As we enter the final stretch of 2014, the deterioration of credit continues to be a concern. While some metrics such as leverage, debt service coverage and the increase in interest-only loans can be easily tracked on a deal-by-deal basis, others are more qualitative, such as location, rollover schedules, and time of ownership. In this report we focus on borrower equity cashouts, one of the options offered to borrowers in an increasingly competitive market. A cashout is any amount of proceeds leftover for the sponsor after earmarking for refinancing, reserves, capex and fees.
In our conversations with investors, the slippage of key credit metrics coupled with equity returned to borrowers (hereafter referred to as an equity cashout) has been on the top of their list of concerns. As such, we sought to quantify the impact that equity cashouts have had in driving our main stressed metrics into riskier territory. In this report, we focused on the Top 20 loans in our KBRA rated conduit universe from 2013, through mid-year 2014, where equity has been returned to the borrower. In determining whether or not a cash out occurred, we eliminated recapitalizations of previously unencumbered properties. We utilized only metrics based off the in-trust first mortgage debt, although we did examine the prevalence of mezzanine loans within this universe, as detailed further below. Our total universe of Top 20 loans comprises nearly $40.0 billion on a balance basis, while the cashout universe is $16.5 billion across 391 loans.
Our main takeaway: slightly less than half of the loans in KBRA’s Top 20 universe have returned equity to the borrower. However, we also individually reviewed characteristics of the loans to gauge the story behind the numbers. Taking that into account, it is evident that a majority of the cashouts have occurred in loans where the borrowers have had long operating histories of at least five years and where equity has built-up in the property either through equity investments and/or asset appreciation. Nearly all these loans were used for refinancing. In the sections that follow, we delve further into these loans, by looking at both qualitative and quantitative attributes.
In 2013, the characteristics of cashout loans closely tracked those of the universe, and in some quarters were better. However, as competition amongst lenders intensified in 2014, the credit metrics of these loans weakened faster than the broader universe. As one would expect, leverage for cashout loans deteriorated faster than the broader Top 20 universe, ending Q2 2014 at 104.3% compared to 101.4% for Top 20 loans. This differential was primarily attributable to the percentage of loans with KLTVs in excess of 100%, which was almost 10.0 points higher than the Top 20 universe in the first half of 2014, at 72.2%. Debt service coverage has maintained a healthy cushion even as leverage has increased, as the weighted average KDSCs for both the Top 20 universe and cashout loans was 1.73x. The percentage of loans with KDSC<1.20x was on average in the single-digits, however Q1 2014 was the outlier, where the percentage was the highest at 17.2% and 23.4%, for Top 20 and cashout loans respectively.
Despite investors’ concerns, the percentage of loans with cashouts has actually declined over time. In the first half of 2014, as other metrics were weakening, borrowers received 20.0% of loan proceeds in the form of equity, down from 25.2% in 2013.
More Qualitative Factors
As one would expect, retail, office, hotel and multifamily properties make up the collateral for at least 85.0% of the loans in the cashout universe. As depicted on the graphs, leverage and debt service for each of these property types are consistent with the Top 20 universe, with the exception of multifamily whose credit metrics were the weakest in 2014 for the cashout universe. Notably, 20 out of the 22 multifamily loans had KLTVs greater than 100% while seven had KLTVs greater than 110%. However, although leverage in both subsets has reached new highs for 2.0/3.0 CMBS transactions, on average, debt service has remained at or above 1.35x for most multifamily loans.
Larger Share of Equity Does
Not Imply More Risk
To further explore the impact a cashout has on metrics, we broke the loans out by percentage of equity returned to the borrower in 10% increments from >0% to 50%+. We broke the loans into these groups to gauge if a larger share of equity returned to the borrower impacts credit metrics more. The chart to the right summarizes the key metrics across the groups as well as the breakout in property types.
As illustrated, there is not a direct correlation between percentage of equity returned to the borrower and shifts in leverage and debt service. For loans with the largest percentage of equity returned, within the 50%+ bucket, the average KLTV was 93.6% with a KDSC of 1.78x, which are more favorable metrics compared to the broader
CMBS universe. There is however, a wide range within that average. Nine loans out of the 39 loans in the bucket (20.5%) carried KLTVs higher than 110%. Of those, two loans were also full-term IO. Nonetheless, no loans in this group had what we would describe as risk layering ie the combination of a KLTV greater than 100%, a lower than average KDSC and an IO term. However, seven of the loans in this bucket were acquired or developed in 2009 or later, exhibiting less of an operating history. It is especially difficult to gauge a property’s true stabilized level taking into account the valleys treaded during the recession.
Next we reviewed the metrics in the 20-30% group, a good proxy for the entire cashout universe, given the average return of equity across the entire universe is 23.8%. The average KLTV for this bucket is 97.3% while the KDSC is 1.71x. The IO index for this bucket is 20.7%, with 37 loans out of 76 loans carrying some IO term. Hotels, retail and office made up 88.3% of the group. This bucket also includes one of the highest leverage points across the universe with a KLTV of 134.9% and a 1.15x DSC. Nevertheless, less than 3% of the loans in this group have pronounced risk layering (KLTV>100%, lower than average KDSC and IO exposure).
Peeling back the layers
To tell the whole story it is necessary to review other characteristics of the loans. In this section, given the degradation of credit metrics, we have concentrated on the qualitative differences in our two subsets in the first half of 2014 starting with property type. We reviewed asset summaries for each of the cashout loans for this section, including the information obtained in our Asset Information Memorandums in our presales.
Although cashout loans tend to have higher leverage, the underlying loans are generally backed by well-performing properties that performed through the downturn steered by borrowers that have provided infusions along the way.
Multifamily properties are predominately (62.6%) made up of Class-A assets located within a mix of urban and suburban markets. The properties are at least 90% occupied and diversified between assisted living facilities, condominiums, garden, high-rise, townhomes and student housing with multiple amenities. The average sponsor ownership is approximately 7.0 years with more than $10 million in borrower equity invested within the properties. However, several properties were Class-C and had significant tenant concentrations in students or military personnel.
Generally, retail loans in this universe were located in urban or KPrime suburban markets. The majority of the properties are anchored retail, with needs-based tenants making up the bulk of the rent rolls. Larger properties also tended to have High Quality Credit-Worthy Tenants (HQCWT) in place. More often than not, the retail properties we examined were more than 90% occupied and were often 100% occupied. That said, there were outliers. Several properties are older and lacked updates and as such received below average KBRA property inspection scores. Others had tenants with above market leases in place, situated in competitive markets. These were the exceptions however and most loans that carried such a negative, had strong individual mitigants (only location for certain tenants for many miles, leases that were set to be executed after closing, strong performance despite tertiary location) in place as a counterpoint.
Office properties were overwhelmingly concentrated in primarily urban or suburban markets, with slightly more than half in or within relative distance to the city’s CBD. Nearly all the properties were Class-A or Class-B, with approximately 42% of loans characterized as Class-B. Sponsors have contributed equity to most of the properties although there are many instances where several years have passed since updates. For properties with serious rollover hurdles during the life of the loan, nearly all were structured with ongoing reserves and cash flow sweep language.
Hotel properties are a mix of limited, full-service and resort properties with the vast majority of the assets carrying national flags. The RevPAR penetration across all loans was 116.4% at securitization. Notably, the percentage of major flags in the 2014 subset was 89.3% versus 70.2% in 2013. More than 61% of the loans have KLTVs >100%, however, the properties have long operating histories as the sponsors have owned the properties on average for nine years. On average, borrowers have spent more than $5 million in improvements. There are three loans that have excessive leverage, above 120% KTLV. None of these loans had an interest only component. One of the three loans had RevPAR penetration rates of 137.6% while the other two were not as strong but were in the 90.0% range.
Risk on Top of Risk
The last attribute we reviewed was in-place mezzanine debt (there were also two instances of subordinate debt at the time of securitization) another example of risk layering. There were only 19 loans with equity returned to borrowers that had mezzanine in-place at the time of securitization. The accumulation of additional debt put an upward pressure on leverage, pushing the weighted average all-in KLTV for loans with mezzanine debt to 119.2%. Likewise, weighted debt service was strained. These loans had weighted average all-in KDSCs at the time of securitization of 1.01x compared to the in-trust KDSC for the loans of 1.62x. These metrics were somewhat skewed by one loan, 375 Park, that had an in-trust and all-in KDSC of 3.42x and 0.78x, respectively. Excluding this loan, the loans with mezzanine debt would have declined in KDSC from an in-trust average of 1.37x to an all-in average of 1.09x.
At first glance, the fact that nearly half of the Top 20 subset returned equity to the sponsors is concerning given that any market mechanism that adds leverage can quickly have a compounding effect. After closer inspection, we have concluded that at this point in the cycle, cashouts are largely being reserved for healthier loans. While returning equity to the borrower inherently shifts metrics, sponsors are simply taking advantage of property appreciation after years of strong operating performance. The ability to tap the CMBS market for that last dollar of proceeds has given the product appeal versus other debt options that may not offer cashouts as frequently. In the hands of the right borrower, it is not necessarily a negative. That said, our analysis revealed exceptions, and as always it is important to take into account the individual nuances that go into the funding of any given loan.