In the case of hybrid ARMs, both the purchase and rate and term refinance (no cash out) risk multipliers are greater than 1.0. At first blush, this seems counterintuitive, especially in the case of the purchase loan, which is generally perceived to be stronger. However, we believe the findings are due to the following:
Purchase borrowers may be first-time borrowers or “stretching” to purchase their home. In addition, they may employ a second lien loan to finance their down payment. In either case, they may be overleveraged.
In addition, purchase borrowers have, by definition, no time in property, and this may influence the propensity to default.
Rate and term borrowers are not extracting equity but rather seeking to lower monthly payments. The reluctance to extract equity or the absence of equity available for extraction may signal a weaker borrower relative to a cash-out refinance.
Fixed rate and term refinance default risk is less than refinance cash out or purchase. This borrower is most likely reducing rate and/or extending term. This, in turn, lowers the borrowers and reduces the probability of default.
Overcollateralization is the excess of the mortgage pool balance over the certificate balance and acts as internally generated credit support. Excess spread is used to accelerate the amortization of the outstanding certificate’s principal balance to a level lower than the mortgage pool balance.
Overcollateralization can either be allowed to build over time or be fully funded at closing. If the OC is built over time, excess spread is used to accelerate the paydown of the AAA classes until the target OC amount is achieved. The target OC amount is usually achieved in the early months of the transaction’s life. Conversely, if the OC is fully funded at closing, then excess spread is used to maintain the OC amount. The target OC amount is generally established as a percentage of the original principal balance. The required OC amount varies depending on the underlying collateral composition, structured used and the level of spreads on the liabilities (bonds) issued.
An overcollateralized transaction can sustain losses equal to the amount of current available excess spread and overcollateralization before incurring principal writedowns in the capital structure. For example, assume the transaction structure and a target OC building to 1.3%. Once cumulative losses exceed the OC amount, and if excess spread is insufficient to cover losses in a given period, then subordinated bond investors will incur principal losses.
Triple X is one subgroup in the broader life insurance securitization market. Total life insurance securitization reached roughly $18 billion in outstanding volume in mid-2007, with issuance of approximately $5 billion in 2006 and $2 billion year to date through June 2007. The life insurance securitization market divides into the following:
- Triple X securitization, which funds the regulatory capital requirement for level premium term life insurance policies. Deal terms may be as long as 30 years,
- Embedded value securitizations (also referred to as value-in-force monetization) release the profits “embedded” in future cash flows from a defined block of business,
- Catastrophic mortality bonds pay the issuer in the event of spikes in general population mortality by referencing a mortality index. These bonds can provide insurers with a level of protection that may not be available in the reinsurance market,
- and New business strain funding has been used to finance the upfront costs of writing new business by raising debt against future premiums.
Aside from the difference in motivation, embedded value and Triple X securitization differ in terms of collateral. Triple X transactions retain cash from the debt raised as collateral in the deal, whereas embedded value transactions are collateralized by future profits from a defined block of business.
Whereas Triple X securitization is specific to U.S. law, embedded value transactions have also been executed in the United Kingdom.
CDOs typically experience three distinct life stages: ramp-up, reinvestment, and amortization/maturity. Ramp-up defines the time between the premarketing phase of an issue and the first cash flow distribution. For cash transactions, the ramp-up phase usually lasts one to nine months; for synthetic transactions, shorter ramp-ups are common. During the reinvestment period, the CDO manager adjusts the collateral portfolio by buying or selling securities, subject to a set of prescribed constraints. The reinvestment period can be as short as three years (for some middle-market CLOs) and as long as seven years.
Equity investors have the right to call the transaction following the noncall period subject to a 2/3 majority vote. This option is most likely exercised when funding costs have fallen and the collateral is trading at or above par. In many cases, equity investors can roll their investment into a new issue and save on underwriting fees. We have estimated this optional redemption clause was worth approximately 61 basis points over the period September 2003 through September 2006. Alternatively, if funding costs rise, this option falls out of the money and has little value.
During the final life stage, the amoritization/maturity stage, a CDO distributes the principal payments from the collateral to the notes, amortizing the latter according to a prescribed schedule. Nearly all older vintage transactions had five-year reinvestment and three-year noncall periods; recent vintage issues typically have longer noncall and reinvestment periods. The final life stage of a CDO can be shortened via a cleanup call (exercised when the collateral’s outstanding balance drops below 10% of its original par value).
An equity investment has none of the contractual certainty or specificity of a debt investment such as a bond (Appraisal Institute, 2001). Ordinary shares in companies can be purchased through a stock market or via a broker. Investors effectively own a share of the company’s assets, that is, its equity, subject to prior claims of operating expenses and debt service, and will receive a regular income or dividend (based on company profits), usually twice a year. The ‘dividend yield’ is similar to the income or running yield on a bond and is calculated by dividing the dividend per share by the market price of the share.
Unlike bonds income from equities is not known in advance as dividends are linked to profits which, in turn, are linked to company performance and economic activity. Also, there is no redemption date so shares must be sold on a secondary market to realise capital. Prices on the secondary market are determined by supply and demand and vary according to future cash-flow expectations and perception of risk (Ball et al., 1998). Equity investments can yield a high rate of return but are more volatile and risky than debt investments such as bonds. Consequently market knowledge is needed if informed decisions are to be made and this incurs fees. Nevertheless, millions of pounds are traded in debt and equity markets daily, traders are sophisticated and well informed, investments are often professionally rated for risk and transaction prices are reported daily. Changes in yields of equities respond quickly to changes in supply and demand due to the efficiency of the equities market. Consequently data from these markets provide an objective basis for property market assumptions, particularly regarding expectations of debt capital performance (Appraisal Institute, 2001).
The majority of subprime loans are refinanced (cash out), meaning that borrowers are extracting equity from their homes. The purchase cohorts differ according to loan type. Specifically, the purchase ARM repayment risk multiplier is 1.06; conversely, the fixed multiplier is 0.97. This suggests that the hybrid ARM purchase borrower demonstrates faster turnover than the cash-out borrower, whereas the fixed rate purchase borrower exhibits a slower turnover.
Rate and term (no cash out) hybrid ARM and fixed rate borrowers exhibit lower voluntary repayment multipliers of 0.89, and 0.83, respectively. This suggests that rate and term borrowers refinance based on an
expectation of living in their homes for a longer period than either refinance cash out or purchase cohort.