Updated: Catastrophe bonds exposed to the Japan earthquake

We’ve performed some analysis on the catastrophe bond transactions which have exposure to Japanese earthquake risks to try to give our readers a better impression of the likelihood of any of them suffering a loss after Friday’s massive earthquake and tsunami. It’s proved difficult, as some of the cat bond transactions have minimal information available publicly, and it’s not possible for us to tell you which deals face losses but we hope you find this of interest.

This transaction utilised the RMS Japan Earthquake Model. The deal utilises the Risk Management Solutions (RMS) Paradex model and in index-trigger and index values are per prefecture in Japan.

The Japan earthquake risk is in the Class E notes.

Standard & Poor’s said about this deal:

RMS’s parametric-based index, Paradex, creates index values (reported on, or converted to, U.S. dollars for this transaction) based on event parameters spectral acceleration for earthquake (using data supplied from the U.S. Geological Survey monitors seismic activity and published in post-event hazard maps). These index values are used to estimate industry losses by line of business. As the event calculation agent, RMS will calculate an index value following a qualifying event. The index value is the sum of the products of Paradex index values and payout.

The Class E notes cover losses from second and subsequent events on an annual aggregate basis. The covered perils are U.S. hurricane and earthquake, Japan typhoon and earthquake, and European windstorm. The initial attachment level is $100 million, and the initial exhaustion level is $200 million. The Class E initial deductible amount for each event is for U.S. Hurricane, $125 million; U.S. Earthquake, $125 million; Japan Earthquake, $100 million; Japan Typhoon, $100 million; and Europe Windstorm, $100 million. The initial maximum loss amount related to any one event is $100 million. Therefore, it would take a minimum of two events to result in a loss on principal on the Class E notes.

So, on Montana Re it would seem unlikely that it will be triggered given that it takes a minimum of two events. However, the deal may have its rating cut as if this earthquake qualifies as one event then the deals risk of default is much higher as only one more event is required to trigger it.

Update: Standard & Poor’s has confirmed that this is a multi-event bond as such there can be no loss of principal to investors due to this event. Of course that does mean it takes less events in future to trigger it.

This Vega transaction utilises EQECAT for risk modelling and analysis and we assume their risk models. The trigger for Japan earthquake risk is measured against a parametric index.

The risk structure of this deal is layered and features a reserve account. It was designed this way to minimise losses to investors as the reserve account must be depleted first, followed by the Class D notes and then Class C. When the deal was issued it was stated that it could take three qualifying events before Class C note investors lost any principal. It also uses an annual aggregate limit per peril which affords more protection to investors. More details on this deal.

It’s too early to say whether this deal will suffer any loss or a qualifying event from this earthquake.

Update: Standard & Poor’s has confirmed that this is a multi-event bond as such there can be no loss of principal to investors due to this event. Of course that does mean it takes less events in future to trigger it.

S&P also said about Vega Capital (and this may apply to the 2008 issuance too):

Only Vega Capital has a trigger mechanism that explicitly allows for tsunami. It applies a 15% uplift to the modeled loss for earthquakes of a magnitude of 8.0 or greater and with an epicenter that is within one of two predefined boxes off the coast of Japan.

Update 28th March; Moody’s said about the Vega Capital transactions:

Vega Capital Ltd. will not experience losses directly associated with this event. This is because these are multi-event cat bonds, i.e., two or more different events are needed in order to trigger losses to investors.

This deal uses the RMS Paradex models for Japanese quake risks and it was the first deal to do so (according to reports at the time of issue). Paradex will be used to create index values based on qualifying event parameters (quake intensity) measured by a network of reporting stations. These index values are utilised to estimate the industry losses from a qualifying event. RMS will calculate the index values. Index values are derived for Japanese quake using the sum of the products of Paradex index values and payout factors per or city code.

“Paradex Japan Earthquake uses near real time ground shaking data available from U.S. Geological Survey ShakeMaps to calculate insured loss estimates.” said Jinal Shah, Paradex manager at RMS. “There was no definitive authority providing industry loss estimates for catastrophe events in Japan, so Paradex was designed to bridge this gap.”

“Paradex allowed us to seamlessly model two major peril regions for this bond,” said Augustin Gas, ILS marketing manager at SCOR. “It provides high-resolution data, which minimizes our basis risk, and contracts settle in fewer than 40 days following an event, which is a major advantage for us.”

This deal provides cover in the region affected but we’ll need to wait for RMS to issue a model report and calculate the Paradex index values before it will be know if a loss will be experienced.

The Vita Capital IV Ltd. Series III notes have recently come to light as a deal with the potential to be impacted by the events in Japan. The deal covers increases in mortality in Japan and has been placed on CreditWatch negative by S&P. The noteholders are at risk from an increase in age and gender weighted mortality rates which exceed a predefined percentage on a predefined index.

For the disaster in Japan to cause a default and loss the indices would have to increase by at least 7.5% over any two  consecutive-year measurement periods between October 1st, 2010, and September 30th, 2014. At the time of issuance of this deal, Risk Management Solutions who provided risk modelling services estimated that an event in Japan which caused 50,000 deaths and was evenly spread across all age and gender groups would cause a loss notes.

Risk modelling and analysis for this deal was provided by EQECAT. The details we have are a little uncertain as to whether the Japan quake component of the deal utilises a parametric or a modelled loss approach. We’ll need to wait for further information to emerge before we will know the outcome for this transaction.

This deal utilises a parametric index based on data gleaned from the Kyoshin network of earthquake monitoring stations in Japan. The index value for the earthquake peril is based on peak ground accelerations (PGAs), provided by the Kyoshin network of Japanese earthquake reporting stations (K-NET). RMS uses the data to determine the sum of weighted shock index values for different earthquake calculation locations.

We’ll need to wait for the index values to be announced by RMS for this deal but as it appears to be triggered based on earthquake severity measured at the Kyoshin monitoring stations it is possible that Friday’s powerful earthquake has triggered this deal.

Topiary Capital uses reports from the Japanese National Research Institute for Earth Science and Disaster Prevention and risk modelling from RMS to create a parametric index against which the quake risk aspect is triggered.

The notes from this deal are second-event notes; which means they are exposed to second and subsequent events during the risk period so are unlikely to be triggered by this earthquake alone. However the risk of losses for Topiary Capital grows if this event qualifies as the first event and as a result we could see the notes downgraded (if it qualifies).

Update: Standard & Poor’s has confirmed that this is a multi-event bond as such there can be no loss of principal to investors due to this event. Of course that does mean it takes less events in future to trigger it.

Update 2: A.M. Best had this to say about the Topiary Capital cat bond which they have ratings for:

Topiary Capital Limited is a non-indemnity second-event catastrophe bond. It has not yet been subject to an event notice; therefore, the Japan earthquake event will not cause note holders to lose any principal.

This deal uses a reserve account and a layered approach to risk bearing (as in the more recent Vega deal above). It uses a modelled loss approach for the Japanese earthquake risk.

So as with the more recent Vega deal, it’s too early to say whether this deal will suffer any loss or a qualifying event from this earthquake.

Update: Standard & Poor’s has confirmed that this is a multi-event bond as such there can be no loss of principal to investors due to this event. Of course that does mean it takes less events in future to trigger it.

Update 28th March; Moody’s said about the Vega Capital transactions:

Vega Capital Ltd. will not experience losses directly associated with this event. This is because these are multi-event cat bonds, i.e., two or more different events are needed in order to trigger losses to investors.

This deal provides indemnity based cover to Flagstone Re. Flagstone Re performed their own risk modelling for the transaction using RMS models, RMS then reviewed their risk modelling aproach. It’s not possible to say whether this deal will be affected by this earthquake event and we’ll have to wait for details of Flagstone Re’s losses to emerge.

Update 2: A.M Best had this to say about the Valais Re cat bond:

The Valais Re Ltd. catastrophe bonds are indemnity-based, and the Japan earthquake peril cover includes fire following, tsunamis and other causes of loss.

Update 28th March; Moody’s said about the Valais Re transaction:

Valais Re Ltd, scheduled to mature in early June, is an indemnity transaction that could be exposed to principal losses due to this event. However, the sponsor’s portfolio is mostly concentrated around large cities like Tokyo, and beacuse of this, losses to the sponsor are expected to be below the cat bond attachment level.

The deal has a parametric index, triggered by the location and peak ground acceleration of earthquakes as reported by a network of seismographs. As the recent earthquake was so powerful and this deal is triggered based on actual intensity of the quake there is a chance that it could experience losses.

Update 28th March; Moodys said about the Muteki Ltd. transaction:

Muteki Ltd. Series 2008-1, scheduled to mature at the end of May, is a parametric transaction. Potential losses to investors are tied to the value of a parametric index that is overweighed around Tokyo. In spite of the marge magnitude of this event, our expectation is that the value of the index will be, at most, closer to the attachment point rather than to the exhaustion point (i.e., losses are expected to be either zero or if any, very small).

The Midori transaction utilises a parametric index. The contract provides for payments to Munich Re if an earthquake with a certain magnitude and depth occurs in a predefined zone in Japan. So again, with this deal triggered by actual intensity of an earthquake event it is possible that it could suffer a loss (although, Reuters has reported that an investor told them this deal was unlikely to suffer a loss).

Update: Standard & Poor’s had the following to say about Midori:

Following a number of questions from interested parties, we can confirm that MIDORI covers earthquakes within a radius of 70 kilometers around Tokyo.

As you can see, it’s not possible for us to predict which deals could suffer a loss. However, we feel it is safe to assume that any catastrophe bond which is a first-event deal (so only requires a single qualifying event to trigger it) and is triggered based on the actual intensity of an earthquake is seriously at risk of suffering a loss.

We will continue to update you as this tragic event unfolds.