There is widespread agreement among economists and politicians today that the economy needs immediate stimulus to minimize the duration and severity of the recession that has probably already begun. In recent months, observers have frequently cited the housing sector as an important remaining pillar of strength in the consumer economy, and one that can provide additional buoyancy to household spending through the mechanism of a large-scale refinancing wave. We have already gone through one such event this year, and now seem poised to begin a second. But the widely recognized need for a timely dose of economic medicine in the form of fiscal and monetary stimulus requires that we re-examine the potential beneficial effects of another round of refinancing. In this context, we find that of the two economic benefits - rate/term refinancing to achieve cashflow savings, and cash-out refinancing to monetize home equity gains - only the latter has the potential to help the economy right away.
The long-term benefit to the economy of cashflow savings in a refinancing wave can be easily illustrated. There was a dramatic decline in the percent of income used to cover mortgage payments in the early 1990s (as rates declined into the 1992/1993 refinancing wave). Since that initial drop, the percentage has generally stayed within a range of 17% to 20% through the present time, depending on the direction and level of mortgage rates. This new lower level of mortgage debt service has allowed consumer debt to balloon from 14.2% in 1993 to 18.1% today, thereby providing one of the sources of fuel for the economic expansion of the late 1990s. The cashflow advantage of a rate/term refinancing does not necessarily provide an immediate benefit; indeed, most borrowers today still pay third-party costs (e.g., attorney fees and title insurance). As a result, most borrowers will need at least a year to reach the breakeven point, after which they will see net savings on their monthly payment.
But in the post-September 11 world, where a rising tide of layoffs and heightened security concerns have combined to deal a body-blow to consumer confidence, a further cyclical reduction in mortgage debt service will probably not be enough to renew economic growth. Enter the cash-out refinancing, a mechanism that can provide immediate cash to the consumer. Just how significant a factor might this be?
To answer this question, we first need to form an opinion on several contributing factors, including the following: the available store of home equity today, the percentage of refinancers that are likely to take cash out today, and the likely uses they will find for those funds. We can then directly compare this potential source of economic stimulus to other sources.
*Available home equity. For most levels of loan seasoning, there is a greater store of home equity buildup today than at any time in the last 10 years (see Table 1). Specifically, for loans seasoned between 1 and 10 years, there was more equity available to extract as of mid-year 2001 than during the last two major refinancing waves in 1993 and 1998. For loans seasoned less than 5 years (the bulk of the outstanding mortgage market), borrowers today have 1.5 and 3 times more equity build-up than they did in 1998 and 1993, respectively. Conclusion: there is more available home equity than at almost any other time in history.
*How many refinancers will take cash out? In general, the percentage of refinancers who take cash out is quite high when rates are high (since rate/term refinancings are so rare). As rates fall, the cash-out percentage falls as well. Thus, according to the latest numbers from Freddie Mac1, an average of 48.5% of refinance transactions were cash-outs in 1998 (when rates were very low), versus an average of 80.5% in 2000 (when rates were very high). Over the first two quarters of 2001, the average was 53.5%. For our analysis, we stay close to the 1998 precedent and assume that 50% of refinancing transactions this fall will be cash-outs, but we regard that as a ceiling on the likely level. Although there are no direct statistics on how much borrowers cash out, we estimate that borrowers increase their loan balance by an average of 8%.
*Spend or save? Earlier this year, Fannie Mae reported2 the results of a survey conducted by the University of Michigan in 1999, which found that 51% of funds received in 1998/1999 cash-out refinancings were spent (33% on home improvement and 18% on consumer expenses). Of the remainder, 28% was used for debt reduction and 21% for saving/investment. Given the state of the stock market in 2001, we believe that a good portion of that 21% could be spent today, yielding as much as 70% of the total for spending.
Starting with these numbers, and using our short-term prepayment estimates for the $5.7 trillion in U.S. mortgages outstanding (assuming rates stay at current levels), we estimate that $21.9 billion could be extracted from home equity and made available for spending over the next 6 months. Will this amount make a difference? The IRS sent out $40 billion in rebate checks over a roughly three-month period in the summer and early fall. Bear Stearns Economics notes that between June and August of this year, the U.S. savings rate rose from 1.0% to 4.1%, the highest level for this metric since the beginning of 1999. This suggests that much of the tax rebate has been saved rather than spent, perhaps because it was a windfall. The assumptions we have used for this cash-out analysis are based on 1998 behavior. To the extent that the current environment of employment uncertainty, security concerns, and falling consumer confidence may decrease the urge to increase home leverage and spend, 2001 behavior may prove to be more conservative. Thus, the actual amount that homeowners elect to extract and spend over the next 6 months may well be lower than our estimates. But even if cash-out refinancers repeat the 1998 experience, a $21.9 billion injection of spending over 6 months seems unlikely to turn the economy around.
A new MBS
landscape in 2002
We expect to see a new MBS market emerge in the aftermath of record-setting prepayments and supply in the fourth quarter of 2001 (see the recent Bear Stearns publications: MBS Prepayment Thresholds and Short-Term Prepayment Estimates for a full analysis). We believe the cash-out phenomenon, a product of unprecedented house price appreciation during the last 5 years, will play a less important role as economic stimulus and driver of MBS prepayments in 2002. Moreover, as in the 1993 and 1998 refinancing waves much of the mortgage market will be recast into very low coupons (primarily 6.0% and 6.5% MBS coupons) creating a powerful "lock-in" effect that will make a cash-out refinance less attractive from a rate and payment perspective. More importantly, we believe that it is highly unlikely that the housing market will repeat its recent price performance, diminishing the equity wealth euphoria that has characterized the housing market. Indeed, the housing market, already showing signs of weakness before September 11, will most likely experience a period of retrenchment. Therefore, mortgages originated in 2001 are expected to be on a very different and slower prepayment trajectory versus other recent vintages.
1 Freddie Mac Quarterly Survey of Refinancing, August 21, 2001.
2 Berson's Housing & Economic Report, February 14, 2001.
Bear Stearns is not responsible for any recommendation, solicitation, offer or agreement or any information about any transaction, customer account or account activity contained in this communication.