Opinion

# Felix Salmon

## Why low-interest payday loans could scale

January 8, 2012

Megan McArdle responds to my post about consumer lending in Missouri by expressing skepticism that it’s possible to lend to people with bad credit unless you do so at extremely high interest rates. She gives a number of reasons why this might be the case, all of which are entirely plausible.

For one thing, she says, the risk of default is high — and particularly invidious when you’re lending money out for short periods of time. Think of it this way: what happens when you lend 100 people \$500 each, for one year, at 10% interest, with a 10% default rate? You start the year with \$50,000. You end it with 90% of the loans paid back — that’s \$45,000 — and another \$4,500 in interest on those loans, for a total of \$49,500. And you also have \$5,000 of defaulted loans, which are worth say 25 cents on the dollar. Which means you make a total profit of \$750.

On the other hand, what if the term of the loan is six months, but the 10% default rate stays the same? Then after six months you’ve got \$45,000 back, plus \$2,250 in interest, for a loss of \$2,750. And if you run the same program again in the second half of the year, you’ll lose another \$2,750. Instead of being down \$500, you’re down \$5,500. Yes, you’ve now got \$10,000 in defaulted loans rather than \$5,000. But even so, you end the year with a loss of \$3,000.

The point here is that defaults aren’t evenly distributed: instead, they’re highly front-loaded. If you haven’t defaulted in the first six months of a one-year loan, you’ll probably pay it off: the probability of default is always highest at the very beginning. And so if you lend for shorter periods rather than longer periods, you have to increase the interest rate you charge, just to make up for the fact that the default rate is not going to fall. Which is the opposite way round to how yield curves normally behave.

On top of that, notes Megan, small loans are labor-intensive: the fixed costs of processing such things are high as a percentage of the principal amount, and the kind of people who take out such loans are, in the parlance of retail banking, “high touch”. These are not the sort of customers who apply on line and repay their loans through online-banking balance transfers.

And so Megan comes to her conclusion. “Credit unions lack expertise and skill in this sort of loan”, she writes, and payday lenders tend not to be particularly profitable, which means that

the reason that the credit unions aren’t putting them into cheaper loans is that they can’t. The cost of an unsecured loan to someone with terrible credit is high because those loans go bad very frequently, resulting not only in the loss of funds, but in considerable overhead expended on collection.

All of this is perfectly reasonable. But I’d point out three arguments pointing in the other direction.

Firstly, what payday lenders are really selling is convenience, at least as much as it is loans. Check cashers, payday lenders, and the like do not keep typical banking hours: they’re open late, they’re open at weekends, and they are generally found in small storefront locations which would not be suitable for a fully-fledged bank branch.

This is entirely rational — you want to be where your customers are, and you need to be able to reach your customers when they’re not working any of their jobs. But at the same time, it’s expensive. And in general, credit unions are already paying for the cost of their overheads, before they start offering any kind of payday loan. So while payday lenders have to cover a lot of overhead from the proceeds of just one product, credit unions have to cover just the marginal cost of the payday loans, which is a great deal smaller. After all, their staff and real estate is already being paid for. If you went up to a payday lender and said that you’d cover the cost of their real estate and their labor, you can be sure that either their rates would come down or their profits would go up.

Secondly, credit unions have a significantly lower cost of funds than payday lenders do. Most importantly, they can take deposits, which pay little or no interest. Payday lenders, by contrast, tend to be small organizations which are largely unregulated. When they borrow money, they have to pay up for doing so.

And thirdly, credit unions are non-profits. There’s only one reason to set up a payday lending shop: to make as much money as possible. Credit unions, by contrast, exist to serve their shareholder-members. If those members need payday loans, then it behooves the credit union to find some way of giving them those loans, even if they’re not particularly profitable.

There’s a pretty strong argument that credit unions should offer payday loans at zero marginal profit — the customers who ask for them are among the credit union’s neediest members, and pretty much by definition they’re also least able to afford expensive financial products. A case could even be made for extending payday loans at a small loss, just because the cross-subsidy is attractive to the membership as a whole. Think of it like an insurance policy: if you join the credit union, then you know that if you ever do fall on very hard times, it’ll be there to help you through them.

Which brings me to the Salary Advance Loan offered by the North Carolina State Emloyees’ Credit Union. Megan wonders whether it’s only possible because state employees’ paychecks are particularly reliable, but the structure of payday loans makes them all relatively safe: they’re fully collateralized by the money coming at the next payday. You can only get one of these loans if you’re employed, and you can demonstrate how much your next paycheck is going to be.

The NCSECU product is very well put together. It involves financial counseling; the creation of a savings account; and a modest interest rate of either 5% or 7%. It lacks the convenience of most payday loans: the credit union surely has shorter office hours than most payday lenders, and the product is more complicated. But the interest charged is so much lower that taking out one of these loans is always going to be a better idea than going to a payday lender who might charge upwards off 400% APR.

And I can’t help but think that with a bit of goodwill, other credit unions across the land could follow suit — even (especially!) those which aren’t specifically designed to cater to a low-income or underserved population. The non-profit business model is a very powerful one in financial services: just look at Vanguard. And I suspect that a lot more credit unions could push the envelope of what they offer, if only they had the will to do so.

Are low-interest payday loans from credit unions “a game changer that will revolutionize payday loans”, in Megan’s words? Probably not — but if they’re not, I think it’s more about credit unions’ willingness to provide these things than it is about their ability to do so. The product exists, and it works. The problem is making it scale.

“So while payday lenders have to cover a lot of overhead from the proceeds of just one product, credit unions have to cover just the marginal cost of the payday loans, which is a great deal smaller. After all, their staff and real estate is already being paid for. ”

A small nit to pick here: if credit unions wanted to offer the same nights& weekends convenience, presumably they would have to absorb some additional staffing cost for that. The daytime service staff isn’t going to want to do that for free.

Posted by Moopheus | Report as abusive

So your original posts had to do with lenders that charge > 100% APR, and McArdle uses an example of 10% interest to show why they have to charge high interest. That’s really not a fair example, it’s more like a straw man argument. Even the regulated banks don’t charge 10% for short term loans (credit card advances), it’s more like 25-30%, so why didn’t she use an interest rate like 30% for her story? My guess is then she wouldn’t have a story, because a 30% interest rate would still yield an acceptable profit with a 10% default rate.

But even a 30% rate is parasitic when you consider the cost of capital is a small fraction of that.

Posted by KenG_CA | Report as abusive

Felix is assuming positive synergies from combining traditional banking with payday lending. They already have the real estate and the staff, so they ought to be able to compete more efficiently. Right?

But I don’t see why this should be the case. The products are VERY different. To operate as a payday lender, you need to have people on hand to immediate process and approve loans. The traditional banking system simply isn’t set up to do this, so you would (at the minimum) need to add staff and systems. Moreover, you probably don’t want to keep the entire branch open until 10 pm — so you’ll need to add a “late night” window as well. Feasible? Sure. But the cost savings aren’t going to be nearly high enough to compensate for the lack of expertise in the business.

Moreover, there is a strong chance of ending up with NEGATIVE synergies from combining the two businesses. Pursue a borrower for defaulting on a payday loan too aggressively and you end up damaging the brand for your more profitable operations. If for some reason a bank wanted to be in both businesses, it would likely choose to operate them under wholly different names for just this reason.

Ken_G, I wouldn’t worry too much about what McArdle wrote. She is good at making up numbers to illustrate concepts, less good at determining whether or not the numbers are realistic. The cost of issuing these loans and processing payments (“high touch”) is surely a big factor in the APR of a small loan — would they be any cheaper if the costs were expressed as “fees” rather than “interest” as the banks are wont to do?

I also question Felix’ assertion that the loans are “fully collateralized”. Does the lender truly have the ability to seize the paycheck? If not, the term is mis-applied.

Posted by TFF | Report as abusive

The Salary Advance Loan …[is] fully collateralized by the money coming at the next payday. You can only get one of these loans if you’re employed, and you can demonstrate how much your next paycheck is going to be … [it] involves financial counseling; the creation of a savings account; [it] lacks the convenience of most payday loans:… the product is more complicated.

With all these caveats, just how available or attractive would such a product be to the consumers of payday loan services? It sounds like it would only be of value to a small minority of credit union members.

Posted by eregibalus | Report as abusive

@Ken_G, I paged over to McArdle’s comments and I think you might be reading her wrong? From the context surrounding, “I’m pretty sure that they’re well north of 10%”, it doesn’t sound like she is talking about an annualized rate. It sounds like she is suggesting that only 90% of the loans are paid back.

On a one-month loan, 12% interest (barely enough to cover a 10% default rate) comes to almost a 300% APR. I have no idea whether or not that assumption of a 10% default rate is realistic or not, but the principle should be plain.

Note also that traditional loans typically use fees to cover their costs (often not included in the APR or amortized over a long payback period). Fees on a mortgage can easily be 2% of the principal — which for a two-week loan would be an APR of 67% even if no interest were assessed. Cash advance fees are also typically 2% plus.

I’m not sure it even makes sense to use APR to analyze very short loans.

Posted by TFF | Report as abusive

TFF. I didn’t read McArdle’s page, I was referring to Felix’s interpretation of her comments. He used examples of one year and six month loans, which I also thought were not typical pay day loans, which have different cost structures like you mention. It is kind of an apples and oranges comparison, so I just was challenging the validity of the example that McArdle provided.

Even on pay day loans, the interest rates are too high, even if some of the costs are considered fees. I really don’t see why the pay day lenders are allowed to operate without any kind of regulation, and traditional loan sharks are not. 12% interest on a one month loan is still usurious. There should be a fee regardless of the amount of the loan, and then an interest rate, which should be capped.

Posted by KenG_CA | Report as abusive

I’m fine with it either way, KenG, but if you cap fees at something half-way reasonable (e.g. 3% on a one-month loan) then I expect you will put the whole industry out of business. These loans are for people who do not have access to “normal” credit. They typically do not have access to normal credit because they are at high risk of default. In most cases they would be better off declaring bankruptcy as quickly as possible, then starting fresh (on a cash basis). But when it comes to paying debts, the poor are typically more ethical (or more fearful?) than the rich. So they try, even when it is realistically just staving off the inevitable.

Payday loans have much in common with loan sharks, including shady collection tactics. Another reason why reputable banks and credit unions won’t want to get involved.

Posted by TFF | Report as abusive

Some information here:
http://www.consumersunion.org/finance/pa ydayfact.htm

Their point #2 appears to be comparing one-month chargeoff rates with annualized chargeoff rates? Still, a chargeoff rate of 3% of loans does not require 15% fees to cover… Wonder what their cost structure is like?

Note that the “cost of funds” for these operations is negligible (even if in the 8% to 10% APR range). The cost of providing these loans is dominated by the overhead and by the chargeoffs.

Posted by TFF | Report as abusive

TFF, the banking industry uses 30% annual credit card interest rates to subsidize an enormous number of negligently granted loans, bad investments, lazy management, and just general incompetence. If you charge 3% (plus an overhead fee) for each month a loan is outstanding, you will still gross 36% (more if it is compounded monthly) each year. If 10% of the loans default, there is still plenty of margin to cover those losses.

Posted by KenG_CA | Report as abusive

Shorter Salmon: payday borrowers value convenience, but I don’t think it’s worth the cost for them. Thus, we should make sure they have access only to cheaper, less-convenient options. After all, if that trade-off is good for a Reuters columnist, it’s good for everyone, right?

Posted by No_Rush | Report as abusive

The point is legit that high interest rates are necessary to compensate for risk of loss, but it’s also true that government has historically defined egregious rates as usury and akin to loan sharking. At a certain rate you risk creating a permanently indebted individual who has no chance of ever becoming debt free. In the long run, it may be better from a societal standpoint to deny the weakest of the weak such credit. For these individuals the problem is most likely a need for training and a better job.

Posted by Sechel | Report as abusive

@Sechel: “At a certain rate you risk creating a permanently indebted individual who has no chance of ever becoming debt free.”

In a society without a bankruptcy statute, I might agree, but under our current code I’m not quite sure how you can say this.

Posted by No_Rush | Report as abusive

@Felix- “The product exists, and it works. The problem is making it scale.” If it exists and it works then it WILL scale.

@TFF- if you cap fees at something half-way reasonable.. then I expect you will put the whole industry out of business.” -Exactly right. Unless we can bring the default rate down to zero. To do that lets lend the working poor their own money. Mandatory federal retirement accounts for all W-2 wage earners. 10% of all wages deposited non-vollentarily on all earnings up to the social security wage base limits.

The money can’t be withdrawn for any reason unless you get a doctors note saying you wife or kid is dying. You can however borrow against your retirement savings just like I can against my 401k. I can only borrow up to \$50,000 or half my balance; but to replace payday lending with a national program the limit could be \$1,000.

A mandatory national retirement savings plan would solve the much larger much more imporntant retirement crisis. Better still with a few easy bolt on programs a forced national savings plan could also solve imporntant problems like the need for low cost short term loans to the working poor better than anything that currently exists.

Posted by y2kurtus | Report as abusive

@Ken_G, you continue to miss the point. If the term of the loan is one MONTH (or less), then you need more than 10% MONTHLY interest just to compensate for default.

Of course the default rate might be well below 10%. That seems to be a number that McArdle pulled out of a hat. The link I posted suggested that just 3% of payday loans default? (But the quality of the data is poor.)

@y2kurtus, that is a very interesting proposal. You might be on to something there?

The key, as you say, is bringing the default rate down to zero. I wonder if the credit union that Felix lauds manages to garnish the paycheck as a guarantee? To manage that kind of interest rate, they can’t afford more than a fraction of a percent of the loans defaulting each month.

Posted by TFF | Report as abusive

TFF, you’re right, if every loan is for one month, and none are rolled over. however, if the distribution of the length of loans is between 1 and 12 months, the interest received will still be 36% of the amount being loaned out, but the total amount subject to default will be less than 10%, even if there is a 10% default rate.

Rather than increasing the interest rate by 10% per month to cover 10% defaults, there should be a 10% one-time fee for each loan to cover the cost of defaults. If you have the loan for one year, it’s just an extra 10% annual interest, not 120%. If the loans are all for one month, then yeah, the interest paid will be > 100%, but most loans won’t be for just one month.

Posted by KenG_CA | Report as abusive

“Rather than increasing the interest rate by 10% per month to cover 10% defaults, there should be a 10% one-time fee for each loan to cover the cost of defaults.”

That would work nicely… Might consider this with other loan products as well, such as mortgages? (In most markets, at least, if a mortgage hasn’t defaulted within five years then it is unlikely to default.) It would lead to a higher cost for those holding mortgages just a few years, but a lower cost for long-term borrowers.

Posted by TFF | Report as abusive

TFF, while I don’t know if such a plan would be good overall for mortgages (it sounds good, but I want to think about it more), the side effect you mention (higher cost for short term borrowers) would minimize flipping of houses in markets where prices are heating up, effectively attenuating the positive feedback loop inherent in free markets that usually leads to bubbles. Not that we have to worry about real estate bubbles for a while, but it would be a better control mechanism than the fed fiddling with interest rates.

Posted by KenG_CA | Report as abusive

If, as Felix claims, defaults are “front-loaded,” then the best “solution” might be a longer-term loan. It is widely accepted that people purchase payday loans because of cash-flow problems–they get the money they need too late. Generally, payday loans are on a short time-frame of a month or so. If, as Felix and Megan claim, default rates are consistent over time, then longer-term loans will have lower annual interest rate.

If monthly loans have a 10% change of defaulting, and so charge a 10% monthly interest rate to cover all costs, then Felix’s calculations imply a 7% gain. This is a 75% interest rate 6 month loan. Now if the default rate stays around 10%, lenders could make a hefty profit off a 30% loan (

Posted by MKCurious | Report as abusive

Clicking Felix’s link to this Salary Advance Loan product, it looks like the APR for most borrowers is 12.0% (not 5.5%), but walk through the math on these loans at 12.0%. Assume an average borrower wants \$300 for 1 week. The max loan is \$500, so \$300 average seems reasonable. One week seems reasonable assuming a 2 week pay cycle, and the loan terms are that a Salary Advance Loan must be repaid from the next paycheck. \$300 * 12.0% divided by 52 weeks (assuming a simple interest calculation of APR) results in \$0.69 of interest. For clarity, that’s less than 70 cents, and it’s 0.2% of the loan balance. For the lender not to lose its shirt, it needs to underwrite to a 0% loss rate, for all intents and purposes. That could be reasonable for this product, since these loans are only available to credit union members who have already established direct deposit of paychecks.

So, assuming that a teller’s salary, benefits, and employer payroll taxes run \$15/hour, this product is maybe break-even if it takes 2 to 3 minutes of time for a teller to process this loan, though that assumes minimal incremental back-office costs to establish and administer the program, as well as a minimal cost of funds for the credit union. The point is that – best case for a credit union – this product pretty much has to be subsidized by other activities. I suppose that a credit union can choose to provide this service at a loss – Felix makes a case for it – but I wouldn’t operate under an illusion that these loans are even modestly profitable on an incremental basis.

Posted by realist50 | Report as abusive

There was a 36% rate cap placed on lenders to military personnel. That didn’t give our troops better options, it gave them less because those lenders closed up shop. The rate cap is designed not for consumer protection, but to shut these businesses down. Rate caps basically serve to say: if you’re poor, you can’t be trusted with your own borrowing, so we’re stepping in and taking credit away. That’s not freedom. That’s not what our troops are defending us for.

Posted by offbeat | Report as abusive

Is a payday advance a good loan option? Learn the truth from a former employee at http://www.advanceamericaonline.blogspot .com/

Posted by marcodmarco | Report as abusive

One thing you mention is that Credit Unions are non-profit. This means they pay no federal or state income tax. You fail to mention that this makes every loan they make – payday or otherwise, taxpayer subsidized to the tune of 40% or so. I could make loans all day if the taxpayers were willing to subsidize every loan I made.

Posted by GWilli | Report as abusive

GWilli, you are confusing “non-profit” and “tax free”. Businesses (and presumably banks) pay taxes on their PROFITS not their REVENUES. Big difference between the two.

Moreover, most non-profit organizations are forbidden to make a profit. If they start to accumulate too much money, they have to find a way to spend it (or cut revenues). Not sure how credit unions manage this, but presumably they have a mechanism to distribute profits to their depositors/stakeholders?

Non-profit or no, you don’t pay income taxes if you don’t turn a profit.

Posted by TFF | Report as abusive

Mr. Salmon,

Would be glad to send to you the PowerPoint given to the CFPB on the NCSECU “payday loan” service (Salary Advance Loan).

SECU has made over 3 million individual loans/advances totaling more than \$2 billion over the last ten years. It’s available 7x24x365 through our Contact Center.

We have “the metrics”! The P&L is on frame #8 on the PowerPoint.

Very much welcome your scrutiny and questions. Even at 12% APR and no fees, it is the most profitable loan we make!

Best regards,

Jim Blaine
President
State Employees’ Credit Union

Posted by JimBlaine | Report as abusive

One other option that is popular is going through a peer to peer loan route, sites such as lending club, prosper, offer consumers real alternatives to borrowing money by letting their peers set the rate based on their credit and borrowing needs, and the rates are certainly not 300% or higher as with most payday loan providers, you can find more information online at http://www.personalloanonline.org for those who might want to explore a personal loan or signature loan from their peers

Posted by kburton | Report as abusive

Mr. Blaine of NC SECU,

As discussed in the North Carolina Commissioner of Bank’s meeting last fall, the assumption of the SECU SALO (payday) loan being profitable was wrong.

The analysis performed by Michael Stegman in his payday paper several years ago and of which has been the source for the claims of the SALO’s profitability is wrong.

Mr. Stegman incorrectly computed the ROA (Return on Assets) by counting the same \$500 loan multiple times (to reach an annualized number). This had the effect of significantly overstating average assets and therefore materially understating the loss percentage.

I also believe there is some question as to the validity of the expense allocation in the ROA calculation. The largest expense of providing a loan is not the cost of funds, nor is it the losses. It is the operating expenses. And whether it is a \$250 loan or a \$2,500 loan, it cost the same to underwrite and managed.

My guess is that if the SECU allowed a proper analysis to be conducted by a legitmate accounting source, the actual results would be considerably less than you have been promoting.

Supporters of “non-profit” lending, Community Development Credit Unions, and progressive advocates have touted this “false” return as a showcase of the cause for attacking market rates.

But the bottom line is that statements concerning this product’s profitability are wrong. Of course that has never stopped some people from continuing to say its something that it isn’t.

Posted by Truthvsfiction | Report as abusive

Mr. Salmon,

Several points being made in your article are incorrect and there are some points concerning the economics of lending that are being ignored.

First, the assumption that losses (defaults) are front loaded is incorrect. While a higher percentage of losses incur in the first 50% of the average term length, losses are none the less incurred throughout the period. And since loan balances very, a higher balance loan defaulting at nine months may carry more impact than a smaller balance at three months. The point being that the statement about “front loaded” losses is incorrect.

Second, if you are lending for shorter periods of time, your are earning less “interest dollars” to pay for the expense dollars you incurred to make and book the loan. And it is interest dollars, not interest rates or APR, that pays the bills. Too often people have lost focus on that truth. Some do so intentionally because it supports their cause.

This goes to the heart of the point Megan was making. That is it is the expense (hard dollars) that goes into making and managing a loan that carries the greatest weight. And therefore you need to earn sufficient interest dollars to cover that expense, then cover losses, then cover cost of funds, before even reaching a profit goal.

Too often in this debate the conversation has centered around what cost of funds are and what is being charged. People then incorrectly assume that if it costs a lender 3% for the money and they lend it at 36%, then the 33% is a large profit.

Yet 36% on a \$500 six month installment loan is just \$53 dollars total interest, or \$8.83 a month. If the allocated operating expense is \$15 to \$20 a month per loan, then providing a loan at that rate is a loss.Numerous studies have shown that it costs \$200 to \$400 a year to book and manage a loan (assuming monthly payments).

Thats why bank’s and credit unions don’t make these loans.The return in dollars is not enough to offset dollars expensed.

The higher rates for small dollar loans are mostly driven by the economies of operating a business everyday, not losses. In fact, most small dollar lender’s losses (annuualized) are lower than credit cards.

Your assumption that credit unions can and should make these types of loans is misplaced. The economic reality is that smaller dollar loans cannot be offered at lower rates because they simply do not produce sufficient dollar revenue to cover costs.

But I do appreciate your focus on keeping the discussion geared towards economics and not making it a totally socialist point of view.

Posted by Truthvsfiction | Report as abusive

If you’re thinking about applying for quick cash payday loans, you have to be at least eighteen, employed and a UK resident. If you fulfil those criteria and have an active bank account, you can apply. You then just need to include a few details about your home, work and banking account, and your application will likely be fast-tracked to a quick decision.

Posted by Piyushr | Report as abusive

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see http://blogs.reuters.com/fulldisclosure/2010/09/27/toward-a-more-thoughtful-conversation-on-stories/
•