESI and EDMC. Initiatives to reduce expenses would ameliorate earnings erosion, but not enough to justify current valuations. While the analysis doesn't consider what APOL, EDMC , COCO and ESI's growth trajectory would be if the Department's proposal on Gainful Employment takes effect, it is worth noting that the debt-to-income proposal removes the key operating lever of price increases - effectively 100% of price increases drop to the operating profit line. Likewise, if promotion and selling costs have to get curtailed to maintain profitability, the growth trajectory of these businesses is also called into question. I estimate that 75%-80% of revenue from new students drops to operating profits.

Catalyst