Last Tuesday at the Value Investing Congress, I gave a 123-slide presentation (posted here) about my largest short position, K12 (ticker: LRN). In it, I shared the many reasons why I think the company (and the entire sector) have run amok and why, at 46x earnings, I think the stock is a fantastic short.
The Bull Case for K12
There are many reasons that explain the bullishness of some investors and nearly all analysts (seven of the eight who follow the stock have a buy or strong buy on it). First, K12 has grown its revenue 32% annually over the past decade, as this chart shows:
And analysts are projecting revenue and EPS growth of 16% and 32%, respectively, over the next 12 months, fueled 10 new schools and enrollment cap expansion of 17,000 new seats. And K12 has less than $1 billion in revenue in a market that it says could become as large as $15 billion annually.
It’s also important to note that online schools sector – schools in which students are supposedly learning by sitting at home all day in front of a computer, interacting with teachers almost exclusively online – can be an excellent option for certain students – for example, children whose pace is extremely accelerated, entertainers, solo athletes, teenage mothers who need to stay home with their babies, victims of bullying, children with cancer, seizure disorders, peanut allergies, etc.
K12 has a well-regarded curriculum, reports very high parental and student satisfaction, and has strong political support, especially among Republicans, for giving parents school choice. Finally, it’s operating in a space, online learning, that has enormous buzz (think MOOCs, etc.).
So what’s not to like? Let me count the ways…
(Note that my critique is specifically of K12, not all online charter schools, for-profit charter schools or blended learning schools. While I think the online charter school sector has, overall, run amok, there are a small number of good online schools – and a few students at even the worst online schools are doing well.)
Weakening Financials and Aggressive Accounting
In my presentation, I show that K12’s profit margins and free cash flows are low and erratic, days sales outstanding has doubled in the past five years, the company uses aggressive accounting techniques to inflate reported profits, and K12’s founder, Knowledge Universe, distributed its entire stake to its investors only a few weeks ago. But I want to draw particular attention to this chart, which shows that year-over-year revenue growth – critical for a stock trading at 46x earnings – has been falling sharply over the past two years, and is projected to fall further in the next year (to 16%):
Year-Over-Year Revenue Growth
Aggressive Recruitment Leads to Vast Numbers of Students Destined to Fail
While online schools can be an excellent option for certain students, it’s a very small number – typically those who have a high degree of self-motivation and strong parental commitment. It’s sort of obvious if you think about it. Do you think you would have learned more during your K-12 educational experience if you’d sat at home in front of a computer, or gone to school and had daily face-to-face interaction with teachers? How many of today’s youth – in a world filled with 500 TV channels, texting, video games, etc. – do you think are better off at home rather than being at school? Lastly, what do you think the answer is for the most at-risk kids, who typically come from poor, single-parent households, in which the parent has little time or ability to be the “parent coach” that’s so critical to online education?
Tom Loveless, an education researcher at the Brookings Institution who did paid consulting for K12 in its early years, said: “The enthusiasts for cyber learning have overstated the potential. What they keep forgetting is we’re not talking about college students here. We’re talking about high schoolers and young kids. The idea that parents go to work and leave their kids in front of a computer—it’s absurd.”
In short, online schools are only educationally appropriate for a very small number of students, which means that K12 and its peers need to be highly selective in recruiting students, warning parents away whose kids are unlikely to succeed at an online school, especially at-risk youth.
So what does K12 do? Exactly the opposite. Since going public in late 2007, numerous former employees have told me that K12 has adopted a growth-at-any cost mentality in order to support its richly priced stock. Jeff Shaw, the former Head of School of Ohio Virtual Academy (which today accounts for 11% of K12’s revenue), told me:
After the IPO, I got discouraged because the company’s priority seemed to shift from academics to growth – it wasn’t so much about academic achievement but on delivering the promised enrollment numbers to shareholders.
[K12’s] call centers that were encouraging enrollment and enrolling students who were obviously ill-suited for learning in a virtual environment. It was apparent to those of us operating schools that parents weren’t being given the whole story. K12 oversold students’ potential to be successful and obligated teachers to do things they wouldn’t likely to be able to do.
Eventually, it seemed as though K12’s enrollment strategy was to cast a wide net into the sea of school choice and keep whatever they caught regardless if the catch was appropriate for virtual learning or not.
…I am shocked that the stock continues to rise. I think it’s a house of cards that is going to collapse. It boggles my mind when I read about and hear stories about what’s going on in schools managed by K12.
Numerous other insiders confirm that what Shaw observed at his school was equally true across the company (notes from these interviews as well as the entire interview with Shaw are included in my presentation).
Deliberately Targeting At-Risk Students
Worst of all, K12 in recent years has been deliberately targeting at-risk students, who are least likely to succeed at an online school. K12 CEO Ron Packard claims that it’s for noble reasons:
It’s just K12’s culture. We want to help kids. It’s just so ingrained all the way through the organization about helping as many kids, doing the right thing for kids.
At another time, he said:
We don’t want to be recruiting kids who it’s not right for. That would be a disaster academically. It would be a disaster for the company economically.
This is nonsense. Luis Huerta, Associate Professor of Education & Public Policy at Teachers College, Columbia University, who has studied K12 carefully and published reports on virtual schools, told me:
The virtual providers like K12 are now mostly going after at-risk kids, kids on their last straw – if they didn’t sign up, many would be dropouts or go back to juvenile court.
K12’s phone banks have figured out a way to target dropouts and special ed kids. They will sign up anyone – as long as that warm body signs in periodically, K12 can draw enrollment money from the district.
It isn’t for some noble reason – it’s because these kids demand the least amount of education. These aren’t kids and parents who will be knocking on K12’s doors saying, “Hey, you need to do more for my kid.”
K12 and Packard use this as an advertisement, saying they’re doing noble things and wondering why they’re being criticized. It’s almost comical. It’s so misleading and conniving.
Again, numerous people confirm Huerta’s account. One person told me that online charters know that “internet advertising leads to lots of students enrolling, but none succeed. Yet online charters are spending big money on this.”
And a former Employment Consultant wrote the following on the web site, Glassdoor:
They push these quotes and “true stories” about all the children they have helped, but the truth is, this product is really only good for about 10% of the market that they target. The only success stories come from homes where there is a large parental support and willingness for the student to learn. K12 markets to low income families who are oftentimes more interested in a free computer and staying out of truancy court than anything else. Not only that, but the curriculum (which is actually very well developed) simply will not work for a low educated family who is having a hard time getting their child to go to a brick and mortar school, let [alone] apply themselves in a home based environment. They push enrollments on families where the adult in the home cannot even read, speak, or write in English, knowing that these students are destined to fail.
While K12 celebrates how many at-risk students it’s serving, I view it as an educational catastrophe in light of the these students’ need for intensive, personalized instruction and hence how inappropriate online education is for most of them.
Low Spending on Teachers, Who Are Harried, Overworked and Unsupported
K12’s online schools spend far less on teacher and administrator salaries than regular schools, as this chart shows:
Per-Pupil Expenditures for Salaries, 2008-09
Source: Understanding and Improving Full-Time Virtual Schools: A Study of Student Characteristics, School Finance, and School Performance in Schools Operated by K12 Inc., National Education Policy Center, 7/12.
In light of such low spending, it’s not surprising that many teachers report excessive class sizes and feeling harried, overworked and unsupported. One teacher I spoke with, who taught English from 2010-12 at Agora Cyber Charter School, K12’s largest, accounting for 14% of the company’s total revenues, told me:
It was a horrible experience. Every teacher had the same experience I did.
Before I started, I was told there would be a lot of support, a low student-teacher ratio, and that if there were students who didn’t show up, they’d take them out and replace them with another. But they took everybody. There was no teacher to student ratio.
When I started, I was assigned 300 students, which was very, very overwhelming. I would try to read each of the essays students turned in a try to grade it and spend the appropriate time, but I was really struggling with that. I couldn’t keep up. I was told to skim over the papers and grade with a rubric.
For each class, I’d have maybe seven out of 30 students in that particular section attend – and even among those seven, just because their name was there showing them present doesn’t mean they were at their computers.
A huge portion of my students never showed up or did anything. I have no clue what happened to them, though I have no doubt Agora was charging the state for them.
When it came time to give grades, I was told, whatever I had to do, I had to pass every student.
I would not say there was much learning going on. If students were doing the program like they were supposed to, it could work – but the majority of students weren’t coming from a family where a parent would help them. (My notes from the entire interview are in my presentation.)
Again, numerous other former teachers confirm this account.
Horrific Educational Outcomes
In light of K12’s growth-at-any-cost approach, targeting of at-risk students, and low spending on teachers, one would expect terrible educational outcomes – but they’re even worse than anyone could imagine.
K12, of course, tries to hide this by producing a slick, colorful 132-page Academic Report (which you can download here), which claims that for K12’s students in reading, “overall achievement was 196% of the norm group gain” and in math, “overall achievement was 97% of the norm group gain.”
These results are virtually meaningless for a number of reasons:
K12 measures its students’ academic growth using the Scantron test, which is not a state-adopted exam, but rather simply a diagnostic tool that is used by K12;
K12 has yet to allow independent external evaluators to both validate its data collection efforts and more importantly evaluate its analysis of student achievement data, across all K12 schools;
The Scantron test is given in the home, unsupervised and untimed, making it easy for students to get help from a parent or the internet;
Only approximately 70% of eligible students take the test (likely to be the better performing ones of course); and
K12’s purportedly strong Scantron results stand in stark contrast to K12 students’ dismal results on state tests.
K12 now admits that the Scantron results can’t be relied on (but that hasn't prevented the company from peddling them to shareholders, politicians, regulators, etc.). K12 Executive Chairman Nathaniel Davis said, "The Scantron tests are optional for K12 students, and about 30 percent decline to take them. That means the company has been comparing a self-selected group of K12 students to the national norm, which isn't appropriate.” The company, he said, needs to find "a more honest assessment" of student progress.
Let’s look at other, independent measures of K12’s educational outcomes. There are various ways to measure them, so let’s start with percentage of schools making AYP (Adequate Yearly Progress) under the No Child Left Behind Act. As this chart shows, barely a quarter of K12’s schools made AYP vs. more than half of all schools:
Source: Understanding and Improving Full-Time Virtual Schools: A Study of Student Characteristics, School Finance, and School Performance in Schools Operated by K12 Inc., National Education Policy Center, 7/12
The same study also showed that 29 of 36 K12 schools (81%) that were assigned school performance ratings by state education authorities failed to earn a rating that indicated satisfactory progress status in 2010-11. Finally, it showed that K12 students trailed state averages for the states in which K12 operates in both reading and math at every grade level (see charts in my presentation).
K12 claims that its dismal academic results are because it serves more at-risk students, but this is questionable for two reasons:
Though its proportion of at-risk students has indeed risen in recent years, it’s from a low base so it’s not clear whether K12 is, in fact, serving a higher proportion of such students relative to state averages; and
Even growth measures show dismal performance.
K12 argues (correctly) that “a more accurate method for measuring student performance is the progress a student makes over the course of a school year, also known as a “growth measure”, so let’s look at these results.
Tennessee, which is well known for having a robust system for measuring student growth, tracked the growth of students in math, reading and science in the 2012-13 school year at all 1,300 elementary and middle schools in the state, including K12’s Tennessee Virtual Academy. As you can see from this scatter plot of math results, TVA students at the beginning of the school year are slightly below the state average (x axis), but their academic growth (y axis) is by far the worst of any school in the state (the results for reading and science are nearly identical, as you can see in my presentation):