Improving Access to Refinancing Opportunities for Underwater Mortgages
July 15, 2013
Federal Reserve Bank of New York, Liberty Street Economics
Joshua Abel and Joseph Tracy
Since the onset of the housing crisis, a focus of policymakers has been
to help underwater homeowners lower their monthly mortgage payments by
refinancing, principally through the Home Affordable Refinance Program
(HARP). This enables households to commit more money to consumption,
debt reduction, and saving. Lower monthly payments also decrease the
risk of mortgage defaults, allowing homeowners to stay in their homes
and reducing expected losses for mortgage guarantors Fannie Mae and
Freddie Mac, which remain under conservatorship of the Federal Housing
Finance Agency. Stanching the flow of defaults also helps to firm up the
housing market and, therefore, the economy as a whole. In this post, we
examine some simple adjustments to HARP that would help to continue the
program’s recent success and provide additional support to the housing
market recovery—in undertaking that has added significance with the
recent
increase in mortgage rates, which could hamper refinancing activity moving forward.
Recent home price appreciation
has benefited homeowners in most parts of the country. However,
millions of homeowners still find themselves with little or no equity in
their homes, the result of the earlier steep decline in home prices.
Without sufficient equity, it can be very difficult to refinance, which
means that these homeowners cannot take advantage of the current low
interest rate environment.
HARP is designed to allow borrowers who are current on their payments
and whose existing mortgages are guaranteed by Fannie Mae or Freddie Mac
to refinance even if their loan-to-value (LTV) ratios are above the
80 percent level that is typically required for prime mortgages. Since
HARP’s inception in April 2009, about 2.5 million homeowners have used
the program. Much of this success has come since a 2012 redesign (the
so-called HARP 2.0). Among other things, this redesign removed the
125 percent LTV limit on HARP refinances, opening the program to
borrowers more deeply underwater. The chart below showing a sharp uptick
in participation in 2012 makes clear how important that policy change
was.
It could be time to consider further program changes. Currently, HARP
can only be utilized by borrowers whose mortgages were obtained by
Fannie Mae or Freddie Mac by June 1, 2009. Additionally, borrowers are
only allowed to use HARP once, meaning that borrowers who have already
done so are now ineligible. We find that altering these rules could
heighten refinancing activity substantially. To evaluate the potential
benefit from relaxing these two program restrictions, we take a random
sample of loans and select those that are
eligible
(backed by Fannie Mae or Freddie Mac, current on payments, and other
criteria). We then say that a loan is in-the-money to use HARP if
refinancing would decrease the monthly payment enough so that the
borrower can recoup the costs of refinancing, such as the appraisal cost
and program fees, within three years. We repeat this exercise using
different eligibility rules, such as moving the cutoff date from June 1,
2009, and allowing “reHARPing” (refinancing a second time under HARP).
The table below summarizes our results as of February. We estimate that
about 1.5 million borrowers are eligible and in-the-money for HARP under
current guidelines. Shifting the cutoff date—to 2010, 2011, 2012, and
then removing it entirely— adds substantially to this number, though at a
decreasing rate. Our analysis shows that eliminating the cutoff date
entirely would add 530,000 in-the-money borrowers, an increase of over
30 percent. If reHARPing were also allowed, this increase would climb to
over 55 percent.
Note that on the whole, the reduction in monthly payments for these
additional borrowers would be lower than for those that are currently
eligible, due to the fact that mortgage rates have declined
substantially since 2009, meaning their contract rates are already
somewhat lower. However, an extra $200 per month could make a
significant difference for many individual homeowners and for the
macroeconomy, when aggregated. Although heightened refinancing activity
will lower returns for investors in agency mortgage-backed securities,
the macro effects of additional money in homeowners’ pockets more than
offset those losses, meaning that refinancing is not simply a zero-sum
transfer from investors to borrowers, as discussed in an
earlier post.
Furthermore, there is reason to believe that these program changes could
provide a bigger boost than our numbers imply. Consider that the table
above provides estimates of how many borrowers have incentive to
refinance, not how many will actually do so. Predicting the “take-up”
rate—the fraction of in-the-money borrowers who will actually complete a
HARP refinance—is a difficult exercise. However, it seems reasonable to
believe that the take-up rate among newly eligible borrowers (if, say,
the cutoff date was moved) will be a lot higher than for those who are
already eligible, but have so far decided not to participate despite the
financial incentive to do so. There are various reasons why someone may
not participate: lack of information about the program, lack of
financial sophistication, or plans to move from the house in the
near-term, among others. All of these are unobservable in the data, but
eligible borrowers, by not responding to the incentive up to this point,
have revealed that some or all of these effects may be at work. Others
have not used HARP simply because they have not been allowed, and it is
likely that many will rush to do so when given the chance. In fact,
there’s evidence for this in our first chart: borrowers with LTV ratios
greater than 125 percent became newly eligible in 2012, with a
tremendous influx of these borrowers into the program soon after.
The recent improvement in the housing market is welcome news, but it has
not erased the negative equity problem brought on by the housing bust.
One of the consequences is that many borrowers are still trapped paying
interest rates far above the current market rate. HARP is a sensible
response to this problem for borrowers with agency mortgages, and
changes to the HARP program in 2012 increased the effectiveness of the
program. Some additional changes to that program could build on the
recent success of HARP and provide ongoing support for the housing
market recovery.
Addendum
(July 18, 2013)
The above analysis assumed that a borrower could refinance into a
thirty-year fixed-rate mortgage with a rate of 4.0 percent. After
receiving reader comments, we have updated the table assuming that a
borrower now could refinance into a mortgage rate of 4.5 percent. This
50 basis point increase in the mortgage rate reduces the impacts
associated with each proposed program change. For example, removing the
origination date deadline would now add an estimated 264,000 additional
in-the-money borrowers, a 50 percent reduction from the earlier 530,000.
For those borrowers who remain in-the-money, the higher assumed
mortgage rate also reduces the extent to which a HARP refinance will
lower their monthly mortgage payment. Again, looking at the case of
removing the origination date deadline, the reduction in the average
monthly payment is an estimated $161, a 12 percent decline from the
earlier $182.
Disclaimer
The views expressed in this post are
those of the authors and do not necessarily reflect the position of the
Federal Reserve Bank of New York or the Federal Reserve System. Any
errors or omissions are the responsibility of the authors.
Joshua Abel is a former research associate in the Research and Statistics Group of the Federal Reserve Bank of New York.
Joseph Tracy is an executive vice president and senior advisor to the Bank president at the Federal Reserve Bank of New York.
Authors: Zoltan Pozsar, Tobias Adrian, Adam Ashcraft, and Hayley Boesky
E-mail alert
RSS feed
Twitter
New York Federal Reserve
Liberty Street Economics
Common Stock Repurchases during the Financial Crisis
Summary: Large bank holding companies (BHCs) continued to pay dividends to their shareholders well after the onset of the recent financial crisis. Academics, industry analysts, and policymakers have noted that these payments reduced capital at these firms at a time when there was considerable uncertainty about the full extent of losses facing individual banks and the banking industry. But dividends are not the only means to return capital to shareholders; stock repurchases serve much the same function. In this post, I examine common stock repurchases by large BHCs during the financial crisis and show that they behaved very differently from common stock dividends during the same period. While dividends remained relatively constant through late 2008, common stock repurchases dropped quickly after the beginning of the financial crisis, consistent with their historically tighter sensitivity to current performance and financial conditions.
BHC Dividends during the Financial Crisis
Many analysts and researchers have noted that large BHCs did not reduce common stock dividends until the financial crisis was well under way. For instance, in a study examining capital at large U.S. and European banks, securities firms, and U.S. government-sponsored enterprises, Acharya, Gujral, Kulkarni, and Shin show that dividends at these firms did not decrease significantly until early 2009. In an October 2008 New York Times op-ed piece, Scharfstein and Stein argued that dividend payments by the largest U.S. BHCs would “redirect more than $25 billion of the $125 billion [in TARP capital] to shareholders in the next year alone.” Both sets of authors argued that continued high dividend payments undercut the capitalization of the U.S. banking industry during a time of stress. In 2011, the Federal Reserve implemented the Comprehensive Capital Analysis and Review (CCAR) to review planned dividend payments and other capital distributions by large BHCs to ensure that they would retain sufficient capital to withstand stressed economic and financial market conditions even after making such distributions to shareholders (see my earlier post for a description of how the CCAR addresses this objective).
Regulatory report data confirm the findings about dividends from these earlier studies. The first chart (below) shows the quarterly history of dividends declared by large U.S. BHCs—those with at least $10 billion in assets—from 2005 to the end of 2009. This period starts two years before the onset of the crisis in mid-2007 and extends until the beginning stages of the recovery at the end of 2009. (For consistency over the period, the sample excludes the large nonbank financial firms that became BHCs in early 2009.) As the chart shows, large BHCs made dividend payments in the range of $9 billion to $12 billion per quarter from 2005 to 2007, and while dividends declined over the course of 2008, they did not fall significantly below this level until early 2009, more than a year into the financial crisis.
Dividends Are Not the Whole Story
Focusing just on dividends misses an important part of the story, however. Stock repurchases—when a company buys its own common stock in public or private markets or by tender offer—are another important way that a firm can return capital to shareholders. Like dividend payments, stock repurchases disperse cash from the company to shareholders. Moreover, repurchases reduce the amount of common stock outstanding one-for-one, just as dividend payments do. Dividend payments reduce retained earnings and thus reduce (potential) common equity, while stock repurchases are a direct reduction in the outstanding amount of common equity.
The second chart (below) updates the first one to include common stock repurchases for the sample of large BHCs. Both dividend and repurchase information come from the quarterly FR Y-9C regulatory reports, which contain balance sheet and income statement information for BHCs. While the Y-9C reports collect dividends declared as a distinct line item, repurchases are not reported directly and must be inferred from other information contained in the report. For the second chart, I’ve used BHC purchases of treasury stock—common stock issued but not held by the public—as a proxy for common stock repurchases. This is a noisy measure of repurchases, as it may omit some repurchase activity that is reported as part of other Y-9C line items. However, it is a reasonable proxy that should capture the broad movements in repurchase activity.
The second chart illustrates several notable aspects of repurchase activity by large BHCs. First, in the period leading up to the financial crisis, stock repurchases were significant, averaging $7.5 billion per quarter and peaking at $12 billion in the first quarter of 2007. Stock repurchases were about two-thirds the size of dividend payments over this period, meaning that total distributions to shareholders were substantially higher than indicated by considering dividends alone. This is not something new in the banking industry. Some of my earlier research, which examined BHC stock repurchases during the 1990s, also found that stock repurchases rivaled dividends as a way for BHCs to return capital to shareholders. Second, share repurchases dropped sharply during the early phases of the financial crisis, several quarters before dividends were significantly reduced. Stock repurchases by large BHCs fell from their peak of $12 billion in the first quarter of 2007 to about $2 billion during the first quarter of 2008, and were at negligible levels after the middle of that year. As noted above, dividends did not decrease significantly until early 2009.
Because of the decline in repurchases, large BHCs’ overall capital distributions to shareholders declined steadily over the course of 2007 and 2008, a somewhat different picture than the one that emerges by looking at dividends alone. By the beginning of 2008, overall capital distributions had fallen below the levels that prevailed during the pre-crisis period. Even so, as the second chart illustrates, it was not until early 2009, when dividends also declined sharply, that overall distributions fell to the very low levels that ultimately prevailed for the remainder of the financial crisis and the period that followed.
Why Are Repurchases Different?
Why did stock repurchases fall earlier and more sharply than dividend payments? Part of the explanation is likely owing to the differing nature of dividends and share repurchases. Dividends are publicly visible actions requiring regular authorization by a firm’s board of directors. In the banking industry, large BHCs typically declare dividends quarterly and announced them publicly in press releases. In contrast, for all firms, stock repurchases are much less transparent. Firms have the ability to select the timing and amount of repurchases flexibly over time, subject to the details of their repurchase programs. Research on nonfinancial firms has documented that stock repurchases generally vary more over time than dividends, with repurchases used more frequently by firms with volatile earnings, especially following periods of higher-than-expected profitability. My earlier research on repurchases in the banking industry in the 1990s also found that repurchases seemed to rise following periods of higher-than-average financial performance. In contrast, dividends tend to be interpreted as signals of long-term profitability, decreasing only when profits seem likely to fall to a lower level for a sustained period of time.
Much further research is needed to come to a full understanding of the decisions BHCs made regarding dividend and repurchase activity during the financial crisis. The charts in this post and the results of previous research provide some insight, but there is significant room to explore issues such as concerns about signaling financial weakness during a time of uncertainty, the role of herding behavior in the timing and extent of dividend reductions, and the relative impact of market-wide versus firm-specific factors in decisions to cut back on stock repurchases. And, looking ahead, supervisory oversight via the CCAR and the adoption of new regulatory capital standards that explicitly require reductions in capital distributions as regulatory capital ratios fall below certain trigger points will significantly affect BHCs’ decisions about capital distributions. In sum, it is important to consider both dividends and stock repurchases in order to understand the impact of capital distributions on the banking industry, since large BHCs actively use both to manage their capital and distribute cash to shareholders.
Disclaimer
The views expressed in this post are those of the author and do not necessarily reflect the position of the Federal Reserve Bank of New York or the Federal Reserve System. Any errors or omissions are the responsibility of the author.