Conduits powering CMBS market.
This article explores the forces behind the growth of CMBS conduits, how rating agencies analyze conduits and our evaluation of the credit risks of CMBS backed by conduit loans. We also examine the relative value of conduit CMBS against other fixed-income securities.
Our review of the market produced the following main conclusions:
* Conduit originations will make up an increasing share of the CMBS market driven by both investor demand and general real estate conditions.
* Competitive pressures will result in a consolidation in the number of CMBS conduits.
* The average credit quality of conduit collateral resembles the average quality of collateral in the overall commercial lending market.
* There has been a trend to larger loan sizes in conduit deals. Tight spreads have allowed conduits to compete with life insurance companies for A-quality collateral.
* Competition for market share has resulted in declines in credit support, but the default risk of investment-grade conduit securities has remained steady and is minimal.
* CMBS conduits have increased the choices available to real estate and fixed-income investors. Investors seeking real estate risk buy noninvestment-grade classes, while investment-grade bonds appeal to traditional fixed-income investors.
* Spreads on CMBS conduits will continue to compress toward corporate-bond spreads and should be more correlated with movements in corporate spreads than in the past.
* Most classes of conduit commercial mortgage-backed securities, including IOs (interest-only strips), currently offer good relative value versus similarly rated corporate bonds on a nominal and default-adjusted basis.
The rise of CMBS conduits
A conduit is an intermediary or facilitator between the providers of capital (investors) and the users of capital (borrowers). A "commercial mortgage loan conduit" is an entity that originates multifamily and commercial loans intended for securitization rather than for holding in portfolio.
According to the rating agency Duff & Phelps, "Combining loans of many different product types . . . creates an optimal pool avoiding borrower or loan concentration risks. . . . Loan originations for conduits are tailored to highly specific underwriting guidelines and can be secured by almost any type of income-producing commercial real estate."
In the first half of 1997, conduit issuance totaled $9.4 billion, or almost 60 percent of all CMBS issuance [ILLUSTRATION FOR FIGURE 2 OMITTED]. Commercial conduit securitizations have risen sharply in the last three years, totaling $10.2 billion in 1996, about one-third of all CMBS issuance. These totals were both up significantly from 1995, when conduits accounted for $4-4 billion in issuance and 23 percent of the total. Prior to that, conduit issuance was negligible.
We believe that conduits will continue to dominate the CMBS market, with a potential market share of 75 percent or more. Conduits have filled the void left by the exit of traditional lenders such as savings and loans and insurance companies in the late 1980s and early 1990s. The residential lending market had a similar experience in the 1980s, as lending by large national mortgage conduits replaced thrift institution activity.
More recently, commercial mortgage conduits have displaced some of the more traditional lenders in commercial real estate, such as insurance companies. Direct insurance company lending on commercial real estate fell by 11 per cent in 1996 and is currently 50 percent below the peak level reached in 1989 [ILLUSTRATION FOR FIGURE 3 OMITTED]. Although some insurance companies are reentering the direct-lending business, many lend only on the high end of the real estate markets and tend to concentrate on large properties.
Conduits, on the other hand, can focus on smaller loans, niche markets or out-of-favor property types and areas. "Large-loan" conduits, which focus on loans larger than $30 million, can lend on properties that present too concentrated a risk for a single institution.
Conduits can and will lend on a broad array of property types. Because of this flexibility, conduits will be able to adapt quickly to changing market conditions. This makes it difficult to precisely define "conduit lending" or the collateral in a "conduit CMBS" at any point in time. The conduit business is perpetually in a state of flux - changing programs, underwriting standards and loan terms to meet the needs of current borrowers.
Conduits are so flexible because they are not constrained by the historical portfolio preferences of traditional lenders such as life insurance companies. Should real estate markets turn down, insurance companies may again sharply curtail lending, but conduits can continue to lend as long as investors buy CMBS.
Securitization allows conduits to parcel out commercial real estate exposure to investors with varying appetites for credit or prepayment risk. For example, insurance companies may shy away from default risk because of regulator or rating agency scrutiny, but a money manager may be willing to take on the risk of lower-rated tranches.
Despite the opportunistic nature of conduit lending, little evidence exists that the credit quality of conduit loans is worse than for traditional insurance company lending. Delinquencies on conduit CMBS are less than 2 percent as of May 1997 and are comparable to the national delinquency rate on member commercial loan portfolios reported by the American Council of Life Insurance (ACLI) [ILLUSTRATION FOR FIGURE 4 OMITTED].
Outlook for growth
The key to conduit growth is ultimately in investors' hands. As long as investor demand for CMBS remains high, conduits will supply the product. If investor demand for CMBS drops, spreads on CMBS will widen and loan rates will rise, choking off new originations.
Currently, the outlook for investor demand for CMBS is good. New, nontraditional real estate investors are continually entering the market, causing CMBS spreads to substantially narrow to Treasuries over the past two years [ILLUSTRATION FOR FIGURE 5 OMITTED]. Investors have realized that CMBS are more liquid than direct real estate investments and carry lower capital requirements for many institutions.
These trends cannot continue forever. At some point, the pendulum may swing away from CMBS and spreads will widen. The forces that spurred the rise of CMBS conduits, however, will remain in place. In our view, the adaptability of conduit lending is a powerful advantage that will make conduits a major factor in commercial real estate lending over the next decade.
The next stage: Conduit consolidation
In its review of mortgage conduits early in 1997, Commercial Mortgage Alert listed 28 active commercial mortgage conduits, ranging from Nomura, with $3.4 billion in cumulative issuance, to Llama Co. and Goldman Sachs, with less than $100 million in cumulative issuance. We believe that there will be consolidation in the conduit lending business over the coming years and that a small number of firms will dominate conduit lending (see Figure 6).
The commercial mortgage lending business is following the pattern set by residential lenders in the early 1990s. Following the failures in the thrift industry, national conduit lenders, e.g., mortgage bankers, increased market share and now dominate residential mortgage lending. Significant consolidation has occurred recently, with a few large national conduits, such as Norwest, RFC and Countrywide, accounting for more than half of all residential mortgage originations. The main difference is that in the case of the residential markets, the conduits are operating companies rather than Wall Street firms.
We think that consolidation of commercial conduits is inevitable, given the economies of scale of larger lending institutions and the desire of CMBS investors for larger-sized transactions. Currently, the CMBS markets are rewarding transactions of $1 billion or more, paying a small premium for their greater liquidity and information flow. Large CMBS deals also save on legal, rating agency and marketing costs. Conduits have also become extremely competitive in pricing loans, a process that we believe will force out smaller lenders with relatively high fixed-cost structures.
The trend toward consolidation in commercial conduits has already begun, with alliances being formed among two or more conduits issuing a single CMBS transaction. Some recent partnerships include Lehman Brothers and First Union; Wells Fargo and Morgan Stanley; Deutsche Morgan, GMAC, ContiTrade, and Boston Capital; and J. P. Morgan, Midland and Smith Barney. We believe that this trend toward alliances will accelerate over the next few years, with more coalitions of conduits to issue a single transaction. Further, we think that many smaller conduits will drop out of the business altogether, forced out by lower margins in a highly competitive environment.
Rating agency approach to conduit CMBS
The credit quality of conduit deals is being pulled in opposing directions by various forces. On one hand, a competitive real estate lending environment, combined with tight security spreads and an increased bid for market share by the rating agencies, has put downward pressure on credit quality. However, the greater degree of diversification in recent conduit deals - both in terms of number of loans as well as underlying property types - has reduced the credit risk associated with this market.
Figure 6 Conduit Lenders Top Conduit Programs to Date Based on collateral contributions to securitizations, 1993-1997Q2 1993-1997Q2 Share (%) Issuance ($Mil.) 1 Nomura $3,439.2 12.5 2 DLJ 2,694.9 9.8 3 Merrill Lynch 1,889.7 6.9 4 First Union 1,862.4 6.8 5 CS First Boston 1,579.7 5.8 6 Lehman Brothers 1,553.9 5.7 7 NationsBanc 1,415.7 5.2 8 Morgan Stanley 1,234.7 4.5 9 J.P. Morgan 1,229.0 4.5 10 Citicorp 1,096.9 4.0 11 Wells Fargo 1,082.1 3.9 12 Heller 1,028.8 3.7 13 Midland 868.4 3.2 14 Daiwa 832.3 3.0 15 Chase 702.2 2.6 16 ContiTrade 653.5 2.4 17 Kidder Peabody(*) 617.2 2.2 18 Paine Weber 572.6 2.1 19 Ocwen Financial(*) 484.1 1.7 20 Amresco 456.1 1.7 21 Smith Barney 359.5 1.3 22 Bear Stearns 345.0 1.3 23 Dillon Read(*) 313.5 1.1 24 GMAC 278.9 1.0 25 Salomon Brothers 212.0 0.8 26 GE Capital 160.0 0.6 27 ING Capital 158.1 0.6 28 Deutsche Morgan Grenfell 150.0 0.5 29 Capital Lease Funding 129.4 0.5 30 Llama Company 54.3 0.2 31 Goldman Sachs 24.0 0.1 Total 27,478.1 100.0 * No longer active SOURCE: COMMERCIAL MORTGAGE ALERT
Conduits loans are also getting bigger [ILLUSTRATION FOR FIGURE 7 OMITTED]. "Large loan" conduit deals have average loan sizes of $30 million or more, compared with the $1 million to $5 million average loan size of the original CMBS conduits. Tighter spreads have allowed the conduits to compete with insurance companies and pension funds for large A-quality properties. At the same time that property quality is increasing, however, some observers note that loan terms are easing. Lenders are now more actively competing on loan proceeds.
Loan underwriting is now concerned more with what investors and rating agencies think about real estate than what the real estate department of a bank, insurance company or pension fund thinks. Loan pricing is being changed to get diversity in a conduit pool to please the rating agencies and investors. Rating agencies prefer large, geographically diverse pools with many property types, so conduits will seek to originate pools with those characteristics (see Figure 8).
Given these trends, we feel it is important that investors understand the rating process and credit trends underlying recent conduit issuance. We discuss in detail the rating-agency approach to rating CMBS conduits in the sidebar "The Conduit Rating Process."
Credit support levels have declined slightly
One threat to CMBS credit quality is that competitive pressures will force rating agencies to lower credit-support levels to gain market share. We think that over time, credit-support levels will start to decline as rating agencies compete with each other for new business, a pattern that occurred with residential mortgages in the early 1990s. For nonagency residential mortgages, AAA-credit support levels for 30-year fixed-rate mortgage pools slid from a range of 12 percent to 15 percent in 1987 to 6 percent or less by 1995.
The decline in CMBS credit-support levels has started to some extent, but from what we believe are extremely lofty original levels. Figure to shows the AAA-credit support level for all conduits since 1995. The solid line shows the unweighted moving average of the past six deals. This moving average has recently slipped from the low between 30 percent and 35 percent to slightly below 30 percent.
For lower-rated securities, the trend also has been to slightly lower credit-support levels, on average. Figure 11 shows that required subordination for BBB ratings has fallen from an average of about 15 percent in 1995 to the current level of about 12 percent. Single-Bs have seen the sharpest drop in credit-support levels. For single-B conduit CMBS, credit-enhancement levels have drifted down from 3 percent two years ago to the current level of about 1.5 percent to 2.0 percent [ILLUSTRATION FOR FIGURE 12 OMITTED].
Collateral has not become more risky
Has an increase in the credit quality of conduit pools occurred that supports the decline in CMBS conduit credit enhancement? We believe not, according to the credit characteristics of the average conduit pool. For example, the moving average debt-service coverage (DSC) ratio of conduit pools has remained between 1.4 times and 1.5 times over the last three years.
The moving average of loan-to-value ratios also has moved in a narrow range between 65 percent and 70 percent over the last three years, although there has been a slight upward trend over the past year.
We also found a slight increase in the percentage of the more risky (as viewed by the rating agencies) property types included in conduit pools. To measure this phenomenon, we assigned risk weights to property types based on what we perceived were the consensus views of long-term default risk. The risk weights ranged from 0.75 for multifamily to 2.00 for hotel properties. We held these weights fixed over time and computed the weighted average risk score for each conduit pool securitized over the past three years. Assuming our risk weights correctly measure the relative riskiness of each property type, we found that there was a slight upward trend in the riskiness of property types included in conduit pools.
The evidence we have collected shows little justification for lower credit-support levels based on collateral quality. We believe that credit-support levels have declined because of market-share competition among the four rating agencies.
Some rating agencies have maintained that the decline in credit-support levels is related to the increased diversification in large conduit pools. We also think that diversification deserves lower credit support, but it is difficult to quantify the amount. Rating agencies have given little guidance on this matter. Our belief is that rating agencies are using diversification to support lower credit-support levels that they could justify by historical and projected default and loss data. Since rating agencies are reluctant to change standards unless new data becomes available, diversification becomes an acceptable premise for lower credit support in the battle for market share.
Little default risk for AAAs
In our opinion, the decline of credit support and the slight decline in pool credit quality has had little impact on the default risk of AAA-rated conduit CMBS [ILLUSTRATION FOR FIGURE 13 OMITTED]. We believe that the AAA credit-enhancement requirements started out at such high levels that recent decreases have had only marginal effects on default risk.
Despite recent declines, subordination remains high relative to historical default risk. For example, with a credit support of 28 percent (the recent average level for large pool CMBS), about 70 percent of the loans in the pool would have to default with a 40 percent severity on each loan before the AAA security holder would suffer a loss. These types of losses are much larger than the worst losses of the commercial real estate recession of the early 1990s and are several times larger than our expectation of losses over the next cyclical downturn.
We believe investment-grade securities are still well protected against reasonable loss expectations over the next real estate downturn. For BBB-rated securities to suffer a loss, more than 30 percent of loans in the pool would have to default, a higher default rate than the worst origination years of the 1980s.
At the noninvestment-grade level, recent trends in subordination and pool credit quality have had a much larger impact than for investment-grade conduit CMBS. Further drops in credit support could have a significant effect on default risk. Last year, the rating agencies placed a minimum of 3 percent subordination for single-B CMBS. That barrier fell this year, with single-B subordination falling to the 1 percent to 2 percent range. If the CMBS market follows the lead set by the residential nonagency market, credit-support levels will fall for higher-rated securities as well.
Relative value of conduit CMBS
We currently have a favorable view of the credit of investment-grade CMBS backed by geographically diversified pools and a mix of property types. We think that many of these securities will gravitate toward higher ratings to the extent that loans pay off and credit support rises as a percentage of the remaining balance. In our view, given the current favorable real estate climate, the positive credit effect of prepayments will more than offset adverse credit selection of the remaining loans.
CMBS conduits should continue to attract investors even at spreads that are significantly tighter than a year ago. CMBS conduits offer several advantages for corporate investors as well as traditional mortgage investors. But it is important to consider both the pros and cons of CMBS backed by conduit loans.
CMBS conduit pros
* Wide nominal spreads relative to comparably rated corporate bonds. Single-A-rated 10-year CMBS are about 30 to 20 basis points wider than comparable corporates. On a default and option-adjusted basis, conduit CMBS look even better than their corporate counterparts (see Figure 14).
* Limited negative convexity. The prepayment lockouts and/or yield-maintenance provisions on most conduit loans make prepayment less tied to interest rates than for residential mortgage loans. This imparts less negative convexity to most CMBS issues than residential MBS. In addition, because CMBS tranches typically pay sequentially, the longer classes have structural prepayment protection. This is especially important for IOs.
* Prepayment protection provides a safe harbor from rising volatility without giving up much spread. The current low volatility environment has made investors in 30-year mortgage product somewhat fearful of a sudden uptick in volatility. CMBS allow investors to improve the convexity of their portfolios while maintaining relatively high yields.
* Improvement in credit quality over time. As conduits season, the LTVs of the underlying loans decrease and the DSCRs generally go up (due to rent inflation). This fact has not been fully priced into the conduit market, and we expect tiering of conduit spreads by origination date over the next year or two.
FIGURE 14 Conduit CMBS Spreads to Treasuries (in bp) (As of 3/12/98) Nominal Spreads to 10-Year US Treasuries Corporate CMBS CMBS Advantage AAA 45 77 32 AA 47 90 43 A 73 100 27 BBB 89 135 46 BB 214 240 26 B 331 450 119 Default Adjusted Spreads (to 10-Year UST) Corporate CMBS CMBS Advantage AAA 44 77 33 AA 43 90 47 A 62 100 38 BBB 47 120 73 SOURCE: MORGAN STANLEY, BLOOMBERG
CMBS conduit cons
* Lack of proven performance. The conduit market is relatively young, and investors have yet to see how conduit loans and structures will perform in a real-estate downturn. Loan diversification mitigates this problem to some extent.
* Balloon risk. The predominance of new issuance makes assessment of the refinancing risk on the balloon date difficult. In fact, extension risk has never been tested in the short history of CMBS conduits. However, an investor could buy a CMBS today and sell it three years from now without ever being in the "balloon situation," which usually comes in 10 years.
* Positive spread correlation to mortgages. CMBS are not completely delinked from the mortgage market. Consequently, a rally in the 10-year Treasury to below 6 percent could cause some widening in pass-through spreads that could spill over into CMBS.
Outlook for the conduit market
The foundations have been put in place for a growing, more liquid and more efficient CMBS market, albeit one with lower profitability for all participants. There will be consolidation in the conduit business as smaller lenders join up with larger conduits or exit the business entirely. In addition, if security spreads remain tight, conduits should be able to outbid insurance companies for high-quality properties. These high-quality loans tend to be larger and will increase the average loan size in conduit deals.
Soon we expect to see new collateral types included in CMBS conduits such as construction loans and new loan features such as rate locks. Documentation on conduit loans should become shorter and less costly over time. Technological change, including electronic submission of loan applications, will make the lending process more efficient.
Also, the conduit market could begin to see call-protection schemes common in the CMO market such as TACs (target amortization classes) and PACs (planned amortization classes). Wall Street will then become an agent for the conduits much like its relationship with mortgage bankers in the residential market. An eventual level of standardization and efficiency could be gained, similar to the residential market, which could even lead to some dealers exiting the business for greener pastures.
Transformation to a public market
The growth of the market share of large mortgage conduits has altered the landscape for real estate-related fixed-income investors. Conduits allow for the issuance of large, diversified pools of mortgages, which issuers can carve up to meet the risk preferences of individual investors. Large, multi-asset CMBS backed by conduit originations are a different product and present different risks and investment profiles than the single-asset or single-borrower transactions of the past.
The more significant turning point for the CMBS market is not the weakening of commercial mortgage credit, but rather the shift to a conduit-dominated market. The transformation of commercial real estate finance to a public market is well under way. We will have to wait to see the implications of this transformation on the real estate cycle.
Finally, we believe that the high level of credit support on investment-grade conduit CMBS minimizes the real estate risk of these securities. Investors seeking higher real estate exposure (and higher returns) can invest in noninvestment-grade CMBS. The creation of conduit securities has allowed investors to customize the level of commercial real estate credit risk in a portfolio, without exposure to the event risk of a single property.
Howard Esaki is a principal in the mortgage research department of fixed income research, specializing in CMBS research, and Joseph Philips is a vice president in the portfolio strategies department of fixed income research at Morgan Stanley Dean Witter in New York.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||multifamily and commercial mortgage-backed securities|
|Author:||Esaki, Howard; Philips, Joseph|
|Article Type:||Cover Story|
|Date:||May 1, 1998|
|Previous Article:||The new mortgage banking business.|
|Next Article:||The conduit rating process.|