Ahack

To eliminate the amount of time wasted on performing audit procedures which can be replaced by machines.

Comment

Inspiration

As students in Accountancy, we learn of the tedious workload auditors constantly worry about. Also, our firsthand experience in vouching and tracing procedures pushed us to explore how this indispensable procedure can be made more efficient. Armed with a curious mind, a passion for digitization and an ideal outcome in mind, our team sought to design a means to automate this tedious process while preserving the purpose of the audit procedure.

The Problem

Audits account for the largest proportion of revenue for most accounting firms but has one of the lowest productivity based on the AECensus 2016/2017 report issued by Singapore Accountancy Commission. This low productivity of audits have been attributed to the nature of work which involves a huge amount of manual labour. Our research has led us to discover that the intensive labour and high pressure arising from the need to produce high-quality work in short timelines led to high turnover rates of approximately 39% in the finance industry belonging to the audit and accounting profession. This creates a manpower issue for accounting firms that rely heavily on audits for revenue, potentially impeding productivity further, driving down audit quality and profitability. The heavy workload on auditors has also caused large firms to appear in news headlines due to auditors who die of overwork, damaging public opinion of the auditing profession.

Why automate vouching & tracing

Vouching and Tracing are tests of details used in most classes of transactions to test a range of assertions. Thus, it is an essential and indispensable procedure in an audit. Yet, these processes involve tedious and extensive manual work that does not demand the expertise of accountants. These highly manual and time-consuming procedures drive the low productivity in audit. To address this issue, our team decided to focus on designing a solution to automate these fundamental procedures to streamline the auditor’s workload and timeline.

Vouching & tracing workflow

Vouching and tracing are substantive audit procedures used to test various financial statement assertions. Vouching tests for occurrence/existence by selecting samples of transactions recorded from a ledger and matching them to source documents showing transactions recorded had occurred while tracing tests for completeness by selecting samples of source documents and checking if transactions were recorded in the ledger. On hindsight, these two procedures involve similar processes which can be executed with a single program where only the input fed to the program is changed based on the procedure carried out. Aside from that, relevant assertions checked in the midst of these procedures include cut-off, accuracy and authorization and these checks do not differ between vouching and tracing.

What it does

In designing our program workflow, we used the relevant assertion tested in vouching and tracing as a basis for the inputs required by the program, data extracted from inputs and the output generated. The output generated will be in the format of an auditor's working papers.

Inputs required:

  1. CSV/Excel file containing sample of transactions in ledger (for vouching) or all transactions in ledger (for tracing)
  2. Image(png) or pdf files of Source documents obtained from clients
  3. Master list of authorised signatures

The program first uses Optical Character Recognition program to convert scanned documents into text to enable further processing to be carried out on the text documents. Following that, the program would provide the following:

  1. Checklist for user to select the assertions to be tested
  2. Choice between 1-way (1 source document), 2-way (where 2 source documents are needed) or 3-way matching of source documents (where 3 source documents are used) to determine the need for program to match between source documents supplemented
  3. Start and end date for the given financial period

In cases where 2-way or 3-way matching is selected, the program would match between 2 or 3 source documents through reference numbers and check that dollar amounts in each document are the same. Following that, the program will then check the relevant data found in source documents against that of ledger records for each assertion to be tested. The following entails the assertions tested and the checks done:

  1. Occurrence/Existence (Vouching): check if document reference numbers found in ledger sample is in the set of source document supplied
  2. Completeness (Tracing): check if document reference numbers found in source documents supplied is in the ledger records
  3. Cut-off: check if dates found in source documents are within the user specified financial period
  4. Accuracy: check if dollar amounts of source documents and ledger records are the same
  5. Authorization: check for presence of signature that matches master list

The program will then generate an output containing 2 tabs in an excel / in csv file. This includes:

Tab 1: Spreadsheet containing the percentage of errors detected in the samples, together with the information from source documents and ledger balances with True/False color coded columns indicating pass/fail of each assertion tested. This tab is for the auditor’s reference to sieve out the entries with errors and to make the necessary adjustments. This tab will be removed prior to submitting the audit workpaper (Tab 2) into the current engagement file.

Tab 2: This is the actual workpaper that will be uploaded by the auditor into the current engagement file. Auditors will refer to tab 1 for the list of errors and make the necessary adjustments (eg. checking the actual documents themselves or clarifying with clients) in this tab before uploading.

How we envision it to be

We envision our software to be a desktop application that seamlessly incorporates the functionalities necessary to automate most of the work done in the vouching and tracing procedure from checking of assertions to producing audit documentation.

Challenges we faced

As students in accountancy, we could not produce a working software within the time constraint as it involved skills that were out of our expertise. We faced this challenge by placing our focus on designing a solution that would cater to the needs of an auditor while keeping in mind the state of current technologies to ensure our program is feasible.

Accomplishments that we're proud of

We successfully mapped the relevant assertions checked during a typical vouching/tracing procedure to the data found in the input files required and incorporated them into a comprehensive IT solution.

What we learned

This project has led us to explore how the different technologies that are currently in the market can be utilised to build a customised solution for the vouching and tracing procedures. In doing so, we have gained insights into the various ways in which machine learning is used in existing technologies and a deeper appreciation for the audit process.

What's next for Ahack

The next step for us would be to look into the existing technologies used in current accounting firms in order to compile and integrate current process into our program as we develop the software.

Share this project:

Updates