Remember CAATs? This was an acronym for Computer Assisted Audit Tools – a general category for all things computery that helped us work with more efficiency and power.
Now that virtually all we do uses a computer, ISA 315 (Revised 2019) does not refer to CAATs but to Automated Tools and Techniques (ATTs).
This kind of thing gets audit software developers like us salivating like Fluffy when the fridge door opens. But let’s stay calm and examine what the standard says first.
So what do we know about ATTs?
The standard doesn’t define ATTs, however the recently issued IAASB First Time Implementation Guide simply calls them “procedures performed leveraging the use of technology”. These may be used for risk assessment procedures, and also for obtaining audit evidence. The IAASB points out that:
The procedures for obtaining audit evidence as set out in ISA 500, Audit Evidence, i.e., inspection, observation, external confirmation, recalculation, reperformance, analytical procedures and inquiry, continue to apply, regardless of whether those procedures are performed manually or using technology.
In matters like this, the new standard helpfully acknowledges that we may be auditing vastly divergent entities. Paragraph 9, titled ‘Scalability’, states:
This ISA (NZ) is intended for audits of all entities, regardless of size or complexity and the application material therefore incorporates specific considerations specific to both less and more complex entities, where appropriate.
It is up to the auditor’s judgement to determine whether to use an ATT or some more manual procedure. For instance, there would be no point in carrying out fancy data analytics for fixed assets additions where there are only a few items. Better to just use judgement. ATTs come into their own where there is so much data, or a level of opaqueness, such that the auditor cannot possibly just ‘eyeball’ the content.
Examples from the standard
Looking at some of the suggestions for the use of ATTs in the explanatory material, paragraph A21 suggests performing “risk assessment procedures on large volumes of data (from the general ledger, sub-ledgers or other operational data) including for analysis, recalculations, reperformance or reconciliations.”
Most of our users tend to do this by entering the trial balance data for up to four years, and then populating analytical review pages that show current to prior year movements, deviations from the budget if required and various key ratios over time. Identified risks may then be flagged and analysed as required directly from the TB or AR pages. Detailed recalculations, reperformance or reconciliations tend to be best done using a spreadsheet and adding to the file as an attachment.
Paragraph A57 suggests that the auditor use ATTs “to understand flows of transactions and processing as part of the auditor’s procedures to understand the information system.” This may provide insight into vendors, customers, and related parties, simply by sorting say a purchases ledger in a spreadsheet by supplier name, or using a search function to look for known related parties.
Paragraph A137 suggests using direct access to the entity’s database “by tracing journal entries, or other digital records related to a particular transaction, or an entire population of transactions, from initiation in the accounting records through to recording in the general ledger.” Typically an auditor is given access to say the Xero ledger and may use the built-in search functions there to drill down into the data for this purpose.
Paragraph A161 suggests that when reviewing journals or ledger accounts in less complex entities inspection of all the entries within a particular account, or all journals may well be possible. But in a more complex entity downloading to a spreadsheet and applying filters and sorting may give a good result.
In Audit Assistant, we provide a built-in sampling tool. A large dataset is extracted out of the client software and then uploaded. A sampling interval based on performance materiality (or adjusted performance materiality) is added. This generates a randomised CMA sample. Appropriate tests are then added to a generated table of results. Alternatively, a random sample of a specified number of samples may be generated, or the auditor may carry out their own sample in the spreadsheet first and then upload it for testing.
The auditor is encouraged to use automated techniques to assist in the identification of significant classes of transactions, account balances and disclosures in paragraph A203. This would typically only be helpful in complex entities. In less complex entities these become fairly obvious by reviewing trail balance and analytical review data as described above.
CAATs are our friends
So CAATS and ATTs are really not some big scary monsters that need to intimidate us. They are our servants – power tools that we pull out when a normal auditing screwdriver or hammer is too slow or not forceful enough. They become dangerous when the auditor uses them ‘just because they can’ without understanding what they are trying to achieve and why.
There is no substitute for learning to do the basics well and always working from first principles, and choosing a tool that we understand and can explain that will achieve our objectives most efficiently.