To minimize confusing a COI with actual coverage, every agent who has frequented an E&O session knows the drill: only use ACORD forms, never alter the language of the COI, do not represent coverages on the certificate — especially in the “Descriptions” section—that do not exist in the actual policies, et al. Follow the righteous path and let the ACORD disclaimers provide the protection for which they were designed, and the agent is in as good a position as possible to defend against COI E&O claims.

But none of those precautions may protect an agent where the COI clearly states a certain coverage has been provided, and it hasn't. It may simply be a clerical error, but to a certificate holder it sure looks like outright fraud. And no state regulation, insurance department memo or ACORD disclaimer, is going to protect you for fraud.

Let's be clear. If an agent, for any reason, deliberately falsifies a COI, they deserve to be slapped down with all the penalties available. It's hard enough to build relationships and esteem for our industry without such charlatans apparently confirming all of the public's and regulator's worst suspicions.

But what of what we might term “inadvertent” fraud? Those situations where the agent truly believed the coverage was correct, but erred? Ah, grasshopper, that is exactly why these humble epistles appear each month. In league with fine insurance education providers everywhere, the mission is simple and direct—help us all get better at finding our way through these coverage jungles with accuracy and effectiveness. That includes recognizing where we are still prone to get lost, and either avoid those areas or take the necessary steps to learn the proper paths.

In the realm of certificates, a key is to be sure that (1) the COI only indicates coverages that are actually in effect at the moment the COI is issued; and (2) if the coverage is not in effect at that moment, nothing indicates that it is.

It seems simple. Yet let's briefly review one coverage area commonly requested for COIs that apparently is often disconnected from reality: additional insureds.

In a recent IRMI [International Risk Management Institute] Expert Commentary, David Dybdahl disclosed that when his risk management firm audited hundreds of COIs and their underlying policies over a period of years, they found that more than 90% had at least one material misrepresentation between the liability coverage shown on the COI and that provided by the actual underlying policy.

A major percentage of Dybdahls's 90% was additional insured disconnects. His firm repeatedly found two key oversights relating to endorsements such as ISOs CG 20 33, CG 20 10 and CG 20 37 that led to gaps in additional insured (AI) status between what the COI stated and the reality:

• Improper usage—and evident misinterpretation—of “blanket” additional insured endorsements such as the ISO CG 20 33. One of my pet coverage peeves is represented by the common reference to this and similar endorsements as “blanket.” That word is not only never used in the actual endorsement (the proper term is “automatic status”) but it's also misleading. “Blanket” implies that every party asking to be an additional insured is covered. But automatic status as an AI only applies to those “for whom you are performing operations when you and such person or organization have agreed in writing in a contract or agreement that such person or organization be added as an additional insured to your policy.” Although that may cover quite a few potential AIs, it also leaves a lot of commonly requested AIs off the list. For example, consider an agreement in which your insured has a written contract with Company B, wherein B also requests AI status for its subs, vendors, affiliates, local municipalities, and so on. Because your insured only has a written contract with Company B, only Company B is automatically added as an AI—none of the rest. Company B is only an AI while your insured is actively performing operations for B. When those operations are completed or suspended—or perhaps, have not yet begun—the ISO CG 20 33 provides no AI status to B either.
• Improper usage of scheduled AI endorsements. For situations in which it's preferable to nail down specific AI status for any person or organization—or, in the case of completed operations, exposures not covered by the CG 20 33—ISO provides other endorsements. Some affirm AI status to entities (such as the CG 2010); others may address specific exposures (such as the CG 20 37 for completed operations). Although these endorsements have their own requirements and conditions for coverage to apply, Dybdahl's firm found an abundance of errors resulting not from coverage technicalities but rather from simple failure to properly execute the endorsements. For example, on far too many policies, the CG 2010 and CG 2037 were properly added, but the schedules left blank! How could any astute agent—or underwriter—overlook the clear and simple declarative sentence that begins the insuring clause of both the CG 2010 and CG 2037: “Section II—Who Is An Insured is amended to include as an additional insured the person(s) or organization(s) shown in the Schedule…”(emphasis mine).

Mary Poppins famously said “Well begun is half done.” It appears clear that Dybdahl has uncovered a veritable plethora of agents who have evidently seized on the well begun and totally forget they were only half done.

So are your COIs more trick or treat? Will your certificate requesters merrily feast upon dozens or hundreds of filed away COIs, only to later discover, unfortunately often only at time of a claim, that easy certificate candy has turned into a burning case of coverage indigestion? Can accusations of fraud or allegations of E&O be far behind?

Perhaps the old Sarge from Hill Street Blues said it best. Whether you are talking Halloween fun or COIs, “Hey, let's be careful out there.”

Chris Amrhein, AAI, is an insurance educator and speaker, and serves as the chief fun officer at insuranceisfun.com.