Household Balance Sheets: Continuing Fiscal Cliff Impact

During the first quarter of 2013, household balance sheets improved with increases in home values and reductions in mortgage debt, thereby boosting household net worth. These are favorable improvements that will help housing demand in 2013. In particular, over the last five quarters household real estate values have risen by more than $2 trillion.

Since the end of the Great Recession such developments have typically been associated with a decline in the personal savings rate. Due to the Fiscal Cliff, this relationship has been disrupted for the last two quarters.

The graph above plots the current value of net worth to disposable personal income (NW/DPI) and the corresponding 25-year historical average (1982-2007). The dashed blue line charts the personal savings rate. Household net worth data are from the Federal Reserve’s Flow of Funds and the savings rate and disposable income data come from the Bureau of Economic Analysis National Income Product Accounts. 

While there has been a general trend of an increasing NW/DPI ratio since early 2009, there have been ups and downs in this process due to stock market and other asset value fluctuations. As of the start of 2013, the NW/DPI measure stood at a value of 5.86, above the historical level of 5.24 and significantly higher than the cyclical low of 4.85 set during the beginning of 2009. The increase since 2009 is a reasonable measure of the improvement in household balance sheets.

However, for the last two quarters, the savings rate and the measure of disposable income experienced non-balance sheet driven movement due to the Fiscal Cliff and its resolution. In particular, the legislation staving off the Fiscal Cliff included a number of tax increases, including an end to the payroll tax cut and rate hikes at the top end of the income distribution.

In anticipation of these higher tax rates, the amount of income paid out and earned by American households jumped in the last quarter of 2012 (the uptick/peak in 2012 for the blue line below). As this accelerated income was due to a one-time cause, spending did not increase at the same rate. Thus, with a rise in income and no corresponding increase in spending, the personal savings rate increased to a revised rate of 5.3%  in the fourth quarter of 2012, marking the highest rate of savings in more than two years.

As we forecasted in March, the data for the first quarter of 2013 showed that DPI fell at the start of year due to the accelerated income payments made at the end of 2012, as well as the enacted 2013 tax hikes. The result was that the NW/DPI measure reached its highest level since 2008. With a drop in DPI, and little change in consumption, the personal savings rate fell considerably.

It is important to note that while we’ve typically associated a rise in NW/DPI as a sign of recovering balance sheets, in this case the rise was drive by the decline in DPI. Similarly, the drop in the savings rate to 2.3% is less a sign of healing balance sheets, and more due to the timing impacts of income shifting due to the Fiscal Cliff. Data from the second and third quarter of 2013 will provide a clearer picture.

Nonetheless, household balance sheet repair is ongoing. Flow of Funds data from the first quarter of 2013 show that total home mortgage debt continues to decline. Since the first quarter of 2008, home mortgage debt has declined 12% or $1.27 trillion. And the value of real estate owned by households has risen for the last five consecutive quarters for an increase of $2.2 trillion.

Growth in Consumer Credit Accelerates

According to the Federal Reserve Board, growth in the amount of consumer credit outstanding, this includes outstanding credit extended to individuals for household, family, and other personal expenditures, excluding loans secured by real estate, accelerated in April. According to the release, the amount of consumer credit outstanding increased at a seasonally adjusted annual rate of 4.7% in April, 1.1 percentage points higher than the 3.6% growth in consumer credit that took place in March. At the end of April the total amount of consumer credit outstanding was $2.8 trillion.

The increase in the amount of consumer credit outstanding reflects an expansion in both the outstanding amounts of non-revolving credit and revolving credit. While growth in non-revolving credit accelerated from the previous month, the monthly increase in revolving credit reversed the decline that took place in March. The amount of non-revolving credit outstanding, which is mostly composed of student loans and auto loans, rose by 6.4% or $124.5 billion on a seasonally adjusted annual basis in April. In March, non-revolving credit rose by 5.7%. Meanwhile, revolving credit outstanding, which is largely composed of credit cards, rose by 1.0% or $8.2 billion on a seasonally adjusted annual basis to $849.8 billion. In March, the amount of revolving credit outstanding fell by 1.3%.

The April expansion in revolving credit largely reflects an increase in the amount of revolving credit holdings at depository institutions. In April, depository institutions accounted for 80.4% of total revolving credit holdings. As Chart 1 illustrates, revolving credit increased by $2.6 billion in April on a not seasonally adjusted basis. Depository institutions contributed $4.0 billion in revolving credit holdings and credit unions accounted for an additional $0.1 billion. However, these gains were partially offset by a decline in the holdings of revolving credit by finance companies and pools of securitized assets. The monthly increase in the holdings of revolving credit by depository institutions may indicate that lending activity is beginning to grow.

Student Loans Conditions Vary Across the Country

Student loan debt reflects the cost of an investment in human capital. The typical return on this investment is characterized by both higher wages and a more stable employment. College graduates and those with some college experience tend to have higher wages and lower unemployment rates than their counterparts with a high school degree or less. However, failure to repay student loan debt could impair a borrower’s access to other forms of credit restricting their ability to buy a home or a car.

Recent research from the Federal Reserve Bank of New York focuses on the geographical distribution of student loan debt. Higher than average student debt per borrower typically occurs in states along the coastal states of the country, while the majority of the country, especially those located in the middle of the mainland, tend to have lower than average student debt. The one exception is Illinois (i.e. Chicago).

Counter-intuitively, with the exception of Florida, Mississippi, and Louisiana, states with higher than average student debt per borrower tend to also have an average or below average percent of student loan balance that is seriously delinquent (90 or more days late). The majority of the states that have above average serious delinquency rates are also states with lower than average student debt per borrower.

Earlier research by the Federal Reserve Bank of New York illustrates that borrowers who are seriously delinquent on student loan are less likely to secure mortgage credit. However, the underlying reasons behind serious delinquency in student loan debt are not immediately clear. Some states that have a high share of seriously delinquent student debt also have an elevated unemployment rate or low employment growth. Conversely other states with strained employment indicators have below-average student delinquency rates. At the same time, some states with higher than average seriously delinquency rates also suffered the worst of the housing bust, but this is not the case across the 50 states, the District of Columbia and Puerto Rico. Moreover, the link between student loan delinquency and access to mortgage credit may not be causally related if those that are seriously delinquent on student debt tend not to apply for mortgage credit. Given the potential implications of student loan debt repayment for future housing demand, additional research on this topic should seek to establish the sources of student loan delinquency.

Rental Price Growth Continues to Exceed Overall Inflation

The Bureau of Labor Statistics reported that its measure of consumer prices declined in April. According to the Consumer Price Index – Urban Consumer (CPI), prices faced by consumers declined by 0.4% on a month-over-month seasonally adjusted basis. This is the second consecutive monthly decline for the index. In March, consumer prices fell by 0.2%. Consumer prices have experienced three episodes of month-over-month declines in the past 6 months and 5 instances of monthly declines over the past twelve months. Over the past year, consumer prices have risen by 1.1% on a not seasonally adjusted basis.

As Chart 1 illustrates, the decline in consumer prices largely reflects falling energy prices. In April, energy prices declined by 4.3% on a month-over-month seasonally adjusted basis after falling by 2.6% in March. Gasoline prices were largely responsible for the decline in energy prices, falling by 8.1% in April. Over the past twelve months energy prices have declined by 4.3%. Meanwhile, food prices, which also display higher than average volatility, rose by 0.2% in April after remaining flat in March. Core CPI, which excludes both food and energy prices, rose by 0.1% in April, mimicking its growth rate in March. Over the past twelve months, core prices have risen by 1.7% on a not seasonally adjusted basis.

NAHB constructs a real rental price index by deflating the price index for rent by the index for overall inflation. This measure indicates whether inflation in rents is faster or slower than general inflation, excluding more volatile food and energy prices, and provides some insight into the supply and demand conditions for rental housing. When rents are rising faster (slower) than general inflation the real rent index rises (declines). Alternatively, the real rental price index also conveys information about the importance of the rental prices faced by consumers relative to their other expenditures and sheds some light on the relative importance of household expenditure items. In this way, an increase (decrease) in the real rent index also indicates that rental prices are a growing (shrinking) share of the overall expenditures made by consumers.

Computationally, the real rental price index and the relative weight calculation are closely related. As Chart 2 illustrates, the real rental price index and the relative weight of rental prices within core CPI follow a very similar trend. The relative weight measure is first calculated using not seasonally adjusted data and overall CPI in order to ensure proper measurement. Then core CPI is substituted for overall CPI and finally the not seasonally adjusted data is converted to its seasonally adjusted counterpart. In April, rental price inflation, 0.2%, exceeded core inflation, 0.1%. As a result, real rental prices faced by consumers increased. This is the third consecutive month that real rental prices have increased. Similarly, seasonally adjusted rental prices as a share of consumers’ overall expenditures also rose for the third consecutive month.

Non-revolving Credit Continues Growing, but Revolving Credit Declines

The total amount of consumer credit outstanding expanded for the 19^th consecutive month, but growth in March occurred at a slower rate than in previous months. According to the Federal Reserve Board, consumer credit outstanding grew at a seasonally adjusted annual rate of 3.5% in March to $2.8 trillion. In February, consumer credit rose by 8.0% and by 5.5% in January. Over the first quarter of 2013, consumer credit rose by a seasonally adjusted annual rate of 5.7%. This rate of growth was slightly lower than then 6.5% growth rate measured in the fourth quarter of 2012, but higher than the 4.9% rate of growth observed in the third quarter of 2012.

The March increase in consumer credit outstanding reflected a 5.9% rise in non-revolving credit. Non-revolving consumer credit outstanding rose to seasonally adjusted $2.0 trillion. Non-revolving credit is largely composed of automobile loans and student loans, but also includes secured or unsecured loans for manufactured housing, boats trailers, and vacations. In February, non-revolving credit outstanding grew by 11.3% and in January it rose by 6.9%. Over the first quarter of 2013, non-revolving credit rose by a seasonally adjusted annual rate of 8.1%. The first quarter growth rate was slightly lower than then 9.3% growth observed in fourth quarter of 2012, but higher than 6.9% growth rate that occurred in the third quarter of 2012.

The expansion in non-revolving consumer credit that was recorded in March was partly offset by a decline in revolving credit. Revolving credit is largely composed of credit cards. In March, revolving credit declined by a seasonally adjusted annual rate of 2.4% to $0.8 trillion. This is the first monthly decline in revolving credit outstanding since December 2012. Following a month-over-month decline in December 2012, revolving credit grew by 2.3% in January and by 0.6% in February. Over the first quarter of 2013, revolving credit outstanding rose by a seasonally adjusted annual rate of 0.2%, below the 0.3% growth rate that recorded in the fourth quarter of 2012 and the 0.4% growth rate that took place in the third quarter of 2012.

Measures of Consumer Confidence Mixed

Measures of consumer confidence were mixed in April. According to Thomson Reuters and the University of Michigan, the Consumer Sentiment Index fell by 2.8% on a monthly seasonally adjusted basis to 76.4. The final reading of consumer sentiment was revised up from the preliminary reading of 72.3 that was released earlier in the month. Conversely, the Conference Board reported that its Consumer Confidence Index rose by 10.1% on a monthly seasonally adjusted basis in April to 68.1. Also, the original March reading, 59.7, was revised up to 61.9. Consumer confidence regained all of the ground that it lost in March. However, over the past twelve months, consumer confidence is lower by 0.8%.

Consumer sentiment and consumer confidence, measures of consumers view of the economy, have diverged in three of the past four months but, as Chart 1 above illustrates, these two measures of consumer confidence have tracked each other over a longer period of time. Despite displaying a similar long-run trend, the Conference Board’s Consumer Confidence Index displays greater volatility. Between 1996 and 2000 growth in the Conference Board’s measure exceeded growth in the University of Michigan and Thomson Reuters’ measure. A similar pattern occurred between 2006 and 2008. Between 2008 and 2009, the Conference Board’s measure of consumer confidence fell more than the Consumer Sentiment Index.

Both measures of consumer confidence reflect consumers’ assessment of their present situation and their expectations for the future. The Conference Board captures these two viewpoints with its Consumer Confidence Index – Present Situation and its Consumer Confidence Index – Expectations. The University of Michigan and Thomson Reuters separates the information in its headline index with its Consumer Sentiment Index – Current Conditions and its Consumer Sentiment Index – Expected Conditions. The source of the increased volatility displayed by the Conference Board’s Consumer Confidence Index is in its measure of consumers’ assessment of their present situation. Chart 2 illustrates that the Consumer Confidence Index – Present Situation tends to exceed the Consumer Sentiment Index – Current Conditions during upswings and it falls below the Consumer Sentiment Index – Current Conditions during downturns. Meanwhile, Chart 3 shows that Consumer Confidence Index – Expectations and the Consumer Sentiment Index – Expected Conditions track each other closely.

The dissimilarity in the trend of the two sub-indexes measuring consumers’ present situation likely reflects the different focus of the underlying question topics. Conversely, the topics addressed by the two indexes measuring consumers’ expectations capture similar information. The Consumer Confidence Index – Expectations asks survey respondents about their view of business conditions, employment and total family income over the next six months while the University of Michigan and Thomson Reuters’ Consumer Sentiment Index –Expected Conditions asks its survey respondents about their financial condition one year from now and business conditions both one year and five years from now. On the other hand, the Consumer Confidence Index – Present Situation asks consumers about their view of present-day business and employment conditions while the Consumer Sentiment Index – Current Conditions focuses its questions on consumers’ personal financial situation today relative to a year prior and on consumers’ attitudes towards purchasing durable goods.

Consumer Credit Expands

The total amount of consumer credit outstanding continues to climb. According to the most recent release by the Federal Reserve Board, the total amount of consumer credit outstanding expanded by seasonally adjusted annual rate of 7.8% in February to a seasonally adjusted level of $2.8 trillion. This is the eighteenth consecutive monthly increase in consumer credit outstanding. Over this period, the total amount of consumer credit outstanding has widened by 8.9% or $228.2 billion. Since reaching a trough in July 2010, consumer credit outstanding has grown by 17.2% or $410.1 billion. However, the increase over this period partly reflects a shift of consumer credit from pools of securitized assets to other categories that was largely due to financial institutions’ implementation of the FAS 166/167 accounting rules.

The increase in consumer credit continues to reflect an expansion in non-revolving credit. According to the release, non-revolving credit, which is mostly composed of auto and student loans, expanded by a seasonally annual rate of 10.9% to a seasonally adjusted level of $2.0 trillion. Meanwhile, revolving credit, which is largely composed of credit cards, rose by a seasonally adjusted annual rate of 0.8% to a seasonally adjusted level of $848.0 billion. The larger role played by non-revolving credit relative to revolving credit in the growth of total consumer credit outstanding over the month of February is consistent with the trend since July 2010. As Chart 1 illustrates, growth in total consumer credit outstanding since July 2010 has largely reflected a widening of non-revolving credit as opposed to an increase in revolving credit. Over this period, non-revolving credit has grown by 28.8% or $436.1 billion, while revolving credit has decreased by 3.0% or $26.0 billion.

Household Balance Sheets: Fiscal Cliff Impact

During the fourth quarter of 2012, household balance sheets improved with increases in home values and reductions in mortgage debt, thereby boosting household net worth. These are favorable improvements that will help housing demand in 2013.

However, since the end of the Great Recession such developments have typically been associated with a decline in the personal savings rate. Due to the Fiscal Cliff, the data do not show this impact for the last quarter of 2012.

The graph above plots the current value of net worth to disposable personal income (NW / DPI) and the corresponding 25-year historical average (1982-2007). The dashed blue line charts the personal savings rate. Household net worth data are from the Federal Reserve’s Flow of Funds and the savings rate and disposable income data come from the Bureau of Economic Analysis National Income Product Accounts. 

While there has been a general trend of an increasing NW/DPI ratio since early 2009, there have been ups and downs in this process due to stock market and other asset value fluctuations. As of the fourth quarter of 2012, the NW/DPI measure stood at a value of 5.43, above the historical level of 5.24 and significantly higher than the cyclical low of 4.8 set during the beginning of 2009.

Unlike most periods after the Great Recession, the ongoing improvement in balance sheets did not result in a decline of the savings rate at the end of 2012. In anticipation of higher tax rates, particularly on dividend income, the amount of income paid out and earned by American households jumped. As this accelerated income was due to a one-time cause, spending did not increase at the same rate. Thus, with a rise in income and no corresponding increase in spending, the personal savings rate increased to 4.6% in the fourth quarter, marking the highest rate of savings in a year and a half.

The data for the first quarter of 2013 will also show a Fiscal Cliff impact, as income will temporarily fall due to the accelerated payments made at the end of 2012. That result, plus the end of the payroll tax cut, will boost the NW/DPI measure and muddy our tracking of balance sheets until the data for the second quarter of 2013 are available.

Nonetheless, Flow of Funds data from the fourth quarter of 2012 show that total home mortgage debt continues to decline. Since the first quarter of 2008, home mortgage debt has declined 11% or $1.2 trillion. The value of real estate owned by households has fallen by a net 12% over the same period, but has now risen for the last four consecutive quarters for an increase of $1.4 trillion.

As a result, total household equity as a share of real estate value has increased to almost 47% as of the end of 2012.

Consumer Credit Expands, but HELOCs Continue Their Decline

Household debt outstanding rose for the first time in two years. According to data released by the Federal Reserve Bank of New York, household debt grew by $31.0 billion in the fourth quarter of 2012. The quarter-on-quarter not seasonally adjusted growth in household debt reflected an expansion in outstanding mortgages, auto loans, credit cards, and student loans. In the fourth quarter of 2012, these household debt products rose by a combined $41.0 billion. However, these gains were partially offset by a $10.0 billion decline in home equity lines of credit. Since increasing by $41.2 billion in the first quarter of 2011, total household debt outstanding experienced six consecutive quarters of declines, falling by $444.4 billion over that time span.

Outstanding balances on home equity lines of credit (HELOCs), which, along with home equity loans, were an important source of bond market growth, expanded significantly between 2003 and 2009. As Chart 1 illustrates, the outstanding amount of HELOCs totaled $242.0 billion in the first quarter of 2003, this amount was 37.8% of the outstanding amount of auto loans and 35.2% of the outstanding credit card debt. However, by the second quarter of 2009, the outstanding amount of HELOCs nearly tripled, growing by 194.6% to $713.0 billion. Over this same period, auto loans outstanding, $743.0 billion in the second quarter of 2009, rose by 15.9% while credit card balances, $824.0 billion, grew by 22.5%. Since the second quarter of 2009, the outstanding amount of HELOCs has contracted by 21.0% while credit card debt outstanding has fallen by 17.6%, although it rose in the latest quarter. The amount of outstanding auto loans, which returned to sustained growth in the second quarter of 2011, is now 5.4% above its second quarter 2009 level.

The number of HELOC accounts also experienced a period of rapid growth. However, despite the rise in the number of HELOC accounts between first quarter of 2003 and the first quarter of 2008, the number of these accounts remained well below the number of auto loans and the number of credit card accounts. In the first quarter of 2008, there were 24.2 million home equity line of credit accounts, 80.8% greater than the number of accounts in the first quarter of 2003. Over this same period the number of auto loan accounts, 87.2 million in the first quarter of 2008, rose by 18.6% and the number of credit card accounts, 474.6 million in the first quarter of 2008, grew by 1.0%. However, at its peak in the first quarter of 2008, the number of HELOC accounts was only 27.8% of the number of auto loan accounts and 5.1% of credit card accounts.

The growing amount of outstanding HELOCs was concentrated in a relatively smaller number of accounts. Despite the 80.8% increase in the number of HELOC accounts between the first quarter of 2003 and the first quarter of 2008, the outstanding amount of HELOCs rose by 174.0% over this same period. As a result, growth in the outstanding amount of HELOCs raised the size of the average account balance. As Chart 3 illustrates, growth in the balance on the average HELOC account, which generally tends to be larger than balances on other consumer financial products, eclipsed account balance growth of both credit cards and auto loans. Between the first quarter of 2003 and the first quarter of 2008, the average account balance on a HELOC account grew by 51.6% to $27,351. Meanwhile, the average auto loan and credit card balance, which grew by 6.3% and 20.4% over this same period, were $9,266 and $1,764 in the first quarter of 2008. In the fourth quarter of 2010, the average balance on a HELOC account peaked at $31,619, 3.6 times the average auto loan account balance and 4.6 times the average credit card account balance. However, since the fourth quarter of 2010, the average balance on a HELOC has declined somewhat, falling by 4.6% over the two year period.

Data from the Federal Reserve Bank of New York that is displayed in the graph shown on page 9 of their report depicts the serious delinquency rate for HELOCs as being the lowest of household debt products. However, the current rate partly reflects that it started from a very low level. Instead, comparing the serious delinquency rate in each quarter relative to its level in the first quarter of 2003 conveys the magnitude of serious delinquencies in HELOCs. As chart 4 illustrates although most household debt products have begun the healing process, HELOCs still languish.

In the fourth quarter of 2007, the percent of outstanding HELOCs that were seriously delinquent was 3.8 times its level in the first quarter of 2003. By the second quarter of 2009, the percent of outstanding HELOCs that were seriously delinquent rose to 11.3 times its first quarter 2003 level. Between the first quarter of 2010 and the third quarter of 2012, the percent of outstanding HELOCs that was seriously delinquent rose from 11.6 times its first quarter 2003 level to 14.1 times its first quarter 2003 level. The percent of outstanding HELOCs that are seriously delinquent fell to 9.9 in the fourth quarter of 2012. However, according to the Federal Reserve Bank of New York, the decline in the delinquency rate that occurred in the fourth quarter of 2012 “can be attributed in large part to unusually high charge-offs of delinquent home equity lines of credit”. Meanwhile, in the first quarter of 2010, the percent of outstanding mortgage debt that was seriously delinquent peaked at 7.3 times its first quarter 2003 level, but has since fallen to 4.6 times its first quarter 2003 level. Over this same period, the percent of auto loans that were seriously delinquent declined from 2.2 times its first quarter 2003 level to 1.7 times its first quarter 2003 level and the percent of outstanding credit card debt that was seriously delinquent fell from 1.6 times its first quarter 2003 level to 1.2 times its first quarter 2003 level.

Refinancings Contribute to A Lower Average Mortgage Debt Burden

Mortgage applications eased for the second consecutive week. According to the Mortgage Bankers Association, mortgage application activity, which includes both refinancing and home purchase demand, was 1.7% lower on a seasonally adjusted basis in the week ending Feb. 15. In the week ending February 8, the market index fell by 6.4%. The most recent decline in weekly seasonally adjusted mortgage applications reflected both a 1.7% drop in refinancing applications and a 1.7% decrease in mortgage applications for purchase.

Applications for a refinancing account for the majority of growth in mortgage applications. The impact of refinancing applications partly reflects its share of total mortgage applications. Refinancing applications currently account for 72.5% of all applications. In addition, applications for mortgage refinancing have grown faster than applications for mortgage purchase. As Chart 1 illustrates, mortgage applications for refinancing have risen steadily since early 2011. Meanwhile, mortgage applications for purchase have remained relatively flat through 2011 and most of 2012, rising noticeably only in the last few months. Between March 2011 and October 2012, total mortgage applications grew by 78.2%. During this same period, mortgage applications for refinancing more than doubled, rising by 115.4% while mortgage applications for purchase grew by only 1.1%. Since October 2012, total mortgage applications have declined somewhat as the drop in applications for refinancing more than offset the increase in the mortgage applications for purchase. Over the past four months, total mortgage applications have decreased by 9.6% as mortgage applications for purchase rose by 6.7%, but mortgage applications for refinancing declined by 12.6%.

The Federal Home Loan Mortgage Corporation compiles statistics on loans it purchases that refinance loans already held in its portfolio. These statistics are based on a sample of properties for which it has funded two or more successive loans. In 2012, 53.8% of these refinancings resulted in a new loan amount that was unchanged from the previous loan, an increase of 7.2 percentage points from 2011 and 8.6 percentage points from its 2003 peak. Meanwhile, the share of refinancings that led to either a higher loan amount or a lower loan amount both declined between 2011 and 2012, by 2.5 and 4.7 percentage points respectively. Conversely, between 2003 and 2006, the vast majority of these refinancings resulted in a higher loan balance for the homeowner. In 2003, 38.5% of refinancings resulted in a new loan balance that was at least 5.0% greater than the pre-refinanced loan balance, while 16.3% of refinancings led to a lower loan balance and 45.2% of refinancings had no effect on the loan balance. By 2006, the share of refinancings resulting in at least a 5.0% higher loan balance grew to 86.3% while the share of refinancing that resulted in a lower loan balance stood at 5.3% and the share refinancing that left the loan balance unchanged was 8.4%. However, between 2006 and 2011, growth in both the share of refinancings resulting in a lower loan balance and in the proportion of refinancings that leave the loan amount virtually unchanged have offset the decline in the percentage of refinancings that led to a higher loan amount.

In addition to its data on loan amounts, the Federal Home Loan Mortgage Corporation also maintains statistics on the ratio of the new to old mortgage rate and reports the median ratio. In this ratio, the numerator is the mortgage interest rate on the newly refinanced loan amount and the denominator is the rate on the old mortgage. The data exclude adjustable rate mortgages. A ratio equal to 100.0 indicates that the average refinancing did not change the mortgage interest rate while a ratio below 100.0 means that a refinancing resulted in a lower mortgage rate on average. A ratio greater than 100.0 indicates that refinancing led to a higher average interest rate.

As mortgage rates rose between January 2003 and December 2005, the difference between the mortgage interest rate on the old mortgage and the interest rate secured after refinancing began to shrink. By January 2006, refinancing resulted in a higher interest rate, on average. As Chart 3 illustrates, between January 2003 and December 2005, the 30-year fixed rate mortgage rose by 0.4 percentage points to 6.3%. Meanwhile, the ratio of the new rate to the old rate rose by 18.5 points to 99.7. In January 2006, the ratio reached 100.7. The ratio of the new rate to the old rate remained above 100.0 until November 2007 and it eclipsed 100.0 again during the third quarter of 2008.

More recently, refinancing has allowed some existing homeowners to lock-in a lower interest rate and lower their monthly mortgage payment. As Chart 3 shows, the recent decline in the 30-year fixed mortgage rate coincides with a decrease in the median ratio of the new mortgage interest rate and the old mortgage rate. Between April 2011 and December 2012, the 30-year fixed mortgage rate fell by 1.5 percentage points to 3.4%. At the same time the ratio of the new rate to the old rate declined by 15.4 points to 66.5. In other words, the new mortgage rate obtained from a refinancing was, on average, 66.5%, of the old mortgage rate. According to research by NAHB, a new interest rate that was both 3.4% and was also 66.5% of the old mortgage interest rate, 5.0%, would result in an estimated mortgage payment savings of roughly $182 per month or $2,184 per year on a $225,000 mortgage loan. Over the span of 30 years, an existing homeowner would save an estimated $65,520 under these conditions.