Notes and Caveats

Both Data Sets:

  • The data is provided to the LMA at a level designed to assist Members with benchmarking, whilst safeguarding commercial sensitivities. The nature of the data and various caveats and the terms contractually agreed between LMA and DXC and the LMA and Lloyd’s means that the data is strictly not to be circulated beyond members' own organisations.
  • Reports represent gross loss ratios and as such are not to be taken as profitability indicators.
  • The data covers all Lloyd's risk codes, but for reasons of commercial sensitivity, risk code data extracted from a population of fewer than five syndicates or where one syndicate writes more than 50% of the premium is excluded from the data.

DXC Data:

  • At individual risk code / class of business level, users should remember that, because the reports reflect only premiums and claims processed through Xchanging, comparison of DXC data with other statistical sources may throw up differences, both in terms of premium amounts and claims ratios.

GQD Data: Basis of Preparation

  • The following is a non-exhaustive list setting out the basis on which the GQD Triangle Data data will be prepared.  This may be updated from time to time in accordance the terms set out above and any changes shall be brought to the attention of managing agents.
  • The GQD Triangle Data consists of gross-gross signed premium, acquisition costs (signed basis), gross claims paid and case reserves.
  • Data is provided at risk code, year of account, development quarter level of granularity only. 
  • Lloyd’s will only provide GQD Triangle Data for a risk code for each year of account if the data for that risk code year of account as at the most recent quarter or at the sixth quarter (whichever is the earlier of the two) is the aggregation of data received by Lloyd’s from no less than five syndicates and where no syndicate accounts for more than 50% of the premium.
  • In addition to risk code level data, Lloyd’s will use its standard risk code groupings to group risk codes. This will be based on the latest hierarchy as published by Lloyd’s at the time the GQD Triangle Data is provided to the LMA.  Changes to the hierarchy will be applied retrospectively across historic years and development quarters.
  • Data provided is for all years of account from (and including) 1993.  For the years of account starting from 2012 the data provided is based solely on syndicate GQD return data.  For the years of account prior to 2012 the data used is a combination of  Xchanging data supplemented by data provided by managing agents in respect of premium and claims not processed through Xchanging and (for quarterly updates from 2012 onwards) from GQD return data. 
  • Figures provided are based on original reported currencies converted to GBP but based on two currency regimes: latest plan (all YOAs at latest SBF planning rates) and plan (each YOA at SBF planning rate for the YOA). The exchange rates used will be specified when providing the GQD Triangle Data.  No other currency level data is provided.
  • The GQD Triangle Data includes data from SPA syndicate GQD returns, for the period from 2012.
  • Syndicates may from time to time resubmit GQD return data and this may result in Lloyd’s retrospectively updating GQD Triangle Data at the next quarter to reflect the revised data. 

Identified Limitations

The following is a non-exhaustive list setting out known limitations with the GQD Triangle Data. This may be updated from time to time in accordance the terms set out above and any changes shall be brought to the attention of managing agents.

  • In preparing the GQD Triangle Data, Lloyd’s makes use of GQD return submissions provided by syndicates.  While Lloyd’s may prompt managing agents to investigate return data that appears incorrect or unusual (including via automatic warnings given by the Lloyd’s Core Market Return (CMR) system), the final decision to submit GQD return data is a matter for managing agents.
  • Syndicates may close their accounts early to meet processing deadlines for financial reporting. This can lead to a small delay in the reporting of data, which in turn can have an impact on the development of the data in the GQD returns provided by syndicates.
  • Prior to 2012, the claims outstanding figures reported are those held by Xchanging. In the GQD, these are the syndicates own estimate outstanding reserves.
  • GQD uses settlement due date. In the data prior to 2012, the processing due date was used; this may cause a limited double counting of data in 2012 calendar year.
  • No SPA and Life syndicate data was collected for the period prior to 2012.  The GQD Triangle Data therefore only includes SPA and Life syndicate data for the post-2012 period.  
  • Including SPAs in the dataset may cause distortions with the initial developments including some double-counting of data, which will drop out for SPAs that subsequently close back into the host. This double-count will have a larger impact on premium due to the different development patterns between premiums and claims. 
  • In 2016 the GQD reporting basis was changed so that the Acquisition Costs information required from syndicates was changed to include additional items to be consistent with reporting required for other returns. The historic submitted data has not been not changed to be on this basis and therefore a discontinuity in acquisition cost patterns may be found.
  • For 2017 year-end reporting, Lloyd’s provided updated guidance for the reporting of loss funds in the GQD, with these payments now being excluded from all reporting until the claim has been settled with the insured. A correction was also inserted into the 2017 year-end GQD to back out any historic payments a syndicate may have made into loss funds yet to be paid to the insured.  For 2016 and prior years of account the payment into loss funds continues to be reflected in the loss development patterns for all incremental changes until the adoption of the new reporting basis.