Evaluating credit enhancement floors in equipment ABS.

First Union [Securities.sup.*]

Since late 1997, the majority of equipment asset-backed securities (ABS) have been structured with credit enhancement (C/E) floors instead of the credit triggers used in previous deals. We see this change as a positive for the transparency of the equipment ABS market, as floors are easier to understand than credit triggers for most investors.

In a typical equipment ABS deal, pro rata payment of the senior and subordinated classes will revert to sequential payment if the C/E floors are hit. If the floors are hit, the senior bonds must be paid off before the more junior bonds, at which point the junior bonds are paid off sequentially among themselves. However, there is normally ample credit enhancement to pay off the junior tranches in full-- the cash flows are simply delayed if the G/E floor is hit. The purpose is to protect the more senior note holders if credit enhancement levels decline significantly.

I. INTRODUCTION

Since late 1997, the majority of equipment asset-backed securities (ABS) have been structured with credit enhancement (C/E) floors instead of the credit triggers used in previous deals. We see this change as a positive for the transparency of the equipment ABS market, as floors are easier to understand than credit triggers for most investors. Furthermore, unlike credit triggers, C/E floors have the following benefits:

* Absolute minimum enhancement levels, which reduce "tail-end risk" if defaults toward the end of the deal cause cash flow volatility

* Incorporation of the benefits of triggers into a less complicated structure

* Irreversible, so once floors are hit there is no risk for senior note holders that floor structural protection will disappear

In a typical equipment ABS deal, pro rata payment of the senior and subordinated classes will revert to sequential payment if the C/E floors are hit. If the floors are hit, the senior bonds must be paid off before the more junior bonds, at which point the junior bonds are paid off sequentially among themselves. However, there is normally ample credit enhancement to pay off the junior tranches in full--the cash flows are simply delayed if the C/E floor is hit. The purpose is to protect the more senior note holders if credit enhancement levels decline significantly.

This article covers the following topics:

* Overview of the C/E floor structure

* Equipment ABS structural payment description

* Historical loss stability of equipment ABS collateral

* Credit stress tests for mezzanine and subordinated equipment bonds

II. OVERVIEW OF THE C/E FLOOR STRUCTURE

As equipment ABS is a comparatively less commoditized asset class, it has benefitted from the innovation of simpler structures. C/E floors are less complex than the credit triggers structured into other ABS classes, which exhibit entrenched structuring standards. For instance, it is more difficult for auto ABS, which have been commoditized for many years, to overhaul structural standards already firmly accepted by the market by replacing triggers with C/E floors.

We have discussed credit triggers in depth in two research reports, Credit-Sensitive Asset-Backed Securities (ABS): Performance, Risk, Structure and Valuation on May 10, 2000, and "Examining Trigger Events in Home Equity Deals," Present Values, February 2000. Depending on the particular deal and asset class, credit triggers are activated when delinquencies, current losses or cumulative losses exceed a specified level or when residuals, the reserve account or excess spread decline to a specified level. Triggers tend to be opaque to the investor, as they often have many moving parts and are often inadequately disclosed in prospectuses. Furthermore, unlike C/E floors, triggers are correctable, which complicates the analysis. Although we explain in our research how First Union's proprietary Bond Analyzer can determine yield impairment from credit triggers, many investors do not have access to this type of analysis, which requires a considerable investment of time.

As diminution in credit enhancement represents lost or delayed cash flow due to delinquencies and losses, the protection provided by C/E floors for the senior tranches is similar to that provided by triggers. However, C/E floors are less complex than triggers and provide similar protection to the more senior note holders. In an equipment ABS deal, the C/E floor is expressed as a percentage of the original pool balance. The C/E floor for the most senior tranche is set at approximately 25% of the initial credit enhancement, which includes subordination, overcollateralization and the reserve account (or some combination thereof). For the subordinated tranches, this percentage is approximately 20%.

As an equipment deal amortizes, the cash flow as a percentage of current balance becomes increasingly significant. As floors remain constant (except as losses occur causing floors to increase pro rata with the losses), there is better enhancement for "tail-end risk." Because the floor is set as a percentage of the original balance, a transaction that experiences no losses will still hit its floors, changing the payment structure from pro rata to sequential. The mechanics of the C/E floor are best explained using the following simplified example in Exhibit 1:

If there are no losses, the floors will be hit when the deal has amortized 75%. At this point, the remaining $2.5 of unrated bonds will start paying down principal only after the remaining $22.5 of rated bonds are paid down. In the scenario of 2% losses, the floors are hit when the deal has amortized 55%. In this scenario, the remaining $4.5 of unrated bonds will start paying down principal only after the remaining $40.5 of rated bonds are paid down.

III. EQUIPMENT ABS STRUCTURAL PAYMENT DESCRIPTION

Exhibit 2 depicts the payment structure of the AAA rated bonds in a typical equipment ABS deal. In such a deal, the senior and junior bonds are typically paid sequentially until the shortest AAA tranche (usually money market eligible) is paid off. Subsequently, the remaining AAA tranches are paid off sequentially only amongst themselves so there is a tiering of average lives to appeal to a number of investors' risk/return targets. However, after the shortest AAA tranche is paid off, these remaining AAA rated tranches are paid off pro rata as a group with the lower-rated tranches.

Exhibit 2 shows how the Al tranche must be paid off before the other three AAA rated tranches. Again, the remaining AAA tranches are paid off sequentially amongst themselves. For instance, tranche A4, the longest bond, does not start receiving principal payments until Month 35, the month in which tranche A3 is fully paid off.

Exhibit 3 demonstrates how the junior tranches in an equipment deal priced in the second quarter of 2000 would be affected in a recession-like scenario in which the C/E floors are hit and losses are taken on the most subordinated tranche. The normal pro rata payment, which starts after the Al tranche is paid off, will be converted back to sequential payment if the C/E floors are hit. Due to the high credit enhancement required by the rating agencies, these floors would normally only be hit close to the clean-up call date. Therefore, the senior bonds have substantial protection from C/E floors, which are considerably more transparent than triggers yet serve the same purpose.

The recession-like scenario in Exhibit 3 assumes a constant delimit rate (CDR) of 4% and a cumulative net loss of 6%. At Month 23, the C/E floors are hit and the principal on all the subordinated bonds remains constant until Month 60. At Month 60, the principal on all the AAA rated bonds has been paid off and the junior tranches begin receiving principal. In this exaggerated scenario, the subordinated tranche D (BBB rated) receives some loss, whereas the B and C tranches are fully paid in Months 70 and 92, respectively. The maximum CDR this deal could withstand before any losses on any bonds are incurred is 3.82%, or a cumulative net loss of 5.73%.

IV. HISTORICAL LOSS STABILITY OF EQUIPMENT ABS COLLATERAL

In the equipment ABS sector, rating agencies have allowed C/E floors to replace triggers primarily due to the smooth loss curves exhibited by the underlying collateral. Exhibit 4 shows delinquencies, allowances and losses (net write-offs since 1983 for the equipment lease portfolio of one of the larger equipment originators, for which we have the longest historical data. Although annual losses were 0.2%-1.3% from 1983 to 1995 during a few recessions, annual losses have been constant at 0.2% since 1995. Compared with other ABS collateral, this loss curve is extremely smooth. Credit card and auto collateral tends to have relatively erratic loss curves. With erratic collateral loss performance, AAA note holders have a higher degree of comfort with triggers that can be activated early in the deal. Because the equipment loss curve is relatively stable, AAA note holders do not require protection early in the deal as there is relatively more certainty about expected losses.

V. CREDIT STRESS TESTS FOR MEZZANINE AND SUBORDINATED EQUIPMENT BONDS

Similar to the analysis performed in our May 2000 report on credit-sensitive ABS, we have stress tested the CDRs for the mezzanine (A rated) and subordinated (BBB rated) tranches of two bellwether equipment ABS issuers. The CDR assumes an annual percentage default rate over the life of the bond. The annualized realized loss rate is the product of the CDR and a severity rate. To be consistent, we assume a constant severity rate of 50% for all the bonds in our example. We use the Bond Analyzer, which uses Intex data on the underlying cash flows, to evaluate at what CDR the yield will break due to principal loss. As we demonstrated in Exhibit 3, C/E floors are hit well before principal losses on the junior tranches would occur. Therefore, although the floor is hit, there is still a high probability the principal will be paid back to the note holders.

Yield breakpoints vary depending primarily on the rating agencies' evaluation of the quality of the underlying collateral (reflected in the initial credit enhancement), the age of the deal and prepayments. Therefore, yields on bonds with better underlying collateral tend to break at lower CDRs than yields on bonds with poorer-quality collateral. This is because better-quality collateral is less likely to experience significant losses so less credit enhancement is required. Furthermore, the closer to the bond's maturity, the lower the impact on losses, because the CDR accumulates over a shorter period of time. Finally, the higher the prepayments, the closer to full amortization the bond is compared with similar vintage bonds.

Exhibit 5 shows the yield breakpoints for the mezzanine bonds of 1999 vintage CIT/Newcourt Credit Group Inc. and ORIX Credit Alliance, Inc., deals. The

Newcourt 99-1 and ORIX 99-A mezzanine yields are not impaired until after 5% and 6% CDR, respectively, which are extremely unlikely default events. Although the Newcourt deal had a significantly higher initial credit enhancement than the ORIX deal, the Newcourt deal has prepaid at a slower speed so that CDR accumulates over a longer period of time.

Exhibit 6 shows the yield breakpoints for the subordinated bonds of the same deals in Exhibit 5. The ORIX 99-A and Newcourt 99-1 subordinated yields are not impaired until after 5% and 4% CDR, respectively, which also represents an unrealistic default scenario. The Newcourt subordinated bond breaks approximately 1% CDR before the ORIX bond for the same reasons cited in the previous paragraph.

Exhibit 7 shows the average life extension in a severe recession-like 6% CDR scenario for the mezzanine and subordinated bonds for the two bellwether deals. Even in the exaggerated scenario of a 6% CDR over the remaining life of the deals, the C/E floors have little impact on the average life of the junior tranches. The Newcourt bonds extend by less than one year; whereas the ORIX bonds extend by well under six months.

VI. CONCLUSION

We believe the replacement of triggers by C/E floors is a strong positive for the equipment ABS market. As structured products can sometimes be relatively difficult to understand, any move in the direction of clarifying the structural protection mechanisms can only be good for the market due to increased investor acceptance. Furthermore, floors have absolute minimum enhancement levels, incorporate triggers into a less complicated structure and are irreversible.

Fortunately, there are tools, such as the Intex-based Bond Analyzer, that can clarify the performance and sensitivity of structured products so investors can perform adequate analysis. The G/E floor is a mechanism to revert the payment structure from pro rata back to sequential payment to protect the senior note holders. The effect on the junior note holders is that principal payments are delayed until more senior note holders are fully paid. Consequently, although the average life on the junior bonds can be extended somewhat by the C/E floor, yield impairment and losses would require a significantly larger deterioration in credit enhancement, which would normally only occur toward the end of the deal.

[GRAPH OMITTED]

[GRAPH OMITTED]

[GRAPH OMITTED]

[GRAPH OMITTED]

[GRAPH OMITTED]

[GRAPH OMITTED]

Exhibit 1

Credit Enhancement Floor Mechanics

Trust Assets                 Trust Liabilities

Deal Inception
  $100.0                       $90.0 Rated
                               $10.0 Unrated

75% Amortization--No Losses
   $25.0                       $22.5 Rated
(floors hit)                    $2.5 Unrated

55% Amortization--2% Losses
   $45.0                       $40.5 Rated
(floors hit)                    $4.5 Unrated

* Assumes all floors are 25% of the credit enhacement.

Source: First Union Securities, Inc.

LANG GIBSON is responsible for the research and strategy group for Structured Credit Products (SCP), whose product areas include credit derivatives, synthetic CDOs and cash GDOs. He writes and publishes the only extensive credit derivative/CDO weekly in the industry as well as numerous topical and primer reports covering SCP's three product areas. Prior to Bane of America Securities, Lang was the structured product strategist at First Union Securities. Prior to joining First Union in January, 1999, Lang brought eight additional years of structured product research and risk management advisory experience from Goldman Sachs, J.P. Morgan and Ferrell Capital Management. In addition to firm research, Lang has published numerous articles and chapters in well-known trade journals and Frank Fabozzi publications. Lang holds an M.B.A. in Finance from the NYU Stern School of Business and a B.A. from the University of Virginia.

Mr. Gibson is currently Director of Research for Structured Credit Products at Banc of America Securities. This article was written while he was with First Union Securities.