Wednesday, October 1, 2014

Interesting discussion of Muslim extremism.

Discussion of Muslim extremism.

I don't believe that I have ever seen a two-on-one argument between two reporters and one interviewee come out so poorly for the two reporters.   The ability of this particular interviewee to stay focussed and to distinguish among specific situations was amazing.

This interview demonstrates that details are important.

Growth of student debt -- a view from the FRB of New York.

I am reading the recent literature on student debt and will provide some reviews of material that I found interesting.  The first paper to be reviewed in this series is a presentation put together by a researcher at the New York Federal Reserve Bank.

Review of Household Debt and Credit:  Student Debt

Briefly, this report provides some useful numbers on the growth of student debt but the discussion of the causes of this debt growth are off.  

Summary:   The Federal Reserve Bank of New York study looks at statistics on the growth of student debt, student debt delinquency, and the size of student debts in relation to other forms of debt. 

The main findings: 

Total student debt tripled between 2004 and 2012.

Number of student borrowers increased by 70%.

Average student loan balance increased by 70%.

Share of borrowers 90+ days past due to increased from 10% in 2004 to 17% in 2017.

Among borrowers currently in repayment, 90+ past due rose from around 20% in 2004 to over 30% in 2012.

Individuals who are 90+ days on student debt are also more likely to be delinquent on other forms of credit like auto loans, credit card loans and mortgage loans than other borrowers.

Non-student debt declined for all borrowers age 25-30 between 2005 and 2012.  

The share of borrowers with student debt who also had a mortgage fell from around 9% to around 4%.

Comments on This Report

Comment One:  On page 9 of the presentation the author speculates about the potential reasons for the growth of student debt.

These include more people attending college and graduate schools, parents taking out student loans, students staying longer in college and attending graduate school, lower repayment rates as borrowers delay payments though deferments and forbearances and difficulties discharging student debt.  

Most of these factors are not significant or even relevant and in some cases the claim made by the author is incorrect.

Between 2004 and 2012 there was around a 19% increase in the number of people attending college or graduate school.   This rate of increase in enrollment was actually smaller than the rate for the previous 8 years; hence, it appears that increases in enrollment had little impact on total student debt.

Many students with debts had parents take out PLUS loans but the PLUS loan program has been around for awhile and I am unaware of any statistic suggesting that expanded use of PLUS loans is why student debt is increasing.   There is some evidence indicating that use of PLUS loans is associated with large overall student debt totals (although this may be tautological) but again the PLUS situation is not changing.  

Students are not taking longer to complete their undergraduate degree.    Around 39% complete their program in four years.    This number does not appear to be changing much.

The request for forbearance on loans and payment delays is likely affected by the business cycle and the general state of the economy.   The author did not provide information on the percent of loans not in repayment or in forbearance in 2004.

It has always been the case that student loans are not dischargeable even in bankruptcy.   There was a rule change in the bankruptcy code of 2004 that made it impossible to discharge private student debt but this rule change only impacts relatively few borrowers and is not a driver of total debt.  The 2004 rule change giving credit cards priority over student debt in bankruptcy may be the more important issue.  (see comment four below.)

Comment Two:  The primary cause of the increase in student debt totals for the new generation of students is the increase in tuition.   Here are two articles documenting the extent of the increase in college tuition.

Comment Three:  Another factor leading to higher student debt totals is the increase use of for-profit colleges.   For profit colleges have low student graduation rates.

The graduation rate for four-year four-profit colleges is around half the graduation rate for all colleges.

A Congressional report has tons of information on the impact of the growth of for-profit colleges on student debt.

These statistics show that for-profit colleges heavily recruit students and provide very little in terms of support in the job market.   Tuition is higher at for-profit colleges.    Students are more likely to have student debt and default rates on loans are high.

Comment Three:  The chart on page 18 of the study showing share of student borrowers with new mortgage originations is very interesting.  It would be useful to study the impact of student debt on the age of a person when he buys his or her first home.  It is possible that increases in student debt have reduced real estate purchases and have GDP growth.

Comment Four:  Rules have always severely restricted the ability of student borrowers to get their debt forgiven.   The 2004 bankruptcy law went further by making it more difficult to declare chapter 7 bankruptcy.   In chapter 13 bankruptcy, credit card debt and consumers loans often have priority over student debt.   This change resulted in lower student debt payments for a substantial number of people.   The reduced payment of debt among people in bankruptcy is one reason why the rate of growth of the stock of outstanding student loans increased after 2004.

A change in the bankruptcy law that allows for bankruptcy courts to accelerate the repayment of student loans would benefit both taxpayers and students.   For discussion of this issue see the post below.

Friday, September 26, 2014

A Primer on Adjustable Rate Mortgages:

A Primer on Adjustable Rate Mortgages:

In general rates on adjustable mortgages are lower than rates on fixed rate mortgages.   The current differential is high compared to previous values.  The median difference between a 30-year FRM rate and a 5-1 ARM rate between 2005 and 2014 was around 0.62 percentage points.   In August 2014 this differential stood at 1.15 percentage points. 

Many households will be tempted to choose an ARM over an FRM because of the lower monthly payment.

Borrowers need to understand the risks associated with the use of ARMs rather than a fixed rate mortgage.  This post attempts to provide insight for borrowers who are considering an ARM rather than the traditional fixed rate mortgage.   

Discussion of ARMS:

The initial interest rate on an ARM is fixed for a period of time after the origination of a loan.  After this initial period interest rates adjust periodically.   The movement of a highly publicized interest rate determines the periodic adjustments on the ARM interest rate.

There are many types of ARMS with many different features.

Features of ARMs:

Initial years of constant interest:  Different ARMS set the interest rate for a different period of time.  After the initial fixed-interest period rates adjust periodically.

This advertisement from Bankrate lists 1-year, 3-year, 5-year, 7-year and 10-year ARMs

I played with the calculator and found that many products were not available in my area.   Most quotes were for the ARM that had an initial term of 5 years.  A one-year ARM was currently not available in the Denver area.  Offers of three-year ARMS were sparse.   Rates were a bit higher when initial period of constant interest rate exceeded 5 years.

ARMS available 9-14 in Denver Colorado Area
Initial Term
Adjustment Period
Low Rate
High Rate

Periodic Adjustment:  After the initial fixed-interest period rates adjust periodically.  The most common adjustment period at the end of the initial fixed-year period appears to be one year.  This was the adjustment period for all ARMS on the BankRate Site.

The Index:  Lenders base the adjusted rate on the ARM to an interest rate on a widely traded asset or index.  The adjusted interest rate on an ARM is index rate plus a margin.   The three common indices used for ARMS are the one-year constant maturity Treasury (CMT) index, the Cost of Funds Index (COFI) for the 11th district and the

ARM Indices -- Current Rates and January 2000
Most Recent 2014 Data
One Year Constant Maturity Treasury (CMT)
Cost of Funds Index (COFI)
London Interbank Offer Rate (LIBOR)

Rates on the index were even higher in the 1980s.

Margin:   The index is a riskless asset.   The mortgage is a risky asset.  Lenders are compensated for their risk through a margin over the index.   The margin is typically 2-3%

Caps:   After the initial fixed-interest rate period the interest rate can rise periodically, typically annually.  However, the rate increase is limited by an annual cap and by a lifetime cap.  I don’t have data on the size of the allowable annual or lifetime caps that are available in the market.    Personally, I would look for an ARM that limited annual changes to around 2.0 percentage points and lifetime changes to around 5 percentage points.

Term of mortgage:  Most mortgages have either a 15-year or 30-year term.  Most ARMS tend to be 30-years.  (Often borrowers who seek ARMS do so because they cannot afford the payment on a 30-year FRM.)  However, some borrowers want the lower rate on the ARM and also desire to pay off their loan quickly.   There are some 15-year ARMS in some markets.  See link below.

An ARM with a 15-year term:

Subprime Loan Features: 

Negative Amortization:  Many ARMS allow the balance on the loan to increase if the actual change in interest rates exceeds the capped interest rate.

Interest-only ARMs:  Also, some ARMs allow the borrower to only pay interest on the loan for a period of years.   Such loans will result in higher future payments.

Prepayment penalties:  Some mortgage products, both ARMS and FRMs include a stipulation that borrowers pay a fee if they pay off their loan early when interest rates fall.

The mortgage brokers who champion subprime lending argue that these features allow them to charge lower interest rates.  

Christopher Mayer recently made this argument quite articulately with a straight face.

My brief critique of this argument was published in a previous policy post. 

A lot more research is needed on this issue.   I suspect that very few recent ARMS have prepayment penalties.  Prepayment penalties are likely very popular when interest rates are above their historic average.  

I do wonder whether people who can only borrow when mortgages have these types of features should be purchasing homes.

The view from the CFPB: The Consumer Finance Protection Board CFPB provides some useful information about financial risks

Consumer Handbook on Adjustable Rate Mortgages:

The CFPB has a general list of questions that consumers should ask mortgage lenders.  In my view, disclosures on ARM features should be routinized and made more accessible.   Lenders should be required to include a forecast of future payments based on the assumption that interest rates move towards a historic average.   Borrowers also need to calculate the amount they need to save in order to pay off the loan around the time the interest rate resets.

Concluding Remarks:  The decision on the type of mortgage and/or whether to buy a house cannot be separated from all of the variables affecting the finances of the household.   My view of the market for ARMS is that there are three types of borrowers.   For one group the choice between ARMS and FRMS is a close call.   This group is not affluent but has very little debt, would benefit from the lower ARM interest rate and could most likely handle payment shocks if rates go in the wrong direction.   The second group has enough cash to payoff the mortgage should rates rise so there is little downsize from the taking out an ARM with a lower interest rate.   The third group has large consumer debts and little cash and probably should achieve other improvements in household finances prior to purchasing a house. 

Thursday, September 25, 2014

Are prepayment penalties actually good for risky borrowers?

In a provocative article, Christopher Mayer, Tomasz Piskorski, and Alexei Tchistyi assert that prepayment penalties are actually good for risky borrowers.   They point out that a loan with a prepayment penalty requires a borrower to remain committed to the same lender.  As a result, borrowers who experience a positive shock in income subsidize borrowers who experience a negative shock in income. This cross-subsidization leads to lower initial interest rates for both borrowers. Empirical results presented in this paper indicate that a prepayment penalty could reduce coupon interest rates 14 basis points for subprime first liens and 62.4 basis points for subprime junior liens.

This blog provides a brief assessment of their work.

Mayer on page 3 of the paper argues that many refinancing transactions appear unrelated to interest rate changes and could therefore stem from positive wealth shocks.

“ Our focus is motivated by data from the last decade showing that an appreciable number of refinancing appear unrelated to mortgage rate declines.  According to the Freddie Mac primary mortgage market survey, in June 2003 mortgage rates hit a local low of 5.23 percent and averages 5.8 percent for the year.  Rates remained relatively flat in 2004 and 2005 (averaging 5.84 and 5.87 percent respectively, before rising to 6.41 percent in 2006.  Yet refinancing activity was remarkably strong even in the face of flat or rising mortgage rates.  According to data reported under the Home Mortgage Disclosure Act (HMDA) about 15 million mortgages were refinanced in 2003 when mortgage rates hit their low.  About 7 million mortgage refinancing were processed each year 2004 and 2005 and another 6 million mortgages were refinanced in 2006.  Improvements to household credit during the economic boom and rising home prices (another source of positive household wealth shocks) are likely important contributors to these large numbers of refinancings.”

This argument is not persuasive.   The decrease in the number of refinanced mortgages from around 15 million in 2003 to around 7 million in 2004 can be interpreted as a major decline in refinancing activity rather than evidence that refinancing activity “was remarkably strong even in the face of flat or rising rates.”    Moreover, it would be na├»ve to believe that all high interest rate mortgages were immediately refinanced once interest rates reached their 2003 trough.  Individuals are busy and at times myopic.  It is the lucky borrower who refinances exactly when interest rates are at a local minimum.   Some individuals in 2003 delayed refinancing their mortgage because they incorrectly believed mortgage rates would continue to decline.  Others may not have had enough cash to close on a new loan.

Mayer argues that banning prepayment penalties could lead to higher interest rates, which will lead to higher default rates.  Mayer’s estimates of the interest rate saving suggests the decline in the payment ratio would be small.   The literature reveals that other factors like FICO scores and home equity have a much larger impact on default rates than payment ratios.   Moreover, the finding that savings from prepayment penalties are highest for subprime junior liens suggests that prepayment penalties helped facilitate subprime borrowings, which tend to have higher payment ratios.

Intuitively, prepayment penalties are likely to increase defaults especially if interest rates fall and the home buyer has no way out but to default.  The costs and benefits associated with prepayment penalties would have to consider whether loans with prepayment penalties are more prevalent when interest rates are high.   This would not be good for risky borrowers or for government agencies that guarantee their mortgages.

The Inefficiency of Refinancing:  Why Prepayment Penalties are Good for Risky Borrowers.
Christopher Mayer, Tomasz Piskorski, and Alexei Tchistyi

Monday, September 22, 2014

The smoking gun on Hillary.

Fox is funny.    The smoking gun at the end of this article is surreal. 

On July 8, 1971, Clinton reached out to Alinsky, then 62, in a letter sent via airmail, paid for with stamps featuring Franklin Delano Roosevelt, and marked “Personal.
Will Fox forget about Bhengazi and concentrate on Alinski?

Hillary’s problem is not that she is too liberal but is instead a lack of conviction.  I don’t know where Hillary stands.  Is she for the Keystone pipeline or against it?   She was never out front for gay marriage until perhaps very recently.   Her views on Social Security reform are unknown.  Would she accept or oppose the COLA cut?   Would she link the COLA cut to other changes in the Social Security system?

These are the type of issues that count not a 1971 letter with an FDR stamp.  (By the way,  In my view FDR was our greatest President.)