Hacker News Reader

Comment by 999900000999 | Hacker News Reader

Good riddance, when I first heard about this I was appalled.

It's basically an extremely high interest loan, combined with bizarro access to banking records. Like for most of these ISAs if you couldn't make the payments, they want it access to your bank account to see that you really couldn't.

You have this really strange setup where if your income increases even marginally, the ISA kicks in. Lambda school, which fortunately also collapsed was notorious for this.

Student loans aren't necessarily bad, they just need harder. Caps. Like federal loans are very reasonable, private ones are not.

Outside of attending medical school or law school, private loans are just a bad idea. On top of that, I think private loans are given out too willy nilly. Anyone can go to a bottom tier law school and take out $200,000 worth of loans.

A good compromise here would be to make private loans fully forgivable, but federal loans are a fantastic deal and help me improve my life.

40k in student loan debt, which is about the max you can take out federally for undergrad, isn't bad.

Anything above 100k can easily be insurmountable, the interest just accumulates way too fast for most people to pay it off.

Then again, I never understood why lambda school needed to be so expensive. You're not running a real school, there's no reason you can't just tell people to take a free class off YouTube, and then have them pay $500 or so to have a project graded. And maybe more adventurous companies would be open to recognizing that project as proof you'd be a good hire.

View on HN

If anyone's interested, I'm a bootcamp grad with a current ISA. The ISA I currently have:

  *pros*:
  - zero interest
  - whatever hasn't been paid after 5 years, I don't owe
  - no payments are due if you don't have a job making over $50k

  *cons*:
  - tuition was definitely overpriced and you end up having to pay 150% of the principal loan (assuming you have a high enough paying job for long enough within those 5 years)
  - a lot of bureaucracy you have to go through to prove you don't have >$50k salary

It definitely sounded scary when I first heard of them. And at the point I was in the bootcamp we were literally dumpster diving and shoplifting to get us through till rent was due and our foodstamps were renewed. Recently got hired with $90k salary so it worked out, but I was very aware of the fact that I was in no position to negotiate and could easily be taken advantage of

EDIT: To clarify, zero interest means zero interest. The 150% is the terms of the ISA agreement. You pay 10% of your paychecks to them until it's either been 5 years or you've payed 150% of the principle amount. Most people that get jobs in tech will have paid the 150% before the 5 year mark

To be clear, this essentially means if I go through this program and end up working at McDonalds for 5 years, all I have to pay is through the labor of bureaucracy

Only if the program doesn't teach you the meaning of the term "opportunity cost"

The unqualified word "pay" is universally understood among English speaker not to include opportunity cost. When someone asks "how much did you pay for that big Mac", people answer "$3", not "$3 plus twenty minutes to walk there and five minutes in the store"

What if you go to a bootcamp, get saddled with an ISA and then work at mcdonalds and somehow get a better job that has nothing to do with the bootcamp? The bootcamp gets your money. Quantity over quality.

> get a better job that has nothing to do with the bootcamp

I'll take it.

Contrast that with a private loan for tuition.

You pay if you have no job, or any job. Debt has no expiration and is not disc arable in bankruptcy.

It is clear that ISAs are worse if you will be successful. It is also clear that traditional loans are worse if you are wildly unsuccessful.

The tradeoff is the tricky part.

Debt does generally have an expiration and is discardable in bankruptcy... unless you live in America. ISAs might be an attractive alternative to how terrible American laws about student debt are, but the real solution is to fix the awful laws.

A better alternative that you can pursue today is just to seek education abroad, Canadian universities are a fraction of the cost for an equivalent service - or be careful about the education you choose, most state colleges are very reasonable for in-state tuition and you can potentially save on room and board.

Ivy-league schools are overpriced and a waste of money.

If we are talking about higher education in general, the US also has a plethora of affordable colleges.

In California, tuition at community colleges is $1,300/year and $5,700/yr for state colleges.

Even the expensive Universities of California are 14k/year.

The real costs are room, board, and entertainment. For example, a shared dormitory room and food plan is easily 20k/year

> Debt does generally have an expiration and is discardable in bankruptcy... unless you live in America.

If we start with other countries, there will be many other catches. Such as personal bankruptcy not being a concept at all in many jurisdictions.

Only around 10% of total loans disbursements are private.

I don't know about other ISAs, but with Lambda school the job had to be related to your bootcamp learning for you to owe them money.

They removed the "you must get a job in software" clause of their ISAs. A general manager at McDonalds in California makes $58k a year. So if you think you will work your way to manager within five years, you can expect to pay them a cut of your salary.

You can call it "zero interest" if you want, but, if you're paying back 150% of principal over 5 years, that's a 19% APY.

https://www.calculator.net/loan-calculator.html?cloanamount=...

To clarify, you can either pay the principal amount (e.g. 20k) or use the ISA which maxes out at 150% of the principle amount. By "zero interest" I just mean it's a fixed amount that won't ever change. This might seem like a frivolous distinction, but it means that you'll never have to worry about getting trapped paying just the interest so I think it makes a big difference

It’s not helpful to calculate APY that way. After all, if they had made the maximum repayment period 50 years instead of 5 years that would result in a lower APY according to your calculation.

In reality the APY will be heavily dependent and on your income (as a high income will result in repayment period less than 5 years, while a low income will result in only partial repayment by the end of the 5 years).

I don't understand, how do you end paying 150% of the principal loan if it's zero interest?

You can either spend $20k up front for the bootcamp, or pay $30k as part of an ISA over 5 years.

That just sounds like interest with extra steps.

actually its interest with the least amount of steps. Amount due isn't variable and no compounding.

Non compounding interest? I think there is a better word for this.

https://en.wikipedia.org/wiki/Interest#Simple_interest

Still has a time component, though. A flat 50% that is due whether you pay it off in 1 month or 60 months is a fee, not interest.

The term is “fee”. It’s a 50% lending fee.

It is a static markup but not interest, as interest varies with time and is dynamic. This is a markup regardless of time.

Because it’s an income share agreement, so whether you land a job making $100k or $51k you still owe 10% of your income to this company for 5 years. Regardless of the amount actually “borrowed”. And I’m guessing there is a cap at 150% of the “loan”. So if you land a high paying job rt off the bat you could end up paying a lot more than you borrowed.

It's not interest, it's... hey, look, something behind you!

See https://en.wikipedia.org/wiki/Contractum_trinius . There are lots of ways to slice and dice cash flows and quibble over what's interest and what isn't. (See also repurchase agreements, Islamic finance, general structured finance, etc. etc.) It's generally more useful to compare cash flows in terms of yield-to-maturity, instead of strictly by interest rate.

see edit. The major difference is you don't have to worry about one of the most common debt traps: getting stuck paying only the interest. The amount you have to pay is totally fixed. 150% of the principle is the most you'd have to pay assuming you've had a high enough paying job for long enough within that 5 year span after graduating (most people do)

Presumably they mean that the total amount paid is not affected by the timing of payments. If they make a lot of money and pay it all off the first year, they pay the same total number of dollars as if they are unemployed for 4 years and make enough to pay everything off the 5th year. It's something like 50% financing fees, but 0% interest, and forgiveness after 5 years.

> I don't understand, how do you end paying 150% of the principal loan if it's zero interest?

That extra 50% is a once-off fee.

A high fee, to be sure, and one that is deferred, but a fee nonetheless.

> - tuition was definitely overpriced

IMO, this is the essence of why ISAs have become so scammy.

ISA-based bootcamps are doing a fantastic job of competing against the perceived cost of a college education. We've all seen news headlines about how private college tuition is reaching $200K. What most people miss is that almost nobody pays full price for college. They have a high headline number for the wealthiest families, but virtually everyone else pays a fraction of this cost or some times nothing at all.

People see the "$30K" price tag for Lambda School and other programs and assume it's a steal, yet most in-state tuitions are in the same range and provide substantially more education.

I've also been completely stunned by how bad some of these programs are. I took the bait about Lambda School back when PG was promoting it all over Twitter, but it has been a terrible disappointment for the small handful of bright students I've known who went through it. Maybe it's changed, but at the time it felt like they were paying $30K of future earnings to have an inexperienced person point them to simple tutorials that they could have found by themselves online. They were basically paying for the certificate that said they completed the program, which is ironically the stereotype that people were using for 4-year colleges at the time.

I totally agree. A lot of people talk about ISA being scams and I basically agree. But I don't think it's a scam in how they assume. If you do your research most ISA agreements are good about being more straightforward than most other types of loans. The scam is usually solely in the principle amount being charged.

That being said, I think ISAs are actually partially successful in that they provided a way for someone broke like me to try something and not have to worry too much about the impact if it doesn't end up succeeding. I wish people in my position had more negotiating power to avoid being overcharged, but that's not a problem unique to ISAs

To put it another way, it’s a high risk loan from the perspective of the ISA lender — which they offset with a high premium to get around the adverse selection are problem. It’s basically a mirror image of the problem that insurance providers face.

> People see the "$30K" price tag for Lambda School and other programs and assume it's a steal, yet most in-state tuitions are in the same range and provide substantially more education.

on the other hand, the time cost of a bachelor's in CS is much higher than that of an 8 month (?) bootcamp. four years is a long time to spend treading water wrt employable skills.

I do feel that the degree is better long-term if it's at all possible. but that's not realistic for everyone. if at the end of the bootcamp you get a decent job that you wouldn't have otherwise, that seems like a good deal. if not, you're out a lot less money and time than flunking out of undergrad halfway through.

> four years is a long time to spend treading water wrt employable skills

Internships during school can be a massive salary increase for many people. Even 25+$/hour is quite a large pay bump for a large chunk of the population, and many internships go well past that.

good point. getting an internship for summer and maybe fall/spring part time is a good way to defray the opportunity cost of not having a full-time job for four years. I think you have to be a pretty strong student and well organized to get an internship before junior year though.

> four years is a long time to spend treading water wrt employable skills.

Only if your sole goal in life is the pursuit of money. Most of the people I know enjoyed their time at university, they got to learn in a structured manner about philosophy and other interesting topics while having freedom to indulge in intellectual curiosity. Life isn't a race to get on a career path.

I do agree that flunking out is probably the worst outcome though, so if academia isn't for you don't feel pressured to pursue it - it certainly won't make any difference in terms of your job prospects by the time you're 30 as a developer. We just need to stop treating university as a vocational training session.

People joining boot camps are often later in life than your typical college student and time is at a greater premium.

> We just need to stop treating university as a vocational training session.

Is it treated as a vocational training session? I thought university was more about signaling some combination of socioeconomic status and ability to learn/motivation. More selective schools and more difficult coursework (such as hard sciences or engineering) signal a potential employee who is able to learn and manipulate the system they are put in, as well as possibly bringing useful connections.

> Life isn't a race to get on a career path.

Presumably life isn't a race at all, which makes you wonder why everyone is in such a rush to attend university when they are young and don't yet have the resources to pay for it. It will be still be there when you are older and the pursuit of knowledge will still be every bit as enjoyable. It may be even more enjoyable as you then have the life experience to appreciate it. Retirement is meant for these types of hobbies.

> Presumably life isn't a race at all, which makes you wonder why everyone is in such a rush to attend university when they are young

It’s the networking opportunities (i.e. “the college experience”) which suffer as the age gap grows between you and the other students. Since networking is the bulk of the value of university, your age during attendance is of disproportionate consequence. I’m speaking from personal experience here — going into a post-grad program, it was immediately obvious to me that the age gap would act as an all but uncrossable chasm. And I’ve seen this chasm in action in undergrad, where it was even more pronounced.

Simply put, university is a lot harder to justify when a significant age gap exists. To your point, maybe most people should wait a few years before attending university, but they don’t, and those who’d otherwise prefer to wait before starting university — which is doubtlessly a wise move for some — have to risk their college experience to do so.

> It’s the networking opportunities

The move to taking university online, and the masses of students following, through the COVID period suggests not. Online university does not provide any unique networking opportunities that aren't already available through every other online service, HN included.

I don't want to spend my whole life waiting for a retirement that might never come - I might die, my mental faculties might degrade, even if everything went right I might still burn out.

If I thought I could take a three- or four-year break to study in my thirties or forties without it hurting my career I absolutely would. But employers seem to be suspicious of that kind of thing (unless it's something like a 1-year MBA). So if I want a sabbatical without hurting my earnings too much, a college degree seems like my best option.

> What most people miss is that almost nobody pays full price for college. They have a high headline number for the wealthiest families, but virtually everyone else pays a fraction of this cost or some times nothing at all.

You're using this statement to claim that college is not actually expensive but I'm pretty sure it supports the opposite claim. Anecdotally, I was accepted into a range of universities and ended up picking ~ the worst one because that was the university which gave me the best scholarships.

If college is extremely expensive then it makes perfect sense that the only people paying full price are the people who have a lot of money: everybody else is filtered out by the high price.

Interesting. What kind of bureaucracy?

  - zero interest
  -  {snip} you end up having to pay 150% of the principal loan (assuming you have a high enough paying job for long enough within those 5 years)

Those two things don't line up, and repeating the 'zero interest' line in an edit doesn't change that. If your repayment is 150%, you've paid interest.

It doesn't really behave like interest though: the amount paid doesn't depend on compounding or your repayment schedule.

That it doesn't compound doesn't mean it's not interest.

That just means it's not compound interest.

It’s a fee, not interest. Interest is charged over time (either compounded or uncompounded, either way you pay more interest if you take longer to pay off the loan). Here they’re charging a onetime 50% fee.

I think the point here is that "zero interest" isn't particularly compelling when the cost of credit is being driven up (to exorbitant levels) by another mechanism. An equivalent loan with a 5-year term would have an interest rate of almost 18% and in this case you don't have the option to save money through early repayment.

Debit is not just big digit, it is an ever deepening well. There are many people that have paid ten of thousand of dollars of their student loans and still saw the total due amount increase.

A loan with a very reasonable 5% overall interest if paid on time can have unbounded interest if paid late enough.

The student dept crisis is dominated by the latter kind of situations.

(still yes, it is zero compound interest + 50% fixed interest)

> There are many people that have paid ten of thousand of dollars of their student loans and still saw the total due amount increase.

The problem is a lack of financial literacy education on one side and a lack of career options for decent-paying jobs on the other side - a lot of office jobs these days require college degrees, and not everyone can work in the trades or join the military.

Personally, I believe that all of this is actually designed the way it is to make money for the banks handing out student loans.

To be pendatic, the fee is just interest that is compounded exactly once.

Instantaneously on contract creation and irrespective of loan term / repayment / etc.

It's such a non-standard contract that the terms loan and interest are kinda fuzzy

> That just means it's not compound interest.

that's not true. simple interest is calculated on the remaining principal, thus it does continue to accrue with time (which this "zero interest 50%" does not)

compound interest means interest calculated on the remaining principal and on accrued interest.

https://www.investopedia.com/ask/answers/042315/what-differe...

Important distinction to make, most people just think "more money means interest" but the mechanics of the two financial instruments differ and this is the entire selling point for these ISAs, and the reason why one may be a good / bad idea in a situation.

-> tuition was definitely overpriced + having to pay 150% of the principal loan

So the actual interest is like 100%? Though it's fixed over a 5 year period, so that turns to around 30% yearly interest rate. Hardly zero.

-> Most people that get jobs in tech will have paid the 150% before the 5 year mark

That's even higher interest for these people. But I guess it's an incentive for the program to get people good jobs.

That's not 30% unless you repay real quick, if you paid it at the end of 5 years in a single payment that's 8.5%, 1.085^5 = 1.50.

In straight amortization using $30k tuition, $45k repayment, over 60 months is $750/mo which works out to 17.25% rate.

It isn't a great deal but it's not terrible vs. other low recourse private loan options.

>It's basically an extremely high interest loan,

Yeah, because there's adverse selection going on: "But people with high credit scores tend to have better, cheaper options than ISAs". All the academically successful student can get student loans/scholarships to traditional 4 year programs, so you're left with people who are genuinely bad mixed with diamonds in the rough. Because of this, you have to charge more to make up for the bad application pool. I'm not sure why this is worth complaining about. It's like complaining that lawyers who work on contingency overcharge compared to ones that work by hourly.

>A good compromise here would be to make private loans fully forgivable

So basically turn them into unsecured loans with sky high interest rates?

Do you have any idea the issues that would solve?

Warning: US centric post below.

If you could file bankruptcy on private student loans, education costs would plummet overnight. Before you claim that education quality will go down, MOST of the money for both private and public colleges/universities ends up somewhere else beyond education. Football stadiums, expensive ad campaigns, fancy buildings, large marketing departments, even political lobbying. Kill off untouchable private loans and place restrictions on public loans then you will see money flow in the right direction.

ISAs are foolish, regardless, because you don't need a degree to write code or manage infrastructure. My income was near $400,000 last year, and I not only didn't go to college, I dropped out of high school due to being homeless. In either W2 or contractual work, I've NEVER had a company question why I did not have a college degree. I understand some folks need the structure, and I think those folks need to be protected. For non engineering jobs or more structured jobs, ISAs are borderline criminal. For everyone else? Unless you like throwing away money that you don't have, stay away.

> ”there's no reason you can't just tell people to take a free class off YouTube, and then have them pay $500 or so to have a project graded”

I think there is a reason you can’t do that. People won’t learn to program with that. I do not defend Lambda School anymore, but I think you are oversimplifying things to make your point. Which, for me, has the opposite effect of thinking less of your point when you oversimplify like that.

As far as I'm concerned, that's more than enough to at least give someone an interview

If they can complete a difficult project, and get though a whiteboard interview, why not hire them as a junior developer.

Just because someone paid 30k for a boot camp doesn't mean they can code.

I think the issue is that there is demand for a reputable (hirable) $3,000-30,000 program which can teach you to program in less time and/or more flexible schedule than traditional university.

I think one good solution is Western Governor’s University which falls in that price range and is extremely flexible on timeline.

But honestly there are a lot of students who want high quality intensive instruction but cannot attend a top-10/20 university. It’s very very hard for most people to find high quality learning outside of those programs unless you get extremely lucky to be in the Bay Area or at a company which can provide an environment full of experts to learn from.

I think the issue being criticized isn't about credentials, which smart companies don't care much about already, but about ability to learn. Your oversimplification of equivalent education being to just watch a free series of videos and get graded on a project has the same flavor as just telling people to teach themselves.

I mean, learning on your own is certainly a thing, especially in our field. My own beginnings were self-taught from a book. But structured environments are important, too, and for many people the only chance they'll have of learning something, because they can't or won't otherwise. Does the "cost" (which we're calling a flat $X as if it were paid upfront, ignoring the ISA details) have to be so high for these bootcamps? Probably not, but if you're going to attempt to replicate something like a structured experience, your costs aren't going to be nothing. But that's beside the point, since in the style of "charge more" advice from startup founders, if you're trying to build a business, your cost is what the market lets you get away with, not what some randos on the internet not in your target customer base think is "fair".

>has the same flavor as just telling people to teach themselves

Worked for me.

Noting bad happens if you open up VS code and write bad Python. I wanted to make games with my friends, subsequently I learned enough to get an entry level job.

What about the Odin project, that's free. Coursera is very inexpensive.

I'd advise at least trying that before blowing 30k. Particularly if these bootcamps are fudging placement numbers.

I agree, it happens, and it's a good idea for people to at least dip their toes in the seemingly infinite free or low cost resources out there to gauge things before spending very much money. (Again ignoring that the whole premise of these ISAs is that you don't spend money except under some reasonable "win" conditions.) But lots of people don't do that. Lots of people can't, even, though with some structure some of them can learn and be as or even more competent than those who took a mostly self-taught immediate-goal route, or self-taught-with-formal-supplementation. Seeing how little pre-college programming experience matters vs. coming in with no programming knowledge is an interesting lesson of doing college. Learning programming is actually complicated, what works for you is not going to work for everyone.

> Seeing how little pre-college programming experience matters vs. coming in with no programming knowledge is an interesting lesson of doing college.

Wait what do you mean by this? I started programming in 6th grade (after begging my dad for a month to spend $50 on a Java textbook) and so the CS portions of college was incredibly easy. When I didn't understand algorithms I just implemented them and learned by doing (and I instantly understood how they could be applied to past projects). I would show up to classes only on the midterm and final days because it was such a headstart. I mostly focused on the humanities because there's nothing the computer science department could teach me that I couldn't teach myself online.

I'm saying on average such experience tends to wash out, at least among those who graduate, though I saw it in individuals by sophomore year. You have two groups, one with some level of pre-college experience, and the other without (not even a HS programming class, though those with basically just that but not any personal learning are an interesting middle ground with parallels to 9-5 employee types that treat programming as just a job). Both groups have those who find the material easy overall (though the second will likely attend lectures; but these days the first too because attendance is mandatory at many places) and both groups have those who struggle on some things and must work and as a side-effect improve their general learning-to-learn abilities. (My own experience was a mix of both depending on the class despite having a general head start where I began in 8th grade with a book and a website dream.) By the end you'll see similar distributions with respect to measures like grades or time-to-complete-assignment or job placement and success, the starting buffs wash out.

Admittedly my sampling is unscientific, and I'm willing to believe pre-college experience can help avoid some of the culling that happens, and it was especially useful to me to land a mostly part time job that wasn't an internship. I also may be biased by my choice of school where most my data comes from (though talking to coworkers and interns from various other schools at my job after college contributed) because my school was a lot more hard-core in terms of course load and especially in later years project focus -- the number of required CS courses typically exceeded other schools known for good CS programs like, say, Stanford, with 5 or 6 overall courses per semester rather than a more common 3-4, unless you did it in 5 years instead of 4.

> As far as I'm concerned, that's more than enough to at least give someone an interview

Are you a person who interviews? If not your opinion is of little weight. If you do hire people I’m sure if you advertise your policy you’ll get lots of people who will be happy to try their luck.

Not only did Lambda School crash and burn (Bloom Tech Institute rebrand = LOL). But those of us who managed to not give up and actually learn in other ways... are still on the hook for that ISA repayment. What a monumental scam.

Since apparently it's not connected to credit score, what stops you from just not paying it?

If small claims court is genuinely their only recourse, that sounds like something I might try my odds in, especially if I have a compelling case that I've been screwed.

Maybe all we need for education is a room and a minimum wage employee making sure we are actually studying. After all, education happens in the students head, not the teachers.

I'm just thinking out loud. Lambda school obviously failed to deliver on its promise but what if we can deliver the same service for very little money?

I never fully understood why students would go to a bottom tier law school and take out 200K in loans, when you very well know the value of a bottom tier law school. I mean, today is not the lack of information age. You can pretty much find out all the data points needed about a school. For instance, bottom tier schools typically don't advertise the graduating salaries or employment data. Just this one data point is an indicator of future earnings potential.

That being said, top law schools or top schools for that matter are within reach for low income individuals as long as they clear the entry criteria.This entry criteria is a whole another debate but Lambda and other schools also have an entry criteria.

There are need based scholarships and I think Stanford/MIT/Berkeley offer free tuition if your parents earn less than 100K per annum or something like that. So opportunity is definitely there but then not everyone can make it to Stanford, yes but that doesn't mean you saddle yourself with a huge debt by going to a lower tier school, knowing that your loan repayment capacity will be very low in the initial years of post graduation employment.

> when you very well know the value of a bottom tier law school. I mean, today is not the lack of information age. You can pretty much find out all the data points needed about a school. For instance, bottom tier schools typically don't advertise the graduating salaries or employment data. Just this one data point is an indicator of future earnings potential.

You're thinking about this from a very different set of life experiences. Not everyone knows to look for things like that. Not everyone even realises that there are different tiers of law schools. If no-one in your neighbourhood has ever met a lawyer socially, why would you?

But people do have an understanding that some schools are better than others. That's why Harvard has prestige and is well known even in those neighborhoods and some random school doesn't.

They might have an understanding that Harvard is something famous. They probably don't know that that translates into earning a whole lot more. I certainly didn't.

I'm not following. Afaik Lambda didn't collapse. It recently rebranded due to a trademark conflict, but they're still alive and well from what I gather. Someone be so kind to explain?

Lambda School leaked documents show poor performance over the last two years https://news.ycombinator.com/item?id=28989092

Lambda School agrees to end deceptive educational financing practices https://news.ycombinator.com/item?id=26946972

There's nothing to follow - most of the comments here are wrong.

ISAs align incentives between teaching institutions and students, they're strictly better than the tuition model. They don't fix all issues because they still require selectivity on admissions (since the school needs to students to succeed in order to survive).

ISAs are also not guaranteed to be good, but the ones originally used by Lambda School were good (Lambda School is somewhat of a third rail topic on HN, so it's worth just considering the terms).

- Only require repayment if students gets a software job making >50k.

- Payment was bounded to 10yr or 30k whatever happens first.

Compare to college tuition which charges huge sums of money often paid for by non-defaultable loans and schools don't really care if students ever get employed by anyone.

From his tweet thread conclusion:

1. Consumers are confused by ISAs

2. And when they take the deal, they often behave poorly

3. And when they behave poorly, you don't have great recourse

4. And there's a looming regulatory threat

5. And the financial markets aren't supportive

My takeaway from this is that it's harder to build an incentive aligned business that really gets students to succeed than it is to take their money as tuition and not have to worry about that so much for business building. I guess I'd argue no shit - that's why ISAs are better, they force the companies to be good at getting positive student outcomes because it's an existential risk if they don't.

That's kind of the entire point.

> ISAs align incentives between teaching institutions and students, they're strictly better than the tuition model.

Well, there's two ways you could operate an ISA:

1. You invest $20k in each student, carefully selecting only the applicants who are most likely to succeed, and giving them the most impactful, high-ROI courses you can, so that they almost all repay at least $20k

2. You invest $500 in each student using prerecorded video classes and assignments graded by unpaid 'mentors', take on far more students, and every time you luck into a student who repays $30k that's pure profit, baby.

Compare to the tuition equivalents:

1. Invest $20k in each student, etc.

2. Invest $500 in each student, and get paid $20k in pure profit right away, baby.

So while the ISA model doesn't prevent scammers, I have no counterargument to "they're strictly better than the tuition model".

This is true and one of the initial criticisms of "there's not necessarily incentive alignment here" I thought of when I first encountered the idea. Hearing about #2 happen in reality is also sad. Though I think over longer periods more competition and reputation trashing has a chance to correct some of #2. Nevertheless I think "University" of Phoenix is still operating and hoovering up tax payer money despite being an internet punching bag for years, so...

You missed a third way which I think is actually most likely, given recent monetary stimulus policies:

3. You invest $500 in each student using prerecorded video classes and assignments graded by unpaid 'mentors', take on far more students, then structure groups of ISAs together into ISA-backed bonds, which you then sell to pension funds and other institutional investors for 15% off par value, or $25.5k each.

Not only do you get paid up front, but whether or not the ISAs eventually get paid back is no longer your problem. Also now instead of recruiting and vetting 200 teachers to teach your 2000 students, you recruit 10 teachers and a couple experienced sell-side traders and analysts.

To be fair, the original post was complaining that he couldn't find suckers to sell the loans to ^W^W^W "there are not mature capital markets for ISA programs to tap".

"ISAs align incentives between teaching institutions and students, they're strictly better than the tuition model."

That's very, very hand-wavey. I'm in broad agreement with you that, at this moment in time, ISAs are probably better than tuition for most people, however, they are an immature financial instrument (relative to traditional tuition loans) and so little effort has been into extracting as much profit from them as possible... but that doesn't mean it's not possible, nor does it mean that it won't happen, and it certainly doesn't mean that there is some inherent alignment between ISA institutions and students.

All the positives about ISAs could have been said about student loans a few decades ago! That's before even challenging the assumption that higher student income is best for all parties involved: if a student discovers that actually, their new career is miserable, there is suddenly a huge incentive mismatch.

The ISAs are being sold to investors. It's in the interest of the teaching company to set the price as high as possible and minimize program spend, not as low as possible with maximum program spend (which is what the student wants).

Lambda’s ISAs are not incentive aligned for most students. Let’s say the income threshold is 50k.

They are not incentive aligned for students who make between 0-50k. Lambda gets no money.

They’re extremely unaligned for students making a little over 50k. The student takes home drastically less money because they hit the threshold.

They’re aligned for students making some narrow range above 50k.

But once you go above that range they are no longer aligned because lambda is capped out.

They are not incentive aligned with most students, incentivizing Lambda to act against the interests of most students, which they have.

> "They are not incentive aligned for students who make between 0-50k. Lambda gets no money."

The are aligned because <50k is a failure for lambda. They're aligned because they need students to make >50k.

> "They’re extremely unaligned for students making a little over 50k. The student takes home drastically less money because they hit the threshold."

The point of lambda school is to get people decent jobs when they didn't have any good options prior to that. A little over 50k where 14% of salary is paid back is the worst case, but probably still better than before. Definitely better than >50k of debt and no job from tuition costs.

> "But once you go above that range they are no longer aligned because lambda is capped out."

Good student returns funnel more students into the program. If you can succeed with this at scale it's a big deal. If they can actually teach people to dev vs. just selling a credential there are even more ways to build on this.

I don't think the problem is that it is a loan. The real problem is that you can manipulate people to sign up for the ISA and then give a low cost bad quality service. If you scale this up high enough, there are going to be people whose income is going to increase above the threshold just through chance alone. If you provide some low quality MOOC that costs you very little per person then you basically get to take a share of someone's income without providing any value.

If the institution was accredited and there were inspections and punishments to keep quality above a certain threshold the model might work but right now it basically invites companies to operate like scammers.

Federal student loans are a major driver behind increases in college tuition costs (1). Like you say, debt isn't necessarily bad - but there's a strong argument that subsidized debt and the lack of risk in granting loans just gives a bigger pool money for institutions to take from.

Point being, federal loans are not a good deal. Each dollar you borrow effectively increases your tuition by 60 cents.

(1) https://www.newyorkfed.org/medialibrary/media/research/staff...

"2. And when they take the deal, they often behave poorly 3. And when they behave poorly, you don't have great recourse"

Yes. As it turns out when you invest in people with no skill, background or money - your investment goes to zero (excluding outliers). For companies like ycombinator the outliers are everything. But in this horrendous business model they can't capture the outliers' profits without being exploitative.

If you honestly sat me down for 4 hours I couldn't think of a worse idea. Uber for cats? Rating people with a general global score? Selling loans to people out of rehab. I can't do it.

First 15 minutes you learn in any loan business is: Income, Debt-to-income Ratio, Collateral. If those are out the window - you are either better than a trillion dollar loan system - or completely insane.

I think we’re between a rock and a hard place. We can’t tighten the student loan spigot. We have held out college as the route to upward mobility for everyone, especially Black people and immigrants, and the demographics of the people who will be excluded from tightening student loans will make it politically toxic.

In view of that I see no alternatives to price controls. Any school that takes federal student loans can charge no more than $X,000 annually. I hate price controls but I don’t see the alternative in a market where government policies otherwise both goose demand and limit supply.

Price controls are being tried out in the UK and it doesn't work - every institution charges the maximum, so it's no longer a ceiling but simply the price everyone pays annually to go to university (£9,250 at the moment).

What about the alternative to ISAs proposed by OP?

Better than an ISA is an income-dependent loan with some minimum amount that must be paid back regardless of the program outcome.

With this setup:

1. You concede and comply with the existing regulation

2. You get access to the existing capital markets so financing costs come down

3. You can use credit reports as a way to enforce the contract so collection rates go up

4. Consumers already understand debt

And, most importantly, the consumer gets a better deal than a classical loan.

Similarly a bad idea relative to plain old federal loans?

Federal student loans already have income-based repayment options, along with forbearances and public-service loan forgiveness options.

I don't see how this proposal is a better deal.

It’s basically a proposal to do the same thing for private loans. Unlike the federal government, the originator of a private loan will care about repayment rates, thus somewhat aligning incentives.

> 40k in student loan debt, which is about the max you can take out federally for undergrad, isn't bad.

The max for undergrad federal loans is $57,500.

That’s the overall max, but the increasing yearly limits mean you can’t reach that unless you take more than 4 years.

If you finish in 4 years, the max is about $45k

> Lambda school, which fortunately also collapsed was notorious for this.

This statement seems to be going unchallenged. It wouldn't surprise me, but if you check @BloomTech and @Austen on twitter, you can see they're still talking it up: https://twitter.com/BloomTech https://twitter.com/Austen

The rebranded Lambda School still offers ISAs:

https://www.bloomtech.com/tuition/income-share-agreements

So they didn't go away and Lambda didn't completely collapse.

The outcome based loans aren't better, they don't have a payment cap and don't expire if you get a job offer above $50,000 a year.

I wonder if they offer every graduate a teachin' position.

I learned from your comment that Lambda School fell apart a couple months ago. And I'm a pretty regular HN reader. Pretty interesting in the context of the founder being super active on social media. The denouement is always quieter I guess.

I think they just rebranded.

Rebranded (to avoid getting sued for using someone else’s trademark)

https://www.bloomtech.com/

> The online coding school that invests in you

> Train remotely to become a software engineer or data scientist and pay nothing upfront until you are earning $50k or more

"Regulatory arb" makes it sound like they did some clever financial engineering to conjure money out of this air. It actually means they were able to skirt regulations meant to put a limit on the interest rates that borrowers pay.

I think everyone in finance knows "regulatory arb" is the polite term for "we found a legal loophole".

If it weren’t for private loans, it would have been impossible for me to go to college.

Did Lambda fold or just rebrand?

The thread admits that savvy financial consumers have better options:

> Second, credit scores were the most predictive variable of good participant behavior for us. But people with high credit scores tend to have better, cheaper options than ISAs.

And then claims that the simplicity of the instrument makes them a better option than debt ("simplicity" isn't a virtue if the alternative is middle school pre-algebra levels of "complicated" and comes at much lower prices):

> Furterhmore [sic], consumers were consistently confused by ISAs and had a vague sense they were exploitative. I think ISAs are less complicated than debt, but consumers don't have experience with then.

His TL;DR has borderline misleading levels of... interpretation. The real story is that ISAs are exploitative and most consumers -- ESPECIALLY the ones who might actually do okay in life -- can't be "educated" out of seeing that an obvious raw deal is an obvious raw deal. So you're left exploiting the folks everyone else considers too risky, who, it turns out, are... too risky.

Like... he says so himself, in so many words.

ISA, from a high level perspective, is a private, capitalistic institution, trying to prove that taxes makes both parties come out on top.

Reply on news.ycombinator.com