Peer & Independent review Feedback and additional guidance paper august 2009

Transcription

1 Peer & Independent review Feedback and additional guidance paper august 2009

2 2 Disclaimer This paper is intended to provide up to date feedback and additional guidance to that contained within Lloyd s Franchise Standards for Underwriting Management, available via lloyds.com. Whilst all care has been taken to ensure the accuracy of the information herein, Lloyd's does not accept any responsibility for any errors or omissions. Lloyd's does not accept any responsibility or liability for any loss to any person acting or refraining from action as the result of, but not limited to, any statement, fact, figure, expression of opinion or belief contained in this paper.

3 3 Contents Executive Summary 4 Background & conclusions 4 Main findings peer review 4 Good practice peer review 4 Main findings independent review 5 Good practice independent review 6 Appendix I Review process 7 Establishing the position 7 Limit of the scope 7 Further information 8 Appendix II Peer review detailed findings 9 Key themes 9 Approach to peer review 10 Appendix III - Independent expert review main findings 16 Key themes 16 Approach to independent expert review 17 Risk management contacts 21 Underwriting performance contacts 21 Underwriting standards contacts 21

4 4 Executive Summary Lloyd s u/w Standards* The franchisee has effective systems and controls over each managed syndicate s underwriting: underwriting decisions are the subject of an appropriate peer review process a representative range of risks underwritten by the syndicate is reviewed and assessed regularly by an appropriately qualified individual who is independent of the underwriter of those risks * as set out on lloyds.com Background & conclusions Lloyd s requires all managing agents to meet the Franchise Standards for the management of underwriting. The implementation of effective peer and independent review processes, as addressed by the underwriting controls section of the Underwriting Standards, are generally regarded as valuable features of Lloyd s market practice. During Q Lloyd s Risk Management and Underwriting Performance Teams conducted structured interviews with a large sample of managing agents to confirm how review activity is being undertaken, with particular focus on the review of slips. This document conveys the results of that review and provides further clarity regarding Lloyd s expectations of effective peer and independent review. In summary, the findings from this review demonstrated that many managing agents adopt robust approaches to peer and independent reviews, taking account of the nature of their businesses. However, a number of agents fell short of Lloyd s minimum standards with regard to peer and independent review and Lloyd s will follow up with those agents individually. All managing agents need to ensure that their processes remain effective and it is hoped that the notes below regarding the main findings and comments on good practices seen in the market will be of assistance. Main findings peer review peer review was regarded as a crucial element in the management of underwriting risk. half the sample adopted a risk-based overall approach to peer review, with most of the balance featuring more extensive or random coverage of risks reviewed. Agents should be able to demonstrate a documented approach to sample selection, and where pre defined criteria are used to select the sample, this should be supplemented by a random element. the approaches to risks selected for review, focus of the review, timing of review, recording of review comments and the sharing of information collected varied by managing agent. Variation in process is expected as there can be no one size fits all approach. However, agents should ensure that their processes are properly documented. there is a general recognition that peer reviewers need to have equivalent or higher levels of experience and authority than the original underwriter. majority of managing agents enforce the same peer review disciplines in their service companies. Where agents do not apply the same standards of peer review in their service companies that decision must be based on a clear rationale. Good practice peer review Lloyd s considers the following to be indicators of good practice in a sound peer review process, and expects agents to be able to meet these benchmarks: the peer review process should be owned and supported by underwriting leaders.

5 5 the review should consider whether the syndicate s pricing/ underwriting policies and other procedures are applied correctly and consistently, and where they are not, issues should be recorded and resolved. if a risk based approach to peer review is applied, prescribed criteria should be used for selecting the risks for peer review, which is preferably automated. A random element should also be included to ensure that any risks and all underwriters may be part of the selection. in setting the criteria for targeting peer review, managing agents should particularly consider more focus on newer classes or historically under performing classes. peer reviewers need to have appropriate experience in the classes that they are reviewing and sufficient time to complete the process effectively, taking account of business volumes and absences. for each risk reviewed the findings should be recorded to a consistent standard and in a means that enables easy reference. It is important that any comments raised during the peer review will be readily available and utilised should the risk be renewed. the timing of peer review should reflect risk and may precede lines being put down, or ideally within 7 days post bind. the process of peer review should contribute to the development of skills of underwriters. the underwriting management team should receive regular written reports on the status of peer review activity and material findings. material findings should be escalated to senior management where appropriate and actions taken and evidenced where necessary. Main findings independent review independent review is generally being undertaken although the degree of value added is variable. Lloyd s expects that all classes of business and underwriters will be subject to the scrutiny of independent review. the majority of managing agents allow the independent reviewer to make risk selections for review activity (reviewers will usually adopt a risk-based overall approach to sampling risks, while ensuring coverage of all classes and individual underwriters). Managing agents should ensure that they have a defined sampling strategy for independent review of risks. most, but not all, independent reviewers examine slips as part of their review process. Lloyd s expects that an appropriate sample of slips should be reviewed by the independent reviewer in order to provide an effective challenge to the underwriters. the independent review process varied regarding timing of review, recording of review and the extent to which findings were distributed. there is a general recognition that independent reviewers need to have high levels of relevant underwriting experience. Lloyd s expects that in order to be able to provide effective challenge to the underwriters, the individuals carrying out independent review will have relevant underwriting experience. where internal reviewers (or reviewers previously employed by the firm) are used the degree of independence is hard to judge. Lloyd s expects that the reviewer must be independent of day to day underwriting. For example. this could be the Director of Underwriting (or equivalent) if that individual is not underwriting themselves. a significant minority of reviewers do not formally report their findings. Much of the value of the independent review process is lost where there

6 6 is no record of the findings. Lloyd s expects that there is a regular documented output from the independent review process summarising work performed and findings. a number of managing agents boards are not seeing the conclusions and recommendations made by independent reviewers. The report should be made available to the Underwriting Committee or equivalent and the Board. Good practice independent review Lloyd s considers the following to be indicators of good practice in a sound independent expert review process: the independent review process should be owned and directed by the active underwriter or Director of Underwriting, with full engagement from the agency board. the independent reviewer should have a terms of reference which specify review criteria, review scope, review frequency, access to information and reporting requirements. the review should identify whether the syndicate s strategies and pricing/ underwriting policies are applied correctly and consistently. the selection of risks to be reviewed should take account of risk and includes a random element ensuring that any risks, all classes, and all underwriters are covered within the process. the independent reviewer should have access to all underwriting records, the underwriting staff involved and the leadership of the agency. for each risk reviewed the findings should be recorded to a consistent standard and any outstanding issues followed up. independent review and reporting should be completed in a timely manner and reports should be produced at least quarterly. managing agents should ensure that independent reviewers have adequate time and access to complete the required process. the findings from independent review should used by the active underwriter to assess the extent to which underwriting is in line with the managing agent s strategies and plans. the independent reviewer s reported findings should be shared with the managing agent s board, with issues recorded in minutes and actions taken as necessary. To ensure that the background and nature of issues is understood, it would be good practice for the independent reviewer to meet with the agency board, or relevant board committee, periodically. The Good Practice notes above will be added to the Underwriting Standards section within lloyds.com for easy access. Any enquiries relating to Lloyd s Underwriting Standards should be directed to Kieran Flynn, Franchise Performance, Lloyd s ( )

7 7 Appendix I Review process Members of the Risk Management and Franchise Performance departments met with a representative sample of 22 managing agents during April and May 2009 Establishing the position Members of the Risk Management and Franchise Performance Departments met with a representative sample of 22 managing agents during April and May Those meetings were structured around a common set of questions covering the areas of peer and independent expert review. Additional documentation was also sought to support the approaches taken in some areas by the respective agents. Responses to our questions were subsequently collated and analysed to identify areas in which we could provide specific feedback to those agents involved, as well as providing a representative picture of the controls in these areas for the market as a whole to consider. We are taking the opportunity to offer additional guidance within the text and will seek to follow-up where that may be appropriate with relevant agents. Our findings are presented anonymously here so that the market as a whole may benefit from a greater awareness of the range of approaches. We would like to take this opportunity to thank those agents who took part in the information gathering element of this work. Limit of the scope A sample of Lloyd s managing agents was invited to participate in this review (amounting to roughly 50% by underwriting capacity). Whilst we consider that sample to be representative of the whole market in terms of size and classes of business written, it should not be assumed that our sample necessarily presents the complete picture. Furthermore, the review was not designed as an exhaustive assessment of the work performed by any of the agents who participated in the review. Our findings are only meant to comment on significant matters that were the subject of the review, together with any other related matters that arose out of the review where we consider that additional guidance to the market may be helpful. It is in the nature of a sample based risk review that only certain aspects of the operations of an agent are examined within a particular time frame and at a certain level of detail. Shortcomings may therefore exist which were not identified during the review. In particular, an inherent part of the process of independent expert review is that it is a qualitative review that is expected to cover a wide range of a syndicate s activities. However, our questioning in respect of independent expert review was aimed principally at gathering information on each agent s approach to the independent expert review of slips. We undertook only preliminary questioning on how the independent expert review of other areas of a managing agent s business was conducted and so we have not presented any feedback on those other areas. In this regard, agents are reminded that the Guidance in the Franchise Standard states: [Independent Expert Review] should ideally include consideration of reinsurance arrangements, systems for modelling risks and aggregate exposures and the records supporting pricing.

8 8 Further information For further information on this review process and findings, or any other area of risk management, please contact a member of the Lloyd s team who conducted the review. Contact details can be found at the back of this paper. Any enquiries relating to Lloyd s Underwriting Standards should be directed to: Kieran Flynn Franchise Performance Lloyd s ( ).

9 9 Appendix II Peer review detailed findings Key themes there was a considerable variation in agents approach to peer review 41% of our sample did not peer review all risks, instead implementing some form of risk based sampling the method of recording peer reviews was split - 52% electronic/ 48% paper based 36% of our sample prepared no reports on peer review - on completion or timeliness rates, nor on systemic findings the majority of agents have a system of peer review that is post-bind for the majority of risks it was not uncommon for backlogs to arise in the event of peer reviewer absence

10 10 Three key elements in a best practice approach to slip review appeared to be: 1 coverage of risks 2 the quality of review 3 the timeliness of review Approach to peer review All agents we met during the review appreciated that peer review was a crucial element in their underwriting risk management. The majority of agents we met had broadly similar approaches to peer review. However, there were some significant differences in those approaches and we have attempted to highlight those below. How does peer review differ between those agents we met? Coverage of Risks We asked agents whether all risks were peer reviewed. Are all risks peer reviewed? Yes 41% No 59% Risk based approaches to peer review are best supported by: a prescribed criteria for selection that is preferably automated a random element of selection to prevent the ability to predict which risks will escape peer review 41% of agents took a risk based approach to the inclusion of slips in their peer review procedures, i.e. those risks that were considered to be low risk were excluded from requiring a peer review. The risk based element commonly allowed underwriters to focus more time peer reviewing larger and more complex risks and waived the need for peer review of smaller but high volume risks. Determinants of whether a risk was required to be peer reviewed or not included: 1 class of business 2 line size 3 whether new or renewal business An added control that agents had not considered when implementing such a risk based approach to peer review is to include some element of random selection. Such a step ought to reduce the risk that underwriters could predict which risks would automatically escape peer review. All but one of the agents who implemented a risk based approach to peer review had an automated system to identify and communicate those risks that were required to be peer reviewed. In most cases, identification was by pre-agreed criteria, programmed into the underwriting risk recording system, which could not be varied by underwriters. In one case, the criteria could be varied by the Class underwriter so as to target particular areas of concern.

11 11 Where agents did not automatically peer review all risks, the total number of risks peer reviewed varied from 50% to 90% (by both policy count and premium income). The Quality of Review Aspects covered in peer review 43% Basic high level review 57% Detailed quality review in some cases the peer reviewer would be considered to be taking equal responsibility for the underwriting of the risk all but one of the agents that took a high level approach to peer review had a process of independent expert review that included a detailed review of slips by an external independent expert (or consultant) agents may wish to consider in more detail their approach to the peer review of endorsements The depth of peer review expected of underwriters varied considerably between agents. 43% of agents expected that underwriters would conduct high level checks of risks they were peer reviewing, covering matters such as data accuracy and the broad appropriateness of the price, terms & conditions, limits and deductibles. 57% of agents expected underwriters to conduct more detailed and qualitative peer reviews. To the extent that in some cases the peer reviewer would be considered to be taking equal responsibility for the underwriting of the risk. Two thirds of the agents we met explained that their syndicates performance of peer review was supplemented by a detailed discussion of risks written during regular (sometimes daily) syndicate meetings. It was interesting to note that those agents that took a risk based approach to peer review did not directly correlate to those agents that stated their approach to peer review was very detailed. However, we did note that those agents that enforced a peer review over all risks had a propensity to perform more high level peer reviews. A further key consideration in this area was how well the process of independent expert review complemented the approach to peer review. In this regard we noted that all but one of the agents that took a high level approach to peer review had a process of independent expert review that included a detailed review of slips by an external independent expert (or consultant). Only one agent commented on the approach to peer review of endorsements, stating that endorsements are peer reviewed where they result in an increase to exposure.

12 12 Two agents stated that higher risk items (identification of which was automated via prescribed criteria) would receive a second peer review from more senior staff. Ensuring peer reviewers have sufficient experience 14% 18% Built into underwriting authority, or other prescribed schedule Peer reviewer must always have equivalent or higher level of experience / authority All underwriters are involved 68% 18% stated that underwriters authority and responsibility to conduct peer review was built into their underwriting authority documents, or another prescribed schedule 68% of agents stated the peer reviewer must always have an equivalent or higher level of experience / authority than the original underwriter. A further 18% stated that underwriters authority and responsibility to conduct peer review was built into their underwriting authority documents, or another prescribed schedule. Where agents stated that all underwriters were involved in peer review it was commonly because the process was performed primarily during underwriter meetings. All agents stated that the peer review process and the communication of findings was a valuable element in the ongoing training of underwriters.

13 13 Method of recording peer review Electronically on the workflow or underwriting system(s) 48% 52% On paper / manual system where peer review notes were also recorded on the system, underwriters were given the benefit of being able to refer easily to those notes when considering the renewal of the risk Commonly, where peer review was recorded electronically, this allowed agents to monitor and report on the timeliness of peer review completion - a valuable key risk and performance indicator. In some cases, where peer review notes were also recorded on the system, underwriters were given the benefit of being able to refer easily to those notes when considering the renewal of the risk. Reporting of peer review 36% Reports on peer review are prepared and reviewed at an appropriate forum No reports are prepared 64% Two thirds of agents prepared reports on peer review. For most of those agents, their reporting capability was facilitated by an electronic system that monitored and recorded the completion of peer review. However, 3 agents had a manual process for reporting and as such had only periodical sample testing that could give rise to reporting of completion and timeliness rates. Where reports were produced they contained matters including: 1 percentage of peer reviews completed against expected timescales, with results sometimes split per team/ class/ underwriter

14 14 2 systemic issues identified during peer reviews. Peer review of service company business 21% Different approach (generally less oversight) Same approach 79% the vast majority of agents enforced the same approach to peer review over service companies as was enforced in the syndicate 36% of the agents we met did not have a service company. Of those agents that did, the vast majority enforced the same approach to peer review over that service company as was enforced in the syndicate. 3 agents had a less onerous approach to peer review in their service companies, but nonetheless had some system of oversight, for example annual audits or automated exception reports. One agent mitigated the absence of peer review by enforcing strict rates and conditions under which the service company could underwrite. The timing of peer review varied amongst agents, from pre-bind, to next day, to within 28 days Timeliness of Review We asked agents when risks were peer reviewed. When are risks peer reviewed? 5% Pre bind Post bind 45% Mix 50% 50% of agents were content to have all risks peer reviewed post bind. However, 45% of agents determined that, here again, it was appropriate to have a risk based approach and to require that, in some cases, a peer

15 15 review prior to binding was appropriate. Criteria where a pre-bind peer review was considered appropriate included: in the absence of the primary person responsible, to help prevent backlogs it may be sensible to Appoint deputies to conduct peer reviews The most effective method of monitoring the timely performance of peer review appeared to be via an automated electronic workflow system 1 binding authorities 2 risks written outside authority limits 3 larger/ more complex risks, sometimes determined by class, e.g. satellite or reinsurance treaty, or 4 risks written by junior underwriters One agent said they had an expectation for pre-bind peer review of all risks, though it accepted that much of the documentation to support that peer review would likely not be completed until post bind. Where risks were peer reviewed post bind, agents had varied requirements for when that should be completed, ranging from next day to 28 days. The majority expected peer review to be completed within 10 days or less. We noted that where agents had nominated specific underwriters for the task of peer reviewing specific pieces of business (or business meeting certain criteria) it was important that they also nominate a deputy to carry out that task in their absence to prevent large backlogs building up. The most effective method of monitoring the timely performance of peer review appeared to be via an automated electronic workflow system that: 1 was linked to the underwriting risk recording system 2 had a means of automatically communicating to underwriters those risks that were awaiting their peer review and 3 was capable of producing reports to show the timeliness of peer review, split by underwriter

16 16 Appendix III - Independent expert review main findings Key themes minimum standards for independent expert review were met by the majority, but not all the process of independent expert review in 3 agents did not include a review of a sample of slips 23% of agents did not document the findings of independent expert review in a formal report over half of our sample employed ex-underwriters as consultants to conduct their independent expert review 84% of our sample ensured the selection of risks to be reviewed was made independently of underwriters

17 17 Approach to independent expert review How does independent expert review differ between those agents we met? Structure of independent review 14% 18% Internal individuals Internal group 14% External consultants Combined approach 54% The use of internal experts was best employed when supplemented by a group/ committee that was composed of other independent experts of suitable stature Over half of the agents we met had a traditional approach to independent expert review and employed a suitably qualified independent consultant. Where internal individuals were used, they were commonly senior underwriting staff who sometimes had involvement in major underwriting decisions and underwriting/ reinsurance strategy. Their expertise in such matters was never in doubt but their independence was not always assured. Such approaches were therefore best employed when supported by a group/ committee that contained other individuals of suitable stature.

18 18 How is the sample of risks selected? 5% 11% Independent selection Selection not always independent Automated 84% Where slip review formed a key element of the independent expert review, 84% of our sample took steps to ensure that that selection was made entirely independently of underwriters. This was commonly achieved by asking the independent expert to make the selection. One agent had an automated system for the selection of risks for independent expert review. We noted the following three areas where a number of agents did not appear to be meeting the minimum standards for independent expert review: There were three areas where a minority of agents did not appear to be meeting the minimum standards for independent expert review 1 Review of a sample of slips Three of the agents we met had no system for the independent expert review of a sample of risks. 2 Preparation of written reports Five agents did not prepare a written report to document the findings of independent expert review. 3 Frequency of written reports Two agents produced independent expert review reports less frequently than quarterly.

19 19 How often do the results of independent expert review get reported? 12% 24% Monthly Quarterly Less often 64% It is clear from the Underwriting Management Standard that independent review should extend further than the review of slips Where reports were prepared on the findings of independent expert review, only a third of agents routinely provided those reports to the board. Other agents preferred to submit them to a nominated subcommittee. Other Areas of Independent Expert Review As stated above, our questioning in respect of independent expert review was aimed principally at gathering information on each agent s approach to the independent expert review of slips. We undertook only preliminary questioning on how the independent expert review of other areas of a managing agent s business was conducted. To that end we take this opportunity to remind all agents of the minimum standards set out in this regard (see above), for it is clear from the Underwriting Management Standard that independent review should extend further than the review of slips. There are many areas within the business of a syndicate and agent in which the benefit of an independent expert s input or oversight can be highly beneficial. Clearly, the area where this is proving most applicable is in the appointment of independent non-executive directors to the boards of managing agents. For further guidance in that area we refer agents to Lloyd s Standard for Governance, found here: ork/franchise_standards/governance/. Further to that, a discussion paper on Board Composition and Performance was issued to the market in August 2008 and provides additional information in that regard. A copy can be obtained from Matthew Rowan ( matthew.rowan@lloyds.com) Conclusion This paper aims to outline the main observations we made during our exercise into understanding the processes of peer and independent expert review amongst Lloyd s managing agents and syndicates. It is our intention to follow up our individual findings with those agents involved in the review and to facilitate the raising of standards where appropriate. In addition, we are circulating our general findings, as

20 20 contained herein, amongst the remainder of managing agents in the market and to explore whether there are any areas where we may be able to assist in raising standards amongst those agents also.

21 21 If you would like more detail on these issues, or further assistance to facilitate improvements in your own processes, please contact a member of the Lloyd s team who undertook this review. Risk management contacts Head of Risk Management Olly Reeves olly.reeves@lloyds.com Senior Risk Executives Ian Bryant ian.bryant@lloyds.com Matthew Rowan matthew.rowan@lloyds.com Risk Executives Michael Shaw michael.shaw@lloyds.com Max Diacon max.diacon@lloyds.com Underwriting performance contacts Any enquiries relating to Lloyd s Underwriting Performance should be directed to: Deputy Head Underwriting Performance Mark Stockton mark.stockton@lloyds.com Underwriting Performance Executives Charles Lamplugh charles.lamplugh@lloyds.com Nigel Ralph nigel.ralph@lloyds.com Underwriting standards contacts Any enquiries relating to Lloyd s Underwriting Standards should be directed to: Head of Underwriting Standards Kieran Flynn kieran.flynn@lloyds.com