The Federal Reserve held a conference in April on research in community development. There were several papers on payday lending, tax refund anticipation loans, prepaid cards and other forms of credit. These non-traditional sources of loans may at first blush seem illogical: Why would someone borrow next week's paycheck for a $30 fee when that amounts to a 360% APR on an annual basis? Or borrow their tax refund in October at a cost of nearly $100 for $1500 in 5 months. Why opt out of checking accounts and use higher cost prepaid cards?
The answer it seems, at least in part, is that people are not well served by financial services and using what is available to as best they can. Jennifer Romich and colleagues followed a small group users of prepaid cards using details interviews. It turns out these consumers liked the limits on prepaid cards and generally eschewed conventional checking accounts because they had been burned in the past. Checking accounts have fees on low balance accounts, charges for check printing and from the perspective of these consumers, banks just seem to be out to get them. For people with just enough income to get by several checks may go out that come close to overdrawing the account. Banks generally will cash the largest check first, bouncing several smaller checks. Consumers see this as capricious and vindictive. So they opt out of the system entirely.
Angela Littwin used similar methods to look at high cost credit cards used by low-income women in Boston. In part because Boston has few short term payday loan options (regulations largely ban these loans), credit cards are a key source of credit. Low income families use credit cards as a form of an emergency fund. But they report being frustrated as the card issuer periodically 'gives' them more credit; they feel they would be better off without the option to borrow more. The author suggests new forms of credit cards with simple fees and low balances that the consumer affirmatively selects--and they can set limits on their own use.
Jeremy Tobacman and colleagues look at pay day loans and credit card use. The authors note many payday loan users actually have credit available on their credit cards, yet still use high cost payday loans. They also find that the use of a payday loan is a precursor to credit card default. Thus consumers may be using payday loans because they know they are in financial trouble. The costs of over charging a credit card and/or late fees or even carrying debt over time on a card are such that consumers are allocating their credit across all available options.
One other paper from the conference that is relevant to this discussion is by Peter Tufano and colleagues. They look at income tax filings and clients using advance refund loans. They then track how clients use their refunds among those who took their refund on a stored value card. The #1 item for which the funds were used? Groceries. For many the purchase happened immediately--within a few days. Less than 10% appeared to buy big items or consumer durables (the stereotype of buying a flat screen TV).
These are all insightful papers and deserve more attention than the few sentences offered here. The overall picture these papers paint is that consumers do make mistakes and suffer from mis-information. But they also are doing the best they can with the tools they have available. And it seems many see mainstream banking-- checking accounts and credit cards-- as more trouble than they are worth. Non-traditional users of these accounts are slapped with fees and charges on a regular basis making even a high-cost payday loan a cheaper alternative. It is also clear these products all interact, as consumer trade off one form for another--probably not something regulators pay attention to. Moreover as financial services consolidates and firms look to raise more revenue from basic transactional and small credit services, it seems likely more consumers on the fringe will opt out of mainstream services.
New products and markets are developing to better serve this population. So long as they offer transparent pricing, consumers are likely to be better off for it. A better understanding of the landscape of consumer services in this segment is clearly needed.
Friday, April 24, 2009
Sunday, April 12, 2009
Why so few loan mods?
The HASP effort from the Treasury/White House/HUD has been in play for over a month now. By some measures, lenders are beginning to use the program and at the least there are refinances being made through Fannie and Freddie for loans with loan to value ratios of 80-105% of appraised values ("shallow under water" loans).
In the last week we heard that FHA loans are performing poorly. This created headlines about an FHA 'bailout'...which is not quite right. FHA charges a premium and has generally added to the Federal revenues every year due to a surplus of premiums paid over payouts (there is no "lock box" for Federal programs). In many ways this has been a tax on homebuyers which returned billions to the Treasury. Due to a slew of loans in 2005-2007 where downpayments were gifted by sellers (a blueprint for disaster to all who watched it happen), FHA has a book of loans that is doing lousy. That program has ended and FHA volume overall is robust. FHA may need, for the first time in history, a subsidy for its premiums. Or it could simply to raise its current premiums. But this is not quite a collapse of an otherwise pretty good program.
But the news on loan mods is more modest. Still lenders are slow to get mods approved and seem to still end up doing more foreclosures than alternatives. New data will soon be out from HOPE NOW and some state programs to see if or how fast things are changing.
A recent paper by Christopher Foote, Kristopher Gerardi, Lorenz Goette and Paul Willen on the Boston Fed website (Public Policy Discussion Paper P09-2) with the pithy title "REDUCING FORECLOSURES" has some interesting implications.
The authors argue that loans originated with high mortgage payments relative to income, or at so at "unaffordable levels," are not the main reason that borrowers default. They suggest it is actually household income shocks such as a job loss or income reduction, combined with a declining house price market.
Because investors usually gain more from foreclosures than modifications, the system is set up to pursue foreclosure options even if the borrower may be able to maintain a modified loan. They argue that PSA and securitization agreements are not the barrier to servicers often suggested, and that investors are not putting legal pressure against mods as much as simply are economically better off foreclosing. It would be better for society, borrowers, neighborhoods, etc for more mods to be made, but probably not for specific investors facing a trade off between forms of losses. Both a mod and foreclosure mean a loss. Mods may just prolong the problem in too many cases.
The authors conclude that mods may need to be re-directed. Rather than reducing payments below 31% of income for 5 years across the board, as in the HASP, a better approach would be to focus on the trigger event to get the borrower through until a better income level is achievable, then revert to the prior terms.
It is a nicely written paper and engages a useful discussion of what we ought to consider affordable or "sustainable" lending. It should be noted they do not find income to mortgage payment ratios are not unimportant, just that income:payment at the time the loans is made is not a main driver of default. From other data we know mods that reduce payments tend to perform better over time. In fact the authors repeat this finding. But the situation each borrower is in varies, as does the availability of data on changes in income over time.
One proposal put forth by Steve Malpezzi and colleagues at the Center for Real Estate at the University of Wisconsin Business School is that unemployment benefits could include a temporary housing voucher which pays any amount over 30% of income for making housing payments. The result is a bit like a short term insurance contract against default.
It seems unlikely there will be much momentum for HASP II or even another round of foreclosure prevention policies in the near term. But as housing policy is designed going forward, better attention to what happens when housing and labor markets fall at the same time at scale will be a hard learned lesson.
In the last week we heard that FHA loans are performing poorly. This created headlines about an FHA 'bailout'...which is not quite right. FHA charges a premium and has generally added to the Federal revenues every year due to a surplus of premiums paid over payouts (there is no "lock box" for Federal programs). In many ways this has been a tax on homebuyers which returned billions to the Treasury. Due to a slew of loans in 2005-2007 where downpayments were gifted by sellers (a blueprint for disaster to all who watched it happen), FHA has a book of loans that is doing lousy. That program has ended and FHA volume overall is robust. FHA may need, for the first time in history, a subsidy for its premiums. Or it could simply to raise its current premiums. But this is not quite a collapse of an otherwise pretty good program.
But the news on loan mods is more modest. Still lenders are slow to get mods approved and seem to still end up doing more foreclosures than alternatives. New data will soon be out from HOPE NOW and some state programs to see if or how fast things are changing.
A recent paper by Christopher Foote, Kristopher Gerardi, Lorenz Goette and Paul Willen on the Boston Fed website (Public Policy Discussion Paper P09-2) with the pithy title "REDUCING FORECLOSURES" has some interesting implications.
The authors argue that loans originated with high mortgage payments relative to income, or at so at "unaffordable levels," are not the main reason that borrowers default. They suggest it is actually household income shocks such as a job loss or income reduction, combined with a declining house price market.
Because investors usually gain more from foreclosures than modifications, the system is set up to pursue foreclosure options even if the borrower may be able to maintain a modified loan. They argue that PSA and securitization agreements are not the barrier to servicers often suggested, and that investors are not putting legal pressure against mods as much as simply are economically better off foreclosing. It would be better for society, borrowers, neighborhoods, etc for more mods to be made, but probably not for specific investors facing a trade off between forms of losses. Both a mod and foreclosure mean a loss. Mods may just prolong the problem in too many cases.
The authors conclude that mods may need to be re-directed. Rather than reducing payments below 31% of income for 5 years across the board, as in the HASP, a better approach would be to focus on the trigger event to get the borrower through until a better income level is achievable, then revert to the prior terms.
It is a nicely written paper and engages a useful discussion of what we ought to consider affordable or "sustainable" lending. It should be noted they do not find income to mortgage payment ratios are not unimportant, just that income:payment at the time the loans is made is not a main driver of default. From other data we know mods that reduce payments tend to perform better over time. In fact the authors repeat this finding. But the situation each borrower is in varies, as does the availability of data on changes in income over time.
One proposal put forth by Steve Malpezzi and colleagues at the Center for Real Estate at the University of Wisconsin Business School is that unemployment benefits could include a temporary housing voucher which pays any amount over 30% of income for making housing payments. The result is a bit like a short term insurance contract against default.
It seems unlikely there will be much momentum for HASP II or even another round of foreclosure prevention policies in the near term. But as housing policy is designed going forward, better attention to what happens when housing and labor markets fall at the same time at scale will be a hard learned lesson.
Subscribe to:
Posts (Atom)
Posts
