Speedwell Weather

Speedwell Weather is the dominant Settlement Agent for parametric weather risk contracts world-wide.

Why use a third party as Settlement Agent?

We are aware that some weather transactions are settled on the basis of data provided by a national met service (NMS). Although data from most NMS is of a high quality, we believe this exposes counterparties to a number of possible problems:

• Delayed Data
  Speedwell Action: chase NMS for data, impute temporary value from other information e.g. alternative/backup data sources, interpolate a value from hourly data, use an estimate based on surrounding proxy sites, use an algorithmically determined value from our in-house software, use satellite/radar to estimate a value until official value becomes available
• Missing Data
  Speedwell Action: chase NMS for data, impute temporary value from other information e.g. alternative/backup data sources, interpolate a value from hourly data, use an estimate based on surrounding proxy sites, use an algorithmically determined value from our in-house software*, use satellite/radar to estimate a value until official value becomes available
• Data Errors. These include keying errors, mis-coded reports, manual reading errors and others.
  Speedwell Action: By default, the data is checked via both algorithmic testing and a human visual inspection. Where an error is identified with a high degree of confidence, use the above methods to calculate a replacement estimate. Chase with NMS to seek correction.
• Non-standard reporting periods such as cumulative rainfall totals over a number of days.
  Speedwell Action: where rainfall has been reported for multiple days (eg Australia) and where those days fall over the start or end of a risk period, we impute the daily rainfall split by using proxy analysis/radar/satellite and other information
• Failure or closure of a site.
  Speedwell Action: generation of synthetic value using established relationship between closed sites and group of proxies, as appropriate
• Inconsistency of data reporting: is the data on the NMS web site correct or is the data on their FTP site correct?
  Speedwell Action: we are familiar with these differences on a case by case basis and based on experience are usually able to identify the latest data. If in doubt, we will enquire with the NMS

Regulatory Scrutiny

We believe the importance of an independent provider of settlement data is highlighted by new regulatory scrutiny relating to providers of financial indices following the recent Libor scandal. Please see here for the consultation document relating to this area.

This consultation document specifically refers to weather in section 1.1:

• Other indices: There is a whole range of other indices such as weather indices ("UBS-GWI" UBS-Global Warming Index) used for damages and parametric weather contracts. Other indices, such as the PMI, Purchasing Managers Index measure business sentiment.

How using Speedwell addresses this developing issue:

• We are experts in weather data relating to financial contracts. Our origins as a data provider in 1999 lie in this area.
• We understand the importance of data provenance and audit trails and have built resilient systems to satisfy this need.
• Speedwell Weather Derivatives Limited (SWD) is our FCA-regulated subsidiary. SWD has been regulated by the FCA or its antecedents since 2000. If required, we are able to provide settlement data through SWD. We believe we are the only weather data and forecast provider able to provide weather-related data for settlement of financial contracts through a regulated subsidiary.

Speedwell is the dominant provider of weather data for the settlement of OTC weather transactions

Examples include:

• 2014: We are acting as data provider, settlement agent, and calculation agent for bilateral margining for the largest public weather derivative transaction to date. This is for the Uruguay national power company, UTE, as arranged by the World Bank. This also involved installing additional instruments.
• We were appointed as settlement agent for a large rainfall hedge for a hydro producer in Chile. This involved installing additional instruments.
• Multiple agriculture-related hedges in the USA
• The weather component of a large UK temperature/gas price hedge
• A number of hydro-related and mining related Australian rainfall transactions
• A sugar-beet related rainfall hedge in UK
• A UK snow depth related hedge
• Multiple snow depth related hedges, USA
• And many others

We are also the settlement provider for the CME-listed Australian weather derivatives contracts

Requirements of Settlement Data

• Accuracy
  Reports from most stations are of high quality. However extreme events or reporting problems may cause inaccurate reports. These may be temporary, but even temporary errors can impact in-deal valuations. By using our standard cleaning methods we are able to pick up such possible errors.
• Reliability
  Reports from most stations are reliable. Indeed, most weather reference sites are selected on the basis that they have been reliable. However closures of hitherto reliable stations may occur within a transaction period and feeds can be delayed.
• Transparency
  Should data points need to be revised, it is important that counterparties understand the tools, method and philosophy that are applied to this process while recognising that it may not be possible to be completely prescriptive about the precise method used to derive a given value. It is also important that there is a full audit trail showing the original data, the cleaned/filled data and any subsequent revisions.
• Independence – avoid any moral hazard
  As a settlement agent we are independent. We see both sides of the transaction as our clients. The issue of independence becomes more serious where weather stations themselves are owned by the company taking out the hedge. In this case our role may be extended. This might involve us installing a weather station to verify the readings taken in the primary weather station being used for settlement. However, our standard QC processes also help verify the data by comparing the primary site to proxies and using other sources such as rainfall radar or interpolated data.

* While we have developed an extensive software infrastructure that is key to this process, we believe this should never be fully automated that quality control should always be qualified by meteorological experience.