A PCAOB inspection team identified matters that it considered to be deficiencies in the performance of the audit work of a Big Four firm, primarily including failures involving fair valuation of asserts and internal control over financial reporting. In some instances, follow-up between the audit firm and the issuer led to a change in the issuer’s accounting or disclosure practices. In some cases, the conclusion that the firm failed to perform a procedure was based on the absence of documentation and the absence of persuasive other evidence, even if the firm claimed to have performed the procedure.
In its response letter to the PCAOB, the audit firm said that it conducted a thorough evaluation of the matters identified in the Board’s report and addressed the engagement–specific findings in a manner consistent with PCAOB auditing standards and KPMG policies and procedures.
The Board’s inspection report highlighted issues around the testing of fair value measurements and the use of third party pricing services. The PCAOB currently has a Pricing Sources Task Force examining issues pertaining to fair value measurements of financial instruments and the use of third-party pricing sources. Recently, the PCAOB’s Standard Advisory Group discussed the role of fair value pricing services.
In seven of the audits, the Board found deficiencies in testing the fair value measurements of, and the disclosures related to, hard-to-value financial instruments without readily determinable fair values, including collateralized mortgage obligations, and other mortgage-backed securities. The audit firm generally failed to obtain sufficient evidence to support its audit opinions.
In a separate audit, the Board found that the audit firm failed to obtain sufficient evidence to support its opinions on the financial statements and on the effectiveness of internal controls related to control and substantive testing with respect to the valuation of the issuer’s financial instruments. The issuer’s traders determined the fair value of financial instruments, noted the Board, and non-trader personnel performed monthly procedures to verify if the traders’ fair values were reasonable.
The price verifiers and the procedures differed depending on the type of financial instrument and its classification within the fair value hierarchy. The price verification procedures for financial instruments classified as Level 1 and Level 2 included an automated comparison of the trader’s fair value to prices for the same or similar instruments provided by pricing services or other third parties, comparing the inputs that the traders had used in issuer-approved models to value certain Level 2 financial instruments to available market data, or comparing the trader’s fair value to the fair value the price verifier had developed.
For Level 3 financial instruments with little or no market transparency, the price verifiers were required to perform some analysis to determine whether the fair value was within a reasonable range or document why such analysis was not possible. The issuer’s control procedures required a third group to investigate and resolve differences between the price verifier’s and the trader’s prices that were in excess of established thresholds.
For control and substantive testing, the audit firm selected the issuer’s portfolios that it considered to be significant. For control testing, the firm concluded that the issuer’s price verification procedures over all the diverse financial instruments in the portfolios constituted a single control that operated monthly. The Board found that the audit firm’s conclusion was inappropriate since it did not take into account the different price verification procedures the issuer performed, which ranged from straightforward automated procedures for Level 1 instruments to highly complex judgmental procedures for Level 3 instruments, and the different inherent risks and the fraud risk associated with the various financial instruments, especially the hard-to-value Level 2 and Level 3 instruments.
As a result of the audit firm’s inappropriate conclusion, said the Board, the firm’s sample size for control testing of the price verification procedures for Level 3 financial instruments was inadequate. The audit firm’s control testing of the price verification procedures for some Level 3 financial instruments also was insufficient because the firm failed to test whether variances between the price verifier’s and the trader's prices that were in excess of established thresholds were identified for investigation and appropriately resolved.
Further, for some Level 3 financial instruments, the auditor concluded that it did not need to change the nature, timing, and extent of its procedures despite issues that came to the firm’s attention regarding controls related to the valuation of these instruments. Moreover, with regard to model-valued financial instruments, the firm’s testing of the issuer’s controls to assess whether the proper approved models and inputs were used was insufficient to support the conclusion that the controls were operating effectively.
For example the auditor failed to identify and test important controls within the process, and failed to evaluate the implications of the firm’s identification of financial instruments that had been valued using models that the issuer had not approved for those specific financial instruments. The firm also inappropriately selected only one item, what the Board called a ``a test-of-one’’ to test the operating effectiveness of application controls that addressed multiple types of financial instruments, and therefore included multiple models and multiple pricing inputs.
The Board found that the firm’s substantive testing related to significant Level 3 portfolios was also insufficient because testing only one financial instrument for certain portfolios was inappropriate given the level of risk of material misstatement associated with such portfolios.
In addition, for certain hard-to-value Level 2 financial instruments, the auditor failed to obtain an understanding of the specific methods and assumptions underlying the fair value estimates that were obtained from pricing services or other third parties and used in the firm’s testing related to these financial instruments. Further, the firm used the price closest to the issuer’s recorded price in testing the fair value measurements, without evaluating the significance of differences between the other prices obtained and the issuer’s prices.