As you say, searching for auditor changes on 8Ks is not as simple as it may seem. This is often the case with SEC filing regulations versus what is actually practiced. While a company must disclose when it dismisses or otherwise parts ways with its current auditor, companies are not required, to the best of our knowledge, to file when they engage a new auditor. Thus, 10% to 20% of 8Ks are incomplete. Also, companies often file auditor-change 8Ks as part of being acquired. For example, the larger Company Z acquires Company A, and thus, Company A becomes audited by Company Z’s firm. This is important for Company A’s shareholders to know. However, from the perspective of the auditor, this scenario is much different than a normal auditor change, despite the similar reporting method. Tracking auditor changes has many other nuances, but these two examples illustrate the need for critical, human review of each case in order to have a definitive data product on auditor changes. Taking the automated approach will come to perhaps a 75% accuracy rate, at best.

Regarding competitiveness, one cannot argue there is a limited field from which to choose. However, we find that nearly 60% of auditor changes involve at least two firms proposing for the work.

Looking at auditor changes on an annual basis, the Big Four firms have collectively had a net loss in public company audit clients every year since 2003. In 2002, they enjoyed large gains but only at the expense of Andersen. Going farther back, only Ernst & Young had a net gain in 2001. The firms will say that this is largely due to “strategic exits,” that is, resigning from risky audit engagements. During 2003 and 2004, even into 2005, this explanation makes sense. The firms began to take a hard look at the clients they took on from Andersen. The full impact of SOX 404 takes hold at company and auditor alike, both realizing they lack the necessary resources. And so, the Big Four began assessing their audit clients for risk in a new way, resigning from engagements deemed risky in relatively large numbers. Firms also left relatively unprofitable engagements to rally the resources needed to handle the 404 workload at their largest clients.

Again, this process makes sense when looking at 2004, and 2003 and 2005 to lesser extent. One can observe the same pattern at each Big Four firm during that time. However, since mid-2005, we have observed a turn-around at two of the Big Four. Deloitte and Ernst & Young have significantly narrowed their net loss due to auditor changes. More importantly, both firms have actually gained clients in the largest segment of companies we track. PricewaterhouseCoopers, on the other hand, continues to have a large net loss, and its losses are greatest at the large-company level. Furthermore, we find that PwC is being invited to propose at about the same rate as the other Big Four, but is converting invitations to engagements at a rate far below its competitors.

You also raise the question of what happened to all of these companies deemed too risky to audit by the Big Four. They went to the second-tier firms—BDO, Grant Thornton and McGladrey. These are good, high-quality organizations. However, they are orders of magnitude smaller than the Big Four. During the 2003-2005 window mentioned above, companies were throwing themselves at the second and third-tier audit firms, hundreds of companies. Some, not all, were those risky companies, clients that needed time and resources to be served properly. At the same time, the second and third-tier firms faced off against the Big Four in trying to add qualified people from the same talent pool.

Looking at the situation from a distance, it seems precarious for the second and third-tier firms: A large influx of clients, in a 18 to 24-month period, some of whom were deemed risky by the Big Four, while competing heavily for new people, and trying to implement new SOX regulations. Fortunately, nothing dire has happened here.