Measuring School Risks

Transcription

1 Measuring School Risks In this course, we'll learn that the risk manager performs both qualitative and quantitative risk assessment when analyzing the exposures that face his or her organization. Qualitative assessment tasks will require an understanding of, among other things, the school's loss experience and appetite for risk. The centerpiece of the risk manager's quantitative assessment is often the loss run, or claims data, provided quarterly or semi-annually by the insurer. The risk manager must determine the full impact of claims on the organization, in order to estimate future losses, and allocate resources. The Certified School Risk Managers Program was created to fill a void in the educational opportunities for school risk managers, and those agents and other service providers who work with them. By taking and passing all five of these courses, you can earn the prestigious CSRM designation. The Certified School Risk Managers Program (CSRM) is a career-building, five-part designation program for risk management education specialists. The CSRM faculty is represented by the field's top practitioners and leaders. Five intensive courses guide participants through the risk management process, emphasizing practice over theory. The five core courses for the CSRM Designation Program correspond to the five steps of the Risk Management Process. A discussion of the entire Risk Management Process is included for study at the beginning of each CSRM course. We do this because, in the actual practice of risk management, activities in one step influence actions taken in another step, creating interdependent processes. Copyright The Society of Certified Insurance Counselors, Incorporated. All rights reserved. This material includes copyrighted material of ISO Properties, Inc. with its permission. This course or any part thereof may not be reproduced, copied, republished, uploaded, posted, transmitted, or distributed in any form or by any means or stored in any information retrieval system without the express written consent of the National Alliance for Insurance Education and Research and the Society of CIC. All trademarks and services marks are proprietary to The Society of CIC or its affiliates or licensees. 1

2 Fundamentals of Risk Management (FUN) The first step in the risk management process is identifying exposures subject to loss, since an exposure cannot be analyzed, controlled, or financed until it has been identified. This course gives an in-depth look at the overall risk management process, delves into the identification step of the process, and examines the function of the school risk manager. Measuring School Risks (MEA) This course will delve into the mechanics of forecasting and trending losses to be used in determining insurance program retentions and deductibles. Handling School Risks (HAN) This course studies the development of a safety and health plan, loss control fundamentals (including the six primary risk control techniques), risk control and mitigation for specific exposures, and managing school claims; a post accident loss control reduction technique. Funding School Risks (FSR) The purpose of this course is to examine the various loss funding techniques available to school risk managers. This will include guaranteed cost programs, deductibles and retention programs, pools, and transferring risk through contracts. Administering School Risks (ADM) This course will cover how a school risk manager implements and monitors the risk management program. Some of the topics will include the school risk management team, information technology for risk management information systems, allocating the cost of risk, ethics in school risk management, and requests for proposals. How to Receive Credit for Measuring School Risks To successfully complete this course, you must pass a final examination in the presence of a disinterested third party proctor/monitor. Some states have special requirements for the proctor/monitor. To learn more about the proctor/monitor requirements for your state, please click on the Proctor Instructions link below. Your proctor/monitor will enter his or her address to sign in to the exam website. He or she must remain in your presence for the entire hour of the exam. Your proctor/monitor will enter his or her address to sign in to the exam website. He or she must remain in your presence for the entire hour of the exam. 2

3 The proctor/monitor certifies your identity, and that you have completed the exam without outside assistance of any kind. Make sure to use your 60 day timeframe to carefully prepare for the exam, and select a person to act as your exam proctor/monitor well in advance. You must pass the exam in order to receive any of the following credit: Credit towards achieving the CSRM Designation Credit towards achieving the CISR Designation 8-hour Annual Update credit for the CSRM Designation State Continuing Education Credits/Hours (for state issued license renewal) Affidavit of Exam and CE Request Form With every successful exam, you and your proctor must sign and submit an affidavit of exam. When you pass your exam, fax the affidavit of exam and CE (Continuing Education) request form to the fax number printed on the cover sheet. Please send the original affidavit by mail (address also printed on the affidavit cover sheet.) We strongly recommend that you keep a copy for your records. Many School Risk Managers will have no need to request Continuing Education credits for an insurance license. Please send the CE Request form even if you are not requesting CE hours/credits for insurance license renewal. Check off the "No CE" box at the bottom of the form. Note: You will receive an reminding you to send in the affidavit if you pass your exam with a score of 70 or above. Do not submit an affidavit for an unsuccessful exam. CSRM Measuring School Risks Final Exam The exam is a challenging 100 point randomized test comprised of 50 multiple choice questions worth 2 points each. When you submit your exam for grading you will receive your score immediately. The score page is sent to your address and to The National Alliance for Insurance Education. The score is recorded in your National Alliance permanent record, but is not certified until you submit the affidavit of exam, signed by you and your proctor. (In New York, the proctor/monitor must mail the affidavit.) 3

4 Once we receive and accept your affidavit, we will send your notification of passing. Curriculum Support Faculty Members from the National Alliance for Insurance Education & Research are assigned to take s from students in the Online Courses. Our faculty are experienced practitioners and teachers in the industry. We ask them to respond to each within 24 hours, or before the end of the next business day. Each course mentor will be happy to clarify any portion of the curriculum that you seek help with. Make sure you have carefully reviewed the course curriculum and clearly note the page or self quiz question number when you contact a Course Mentor. Help Desk The mentors will refer any computer issues you have to the Online Help desk. The National Alliance staff is available by phone or for technical support issues. cisronline@scic.com. To phone the Online Help Desk call: and select the online option. Monday through Friday 8:30 am to 5 pm Central Standard Time Course Study and Exam Preparation Have you ever thought about how you learn? The study aids listed below will help you determine your progress and test your understanding of concepts and examples presented in the course. Learning Objectives: Designed for managing your own learning. The learning objectives for the course are listed at the beginning of each topic. The learning objectives are indicated throughout the course pages as well. At the end of the course, you will have the opportunity to read the learning objectives again, and see how confident you feel about each one. 4

5 Self Quizzes: Another learning management tool. You are required to pass each self quiz with a score of 70 or above before moving forward in the course, and you can launch a self quiz as many times as needed. To print the score page of your self quiz, click on Assessment Results, then right click on the page. The Assessment Results page makes an excellent study aid. Glossary: Glossary terms and definitions are critical to insurance professionals, and a key study aid for your online course. To define a term, click on the Glossary link above. Definitions of newly introduced terms will also be included on the course pages. Knowledge Checks: Application level questions. By attempting to apply the concepts of the course, you will better prepare yourself for the final exam. Make sure you attempt each knowledge check in the course. Course Mentor: Don t forget to the Course Mentor with your questions about the curriculum. Our faculty members are distinguished producers and risk managers who currently work in the insurance industry. The mentors are happy to explain and clarify the concepts in the course. They will return your on the next business day. A note about the CSRM Online Glossary, our online glossary is extensive and features many terms and concepts used by risk managers. Glossary terms that you see on a course page may be a part of the final exam. However, when taking the final exam, you will not be tested on the entire glossary. 5

6 Course Objectives Overview Section 1 Introduction and Qualitative Analysis for School Risk Section 1 starts with an overview of the five step Risk Management Process, a powerful set of tools for proactive risk management. How does the Analysis Step fit in to the organization's overall Risk Management Program? Section 2 Qualitative Risk Assessment and Loss Run Analysis Qualitative risk assessment is quite different from quantitative. In this section you will study the seven areas of qualitative risk assessment and then move on to understand the purpose and process of the loss run analysis. Section 3 Quantitative Analysis: Tools and Forecasting Section 3 is about the numbers, or quantitative risk analysis. This section provides an overview of tools used to develop loss data, index data to inflation and estimate a loss pick for a future period. Section 4 The Risk Analysis Process Five factors influence a school's financial capacity, thereby affecting the risk manager's decision making. An overview of cost vs. benefit analysis is presented in this section, along with the concepts of present and future value and net present value. Course Summary 6

7 As a new risk manager, you have many opportunities to learn from experienced risk managers. As an experienced risk manager, you will have opportunities to share what you know with others. Four Ways to Develop your Career as a Risk Manager: 1. Network school principal experienced risk manager board member 2. Join a professional state association that supports school risk management Public Risk Management Association (PRIMA) Your state s school board association 3. Become involved in a qualifies risk management training program Obtain risk management certifications and knowledge 4. Develop rapport with external insurance agents, brokers and/or consultants Meet with insurance and risk management vendors Florida Residents Only The state of Florida has taken a very strong position on the issue of Unauthorized Entities. An Unauthorized Entity is an insurance company that is not licensed by the Florida Department of Financial Services. Agents and brokers have responsibility for conducting reasonable research to ensure that they are not writing policies or placing business with Unauthorized Entities. Lack of careful screening can result in significant financial loss to Florida residents due to unpaid claims and/or theft of premiums. Agents may be held liable when representing these Unauthorized Entities. It is the agents and brokers responsibility to give fair and accurate information regarding the companies they represent. Any question about the authorized status of a company can be checked by calling the Florida Department of Financial Services at or We urge all agents and brokers to adhere to this admonition. 7

8 Roger's First Day on the Job In this course we will sometimes look at things from the perspective of an inexperienced school risk manager. We've named him Roger. Roger is the new risk manager of an urban school district. In his previous position, he served as the business manager of a college in another community. He is an excellent communicator, is skilled at managing large groups of people, and is familiar with insurance. In addition to learning his new job, he's facing the immediate task of preparing the cost of risk section of the annual risk report, which will be submitted to the school board. To assist him in getting started, we will introduce him to Jamal, an experienced risk manager working for a large metropolitan school district. Roger: I've got to get organized to complete this report. Where should I start? Jamal: Well, let's make sure you have all your analysis tools lined up and that you know how to use them effectively. 8

9 Section 1 Introduction and Qualitative Analysis for School Risk Section 1 starts with an overview of the five step Risk Management Process, a powerful set of tools for proactive risk management. How does the Analysis Step fit in to the organization's overall Risk Management Program? Learning Objectives By the end of this section you will be able to: 1. Identify and describe the five steps of the risk management process. 2. Understand the role of analysis in the risk management process. 3. Understand the benefits of a proactive approach to risk analysis. The Risk Management Process The Risk Management Process is a powerful five-step guide that can dramatically increase your effectiveness as a risk manager. 1. Risk Identification: The process of identifying and examining the potential courses of losses faced by the school 2. Risk Analysis: The assessment of the potential impact (costs) that various exposures can have on the school 3. Risk Control: Any action to minimize, at optimal costs, losses which strike the school 4. Risk Financing: The acquisition of funds, at the least possible cost, to pay for the losses that strike the school 5. Risk Administration: Planning, implementing, and monitoring the risk management program. Risk Administration also includes the development of the risk management team, utilizing both internal and external sources. 9

10 Step 1: Risk Identification The process of identifying and examining the potential sources of losses faced by the school. Before you can manage risks, you first need to identify them. Risk managers are, among other things, risk detectives. They find risks and bring them into focus. Risk Identification makes risk management possible. Step 2: Risk Analysis The assessment of the potential impact (costs) that various exposures can have on the school. In Risk Analysis the risk manager is often looking for patterns. In the first step of the Risk Management Process, Risk Identification, you collected data. In the second step, Risk Analysis, you analyze this data to find important patterns. You examine these patterns to assess what impact they may have on the cost of loss. This is done through qualitative and quantitative analysis. Risk Analysis is measuring and evaluating risk. 10

11 Step 3: Risk Control Any action to minimize, at optimal cost, losses which strike the school. Risk control is the action step of risk management. You apply the information you identified and analyzed in the first two steps of the Risk Management Process and take actions that minimize risks. Risk Control: Taking action to avoid, prevent, or reduce the impact of losses. Risk Control is handling risk. Risk Control Techniques There are six techniques of risk control for school districts: 1. Avoidance 2. Prevention 3. Reduction (pre-loss and post loss) 4. Segregation (separation or duplication) 5. Transfer (contractual, physical, or both) 6. A combination of techniques If a violent act occurs on campus, what contributed to the loss? Was there a failure in campus security? Did door locks or broadcast systems fail? If the incident involved a student, should the administration have responded to warning signs? Consider two approaches to why accidents occur. Understanding the causes of accidents will help you control future risks. Human approach: People cause accidents. Engineering approach: Things and energy cause accidents. 11

12 Step 4: Risk Financing The acquisition of funds, at the least possible cost, to pay for losses that strike the school. An earthquake hits a California school, damaging buildings and injuring staff members and students. How will the school pay for this catastrophe? Did it set aside enough emergency funds? Does it have adequate insurance? Whatever risks your school faces, addressing funding before the risks occur keeps the school s expenditures in check. The alternative can be extremely costly both financially and socially. Risk Funding: Funding losses Ways to Finance Risk A school can finance risk through either retention or transfer methods. Some exposures are financed using a combination of these. Retention The acquisition of funds from within the school to pay for losses. Retention can either be planned (active) or unplanned (passive). Transfer Contractual or other arrangements whereby losses are financed from outside the school. Schools can either transfer risk by purchasing insurance or by using one of several noninsurance transfer techniques. In the area of non-insurance transfer of risk, risk managers sometimes refer to "physical transfer," or "contractual transfer" of risk. Two examples of these distinctions: 1. A district outsources a function, such as landscaping services. The risks associated with this function have been physically transferred to another entity. There's a contract involved and the landscaper will take out insurance to meet his obligation. From the district's perspective, this is a physical transfer of risk. 2. Clauses in a contract, such as hold harmless agreements, are often referred to as contractual transfer of risk. The contract states that if a liability arises one of the contracting parties will assume the consequences of that liability. 12

13 Step 5: Risk Administration The process of planning, implementing and monitoring the risk management program. A risk manager is highly dependent on his or her risk management team, a group of dedicated individuals from within and outside of the school who help the risk manager control losses. Risk management team members include: the risk management department staff, insurance professionals, risk consultants, experienced risk managers, concerned community members, and others. In the final step of the Risk Management Process, Risk Administration, the risk manager works with his or her team to plan, implement, and monitor the risk management program in other words, to move it forward. Effective Risk Administration requires excellent communication, teamwork, and project management. 13

14 The Role of Analysis in the Risk Management Process The Risk Analysis Step: Provides an understanding of exposures identified in the first step of the Risk Management Process and the impact on the school district. Provides a method of analyzing exposures so they can be substantiated and prioritized. Provides information to choose control and financing techniques. Proactive vs. Reactive Risk Analysis The goal of risk management is to move from reactive mode to a proactive approach. This means that elements of the risk management program are developed to respond to losses in a timely manner (active management) and the risk management program as a whole addresses exposures before they become losses (proactive management). Proactive risk analysis - Timely and forward thinking analysis of exposures and losses. Proactive risk analysis anticipates risks. Reactive risk analysis - Analyzing issues and concerns as they arise. Reactive risk analysis is response driven. 14

15 Benefits of a proactive approach include: Provides an understanding of issues and concerns before they manifest themselves as losses. Provides an understanding of loss trends to develop more accurate predictions of future losses. Provides an understanding of exposures so loss control and financing activities can be effectively addressed. If you watch shows about crime scene investigation, you can see how being a crime scene detective might be fascinating. Being a risk manager is very similar to this. Like crime scene investigators, risk managers: Hunt down facts Analyze data Interview witnesses Talk to experts Talk to other risk managers Use their intuition Look for anything they may have missed Put the pieces together Write reports 15

16 Roger: I need your help on something. Carol: Sure. Roger: The school's loss data shows that we've been having a lot of slip-and-fall accidents in the cafeteria. The majority of them happen on Wednesdays. Can you think of anything that might be causing this? Carol: Wednesday is Taco Day. But the floors don't get any more spills on Wednesday than any other day...i am not sure what could be causing this. Roger: Anything else you can think of that might be causing more accidents than usual? Carol: Now that I think of it, Tuesday nights are when the floor gets a thorough cleaning. You might want to talk to Samantha about that... Roger: Carol told me that your staff cleans the cafeteria floor on Tuesday nights. We've been seeing a lot of slip-and-fall accidents in the cafeteria on Wednesdays. Is there anything you can think of that might be causing this? Samantha: The cleaning solution that we use for the cafeteria may be the problem. It does seem to leave the floors a bit slippery, even after it dries. Roger Could you switch to a different one? Samantha: I'm sure we could. I'll look into it. Please refer to the end of Section 1 to go over the section exercises. 16

17 Section 2: Qualitative Risk Assessment and Loss Run Analysis Qualitative risk assessment is quite different from quantitative. In this section you will study the seven areas of qualitative risk assessment and then move on to understand the purpose and process of the loss run analysis. Learning Objectives By the end of this section, you will be able to: 1. Understand the purpose of qualitative risk assessment and how it differs from quantitative risk assessment. 2. Identify seven areas of qualitative risk assessment. 3. Understand the purpose and process of loss run analysis. 4. Understand how to establish analysis objectives. 5. Identify and discuss various methods of assessing the quality of loss data. 6. Know how to effectively sort and organize loss run data using meaningful categories. 7. Understand the value of data stratification, tables, and graphs. 8. Know how data is used to provide answers to meaningful questions. 9. Discuss how the relationship between frequency and severity is used in determining risk management priorities. Quantitative vs. Qualitative Data Analysis There are two types of data analysis: quantitative and qualitative. Note: This section discusses qualitative data analysis. Quantitative data analysis is discussed in Section 3. Quantitative What people normally think of when they hear the term data analysis. It is the analysis of numerical data, such as the number of back injuries or the cost of workers' compensation claims. This type of analysis deals with numbers, charts, and graphs anything that is measurable. Qualitative The analysis of descriptive data, i.e., data that can be put into words but not necessarily into numbers. Collecting qualitative data includes interviewing employees, talking to other risk managers and experts, and assessing the politics of the school board and community. It relies on judgment, insight, reflection, and experience. 17

18 It is important to understand the difference between qualitative and quantitative analysis. Quantitative Risk Analysis Is concrete Uses numbers Answers the questions: How many? How much? Can be precisely measured Qualitative Risk Analysis Is descriptive Usually uses words Answers the questions: Who? What? When? Where? Why? Cannot be precisely measured What tasks are associated with qualitative and quantitative risk analysis? Qualitative Risk Analysis Review incident reports Contract analysis Safety audit review Policy review Quantitative Risk Analysis Claims development Indexing for inflation Loss pick calculation Ultimate total loss information QuaLitative = Language QuaNtitative = Numbers 18

19 Qualitative Risk Analysis: Is used to understand aspects of loss exposures that are not easily measured by traditional data analysis and statistics. Is heavily influenced by experience. Benefits from the help of subject matter experts, supervisors, experienced employees, and consultants. This is especially helpful for inexperienced school risk managers. Example: The number of trips, slips, and falls experienced in a school in one year is a quantitative measure. Sorting by time of year and day of week is qualitative. Seven Key Areas of Qualitative Risk Assessment in Schools There are seven qualitative risk assessment areas to consider when you perform risk analysis "detective work." Organization's Appetite for Risk Social Responsibility Citizenship Analysis Contractual Analysis Compliance & Regulatory Analysis Security, Health, & Safety Reviews Internal Policies Review of Other Programs or Issues 19

20 The Organization's Appetite for Risk A school's willingness for risk is strongly influenced by its: History of loss-related incidents Long-term objectives Appetite for risk versus financial ability to tolerate risk Financial stability (tax base, annual giving, fundraising, endowments) Access to insurance, including the cost of insurance Exposure to catastrophic events (natural or man-made) Internal politics School Districts tend to be very risk averse. For example: The fiscal year for schools in a state ends August 31 and begins September 1. Property taxes are paid in January. For these school districts, four months remain after the doors open for the fall semester before their operating funds are replenished. In some cases, they may have to arrange for a short term, tax anticipation note to cover expenses. An environment such as this does not encourage the district to take on costly new projects or accept costly risks. 20

21 Social Responsibility and Citizenship Analysis School districts are inherently in the business of teaching social responsibility and citizenship. Understanding how your school district addresses social responsibility and citizenship issues can help you better control and fund your school's risks. When planning risk management strategies, ask yourself: What are the risk implications if the school district fails to meet its social responsibilities? What if the school district is perceived as being uncommitted to community citizenship? What are the risks to the school district if it addresses certain controversial social issues? Consider how the school district's goals mesh with the social standards of the larger community. Schools are expected to be both the scientist and the experiment when it comes to social responsibility. Imagine the financial impact of the desegregation orders of the 1960's, or the recent Supreme Court case concerning personal searches of students suspected of drug possession. Business managers and risk managers must be alert to new external initiatives that affect reserves, budgets, or even the organization s strategic plan. 21

22 Contractual Analysis Contracts should be reviewed to determine whether clauses in them are appropriate for the school district and to understand the impact that the contract may have on the school district s risk profile. Review state regulations to verify how contracts must be handled - i.e., must certain contracts be put out for competitive bid? For example, review contracts to ensure that they do not contain hold harmless clauses* review insurance contracts for coverage gaps or inappropriate property valuation. *Schools are not permitted to hold another party harmless. However, it is permitted for another party to hold the school harmless. Hold harmless agreement: A portion of a contract that is a promise by one of the parties (first party) not to hold the other party responsible if the other party carries out the contract in a way that causes damage to the party. 22

23 Compliance and Regulatory Analysis Understand how governmental and other regulations may impact your school: Federal, state, and local regulations and administrative procedures Voluntary regulations Violations or fines Regulation enforcement history Regulations are contributed from the federal level all the way down to the municipal and county level, from building codes to teacher/student ratios. An excellent example is the emergence of laws requiring school organizations to submit a security plan and threat assessment to the state legislature every 3 years. Security, Health, and Safety Issues Reviewing school security, health, and safety programs can yield important insights into which areas of risk require the most attention. Review and analyze: Nurse case management reports Loss control reports Inspection reports Accident investigation reports Safety training materials Audits of the school's security plans, safety programs, and ergonomic equipment programs Other in-house documents related to health and safety programs and training materials An example of analysis in this area might be following up on a pattern discovered during the analysis of claims data. After determining that the trend is real, can we figure out what is causing it? 23

24 Internal Policies School risk managers play a pivotal role in improving school policies. Review your school's internal policies with an eye towards improving them. Review policies related to: Audits and oversight o Internal o External o School board involvement Employment issues o Contract employees o Seasonal employees o Educators' legal liability Regulation-based internal policies o Safety and evacuation policies o Purchasing policies - Hazardous chemicals, playground equipment, furniture ergonomics Internal policies can affect internal politics. Proactive risk managers are better positioned to handle the politics of the school community, from the auditors to the local newspaper editor. Why? Because they have identified the organization's internal policies, they are making sure the policies are correctly implemented, and they are continually striving for better performance. Other Campus Programs or Issues Other qualitative risk factors to analyze include: Job descriptions Recruitment and retention of the school's workforce Employee productivity Facilities and grounds programs Why review the organization's job descriptions? Can you name the logical classification of risk that exposures in this area would fall into? The answer is Human Resources, a very large exposure classification for schools. Making sure that job descriptions are accurate and applied equally to all employees is necessary to reduce the risk of an employment practices claim. 24

25 Review of the Seven Key Areas of Qualitative Risk Assessment 1. Organization's Appetite for Risk 2. Social Responsibility 3. Citizenship Analysis 4. Contractual Analysis 5. Compliance & Regulatory Analysis 6. Security, Health, & Safety Reviews 7. Internal Policies 8. Review of Other Programs or Issues The Loss Run Analysis Purpose and Process The purpose of a loss run analysis is to apply various methods of assessing loss data to understand the impact of the losses on the school district's risk management program and the ultimate cost of risk. A loss run analysis also helps with prioritizing loss control programs. Below are the stages of the loss run analysis. 1. Set Analysis Objectives: Decide which questions you want your analysis to answer, then set your analysis goals. 2. Determine Data Quality: Make sure that your data is accurate and relevant to your goals. 3. Organize and Sort Data: Try rearranging your data in different ways to uncover meaningful patterns. 4. Visualize Data (chart and graph): Use data stratification, tables, and graphs to further emphasize patterns in the data. 5. Interpret Data: Review the data to identify and analyze meaningful patterns. 6. Write Report: Incorporate the data into a report that describes the school s losses and risk. 25

26 Step 1 - Set Analysis Objectives The Loss Run Analysis Process The goal of loss run analysis is to answer questions. What are the objectives of your analysis? Retrospective Objectives: Designed to uncover general patterns and trends in historical data to answer questions about the present. Sample Questions How did the previous increase in faculty impact last year s losses? What have been the most significant causes of losses over the past year? Discovery Objectives: Possible causes underlying general patterns and trends (questions about the past). Sample Questions Which activities are causing the most injuries? Which types of injuries are the most expensive? Prospective Objectives: Designed to uncover unknown or hidden exposures, concerns, and issues (questions about the future). Sample Questions Do patterns in the current year s data reveal new or changing exposures that need to be addressed? How will adding a new building to the school impact losses? If our risk manager, Roger, wanted to do an analysis of his workers compensation injuries to check on the numbers of accidents by location, his analysis objectives would be discovery objectives. If he wanted to find out trends in injury types and number of accidents, his objectives would be prospective. An analysis of the impact of adding an additional building to a campus in the school district on the # of injuries would be. Answer: A prospective analysis objective 26

27 Step 2 - Determine Data Quality You must determine the quality of your loss data in order to make accurate projections of loss trends for your school. There are four methods of assessing the quality of loss data: 1. Completeness 2. Consistency 3. Integrity 4. Relevance Data Quality Completeness In order to maximize the accuracy of loss trend projections, review your loss data to determine its completeness. Having data that is complete simply means having enough data to perform a meaningful analysis. Only complete loss data should be used. The requirements for complete data are having: At least five years of data; preferably 10+ years. Enough data about each claim or accident to make the analysis meaningful. An understanding of what is included in the paid and reserve amounts. Reserve amounts should include ALAE*, IBNR*, and defense costs. Allocated loss adjustment expense (ALAE): Includes all expense directly assigned to or arising out of a particular claim any expense assigned and recorded directly to a particular claim. Examples of ALAE include court fees and the expense of outside legal counsel. May also be noted as allocated loss expense. (CSRM Glossary) Incurred but not reported (IBNR): Represents the liability for unpaid claims not reflected in the case reserve estimates for individual losses. The two components of this liability derive from additional development on known cases and the reporting of claims that have occurred but not yet been recorded as of the evaluation date. *ALAE and IBNR are discussed at greater length in Funding School Risks. 27

28 Step 2 - Determine Data Quality continued Data Quality - Consistency Design your analysis so that you are looking at data units that are consistent with one another. Consistency means the same type of data for each type of claim or accident over a period of time. The same types of data should be provided for each claim. Note that if the claims data is reported and recorded in different ways, you will have a consistency problem when you collect it for analysis. The data should be grouped consistently. There should be no change in the policy year, accident year, or calendar year. Placing data in the wrong group can skew, "or throw off," your results. To understand the difference between data completeness and data inconsistency, introduce some errors into the chart shown. Staff Injuries (Example of data that is incomplete) Years Type of Injury Number of Injuries Back Strain 1, * Wrist Strain Neck Strain 2 *FY 1997 injury records not available for Campus A Staff Injuries (Example of data that is inconsistent) Years Type of Injury Number of Injuries Back Strain 1, Wrist Strain Neck Strain 2 Example: Another example of a consistency problem would be a change in the dates of the fiscal year. 28

29 Step 2 - Determine Data Quality continued Data Quality Integrity Integrity means accuracy. In order to get data integrity under control, the personnel who are recording the data should always follow standardized methods set forth by the organization. These methods must not change when a new employee takes over the task. If recording procedures are consistently accomplished and considered to be reliable, the data should then go through a final check for accuracy before it is developed. The following are the basic requirements for data integrity: 1. Data must be promptly reported so that it is more likely to be accurate. 2. Data must be recorded accurately and reliably, and checked after input. 3. The figures for allocation of reserves must be reviewed for accuracy prior to analysis. An important factor in data integrity is the Third Party Administrator (TPA). Make sure when changing to a new TPA that data collection procedures do not change. Third party administrator: A claims administrator or insurance company that processes claims on behalf of a self-insured organization. Data Quality Relevance Loss data is relevant if there is a connection between the data and the matter/question at hand. To ensure that your data is relevant, be careful with each the following situations: Data from Discontinued Operations: Do not include data from operations and activities no longer a part of the school district. Data from Acquired Operations: When adding an operation, include loss data from the portion of the acquired operation only. Irrelevant Data: Do not include extraneous data even if it is consistent, accurate, and timely. Commingling of Data: Do not combine data from diverse operations when trying to analyze losses. If data from one operation is not relevant to the other, the frequency, severity, types, and causes of losses will not be consistent. 29

30 Step 2 - Determine Data Quality continued "Determine the most common injury among staff members during the 10 year period 1990 to 2000." Will this chart provide the information for the following analysis objective? Staff Injuries Years Position Number of Injuries Administrator Maintenance Employee Teacher 532 Answer: The answer is No. This data could be used to compare injury rates among different staff positions. However, the analysis being performed is a comparison of different types of injuries. While the data in this example may be useful for answering a different question, it is not useful in this particular analysis. "Determine the most common injury among staff members during the 10 year period 1990 to 2000" Why is this considered high-quality data? Years Type of Injury Number of Injuries Back Strain 1, Wrist Strain Neck Strain 84 Column 1 Column 2 Column 3 Column 1: Completeness - There is enough data (10 years) to perform a meaningful analysis. Consistency - The three categories cover the same time period and report on the same type of data (injuries). Column 2: Relevance - The data answers the question posted by the loss run analysis objective, i.e., Which type of staff injury was most common from ? Column 3: Integrity - The actual data used has been thoroughly checked for errors prior to development. 30

31 Step 3 - Organize and Sort Data Most data needs to be grouped and sorted in logical ways before it can be properly analyzed. Raw loss data is typically submitted chronologically. Categorizing and separating data in different ways will help you discover patterns. Patterns, or possible correlations, do not necessarily indicate a cause, or even a problem. However, the risk manager does want to find them and subject each to analysis. The insurance carrier or TPA should send loss runs to you in an electronic format. Note: If a paper report is all you have, the loss runs should be transferred onto a spreadsheet program so that data can be manipulated and sorted. Listed below are effective ways to sort and organize data into categories. Organize data by determining category criteria. Sort data according to the logical sequencing of your criteria. Sort data by categorization and separation. Common category criteria for organizing and sorting loss run data include: Last name - identifies accident repeaters Date of loss Length of employment when accident occurred Type of loss - strain, sprain, fracture, etc. Cause of loss - lifting, slip/fall, hit by person/object, trimming tree, etc. Affected body part - back, shoulder, leg, arm, etc. Large losses - regardless of the cause or type Department or location of loss Any other criteria gleaned from the loss runs Sorting and calculations are accomplished with spreadsheet software. The most prominent is Excel. See the Appendix for more discussion regarding the use of Excel. 31

32 Step 4 - Visualize Data Organizing data into appropriate groups may not be sufficient to determine the real issues. Creating visual representations of the data may allow you to more clearly identify areas of concern. This can be accomplished by applying the following techniques: Stratify data Organize data into a Table Graph data Data Stratification: Stratifying data means grouping it into categories that emphasize trends. Effectively displayed data tells a story. If the story is not clear, you may not have the most relevant or meaningful relationships represented. Try sorting the data differently, then graph it again to see if the "story" is clearer. Stratifying Data Sometimes a data set contains logical breaks that can be separated into meaningful groupings. Stratifying data means grouping it into categories that emphasize trends. For instance, look at the table that lists staff injuries per 100 employees. The data has been stratified (grouped) to expose patterns. Staff Injuries in 2008 Length of Employment # Injuries 0 4 years years years years 56 *Per 100 employees What pattern(s) do you see? What might this indicate? There are larger numbers of injuries per 100 employees when length of employment with the district is between 5 and 15 years. The risk manager will be interested in investigating this possible correlation between injury rate and length of employment. 32

33 Step 4 - Visualize Data continued Tables Clear, readable tables can help you emphasize key relationships and recognize patterns more quickly. Keep your tables simple. Tables that attempt to present large amounts of information at once can be overwhelming to read. Table 1 Table 2 Staff Injuries in 2008 Length of Employment # Injuries 0 4 years years years years 56 *Per 100 employees Staff Injuries in 2008 Length of Employment # Injuries 0 1 year years years years years years years years years years years years years years years years years years years years 6 *Per 100 employees 33

34 Step 4 - Visualize Data continued Graphs Graphs and charts can also expose patterns and make data trends easier to recognize. Spreadsheet software simplifies the creation of graphs. For example, the graph below was created from the data set on the previous screen. 34

35 Step 4 - Visualize Data continued The staff injury data is graphed three different ways below Bar graph Pie chart (or circle graph) Line graph 35

36 Step 4 - Visualize Data continued Which type of graph displayed the data on the previous page most logically? The bar graph works best for this type of data. Along the x-axis, it emphasizes the chronological order of the categories (the number of years employed). Along the y-axis, the height of the bars indicates the quantity of losses per category. The pie graph doesn t emphasize the chronological order of the categories (the number of years employed), although it does clearly indicate the quantity of losses per category. Pie graphs are best for displaying data sets in which the categories are not in chronological order. For example, you might use a pie graph to show the number of injuries in different departments. Line graphs are best for displaying continuous data, such as seasonal variation in temperature. They are especially effective for graphing data sets that include many data points. Since the data in this example is grouped into discrete categories (i.e., is not continuous) and has relatively few data points, it is not a good candidate for a line graph. 36

37 Step 5 - Interpret Data Imagine that you have access to a large amount of data covering years of injuries and accidents at your school district. You have both qualitative and quantitative data. Your data is high quality - it is complete, consistent, has integrity, and is relevant. You have examined the qualitative data, taken notes on it, and consulted with employees and stakeholders. You have looked at the data in various ways: stratified it into groups, created tables, and graphed it. Now that you have a clear picture of your data, the next step is to ask questions about it. The risk manager needs clear, data-based answers to his analysis questions: 1. What activity causes the most injuries? What types of injuries cost the most? Are any trends apparent? 2. Who are the accident repeaters? Where and when are most of the accidents occurring? 3. What are the priorities and focus of the loss control programs? 4. Can relevant baselines be established from the data? (Baselines can be used to benchmark against yourself in future years.) 5. How will the result of this analysis affect the school district s appetite for risk? 6. How is the loss experience important to the school district risk manager and the administration? Baseline: Initial data that is used for comparison with subsequently acquired data. Benchmarking: Measurement of the quality of a district s policies, products, programs, strategies, etc., and their comparison with standard measurements, or similar measurements of the best-in-class districts. The objectives of benchmarking are (1) to determine what and where improvements are called for, (2) how other firms achieve their high performance levels, and (3) use this information to improve the firm's performance. 37

38 Step 5 - Interpret Data Continued The frequency and severity characteristics of exposures are the driving factors in prioritizing activities and allocating resources within the risk management program. Why? The principle is this: The frequency and severity characteristics of an exposure indicate the control and financing mechanisms which will be applied to it. For example: A risk that has low frequency and low severity will likely be retained by the school. For cuts and scrapes on the playground, the school will keep a first aid kit, train teachers, and enforce playground safety rules. A risk that has high frequency but low severity might become the subject of insurance, and will certainly become the target of a prevention campaign. You can see from these two examples that the nature of the exposure influences the risk manager's choices of how to manage the exposure. Note: The CSRM Handling School Risks course covers this topic in greater detail. 38

39 Step 5 - Interpret Data Continued Having analyzed the frequency and severity of a given peril, the risk manager can use this chart as a tool to help determine the best way to address it. Frequency = number of claims or claim count Severity = dollar value of claims or lost time days Once you evaluate the frequency and severity of the interpreted data, you should ask yourself questions to help determine your school's priorities: 1. Does the school district have risks so low in frequency and severity that retaining them is an option? 2. How would you handle risks of slight or moderate frequency and high severity that are difficult or expensive to reduce or prevent? 3. How would you handle a high severity, high frequency risk that cannot be avoided? 39

40 Step 5 - Interpret Data Continued The Pareto Principle: The "80/20 Rule" Typically, the majority of claims are the result of a small number of causes. It is common to find that about eighty percent of claims come from twenty percent of causes or claimants. The phenomenon occurs in many areas of life. It is called the "Pareto Principle," after the Italian economist Vilfredo Pareto, who observes that 20% of the people in Italy held 80% of the wealth. You may have heard reference to this as the "80/20 Rule." The Pareto Principle is a powerful tool for interpreting and prioritizing data. If your loss data shows that this is the case, you can significantly reduce your cost of risk by focusing extra effort on the most costly 20% of causes. Step 6 - Write Risk Report The final step in the Loss Run Analysis Process is to communicate the results of your analysis to those involved in the decision-making process. These people may include the school administration, your risk management team, the superintendent, etc. You do this by writing a risk report. A risk report can be a formal written document, a PowerPoint presentation, or another communication method. The document should be configured to meet the needs of the intended audience. Please refer to the end of Section 2 to go over the section exercises. 40

41 Section 3: Quantitative Analysis: Tools and Forecasting Section 3 is about the numbers or quantitative risk analysis. This section provides an overview of tools used to develop loss data, index data to inflation, and estimate a loss pick for a future period. Learning Objectives By the end of this section you will be able to: 1. Understand the purposes of quantitative analysis tools. 2. Know examples and purposes of common tools used for quantitative analysis. 3. Estimate subsequent years losses given an obvious trend. 4. Calculate or determine the three measures of central tendency mean, median, and mode. 5. Understand how to calculate the range of a group of data. 6. Understand how losses are predicted with a certain degree of confidence. 7. Understand how past loss data may change over time as additional claims or costs are realized and how development factors are used to account for these changes. 8. Decide when it is appropriate to use a development factor derived from school loss data instead of industry loss data. 9. Understand the impact of inflation on loss data and how to interpret data indexed to inflation. 10. Understand how indexing to exposure bases provides more accurate loss projections. Purpose of Quantitative Analysis Tools The analysis of loss and exposure data ultimately provides the information needed to allocate resources. Remember, the risk manager's objective is to protect the assets and enable the strategic plan of the school organization. Determine loss picks and assist in forecasting losses for subsequent years. Example: Loss Trends Our risk manager, Roger, intends to have an estimate of his workers comp injuries for next year. He also will "develop" claims amounts that have already happened to demonstrate the true cost, in present dollars, of claims to the school district. 41

42 Determine retention, deductible, and transfer levels for decision making. Example: Retention & Transfer Decisions While examining the district's vehicle damage records over the past 10 years, Roger noticed that there were no losses less than $1,000 and no annual losses greater than $50,000. As a result, he was able to save the district money by selecting vehicle insurance coverage with a $1,000 deductible and a $50,000 aggregate stop loss. Compare funding programs for cost effectiveness. Example: Cost Effectiveness of Funding Programs Through analysis of the vehicle losses, Roger had chosen to retain all vehicle damages that were less than $1,000 and to transfer to the insurance company and aggregate annual losses that exceeded $50,000. Determine which projects to fund. Example: Funding Choices The risk manager was given a choice to fund only one of three potential projects due to limited availability of funds. By conducting cost-benefit analysis, Roger was able to determine that by investing $70,000 in a sprinkler system at the high school, the district would recover the installation costs within a five year period by realizing savings in insurance premiums during that time. Determine appetite for risk. Example: Appetite for Risk The superintendent of schools has asked that Janet, the risk manager for the district, take every possible action in order to assure that no district funds would be paid out in litigation settlements stemming from actions of the school board. Consequently, Janet decided to purchase a comprehensive Director's and Officer's policy for the district with several additional coverages and features at a higher premium than other available coverages. 42

43 Statistical Tools When you analyze your school s loss data to find trends, you are using statistics to analyze and make predictions about groups of people. The scope of this course will include an overview of some of these tools, but not a detailed study of statistical analysis. In many cases, loss data is prepared by actuaries or insurance companies on behalf of the school. However, the school risk manager should have basic understanding of the statistical tools that are used. Common Tools Used in Quantitative Analysis These tools are often combined, depending upon the analysis project. Combining methods should result in improved decision making. Several tools are available to risk managers for risk analysis. Often, a combination of the following is more effective in displaying, discussing, and making decisions regarding the district's mitigation efforts. 1. Graphs used to establish and visualize trends and identify anomalies. 2. Measures of central tendency used to forecast losses and make decisions on retention/deductible levels; they should always be used in conjunction with other tools. 3. Measures of dispersion and confidence intervals used to forecast a range of losses and make decisions on retention/deductible levels. 4. Development factors, inflation index factors, and exposure bases used to estimate indexed ultimate total loss allowing for more accurate comparisons from year to year. Graphs By expressing data in a line graph, pie chart or bar graph, it often makes it easier to recognize patterns. Sometimes it is practical to use the "eyeball" method of trending. In the "eyeball" method, data is plotted to visually determine the position used to estimate an unknown future event, such as the forecasted frequency of losses for next year. 43

44 Below are claims count graphs for two school districts, Jumping Jack ISD and Smoothly ISD. When you look at the data, you can see why the districts are named this way. For each graph, consider whether you would feel comfortable predicting the losses for year 11. Do you feel that you can predict this by looking at these graphs? Answer: Jumping Jack ISD reveals the unpredictability of its claims when viewed over a 10 year period. There is no obvious trend, making a prediction of next years' losses (X11) highly speculative. Answer: Smoothly ISD, on the other hand, has had a predictable increasing number of claims over the 10 year period. The annual increase in claims has increased so steadily and predictably that the obvious trend could be used to predict next years' # of claims (X11) with a fairly high degree of confidence. 44

45 Measures of Central Tendency Measures of central tendency seeks to determine how the values in a data set tend toward a middle value. The three most common measures of central tendency are mean, median, and mode. In other words, Looking for the middle of a data set Mean (average value) There are three commonly-used measures of central tendency, mean, median, and mode. First, we will define mean (average): There are several variations of mean that can be used to calculate central tendency depending upon the characteristics of the data. The arithmetic mean, used for our discussion purposes, is simply the sum of all values in a data set divided by the number of data items n. We will limit our examples to arithmetic mean only. Greater discussion of various types of mean is conducted in the Certified Risk Manager Analysis of Risk class. ( ) 5 = 24 (Mean) 45

46 Median Median is the middle point of a data set. Determining the median requires that the data first be sorted from highest to lowest or lowest to highest. In a data set containing an odd number of data items (for example, 5 items) there is a precise mid-point. For data sets containing an even number of data items (for example, 4 items), there is no precise middle point. Therefore, the two data items in the middle must be added together and divided by 2 in order to establish the median. Data set with off number of values: 10, 10, 20, 30, 50 Data set with even number of values: 10, 10, 30, 50 = = 40 / 2 = 20 Mode Mode is the most frequent value. There may be no mode within a data set or more than one mode within a data set. While reviewing the workers compensation payments for employees last year, the risk manager noticed that the mode payment for back strains was $500. By further analyzing the items comprising the mode, it was determined that the injured employees were treated by the same doctor who used the same billing amount for all injuries. This value is often used as a trigger for further investigation, as in the workers compensation example above. What is the mode of the following data set? 10, 10, 20, 30, 50 A. 20 B. 50 C. 10 D. 30 The correct answer is

47 Measures of Dispersion: Range & Outliers Measures of dispersion are used to determine how data is distributed. They show how far data is from some measure of central tendency, such as the average. Range and standard deviation are two common measures of dispersion Range is the difference between the smallest and the largest number in a data set. Standard deviation measures how far items in a data set vary from the average or mean. The distance can be less than the average (negative standard deviations) or more than the average (positive standard deviations). Range: 3 1 = 2 47

48 Range The range of a group of data is the difference between the largest and smallest numbers. For example, if the largest possible property loss for a school is $500,000, the range of possible losses is $0 - $500,000 or $500,000. What is the range for Jumping Jack ISD below? For Smoothly? # Employee Injuries during Policy Period Policy Year Smoothly ISD Jumping Jack ISD (10 Years Ago) (Last Year) Total Jumping Jack ISD: = 148 Smoothly ISD: = 90 48

49 Knowing the range can provide insight into how widely the data is dispersed. However, a value within the data set may be an "outlier." An outlier is a value which lies outside the general pattern established by the remainder of the data. When not accounted for, an outlier can skew the distribution of the entire data set and can lead to misinterpretation of the data set and misapplying the data in estimating next year's loss pic. Which number is the outlier? Answer: 105 FYI Outliers Outliers are adjusted for using a variety of statistical methods. The scope of this course includes only the definition of the term. Statistical calculations used by the risk manager are presented in further detail in the Certified Risk Manager Program. 49

50 Measures of Dispersion: Standard Deviation & Confidence Intervals Standard Deviation Standard Deviation (S) measures how far data points are from a central tendency such as the mean. Understanding standard deviations can assist the risk manager in understanding data and making predictions about unknown future events (i.e., loss pic). You will notice that the graph on the left has less spread and a smaller standard deviation than the graph on the right. Note: Standard deviation is used to determine confidence intervals, which your agent or broker will use to determine your deductible or retention. You should know the meaning and purpose of a standard deviation so that you can understand the process that your agent is using. You are not required to know how standard deviations are calculated. Confidence Interval Standard Deviation is useful when trying to determine a confidence interval. A confidence interval is the range of values within which we think a predicted value will fall. The probability that our prediction will land within the range we've calculated is expressed as confidence, such as 95.4% confidence. Example: The insurance company predicts what losses will be for next year with 95.4% confidence. The prediction is expressed as a range of numbers. The insurance company needs to calculate the standard deviation for the data set. This is a measure of how all the values are dispersed around the mean value of the data set. Recall that the mean is the average value (in this course we use the arithmetic mean). In normal distributions, actuaries have learned that the following intervals contain the stated percentage of values: 50

51 When we say that we are 95.4% confident that our loss pic will be between one number and another (low/high) we are using ± 2 Standard Deviations to estimate the low (-2S) and high (+2S) numbers. ± 1S means plus or minus 1 Standard Deviation. Mean ± 1S 68.2% of the time Mean ± 2S 95.4% of the time Mean ± 3S 99.7% of the time For our example, the high and low end values corresponding to +2S or -2S from the mean determine the high and low values of the range into which your prediction should fall. We've depicted this visually below. A vertical black line on this graph represents 1 standard deviation from the mean. Notice that 95.4% confidence is shown as having a correlation with ± 2 Standard Deviations. According to this graph, what is the range of values falling within +2 or -2 standard deviations of the mean value of 350? If the costs are normally distributed, (in other words they match the curve shown to the left) the 1,000 losses would be distributed as noted below. Your school district had 1,000 losses in a particular exposure category last year. The average cost of the losses was $350. Looking at the data, you find 95.4% of the losses are between $50 and $

52 In this example, we ve used the concept of a confidence interval to describe a known data set, last year s losses. Confidence intervals have their greatest value in allowing a prediction of an unknown or future data set, such as next year s expected losses. To summarize this concept, we can use the example of the insurance company's estimate of a school's losses for next year. The insurance company has a lot of data about the organization's losses, the average, the range, the mode, etc. In making this estimate, an actuary would determine how values within a certain data set are distributed around the mean or average value. Standard deviation and mean are two calculations that will be needed to make this prediction. The result will be a range of values within which the estimate should fall. See the Appendix to review an example of how standard deviation would be calculated for 10 years of developed claim count data. The information in the Appendix is for informational purposes only, you will not be tested on it. 52

53 Raw Loss Data In this section we are going to apply development factors, inflation indices, and exposure bases to raw loss data. The purpose of doing this is to estimate losses for a future period. Click on each step of the Loss Data Development Process. By the end of this section, you will understand each step. Incurred (Known) Loss Data Incurred losses are: (1) the total amount of paid claims and loss reserves associated with a particular period of time, usually a policy year. Generally, incurred losses are the actual losses paid and outstanding, interest on judgements, expenses incurred in obtaining third party recoveries, and allocated loss adjustment expense for employers liability losses. (2) paid losses, case reserves, and IBNR reserves until ultimate incurred losses are reached, at which time there is not remaining IBNR. Development Factors Development Factors are ratios applied to losses to account for potential ongoing costs as the losses develop over time. Inflation Indices Losses are indexed for inflation to account for changes in purchasing power of the dollar over time. Exposure Bases An exposure base is applied to losses to account for changes in units exposed to potential loss, such as number of buildings, number of buses operated, etc. Differentiating Between Raw and Developed Data Loss data is described as being either raw or developed. Raw loss data comes from the loss runs the organization receives quarterly or annually. When trying to predict future losses it is only the starting point. Developed loss data is data which reflects the application of development factors, inflation indices, and exposure bases. This allows the risk manager to calculate loss trends to enable more effective predictions of future losses. 53

54 Applying Development Factors What are Development Factors? Development Factors are ratios determined by calculating how historical loss data has changed from year to year over the past several years. Ratios are applied to a current valuation of losses to determine an estimate of ultimate incurred losses. It is assumed that current losses will be paid according to the same pattern as prior losses at similar states of development. Development factors are frequently calculated separately for incurred losses, paid losses, and claim counts. Why do Claims have to be Developed? Many types of claims, such as workers compensation claims, are paid out over long periods of time. Often, these payouts occur in different calendar and fiscal years. Additionally, some claims are not reported during the year when the loss occurred. These two characteristics of claims development are called Incurred but Not Reported (IBNR) and must be tracked in order to assure that sufficient funds are available to pay the claim as an expense occurs. How are Development Factors Calculated? The process of calculating a development factor is called triangulation because the data forms a triangular shape in a spreadsheet. Triangulation is not within the scope of this course. An example of triangulation is included in the Appendix. The information in the Appendix is for informational purposes only, you will not be tested on it. 54

55 A claim's full financial impact on the organization develops over time. Claims are said to have tails. A claim s tail refers to how long it takes for the claim to develop (come to a close). 55

56 Step 1: Start With Raw Loss Data Roger's insurance agent, Bill, sent loss data for the school district s workers compensation claims. Recall how we define incurred losses: Incurred (Known) Loss Data Incurred losses are: (1) the total amount of paid claims and loss reserves associated with a particular period of time, usually a policy year. Generally, incurred losses are the actual losses paid and outstanding, interest on judgements, expenses incurred in obtaining third party recoveries, and allocated loss adjustment expense for employers liability losses. (2) paid losses, case reserves, and IBNR reserves until ultimate incurred losses are reached, at which time there is not remaining IBNR. Year Total Incurred Losses 1999 $371, $367, $360, $335, $323, $309, $298, $217, $194, $164,986 56

57 Step 2: Apply Development Factors Roger needs to estimate what the ultimate cost of his known losses will be for next year. The ultimate cost of his losses is larger than the raw data indicate, because IBNR, losses that have been incurred but not yet reported, are not accounted for in the raw loss data. He will use development factors in order to complete this step of the loss trend analysis. Year Total Incurred Ultimate Losses Total Losses 1999 $371,200? 2000 $367,200? 2001 $360,400? 2002 $335,400? 2003 $323,000? 2004 $309,100? 2005 $298,580? 2006 $217,300? 2007 $194,630? 2008 $164,986? Roger applies development factors to each year's data. Development factors are multipliers. We look closer at a few entries on the next page. Year Total Incurred Development Ultimate Losses Factor Total Losses 1999 $371, $404, $367, $422, $360, $443, $335, $489, $323, $552, $309, $584, $298, $641, $217, $686, $194, $741, $164, $755,636 57

58 Here is a closer look at the previous table. The ultimate total losses have been calculated for the first three years. Do you see that the most current years show a higher value for the development factor? Year Total Incurred Development Losses Factor Ultimate Total Losses 1999 $371,200 X 1.09 = $404, $367,200 X 1.15 = $422, $360,400 X 1.23 = $443, $335, $489, $323, $552,330 Why Do Development Factors Decrease With Time? The development factor applied to 1999 is lower than the development factor applied to the most recent data, in our example, the data from Why do you think development factors decrease with time? Year Development Factor Answer The newest claims have only begun to develop. They must be multiplied by a high number to account for the future costs that they will incur in the future. The oldest claims are closed or almost closed. The ultimate losses are almost the same as the incurred losses. When the development factor reaches 1.00 (100%), it can be safely assumed that the claims for that period have all been paid (fully developed.) 58

59 Using In-House vs. Industry Development Factors You can use development factors calculated from your school s loss data when you have at least five years of relevant data (preferably 10+ years). If your school s loss data does not span a sufficient number of years, you will need to use industry development factors those derived from a composite of many schools loss data. Sources of industry development factors Rating bureaus Insurance Service Office, National Council on Compensation Insurance, school associations, etc. Consultants Actuaries, accounting firms, etc. Other Conning & Co., A. M. Best, insurance pools. Some potential problems with using industry development factors Industry factors may be too broad. They may include junior colleges, universities, and private schools, as well as public K-12 schools. The data from industry factors may be unreliable or irrelevant to your school. The classifications within industry factors may not apply to your school Notice the difference between ultimate total losses (developed losses) and total incurred losses (undeveloped). Using the undeveloped data would lead one to believe that losses are growing smaller when, in fact, the developed data makes it obvious that the losses are growing and we will need to acquire more funds for payment of future obligations. 59

60 Step 3 Indexing Losses to Inflation Inflation A persistent, substantial rise in the general level of prices related to an increase in the volume of money and resulting in the loss of value of currency. Because of inflation, a loss settled in 1996 for $5,000 would settle for much more in 2006; consequently, to be able to tell if the school district's cost of losses are decreasing relevant to pricing indices, the losses must be "indexed" or adjusted for inflation. Historical dollars must be inflated to account for the reduced purchasing power of today's dollar. Doing so assures that our severity loss pic (which was based upon historical dollar trends) is sufficiently funded in today's dollars (Present Value). Different Inflation Rates for Different Types of Goods Inflation rates are different for different types of goods. For example, the inflation rate for medical expenses may be higher than the inflation rate for construction expenses. Food inflation may be higher than either of these. Inflation of oil prices may be higher than all of these. The Consumer Price Index (Bureau of Labor Statistics) is often used to make general adjustments for historical purchasing power of the dollar. 60

61 The table shows Jumping Jack ISD's ultimate total losses adjusted for inflation. Year Ultimate Total Losses Inflation Index Factor 7% Indexed Ultimate Total Losses 1999 $404, $797, $422, $776, $443, $762, $489, $788, $552, $828, $584, $817, $641, $840, $686, $844, $741, $845, $755, $808,531 Here is a closer look at the previous table. Notice that as you go back in time, greater adjustments for inflation have to be made. The Inflation Index Factor column of the chart reflects the compounding effect of a 7% inflation rate over time (i.e. 7% x 7% x 7%). Inflation factors are multipliers just like development factors. However, instead of decreasing with age, they increase. Year Ultimate Total Losses Inflation Index Factor 7% Indexed Ultimate Total Losses 1999 $404, $797, $422, $776, $443, $762, $489, $788, $552, $828,495 61

62 Why Do Inflation Factors Increase With Time? In the previous example, the inflation factors increased with time. This is almost always the case. (The exception would be when the economy is not in an inflationary period, which is rare.) It might seem like the opposite would be true. However, there is a good reason that inflation factors decrease. Since inflation factors are used to counter the effects of rising costs, the factors must decrease the values (the buying power) of costs within the data set. This allows data from different years to be compared on a like-for-like basis. Year Inflation Factors Step 4 Indexing Losses to an Exposure Base In a changing environment, the prudent risk manager will index ultimate losses for future changes in exposure. For example, if the loss pic was based on developed losses for 1,000 teachers and the risk manager knows that 200 additional teachers will be hired next year, the loss pic needs to be adjusted to account for the increase in the number of teachers. Selecting an appropriate exposure base can help you adjust your loss data so it can be more effectively used to predict future losses due to changes in risk exposures. 62

63 Usually exposure bases are chosen from causes of loss or exposures. Exposure bases can be measured and quantified as follows: Headcount, (or segment of overall headcount, such as student body or faculty/staff, types of employee classifications.) Payroll Tax revenues Number of buses or miles driven Hours of facility usage (adding evening classes or leasing out to the Boy Scouts or to churches) Square footage of buildings or number of classrooms Others - geographical size, physical hazards Indexed Ultimate Loss Rate Indexing losses to an exposure base, gives you an indexed ultimate loss rate also called simply a loss rate. The example below uses payroll as its exposure base. In this example, a loss rate would be calculated per $1.00 of payroll. In the table below, indexed ultimate loss rates have been calculated for the first three years. Indexed Payroll Year Ultimate Total Ultimate Total Losses Losses 1999 $797,078 / $2,620,500 = $776,995 / $2,751,525 = $762,462 / $2,916,617 = $788,391 $3,033, $828,495 $3,154,

64 Indexed Payroll Year Ultimate Total Ultimate Total Losses Losses 1999 $797,078 $2,620, $776,995 $2,751, $762,462 $2,916, $788,391 $3,033, $828,495 $3,154, $817,879 $3,280, $840,951 $3,412, $844,602 $3,548, $845,356 $3,690, $808,531 $3,838, Column 2 Column 3 Column 4 $8,111,740 Total Losses $811,174 Average Annual Losses Column 2 represents Indexed Ultimate Total Losses losses after development factors have been applied and after adjustment for inflation (results from the three steps covered earlier in this section.) Column 3 represents the exposure base. For this example, payroll is used as the exposure base. Notice it changed from 2.6 to 3.8 million dollars over 10 years. Column 4 represents the Indexed Ultimate Loss Rate. This is calculated by dividing the Indexed Ultimate Total Losses (column 2) by the Payroll Exposure (column 3) resulting in a ratio expressed as a decimal. For example, in Year 1999, the Indexed Ultimate Loss Rate was 0.304, or the result of dividing $797,078 by $2,620,

65 Graphing Indexed Ultimate Loss Rate The Index Ultimate Loss Rate for the 10 year period is displayed below. This is the data from our chart on the previous page, column 4. Depicting this graphically will help show how the rate is trending. Indexed Ultimate Loss Rates (from Column 4) Year Ultimate Total Losses The graph indicates Jumping Jack ISD's loss rate has been declining over the last ten years. In other words, losses (developed and indexed for inflation) are declining relative to payroll. Average Indexed Total Losses From our table, the mean (or average) of Indexed Ultimate Total Losses for the last year is $811,174. Current payroll is $3,838,068. $811,174 $3,838,068 =.211 This loss rate will allow Roger to make his loss pic for the next fiscal year. 65

66 The Loss Pic Mike (District CFO): Have you worked with the claims data that we received this week? Roger (Risk Manager): Yes, I've already done my analysis. Take a look. The average cost of a claim is increasing. However, the complete picture is not bleak at all. Mike: I can see that. Your claim costs are increasing. But when you look at losses indexed to payroll, your loss rates have been decreasing. Roger: Since we have added a number of buildings over the years and increased the number of employees, I'm pleased to see loss rates continuing to decline. We ve really worked to continually improve control processes. Mike: That would explain why the developed losses are going up but the loss rates are going down. You and your team are doing a great job. Have you started working on the loss pic for next year? 66

67 Review the process that will allow Roger to make a "loss pic." Start with Incurred Loss Data The loss run analysis process itself starts with what is known: incurred losses. (However, the risk manager often makes his decisions with data that has already been developed, adjusted for inflation, and indexed to exposure base by consultants or insurance companies.) Develop Losses Developing losses means applying development factors to raw loss data. (This is often accomplished as a service of the insurance company or agency or by an actuary.) Development factors are multipliers (ratios) which account for the changes in the loss data over the period of time being analyzed. These factors are derived by a technique known as triangulation. Development factors are ideally derived from the school organization's own loss data. If there is not enough data (typically a minimum of 5 years) to calculate development factors, industry factors may be used. Index Losses to Inflation After applying the development factors to the loss amounts, it's necessary to adjust the developed loss amounts for inflation. This will result in values which may be termed Indexed Ultimate Losses. These are the amount which are indexed to an exposure base, to obtain a loss rate. Index Losses to an Exposure Base An exposure base is applied to losses to account for changes in risk exposures. By indexing to an appropriate exposure base, the risk manager calculates an indexed ultimate loss rate. This rate can be applied to the exposure base for the future period (next fiscal year) to calculate anticipated losses. 67

68 The Loss Pic Now, to Roger's loss pic. Recall that he analyzed data about employee injuries. Payroll or headcount could be considered a relevant exposure base, and we've chosen payroll for our example. Payroll during the ten-year period increased at a steady pace of 4-6 %. The district plans to open two new campuses next year with an anticipated increase in payroll of 20%. Opening two new campuses is likely to produce a change in risk exposures for the school district. All of these changes in the exposure base will be reflected in the loss rates produced by Roger's analysis. To calculate next year's loss pic using payroll as an exposure base, we first need an estimate of payroll for next year. Last Year s Payroll X 20% =? 3,838,068 X 20% = $4,605,682 Predicting Future Losses Using Developed Loss Data To calculate the projection of future losses (the loss pic) Roger will need an estimate of next year's payroll, the organization's current indexed ultimate loss rate. Human Resources is projecting a 20% increase in current payroll for the next fiscal year, or $4,605,682. Our average annual loss rate, from the previous page is.211. Next Year s Est. Payroll X Loss Rate = Loss Pic $4,605,682 X.211 = $971,799 Please refer to the end of Section 2 to go over the section exercises. 68

69 Section 4 The Risk Analysis Process Five factors influence a school's financial capacity, thereby affecting the risk manager's decision making. An overview of cost vs. benefit analysis is presented in this section, along with the concepts of present and future value and net present value. Learning Objectives By the end of this section, you will be able to: 1. Understand the financial and business environment in which the school district risk manager operates. 2. Understand how a school district risk manager uses cost benefit analysis and time value of money concepts to make financial decisions. 3. Understand the role of an actuary or insurance consultant in Risk Analysis activities. Financial Assessment We've studied the following types of analysis in this course: Qualitative Assessment Areas Qualitative analysis can be applied to a number of different areas that effect the risk management program, including the school district's appetite for risk, contractual analysis, safety and health issues, regulations, and internal policies. Loss Run Analysis Applying methods of organizing, sorting, and visualizing loss run data to understand the impact of the school district's losses on its risk management program, and the ultimate cost of risk. Forecasting or Loss Trending Applying development factors, inflation indices, and exposure bases to loss trend data to more accurately estimate a loss pic. One more important area of analysis to look at is financial assessment. An understanding of the financial capabilities of the school district is essential to making good risk management decisions. 69

70 Financial Capacity Financial capacity is an organization's ability to meet its financial objectives and obligations. Financial capacity affects the schools willingness and ability to take risks in order to receive additional benefits. This is known as appetite for risk and is an important consideration for the school risk manager. Schools are obligated to fund programs that benefit students. To protect this part of the district's mission, risk managers need to be proactive in finding ways to save costs and secure funding. The risk manager can aim his activities towards five factors which directly impact school finances. Factors That Can Affect a School s Financial Capacity The following five factors can affect a school s financial capacity: 1. Tax base: Value of real property and other assets within the taxing jurisdiction. 2. Current needs & long term debt of the district: Current operating expenses and capital requirements of the district. Future opportunities to expand or improve programs and services. 3. Retention and transfer options: The decision to purchase insurance, fund risk internally, or use a combination of both. 4. Liquidity and cash flow: Having sufficient funds available to manage the highly variable cash flows that are typical of school districts. 5. Long term debt and cost of capital: The risk manager must be aware of the school s ratio of long term debt to annual income and the interest rate charged when the district borrows money. 70

71 Factor 1: Tax Base A school s tax base is the money available to the school from local and state taxes. Its tax base is typically the main source of a school s income. The amount of local taxes that a school receives is determined by several factors. Point your cursor to each one to learn about it. 1. Total value of homes in the community: Value of real property and other taxable assets within the school district s taxing jurisdiction. 2. Number if homestead exemptions: Certain homeowners may be able to claim a homestead exemption, which decreases their tax contribution to schools. Communities with older populations tend to have more homestead exemptions. Not everyone who is eligible applies for the exemption. 3. Revenue received by local businesses: School districts assess a tax on local business profits as a source of revenue. 4. Tax waivers for businesses: Some districts grant a waiver of business tax for a period of time. This is done to entice large companies to locate within their community and is justified by the long-term benefit to the community outweighing the short-term loss of tax revenue. Property taxes are a primary source of funding for schools. 71

72 Factor 2: Current Needs & Future Opportunities Capital needs refers to capital assets which typically are of significant dollar amounts and render service for long periods of time. Examples of capital assets include computers, buses, and riding lawn mowers. In contrast, operating costs are items or services that need to be purchased often, such as fuel, electricity, and repair costs. In assessing your school's needs, think about what those needs may be five to ten years into the future. Will new buildings or schools need to be built? Will the number of students increase? What will these new students need in terms of capital goods, such as school furnishings and equipment? Factor 3: Retention and Transfer Options When you look at the dynamics of your school, such as whether the student population is increasing or decreasing, consider the effects that these changes may have on your school's insurance needs. In some situations, a school may choose to retain risk. In other situations, it may choose to transfer risk. Often, the most appropriate option is a combination of these. There is also the option of joining an insurance pool. The growing school will need additional bus routes. There will be more acreage to take care of. Additional equipment and personnel will be required. Assessing new buildings will translate into higher insurance premiums. This usually calls for a review of the district's retention and transfer options. Performing an insurance review often leads to significant savings in the overall cost of risk. Retain Risk: The conscious or unconscious act of assuming the performance and/or financial responsibilities of a loss, should it occur. Transfer of risk: A risk management technique whereby risk of loss is transferred to another party through a contract, e.g., a hold harmless agreement, or to a professional risk bearer, i.e., an insurance company. 72

73 Insurance Pool: Combining together of several insurers to share premiums and losses so as to spread risks. Factor 4: Liquidity and Cash Flow Liquidity Liquid assets are those that can be converted to cash quickly. They include cash, demand deposits and time savings deposits. Cash Flow Cash flow is the inflow and outflow of dollars to finance school operations. Financially sound schools replenish their reserves from local taxes and other sources of revenue on an annual basis. Factor 5: Long Term Debt & Cost of Capital Long Term Debt Long term debt is defined as any debt owed for a period longer than one year. General Obligation bonds are issued by school districts to pay for construction of new buildings and can be paid off for periods of up to 30 years or more. Cost of capital - The interest rate that a party is charged to borrow money. School districts, like businesses, are rated according to the letter based scale used by rating companies such as Moody's: A++, A+, and A, B etc. Schools with lower ratings, such as B, will be charged higher interest rates. Obtaining a good rating and the resulting lower interest rates can be of critical importance to school districts, especially those contemplating issuance of general obligation bonds. 73

74 Making Financial Decisions Most decisions that school risk managers make are financial decisions involving the inflow and outflow of dollars. The following represent examples of opportunities for potential cost saving, that would require the risk manager to compare the overall cost of the project with the overall benefit derived from the project. In other words, he must complete a cost-benefit analysis prior to making a decision. 1. Increasing the deductible on a particular insurance policy to $1000 would reduce the annual premium for that policy by 10%. 2. Installing a sprinkler system on a campus in the school district will result in significant credits applied to the insurance premium. 3. Joining a pooling arrangement with five school districts in the same region to reduce insurance premiums. Cost benefit analysis: Process of measuring the Present Value of Benefits against the Present Value of the Costs. Each opportunity requires that the risk manager or business manager compare actual benefits to actual costs before going forward with decision making. Click on our deductible example below. Premium Discount for Increased Deductible How many times might the organization incur this deductible annually? What is the future value of premiums with the discount? What is the future value of our deductible expense under the discount program? What is the future value of our current deductible expense? How do we expect premium levels to trend going forward? 74

75 Cost-Benefit Analysis A common approach to making financial decisions on risk management projects is to compare benefits to costs. Cost-benefit analysis: The process of weighing the benefits of a project against the costs to determine whether the project is financially advisable. The return on investment (ROI) of the project equals the project's financial benefits minus its costs. Benefits - Costs = ROI (Return on Investment) If the value of the benefits appears to be greater than the value of the costs, the project is probably a good decision. If the value of benefits appears to be less than the costs, the project is probably not a good decision. Factors That Can Complicate a Cost-Benefit Analysis Why did we say, "if the value of benefits appears to be greater...?" As you might expect, cost-benefit analysis in schools can be complicated by a number of factors. Remember that not all decisions are made purely on the basis of a cost/benefit analysis. Good judge Benefits and costs are estimates Hidden costs, such as lost work days due to employee injuries, are often hard to quantity, but their financial effects are often substantial. Hidden costs should be taken into account even if they do not have a set dollar value. The political realities of the school, school board, or community If the school board or community expects a certain measure to be put in place, it may be in the school s best interest to fund it, even if it does not make sense from a cost/benefit perspective. The possibility of injuries and/or fatalities The financial cost of losing one life or injuring a few people may not make a protective measure seem cost-effective. However, from a human perspective, potential injuries and fatalities need to be prevented whenever possible. A school district s reputation Even is a decision may not make sense from a cost/benefit perspective, it may still be a wise decision if it protects the school s reputation in the community. 75

76 Time Value of Money The time value of money is the principle of finance that states that a specific amount of dollars available in the present are worth more than the same amount of dollars in the future. Dollars are worth more because of their earning capacity. One dollar today invested in a simple savings account is worth more in a year or two than if it had been stored under a mattress. The time value of money incorporates the concepts of discount rate, future value (FV), present value (PV), and net present value (NPV). A risk manager uses these calculations when completing a cost-benefit analysis. Cost benefit analysis must account for the time value of money (TVOM). Suppose the insurance company will reduce school district's annual premium by a 10% for the life a sprinkler which is estimated to be 20 years. The school invests in the sprinkler system, with payments made over 4 years. The value which represents the cost of the sprinkler system will have to be adjusted for TVOM. The value of the benefits of installing the sprinkler system will also have to be adjusted for TVOM. Note: A discount rate and an interest rate are numerically the same thing. The only difference between the terms is the context in which they are used. Time Value of Money: Recognizes the change in the value of a dollar due to time. TVOM calculations involve both Present Value and Future Value. Discount Rate: The compliment of the interest rate used to determine the Weighted Average Cost of Capital (WACC). The discount rate is used to compute net present value from amounts to be realized in the future. Future value: The value in the future of a payment or payments made in the present. Present value: The value today of a future payment, or payments, discounted at the appropriate discount rate. Net Present Value: The present value of a series of future cash flows. It is a standard method for using the time value of money to assess the future value of a long-term project in terms of their present (current) value. 76

77 Present Value, Future Value and Discount Rate Present Value Calculations Roger needs to set aside funds to pay for a computer system upgrade in 10 years. How much would he set aside now? In ten years, he will need $150,000 and he can assume that the dollars he sets aside now will earn a 5% annual interest rate. Because the number of payment periods n, the interest rate i,and the Future Value, $150,000 are known, the Present Value can be calculated. PV= 150,000 x.952 PV=$92,100 Note: The scope of this course does not included working problems using PV or FV formulae. Your goal is to be able to recognize this formula and understand how it fits in, when completing a cost-benefit analysis. 77

78 With Future Value (FV), present value, interest and number of periods. Cost-Benefit Analysis Example When a cost benefit analysis takes into account a series of future values, it is often called Cash Discounting. On the next few pages, you will review a simple example of a cost benefit analysis. Making these types of calculations are not required for this course. However, the example provides an illustration of the following concept: COST BENEFIT = ROI (Return on Investment) (or) PV benefit PV cost = NPV (Net Present Value) 78

79 Property losses caused by fire total $10,000 per year. The organization estimates that losses would be reduced to $2500 annually with the installation of a sprinkler system. The school district intends to pay for the $80,000 sprinkler system over the next four years, with the first payment due now. The insurance carrier will give a 10% premium discount (Premium is $12,000) on the property insurance if reduction in losses can be shown after the fourth year. Value of Costs Cost of Sprinkler $80,000 Discount Rate Discounted Benefits (10%) Payout over 4 years (now & beginning each year) $20, $20,000 $20, $18,180 $20, $16,520 $20, $15,020 Present Value of Costs $69,720 The present value of the costs is $69,720. Now we need to bring the value of the estimated benefits of the project into the present as well. Annual losses prior to installation Estimated losses after installation Savings for 20 years of sprinkler life (end of year) Value of Benefits $10,000 $2,500 Discount Rate Discounted Benefits (10%) $7, $63,855 $1, $9,626 Present Value of Benefits $73,481 The savings for the sprinkler system and premium were calculated using an annuity table, which can be used when the same amount is being either paid or received over a period of time. The alternative method would be to list all of the present value or discounting rates separately for each year as illustrated in the Costs portion on the previous page. 79

80 NPV of the Sprinkler Project Recall that: PV benefit PV cost = NPV (Net Present Value) The estimated present value of the costs of the sprinkler system is $69,720, and the estimated present value of the benefits of the sprinkler system is $73,481. The NPV of the project is $3,761. Since the NPV is a positive number (PV of benefits are greater than PV of costs), the project may be a "go" from a financial standpoint although other non-financial factors may come into play. If NPV 0, the project can be accepted (from a financial perspective). A positive NPV indicates that the project is likely to make more money that it will cost. The Role of Actuaries and Independent Consultants Risk analysis is a complex process that often requires actuarial or consulting assistance. Actuaries and consultants are experienced professionals who can provide the following services to schools: Compare and contrast approaches to retention, reserving levels, and the discounting of reserves. Incorporate qualitative factors into reserving and retention levels, such as legal changes, coverage changes, changes in the nature of the school district's operations, changes in risk control/safety programs, and changes in claims-handling procedures. Maintain development data and other data useful for comparison purposes, such as determining confidence intervals. Evaluate alternative risk financing methods. In addition, an actuary or independent consultant can often provide an outside opinion to lend credibility to your work. Retention: 1) Budgeted losses + tolerance corridor. 2) Assumption of risk of loss as through the use of non-insurance, self-insurance, or deductibles. This retention can be intentional (active retention)or, when exposures are not identified, unintentional 80