h
Login
Forgot Password

Please enter your email address to reset your password.

Signup
Incorrect password combination.
I agree to the terms of use
Chat
How to continuously audit journals while incorporating organizational context

How do you continuously audit journals and run routines over journals while incorporating organizational context and journal processing culture? Read this use case to find out more...

The Scenario

Journals represent an essential mechanism to account for financial outcomes of events. Journals can be automated and embedded in system functionality as part of the end-to-end processing of transactions or performed by employees where operational process design or limited system capabilities necessitate this. 

By their nature, journals also lend themselves to the processing of non-routine transactions during the preparation of financial accounts.  They are regarded as a “high risk” item by both management and the audit function. This is because journals have the potential to become a mechanism for financial statement misrepresentation or even to commit fraud. Yet, so many auditors only rely on a couple of standard procedures to identify and select non-routine journals for testing. These include journals with large amounts, journals ending in ‘999, journals with round numbers, journals processed on weekends/ public holidays and journals with specific descriptions/ blank descriptions. 

All these procedures are applied without considering the organizational context or the standard business processes.

This ALICE use case applied to a large manufacturing and distribution company that processes a high volume of journals.  This process is usually audited on an annual basis by the Internal Audit function – which typically focused only on material journals. 

The client is also part of a group of companies that are managed through a federated model. Therefore, group-level management had very little visibility into the journal culture of its organization.

ALICE Procedures & Outcomes

ALICE analyzed and identified journals that were proactively flagged for further investigation by management and the audit function. 

  • The journals were flagged based on intelligent criteria combinations (contextualized risk) applied to the data asset. 
  • These outcomes also provided management with continuous insight into the journal culture and processing behaviors inside the organization. 

Data Assets Used

In this digital service, a curated data asset was created using extracts from General Ledger containing various fields and flags uploaded to ALICE via the Bespoke Generic Connector. 

Examples of the fields and flags included: 

  • Journal reference 
  • Manual versus automated 
  • Value 
  • Posting date 
  • Effective date 
  • Reversals 
  • Financial statement line item (“FSLI”) details and mappings

The ALICE Lab enabled easy access to data and collaboration. Predictive insights could be drawn for various scenarios, sorted based on multiple parameters. 

Human Intelligence

Of course, audit and assurance leaders were crucial to the entire process. “human intelligence” was much needed for creating reusable algorithms. Once the “intelligence criteria” was decided, it had been converted into a method for ALICE to follow, as detailed below: 

  • Design of the intelligent criteria combinations, for example:

Identification of “normal” versus “abnormal” for this industry;

Items posted in the current period for the previous period and then reversed; and

Abnormal FSLI combinations.

  • Identification of journals meeting the intelligent criteria combinations.

The Outcome

The ALICE Lab used data science and artificial intelligence to derive critical insights for the audit function concerning these journals. 

Data Science 

  • Data modeling over indicative flagging of “peaks” and “troughs” for further investigation purposes. 
  • Data modeling on “high value” representations based on the routine journals processed by the organizations.
  • Data modeling to identify “normal” journals and flagged journals as outliers.

Artificial Intelligence 

  • Machine learning applied to journals flagged for further investigation but resulted as a “false-exception” to train the data model mentioned above to more accurately identify valid exceptions.

Do you need assistance with obtaining insight into your organization's journal culture and continuously testing non-routine journals? Employ ALICE as part of your internal audit or governance team.

2020 | 8