An Actuary in the World of Six Sigma

An Actuary in the World of Six Sigma

by David W. Dickson

This article shines a light on the shroud of confusion that surrounds Six Sigma.

An actuary, a black belt and a process owner walk into a meeting room to discuss a Six Sigma project. Sounds like the start of a bad joke, doesn't it? What would an expert in financing risk protection products, a person with a martial arts title and someone who owns a process (What does that mean? Is there a deed?) have in common with six Greek letters? What is this thing called Six Sigma? Why does it require Black Belts and Master Black Belts? Most importantly, why should an actuary care?

To unravel this riddle, a web search may be a helpful tool. Choose your favorite search engine and type "What is Six Sigma?" The only problem is you'll get back at least 1,950,000 search results. It seems everyone is an expert. Well, here is a sampling (something done frequently in Six Sigma) of the results.

"Six Sigma is a total management commitment and philosophy of excellence, customer focus, process improvement and the rule of measurement rather than gut feel." Peter Pande and Larry Holpp, What is Six Sigma?, 2002 McGraw-Hill, page 3.

"It is not a secret society, a slogan or a cliché. Six Sigma is a highly disciplined process that helps us focus on developing and delivering near-perfect products and services."

"(Six Sigma is) a movement, method and set of techniques focused on improving business processes. Relies heavily on statistical techniques to measure success. There are multiple Six Sigma methods, some designed for process improvement and some for designing or redesigning business processes."

"Six Sigma is many things, and it would perhaps be easier to list all the things that Six Sigma quality is not. Six Sigma can be seen as: a vision; a philosophy; a symbol; a metric; a goal; a methodology."

Six Sigma: SPC and TQM in Manufacturing and Services, by Geoff Tennant, 2001 Ashgate Publishing.

Six Sigma really is a magnet for ideas across a broad range of applications brought together to change the way business decisions are made (see Table 1).

Put another way: Six Sigma is a set of problem solving techniques that ties together: good project management techniques (i.e. managing scope, schedule and resources) with basic data mining and statistical analysis and prediction techniques.

At its heart, Six Sigma focuses on one of two things:

  1. Using process maps, data mining and hypothesis testing to find root causes of defects, which if addressed will really make a difference in your product or process (then controlling it) or
  2. Predicting efficient frontiers or regions of operational efficiency where if you can control your product or process within those frontiers or regions your results will be as good as possible, given the uncontrollable noise factors.

The History of Six Sigma

Six Sigma's history dates back to Dr. Walter A. Shewhart's work in the 1920s and 1930s in using applied statistical methods to control processes at Western Electric and later Bell Laboratories. He is sometimes known as the father of statistical quality control.1

Shewhart's greatest influence was on W. Edwards Deming who used Shewhart's ideas and applied them to WWII war production and to help reconstruct and modernize Japan during the 1950s and 1960s. In 1987, Deming published Out of the Crisis, which was the basis for what became known as Total Quality Management (TQM).2

During the 1980s and 1990s, Shewhart's work was incorporated by a new generation of industrial managers in what is known today as the Six Sigma approach.3 A Motorola engineer named Bill Smith is credited for coining the term "Six Sigma." Motorola developed a new standard of quality (3.4 defects per million opportunities) and created the methodology and the cultural change needed to ensure its success. Six Sigma helped Motorola achieve such significant cost savings that other industrial leaders took notice and adopted it for their companies, such as Larry Bossidy of Allied Signal (now Honeywell), and Jack Welch of General Electric Company. Now, Six Sigma has expanded into the services and financial communities; companies are adopting it almost daily.4

GE's Six Sigma initiative is certainly the most publicized if not the best example of how the method changes the way business decisions get made–letting the customer and the numbers drive business decisions. It stands in stark contrast to the traditional method of relying on opinion, "experts" or those with the loudest voice or biggest title.

At GE, Six Sigma is referred to as "The Way We Work" and the corporate culture that evolves from the consistent application of Six Sigma is called a "Meritocracy." GE employees joke, "In God we trust, but everyone else has to bring data."

Six Sigma is ever changing and seeking to improve. It takes the best of quality systems like TQM or ISO, adds project management and other business and statistical disciplines to create a new way of doing business.

Although Six Sigma continues to evolve, there are two ideas that form the foundation of Six Sigma that have not changed:

  1. The "customer" of the output of a particular process or product defines what quality is and what it is not. This is known as the Voice of the Customer (VOC), and
  2. Only that which is measured can be managed. The metrics that measure the process or product should flow down directly from the analysis of what the customer and subject matter experts say is important to measure quality. The analysis of these metrics and what drives them is known as the Voice of the Process (VOP).

By listening to the VOC and the VOP, teams can determine root causes of less than perfect outcomes or outputs and devise solutions and control plans to eliminate defects, mitigate risks or optimize systems.

Implementing Six Sigma

In order to initiate something that is described as a vision, a movement, a philosophy and a management commitment, companies have discovered that introducing Six Sigma requires a complete culture change that takes time to implement. To accelerate that change requires an infrastructure and language to make the enculturation permanent.

The Infrastructure

Champions: Fully-trained business leaders who lead the deployment of Six Sigma in significant areas of the business.

Sponsors/Process Owners: Fully-trained process leaders who together with Black Belts lead the improvement teams and take responsibility for implementing the improvement and control plans developed by the team.

Master Black Belts: First and foremost teachers. They also review and mentor Black Belts. Selection criteria for Master Black Belts are quantitative skills and the ability to teach and mentor. Master Black Belts are full-time positions.

Black Belts: Leaders of teams responsible for measuring, analyzing, improving and controlling key processes that influence customer satisfaction and/or productivity growth. Black Belts are full-time positions.

Green Belts: Fully-trained individuals who apply Six Sigma skills to projects in their job areas. Similar to Black Belt, but not a full-time position.

Team Members: Individuals who receive specific Six Sigma training and who support projects in their areas.

The Language

Six Sigma: A vision of quality that equates with only 3.4 defects per million opportunities for each product or service transaction. Strives for perfection.

Customer Needs, Expectations:

Needs, as defined by customers, which meet their basic requirements and standards.

CTQ–Critical to Quality (Critical "Y"): Element of a process or practice that has a direct impact on its perceived quality.

Defects: Sources of customer irritation. Defects are costly to both customers and to manufacturers or service providers. Eliminating defects provides cost benefits.

Control: The state of stability, normal variation and predictability. Process of regulating and guiding operations and processes using quantitative data.

Control Chart: Monitors variance in a process over time and alerts the business to unexpected variance that may cause defects.

Defect Measurement: Accounting for the number or frequency of defects that cause lapses in product or service quality.

Pareto Diagram: Focuses on efforts or the problems that have the greatest potential for improvement by showing relative frequency and/or size in a descending bar graph. Based on the proven Pareto principle: 20 percent of the sources cause 80 percent of any problems.

Tree Diagram: Graphically shows any broad goal broken into different levels of detailed actions.

Process Mapping: Illustrated description of how things get done, which enables participants to visualize an entire process and identify areas of strength and weaknesses. It helps reduce cycle time and defects while recognizing the value of individual contributions.

Root Cause Analysis: Study of original reason for nonconformance with a process. When the root cause is removed or corrected, the nonconformance will be eliminated.

Statistical Process Control: The application of statistical methods to analyze data, study and monitor process capability and performance.

There are two basic Six Sigma methodologies:

  1. DMAIC–Define, Measure, Analyze, Improve and Control. It is used to "fix" existing processes when there is one, or at most, two ways to measure the capability or success of the process.
  2. DFSS–Design For Six Sigma. Define, Measure, Analyze, Design, Optimize and Verify. This method is used to design new products or processes or for redesigning existing processes when there are multiple ways to measure success, some of which may conflict with each other and therefore need to be balanced or optimized.

Most actuaries find that the problems they encounter are best suited to a DFSS approach.

A good example is the problem of maximizing returns while minimizing return volatility and while matching asset and liability durations.

The key output of a DFSS solution is the ability to predict performance within some confidence limits. To do this, the design must include a model for the expected results and a model of the expected variability of those results due to sampling error and due to noise parameters. This exercise can have many starting points ranging from subject matter expert opinion to existing models to detailed production databases from which a unique specific model may be developed. Figure 2 illustrates the variety of approaches and the trade-offs to each approach.

At GE Insurance Solutions, Six Sigma has been used to:

  • Statistically determine the sensitivity of a rating model to various rating factors, and to determine the most predictive parameter values and distribution curves to use in rate modeling.
  • Determine optimal underwriting strike-zones.5
  • Optimally allocate capital by region and reinsurance layer within a product portfolio.
  • Optimize the interest crediting strategy on a run-off book of annuities.
  • Optimize the use of reserving methods in short, medium and long tail books of business.
  • Develop confidence intervals for reserve estimations.
  • Optimally predict the one-year forward interest rate for ROE calculations, etc.
  • Design software that has optimal response time and meets users demands.
  • Design new products that meet customers needs and our profit goals.

The Value of Six Sigma to the Actuarial Profession

Actuaries have done each of the above tasks without Six Sigma. So what is the value Six Sigma adds to actuarial training and expertise?

Six Sigma insists that:

  • Every process (even thinking processes) can be mapped.
  • Decision points can be described.
  • Data collection and measures can be put in place.
  • Standard operating procedures can be documented.
  • Assumptions should be documented.
  • Monitoring, controls and risk mitigation can be applied.

This applies to the actuarial decision-making, pricing, and reserving processes as well as it does to a manufacturing process. Many actuaries like to think that every analysis is unique, but there are standard ways that actuaries approach each unique problem. If there were not, then peer review would not make any sense. The American Academy of Actuaries has a suggested peer review structure for actuaries ( pdf/prof/peerrevi.pdf). Sarbanes-Oxley is forcing more controls on what actuaries do. But Six Sigma preceded both of these by teaching that a "process mindset" is like oatmeal–"It is the right thing to do."

Team Decision Making

An actuary's strength is quick analysis. Given a problem or a set of circumstances an actuary can immediately devise or envision a solution. This strength can also be a weakness; sometimes a quick decision can leave out other valuable input. Most actuaries who have had any connection with Six Sigma will typically complain that it slows them down. However, actuaries who have been on a Six Sigma project team to the project's completion will admit that though the solution took longer to develop, using the methodology had the following unexpected benefits:

  • The methodology forced them to document along the way.
  • The solution was better. It included aspects that they would not have considered on their own.
  • The solution was more widely accepted by all stakeholders, and therefore is more likely to be sustainable.
  • Controls and data collection plans are now in place to easily update or replicate the analysis.

Six Sigma is a team decision-making methodology. It focuses on getting input from all stakeholders and then using that to build a data collection plan to prove or disprove theories, opinions or hunches advocated by the "experts"–and that includes actuaries. So it requires everyone to check their titles and credentials at the door to the team room and function on an equal basis with other team members. It teaches actuaries humility.

Project Management Skills

The two Six Sigma methodologies are essentially standard work breakdown structures with tollgates and peer review. Actuaries who have worked in consulting may be more familiar with these basic project management tools, but for many of us who have only worked for insurance companies, we have had little exposure or instruction on how to bring a project to conclusion:

  • On time
  • Within scope
  • Within resource constraints (dollars and people)

Six Sigma automatically instills a project management mindset to any project. It forces the project team to define:

  • Business Case–The threat or theopportunity.
  • Problem and Goal Statements–How bad is it according to the data and what will define measurable success?
  • The Scope–Start and end points, what is included or excluded?
  • The Timeline–When are interim deliverables to be completed, when are peer review tollgates to be passed?
  • The Resources–Budget, team members, stakeholders, roles and responsibilities.

Risk Management Tools

Six Sigma teaches basic risk management techniques and tools, especially how to quantify qualitative measures that are essential to enterprise risk management. It looks at risk from both a macro (enterprise or market) level and micro (process step or sub- system) level, and it divides risk analysis into discrete categories that facilitate the ability of an organization to prioritize risk mitigation actions.

The first step in analyzing risk in a typical process improvement project is to identify the opportunities for risk or failure that exist within the process. A process map is typically a good place to start in order to identify points particularly susceptible to risk or failure. The next step is to describe the effect of the failure if it occurs and the probable causes of the failure. A cause and effect (fishbone) diagram is a useful tool when discussing this with the team.

Once the causes have been verified, the next step is to score the risk on at least two of three levels: severity, frequency and the ability to know when a failure has occurred– detectability. The first two are scored on both micro and macro risks. The third is typically applied to only risks identified at the process step (micro) level. The product of these scores is calculated and ranked against other risks in order to identify the risks with the greatest need for mitigation.

Before this scoring can be accomplished, the team or company must agree upon a scale to be used. An example would be a one-to-five scale for both severity and frequency when ranking macro risks. The maximum product would be 25 and any risk with a score of nine or higher would require some mitigation action. The product of the severity, frequency and detectability scores is known as the Risk Priority Number or RPN. After the RPNs are ranked, current controls or mitigation efforts are identified.

Agreement must be reached on any further actions that are deemed necessary along with the person(s) responsible and a timeframe for completion. After the mitigation actions are complete or implemented, the RPN would be recalculated and the process would continue.

Six Sigma also teaches to differentiate between "alpha" and "beta" risk. Table 2 illustrates the difference between the two.

Alpha risk is usually deemed to be more catastrophic in the near term. Beta risk is important in the long term. However, the decision regarding which risk type should be the focus of mitigation efforts is a case-by-case, project-by-project discussion by the team. Alpha and beta risk are mutually exclusive, but they are related. The more protections instituted against one of these risks the more likely the other is to occur. Both must be managed and balanced. The key to balancing both of these risks is more and better data.

The Expansion of Statistical Thinking

Six Sigma teaches everyone to think statistically. On the first day of Six Sigma training everyone is taught that every practical problem, if properly measured, can be described as a statistical problem of either:

  • Being off center relative to a target,
  • Too much variability (spread) relative to customer specifications (Remember the actuary that drowned in the river that was "on average" only three feet deep), or
  • Both off center and too spread out.

Then using process maps, cause and effect diagrams, risk analysis, Pareto charts and hypothesis testing, teams can discover drivers of the centering and spread problems, brainstorm solutions, consider trade-offs and arrive at a set of statistical solutions. If the company is able and willing (note the order) to implement these solutions, then the statistical solutions become practical solutions from which to choose based on cost and ease of implementation. For actuaries this creates a constituency that is speaking our language, so it should be easier to explain our own analysis.

Influencing Skills

As a project leader in Six Sigma, either as a Black Belt or a Green Belt, an actuary will typically have a team of cross functional resources, including those who do not report directly to anyone on the team. This structure forces each team member, including the actuary, to persuade based on the power of ideas and their ability to back them up with data or demonstration. Six Sigma exemplifies the SOA motto:

"The work of science (Six Sigma) is to substitute facts for appearances and demonstrations for impressions"–John Ruskin.In a company that has adopted Six Sigma, the senior leadership team will have also been Six Sigma trained. The senior leaders will gain more confidence in the actuarial team's work if they know the work was accomplished using the Six Sigma methodology and if the team can describe to the senior leaders the impact in Six Sigma terms. So, for the actuary, Six Sigma becomes a communication tool that everyone in your company will understand.

What You Learned in College and on the Exams is Practical

Scene 1: A young, new actuary gets her first big project and is eager to apply what she remembers from her stochastic modeling college course and Course 3 of the actuarial exams. Her mentoring actuary advises her that those methods are not always practical, and shows her the method the company has always used and suggests that all that is needed is a parameter update.

Scene 2: The valuation actuary presents the quarterly reserve analysis to the CFO and makes a recommendation to book the expected plus a 10 percent margin. The CFO asks what the probability is that the actual reserve needed will be within the +/- 10 percent range around the expected value. The valuation actuary cannot give a probability, but shows the CFO two other calculations, one pessimistic and one optimistic and says "I am relatively confident the actual reserve will fall between these two estimates."

Does this sound familiar? Six Sigma does not teach actuaries any new analysis tricks, but it appeals to the researcher in each actuary, and asks:

  • How can the latest technology and knowledge be applied to this problem?
  • How can the model be made more predictive?
  • How can the variability around the model estimates be modeled due to;
    • Sampling error,
    • Noise parameters over which there are no controls?
  • How can probabilistic statements of confidence be made regarding our estimates?
  • In other words, how can the actuary apply the education in which he/she has so much time invested?

Joke or New Paradigm

So now, an actuary, a black belt and a process owner walk into a meeting room to discuss a Six Sigma project. Sounds a lot less like a joke and more like a whole new paradigm for an insurance company to do business, doesn't it?

David W. Dickson, FSA, MAAA, is actuarial project manager for GE Insurance Solutions.