Imagine you just received a request from a vendor, Jones Office Equipment, that the Acme Distribution Company wants to lease a new copier. The copier is the newest model with all of the latest features. You have never heard of Acme Distribution, but their credit application looks good. The application shows five years in business, 18 employees, $1.8 million in revenue last year and growing. You wish every new applicant looked this good. The vendor tells you that the deal is a competitive one, so you need to act quickly. You do a quick review of the application and a credit check and approve the lease. Six weeks later, the first invoice is returned as undeliverable. You call the phone number on the application and it is disconnected. Only then do you begin to realize that you have just been scammed by a professional fraudster.

Fraud Hurts in Many Ways
The facts in the story above are fictitious, but the story is real. Moreover, it is happening every day with greater frequency.

The most obvious cost of falling victim to commercial fraud is the 100 percent loss of an asset. You may need to book dozens of additional profitable leases to make up for the bottom line hit from just one new fraudulent loss. Or you may try to offset the loss by raising your prices, possibly hurting your competitive position.

Fraud can create other potential damages than are less obvious or are more infrequent. Moreover, these damages can be even more devastating than the obvious loss of dollars. After you have been victimized by fraud, there are risks in doing too much to prevent future occurrences. You may increase documentation requirements or conduct additional validation and investigative efforts that add cost and time to approvals. The risk is that the vast majority of customers are legitimate and you don't want to implement new fraud policies that will reduce your ability to attract and book new accounts.

Fraud also has the potential to damage a lessor's reputation in the industry and community. For example, a lessor may see its good name become damaged by being associated with criminals. Or it can become identified in the media as negligent, an easy target for identity theft.

A 20/20 View Into Commercial Fraud
Commercial fraud, defined here as the act of intentional deception for financial gain through the use of a business entity, is a problem that is vast in size and growing rapidly. While most lessors recognize this and have begun to take appropriate action to protect their businesses, many have not since fraud is a complex problem that is not well understood. If a 20/20 view existed into the size of the problem, commercial fraud would be recognized as:

• Contributing $20B in losses to commercial credit extenders annually.1
• Representing 20 percent of annual commercial charge-off dollars for the typical credit extender.1

Difficult to Detect and Prevent
The fundamental challenge of detecting and preventing fraud is in recognition. Virtually all fraudulent lease applications look good on paper, just like every other legitimate customer application. Fraudulent applications cannot be uncovered by standard screening and scoring processes, which are in place to identify more traditional financial risk and potential for delinquency. The typical fraud will look normal at booking and funding. In fact, without any post-loss investigation, a fraud loss will typically be misclassified as a risk loss.
Though a lot of work has been done regarding recognition, awareness and management of consumer fraud, most of it is not directly relevant in commercial space. This is because of additional complexities:

• Unique identification and aggregation of data is a problem, as there is no wide-spread use of business IDs like the Social Security number in the consumer space.
• Turnover in the borrowing population is much greater in the commercial arena due to the faster pace at which small businesses are formed, dissolved or merged. It is far easier to create a fictitious business than to create a fictitious individual.
• Business entities are inherently more complex, as one business can have multiple associated units (branches, subsidiaries, affiliates, headquarters), multiple addresses (physical, mail, ship-to), and multiple names (legal, trade, DBA,2 acronym). Control and ownership changes are unique to business entities.

For these reasons, as well as the larger average commercial ticket size relative to the consumer space, fraudsters are finding the commercial space an attractive new target.

Characteristics of Fraud in Leasing
In leasing almost all fraud occurs at origination where every day an estimated 30,000 to 40,000 new leases are written.3 Of these leases an estimated 100 to 2004 are fraudulent. Several types of fraud can be perpetrated by both clients and vendor partners. For example, establishing a fictitious business identity or a shell entity, stealing the identity and credit of an existing business or purchasing the identity and finally so-called "soft" frauds in which the business is misreporting its financials or line of business to obtain a more favorable deal.

What Can Be Done To Detect And Prevent Fraud
The tools and techniques that can be employed to combat commercial fraud range from basic traditional investigative methods to sophisticated predictive modeling. These solutions vary from easy to complex in their implementation and development. While each of these methods can reduce the risk of fraud, easy-to-implement techniques are also the simplest for a sophisticated fraudster to overcome.

Traditional Tools and Techniques
Today many lessors underwrite utilizing scoring systems and must rely upon these systems and their internal risk team to provide fraud risk protection. However, since much of this is an automated process, especially in small ticket deals, many "fraud signs" can be missed. Some lessors do have a manual fraud review step in their process that can handle a small percentage of new applicants. Typically this review starts by checking the data in each application against established sources like state business registration and licensing files, yellow and white page listings, and other third party data sources. Anything considered questionable would trigger calls to credit and bank references (be careful here as these references could be fraudulent as well).

However, widely implementing these strategies can be very costly and can produce inconsistent results. Since the review work performed by a fraud analyst is highly manual and time consuming, reviewing any meaningful number of applications will require a significant investment. And with the monthly peaks and valleys in application volumes, fraudsters know that applications submitted near the end of the month have a higher likelihood of success. Perhaps most importantly, this review process will delay your decision response time, time that is critical to your ability to compete effectively.

The ability of a traditional manual process to detect fraud is limited in today's new economy. Avoiding detection in the initial checking process is easy for most fraudsters due to their experience and knowledge of the lessor. And if a fraudulent applicant passes the normal basic checks, you will never get to the more intensive verification that might root out the fraud.

Advanced Tools and Techniques
The desired solution for effectively detecting and preventing fraud should start with an automated, statistically based approach to initial applicant screening. This initial screening will allow you to separate applicants with higher fraud risk from the general population. These high-risk applicants should then be fed through the manual fraud checks in credit origination, booking and funding. By rapidly screening new applicants, identifying and segmenting the few percent that are of highest risk, you will offset many of the negatives of the more manually intensive across the board process. Using this approach, the majority of your new applicants can flow through your decisioning process in less time and at a much lower cost.

Internally developed, statistically based automated approaches to initial applicant screening will require an investment in infrastructure, staff, time and data. Of these, the biggest challenge is lack of good data. Having robust data on known frauds is necessary to build predictive models. Losses are not typically investigated to identify frauds, and quality of data is further compromised by deficiencies in tracking customers through various systems. These difficulties typically result in a low count of frauds. Furthermore companies don't have access to predictive data like cross industry/company financing behavior (fraudulent companies hit multiple providers and show a distinct pattern), which a large commercial bureau like D&B has. Both of these factors will severely limit your ability to build strong and reliable fraud models independently.

If you do not have the necessary internal resources, time or, most importantly, data to build your own automated screening system, you should consider broad-based commercial fraud risk screening tools that are being developed by companies like D&B. D&B has been developing its solution in collaboration with fraud experts from 30 cross-industry companies including several from the leasing industry. These organizations have shared their unique insight into the problem, as well as provided data (over 15,000 frauds) to the development effort. D&B has contributed its own proprietary sources of fraud-predictive data and data management tools. The data was then combined with the analytical resources of Inductis and D&B to quickly and cost-effectively create fraud risk screening tools. These tools leverage sophisticated matching and analytical capabilities to:

• Profile high-risk patterns of recent credit seeking activity.
• Screen against D&B's Confirmed Misrepresentation database.
• Check consistency, validity and fraud risk of the address and phone.

These new screening tools, which can be used by any commercial credit extender, will be available in Q4 of 2004. Solutions such as this are extremely adaptable to the individual lessor's needs. For example, lessors focusing on turnaround time and booked dollar volume could set their threshold to review a very small percentage of applications (say 1 percent). Another company, for whom ticket size or fraud incidence rate was high and the turnaround time was less of an issue, could set their threshold to manually review a more substantial group of applicants (perhaps 20 percent to 30 percent) and have the opportunity to prevent more fraud losses. Figure 1 illustrates the effectiveness of D&B's fraud solution at various review thresholds - for example 52 percent of frauds are contained within the riskiest 5 percent of applications.

D&B is also working with a number of lessors and companies from other industries to create the largest repository of known commercial frauds and victims of identity theft. This tool will provide an additional level of screening to its users and is expected to become operational in the second half of 2005.