January 2003 • Table of Contents

Toe the Line: No More WorldComs

How Six Sigma metrics can reliably track an organization's financial results

by Donna M. Faltin and Frederick W. Faltin

REVELATIONS OF MISMANAGEMENT at WorldCom, Enron, Adelphia and other companies have led to an unprecedented erosion of confidence in corporate America. Amid calls for more effective oversight of financial reporting and corporate governance, a long list of public and private entities including Congress and the Securities and Exchange Commission (SEC) are enacting systematic changes, with the goal of bringing about greater openness and accuracy in financial management.

Their initial efforts have focused mainly on enforcing senior management accountability and ensuring greater transparency into traditional financial measures. These changes are a good start, but they rely heavily on deterrence as a tool and on financial metrics that lag actual corporate performance.

As America's institutions try to define a new environment of corporate responsibility, it's important to create a business culture that goes further--to offer greater assurances to investors, regulators, analysts and creditors, as well as management. In particular:

• Corporate financial monitoring needs to be more timely.
• Metrics that are more difficult to falsify must be devised.
• Malfeasance must be detectable even at levels below the corporate office.
• Common practices are needed across business units and across companies.

Six Sigma, when augmented with techniques from economics, finance and operations management, can address these issues and provide a uniform and disciplined framework for reliably tracking financial results and corporate governance metrics. Our experience shows applying the define, measure, analyze, design and verify (DMADV) paradigm is the best way to develop a corporate financial quality system (see Figure 1, p. 30). To illustrate certain key points of our approach, we will make reference throughout the article to the largest corporate bankruptcy in U.S. history--the demise of WorldCom.

The principle issue at WorldCom is the reclassification of routine expenses as long-term capital investments. Investment dollars receive different accounting treatment than expenses and therefore affect corporate performance measures differently. Information about WorldCom in public filings is less detailed than that available internally as part of a corporate financial quality program. Still, it's interesting to look at WorldCom's data to see what might have been found had a Six Sigma monitoring system been in place.

Applying Six Sigma to finance

Define
Financial analysts have traditionally based assessment of an organization's performance on numbers in its financial reports and a relatively small set of derived quantities, such as growth rates and expense to revenue ratios. Wall Street especially has often focused on only a few simple financial metrics as a barometer of an organization's performance.

In the telecommunications industry, for example, the metric that gained greatest favor during the 1990s was earnings before interest, taxes, depreciation and amortization (EBITDA). The government alleges the WorldCom's CFO determined the amount of improper expense restatements to keep EBITDA numbers more in line with company guidance and analysts' expectations.

As we'll see, other related metrics would likely have deviated from their nominal values if a Six Sigma system of financial tracking had been in place. So it's clear from recent events that what is needed is not one or even several independent measures, but a set of related metrics that capture key results and drivers of corporate performance.

Figure 2 shows the plot of a simple metric that calculates the approximate total new debt assumed by WorldCom each year from 1995 to 2001. Figures 3 and 4 show the ratio of WorldCom's short-term debt to its long-term debt annually in 1999 and 2000 and quarterly in 2001 and the first quarter of 2002.

The nature of corporate governance issues and the key constituencies to be served will likely differ from one organization to the next. For example, a financial quality program might be put in place mainly to protect the senior officers who must now personally certify the firm's financial results. Other companies may want to establish a monitoring program to track and validate changes in business or accounting practices required by government regulators. Still other corporations will have dominant shareholders who might want reassurance that management is responsibly managing their investment.

The best way to design a corporate monitoring system is for senior management, including the board of directors, to clearly define its purpose and identify the key constituencies to be served. The next step is to conduct a facilitated customer focus group with representatives from all key constituencies. This meeting's purpose is to define the specific high level issues to be addressed. Potential targets of Six Sigma monitoring include not only corporate accounting, but also items such as compensation practices, stock trading patterns, earnings of foreign subsidiaries and relative contributions from various revenue sources or business divisions.

The outcome of the focus group is a multifaceted set of key corporate metrics that reflect the interests of all constituencies. These metrics will be the top priority and principal focus of the corporate financial quality program.

Measure
The principal objectives of this phase are to determine baseline values for these key corporate metrics (known as big Y's) and identify the component data streams needed for their management.

In Six Sigma, as in traditional financial statement analysis, there are two ways to determine baseline values for oversight metrics: externally, by comparing values of the metric at the host organization with values from a group of similar companies, and internally, by observing the historical performance of the metric over time.

For external comparisons, standard financial references can be called on to help define the peer group of the organization under study. Typical points of comparison include:

• Industry, as defined by the North American Industry Classification System (NAICS). NAICS codes are a standard U.S. Census Bureau classification system that identifies an organization's principal business activity.
• Size, as ranked by Fortune magazine.
• Creditworthiness, as determined by credit rating institutions, such as Moody's or Standard & Poors.
• Other general business characteristics, as reported by business data research firms, such as Dun & Bradstreet.

Time is also a consideration, since rapid growth or merger and acquisition activity may render current comparisons inappropriate in either the past or future.

At first glance, the simplest comparison for any organization would appear to be with its past performances, but historic performance norms are surprisingly difficult to establish in financial quality applications. The merger and acquisition mania of the past 15 years has left few companies untouched, so the organization's functional definition may have changed from year to year.

Elsewhere, replacement of senior leadership may have led to fundamental strategic realignments that have shifted internal norms. Growth, employment and other patterns established during a period of prolonged economic expansion, for example, might be inapplicable during a contraction. As a consequence, appropriate historical comparisons are difficult, and the use of traditional statistical process control monitoring techniques is rare.

Taking another look at the WorldCom example, Figure 3 shows the plot of WorldCom's debt ratio vs. time, and Figure 4 compares it to that of its five most similarly sized competitors. Unfortunately, historical comparisons of this ratio to prior years are not meaningful due to extensive merger activity in 1998.

Tracking down the origins of information needed to calculate key corporate metrics has an importance beyond mere data gathering. This exercise establishes the logical and arithmetic relationships between the key metrics (big Y's) and their precursors at lower levels of the organization.

The components that are arithmetically combined to produce the big Y's are often called little y's.

For example, total corporate revenues or earnings (big Y's) are commonly the sum of revenues or earnings from individual business units (little y's). This data hierarchy helps individual unit managers track how their performance impacts the corporation's key business metrics and assures alignment of business components' objectives with broader organizational goals (see "Management Dashboards").

Exploring relationships in the data presents an important opportunity to review and assure the accuracy and transparency of component data streams. Certain corporate data will be sensitive and, therefore, held confidential within a limited circle. Nonetheless, policy must permit free access to data within that circle to at least some staff beyond those who routinely compile and manage it. This was apparently not the case at WorldCom, where federal prosecutors allege that in a company with more than 60,000 employees, a group of perhaps six officers and staff were able to conduct the improper reallocation of nearly $4 billion in expenses.

There is often a tendency to rush through the define and measure project phases. But in financial quality applications, the importance of going through these phases in a methodical and disciplined manner cannot be overstated. Not only is the attainment of the project's purpose at stake, the thoroughness of the set of target metrics under study and the appropriateness of the baseline values are also determined during these phases. Six Sigma tracking against these established norms makes falsification or mismanagement difficult to hide provided the define and measure phase tasks have been done well.

Analyze
The analyze phase affords the opportunity to identify fundamental relationships with key drivers that move the corporate metrics (the big Y's and little y's) and to discover leading indicators that can provide advance notice of impending issues.

This phase offers the best opportunities to:

• Provide early warning of impending issues.
• Produce new insights of value to management.
• Detect manipulation of component data streams.

We look for relationships of the Y's and y's with two kinds of explanatory variables: key drivers and leading indicators. Both have their roots in the familiar Six Sigma concept of a transfer function, an expression that quantifies the impact some set of explanatory variables has on an outcome of interest (the Y's and y's).

Although many common Six Sigma tools are applicable, the analyze phase is where other tools not ordinarily taught in Six Sigma come most into play from related fields, such as economics. Finance professionals with expertise in activity based costing (ABC) will also find the approach familiar because Six Sigma applies and extends many of the same techniques adopted by ABC. Senior managers will find additional returns on their investment in ABC training and systems, as these provide a jump start in searching for and tracking key drivers.

The distinction between key drivers (called X's) and leading indicators is not entirely crisp, but X's are generally characteristics internal to an organization that directly influence the Y's and y's of interest (see Figure 5). For example, market share or bid win/loss ratio (X's) might be determinants of revenue (a Y); similarly, productivity (an X) may determine operating margin (a y).

Key drivers (X's) typically have an immediate influence on the Y's and y's, or they slightly lead them. In the above examples, items such as utilization, productivity and market share usually have an immediate effect on the corresponding Y's and y's. On the other hand, items such as sales force effectiveness and bid win/loss ratio generally impact financial results the following month or quarter.

But even those X's that have a contemporaneous effect may be measured sooner than the y's and therefore have a de facto lead effect. Utilization or productivity may be recorded weekly or monthly, whereas the financial metrics it impacts are summarized monthly or quarterly. The financial reporting process usually takes days or weeks to complete, adding still more time to the lead provided by the X's. On average, tracking even coincident X's provides one to two months' lead time when the corresponding Y's and y's are financial metrics. So tracking key drivers allows us to anticipate financial results in advance and provide early warning of impending issues.

Leading indicators represent the greatest potential added value that financial quality can offer senior management. They are generally external factors that indirectly influence or correlate with organizational results, but with greater advance notice. An aircraft leasing organization, for example, can anticipate revenues several quarters out based on the volume of business and personal travel today. Changes in the price of oil may not only impact shippers' margins for several quarters, but may also predictably shift the market equilibrium among rail, truck and air. It is often possible to find leading indicators that will predict major financial business metrics six to nine months ahead or more.

An understanding of key drivers, leading indicators and the transfer functions through which they influence the key corporate metrics strikes directly to the issue of corporate governance, for it greatly enhances our ability to detect malfeasance in financial reporting. Fudged Y's and y's that no longer obey known relationships with key predictors are easy to detect. This creates a network of interlocking metrics, key drivers and leading indicators through which we can monitor relationships, not just metrics.

Design
A major challenge in providing transparency into corporate finances for the key constituencies is to clearly communicate financial information in a usable format. Today, this information is contained in SEC filings and a company's annual reports. Financial quality programs must address the need for a more flexible process that conveys to each of the stakeholders the information he or she seeks. Toward this end, the principal activity in the design phase is to identify which Y's, y's, X's and leading indicators will be tracked for review on an ongoing basis, and how.

Some quantities, including financial metrics useful in operations, will be selected for routine report-outs. This is the case with General Electric's online senior management dashboards, which are specifically targeted toward operational and financial management. Others, such as Y's relating to corporate finance or governance metrics, will be updated regularly, but reported only by exception.

Dashboards and cockpits provide a convenient environment for communicating results in a uniform and timely manner to all constituencies. Through use of green, yellow and red highlighting or by limits printed on the chart or screen, dashboards and cockpits provide a simple visual comparison of the metrics to their norms. Digitization, where applied, automates and facilitates preparation of the dashboards, integrates data streams and enables roll-ups from individual business unit dashboards to corporate level summaries.

The design phase is also the time to revisit priorities and get rid of unnecessary detail and redundancy. Perhaps the greatest danger to a corporate monitoring system is to tackle too much too soon. From the choice of the Y's through identification of little y's, X's and leading indicators, conciseness is second only to thoroughness in attaining the program's stated goals. And digitization, though desirable, must not be an excuse for postponing action if the required infrastructure is not yet in place. Maintain focus on the Y's and limit the scope to match resources. Monitoring a few key metrics with a project that succeeds is far more effective than monitoring a great many with one that doesn't.

Verify
The principal tasks in the verify phase are to assure disciplined reporting and response plans are in place and to validate the total program that results.

The response plan is established through a disciplined set of standard operating procedures, as are typically needed for the control and verify phases of many Six Sigma projects. The essence of a response plan is root cause analysis. To assure proper oversight, responsible management should be notified whenever an inquiry is under way, even though there is no presumption any impropriety is involved.

Evidently, none of these practices were in place at WorldCom. Figures 2, 3 and 4 (p. 31) make it clear there was a crescendo of major structural changes in WorldCom's finances in 2000 and 2001. This fact alone does not demonstrate any wrongdoing or point to any root causes. In fact, these events suggest the alleged fraud at WorldCom was a symptom of the company's financial collapse, not its cause.

Six Sigma monitoring of these metrics might have prompted inquiries well before any fraud actually took place. The changes in WorldCom's debt structure were calculable from public financial reports filed on March 30, 2001, and Aug. 14, 2001. The company's own internal audit first brought evidence of fraud to light on June 25, 2002.

Six Sigma corporate monitoring systems will from time to time detect changes in financial relationships that are positive results of shifts in management strategy or economic conditions and will, therefore, often be harbingers of good news. Nonetheless, the purpose of posing a barrier to fraud or mismanagement requires effective oversight procedures and a disciplined response and reporting plan to be in place at all times.

What may differ from case to case is the reporting path to be followed once an explanation has been found. A variety of constituencies will likely be served by a corporate monitoring system, so different groups may be notified at different times, depending on whose Y was affected. Many facts, such as news of a change in business circumstances signaled by a leading indicator, will be internally reported to appropriate management. At the opposite extreme, concerns regarding proprieties of corporate governance would be raised with the board and, if necessary, presented for regulatory review.

The final implementation task is validation. Scenario analysis can be used to test situations that management and the customer focus group identified at the outset as being the most important. Simulated data reflecting these circumstances should be submitted to the monitoring scheme to verify detection, and staff members with oversight responsibility should be interviewed to confirm they understand their role and know the procedures to follow.

Six Sigma can make a difference

The disturbing course of recent events in American business has identified a need to revise corporate financial reporting practices and provide regulators, analysts, boards of directors and shareholders with the information they need to apply more rigorous and quantitative standards in evaluating an organization's performance. New certification requirements for senior management and new expectations of corporate boards impose heightened personal liability at a company's highest levels and compel senior leaders to have reliable means for seeing into the inner workings of their organizations.

Financial quality practices, rooted in Six Sigma, can help define a culture that will respond to these needs, promote legal and regulatory compliance, facilitate better management and restore investor confidence in America's business institutions.

ACKNOWLEDGMENT

The authors would like to thank Frank J. Forkl Jr., certified public accountant and controller at Wilson Greatbatch Technologies Inc. for his valuable insights in commenting on earlier versions of this article.

BIBLIOGRAPHY

"Beyond Six Sigma: A Discipline for Financial Quality," training materials, The Faltin Group, 2002.

Criminal Complaint, United States of America vs. Scott D. Sullivan and David F. Myers, United States District Court, Southern District of New York, July 31, 2002.

Indictment, United States of America v. Scott D. Sullivan and Buford Yates Jr., United States District Court, Southern District of New York, Aug. 28. 2002.

"WorldCom: Could Six Sigma Have Made a Difference?" The Faltin Group, 2002. Case study available by request online at www.faltingroup.com.

WorldCom Inc., annual reports to shareholders, 1995-2001, www.worldcom.com/global/investor_relations/annual_reports.

WorldCom Inc., HO-09440, Revised Statement to the SEC, pursuant to Section 21(a)(1), Securities Exchange Act of 1934, July 8, 2002, www.hoovers.com.

"WorldCom SEC Form 10-K filings, 1995-2001," U.S. Securities and Exchange Comission's EDGAR website at www.sec.gov.

"WorldCom SEC Form 10-Q filings, 1Q-3Q2001, 1Q2002," U.S. Securities and Exchange Comission's EDGAR website at www.sec.gov.

DONNA M. FALTIN resides in Ballston Spa, NY, and until recently was senior economist at General Electric's global research center. She has a doctorate in economics from the University of Rochester.

FREDERICK W. FALTIN is managing partner of the Faltin Group in Ballston Spa, NY, and before that was manager of strategic enterprise technologies at General Electric. He earned master's degrees in mathematical statistics and operations research from Cornell University and is a winner of ASQ's Shewell Prize.