Golden State Re Ltd. (Series 2011-1)

This is the first catastrophe bond which aims to provide coverage for workers compensation losses resulting from earthquakes. Golden State Re Ltd. is a Bermuda domiciled special purpose insurer set up for the purpose of providing $200m of coverage to the California State Compensation Insurance Fund (SCIF) for claims they may have to pay resulting from earthquakes. This is the first time the SCIF have issued a cat bond.

The SCIF provides workers compensation coverage to employers in the State of California and so can be liable for claims such as injuries and fatalities which could result from an earthquake damaging a covered workplace.

Golden State Re Ltd. is issuing a single tranche of $200m of Series 2011-1 Class A notes with the aim of securing a source of index-based risk transfer via ISDA-based reinsurance agreement for SCIF. The deal provides cover to the SCIF on a per-occurrence basis over just more than three years, with maturity scheduled for January 2015.

Golden State Re Ltd. uses a modelled loss approach to calculate an index to determine whether an earthquake has triggered the notes. It will use a notional portfolio of workers compensation risks, earthquake severity factors (ground motion) and time of day and the day of week an event occurs as weighting factors. This is a unique approach as places of work are plainly busier during working hours and on weekdays rather than weekends, so events outside of average normal working hours will result in a lower impact on the modelled loss calculation.

An index will be created with weighting of the above factors using RMS’ U.S. earthquake model. The notes cover a pro rate share of losses from the index attachment point of 1,000 up to the index exhaustion point of 1,447. Losses will be modelled deterministically using the earthquake event parameters (ground motion etc) and modelled against the notional portfolio using the day and time-of-day weighting to determine an index value and notional modelled loss calculation. Standard & Poor’s call this a parametric modelled loss trigger and note it can help to reduce modelling errors and implies a lower level of modelling risk than industry-loss or indemnity triggers.

The covered area under the terms of the deal means that a qualifying earthquake event can occur in any of the 50 U.S. States and the District of Colombia. However 99.99% of the SCIF exposure in the notional portfolio are in California. This means any losses resulting from a severe quake in neighbouring states could be covered under the terms of the deal and possibly any workers comp claims from covered workers travelling to other States may also be protected.

The SCIF is an unrated entity, which isn’t the norm in cat bond deals, so they will be depositing two quarterly interest payments in advance to mitigate noteholders exposure to them.

The deal features an annual reset on 1st November at which point the most recent notional portfolio and payout factors will be included in the modelling by reset agent RMS.

Initial probability of attachment is 0.55%. After each reset it will be no higher than that except the final year when the term will be more than a full year when it will be 0.56%. Initial annual modeled expected loss is 0.36% and probability of exhaustion is 0.22%.

The payout factors mentioned above are the model’s assumptions of the cost of various types of injuries simulated by the model.

S&P say that three historical events (the 1906 San Francisco earthquake, the 1857 Fort Tejon earthquake, and the 1994 Northridge earthquake) would have breached the attachment point if they had occurred between 8am and 4pm. The reason for this is the time/day weighting which means events that happen when the most people are in the workplace have the greatest loss potential. 2pm on a working weekday is the time of the greatest exposure for this cat bond (an event occurring at 4am for example, or even 2pm on a public holiday, would have a markedly different modelled loss amount). So actually, no known historical event would have caused this deal to trigger. However modelling does show that an event of under M6.5 could trigger the deal if it hit in the most exposed locations at the most exposed times.

Collateral from this deal will be invested in highly rated Treasury money market funds.