Frequency of manual controls performance: 1818



  • To determine sample sizes for any given control the frequency of the control’s performance must be established. My question then is: is the frequency based on the number of transactions that occur, or the number of times the control is applied to those transactions.
    To illustrate: payment on expense reports is verified weekly, however, there are 10-15 expense reports process every week. Is the frequency of this control weekly, or multiple times per day?
    Thanks.



  • It’s the number of times the control is applied to the transaction.
    In your example. it will be weekly, a you are testing that the control takes place and is effective as opposed to the actual transactions.



  • Thanks for the quick reply.
    Based on what you’re saying then, the samples taken when testing this control should be (approximately) 5-15, on an annual basis, or between 2-4 every quarter, depending on the level of risk associated with the control. These sample sizes are based on guidance from one of the big 4.
    An interesting point to make is that in my last job, the control in my example would have been assigned a frequency of multiple times per day. Do you think there will ever be a set of universal rules applied to the 404 assessment?
    Thanks again.



  • yes, most guidance would advise about 5 -15 samples on an annual basis at a minimum. my experience of what would normally happen if testing on a quarterly basis would be to take a full sample of 5, 10 or 15 based on how risky any non-occurence of the control activity would be, and, if 100% of samples pass, your sample size can be reduced when testing is completed at year-end.%0AMis-interpretations of control frequencies are quite common. it is vital that when determining the frequency that it is the actual control activity that is reviewed and not the base transaction. I think that what is needed to clarify these issues is further guidance from the PCAOB and AS2.



  • Hi and welcome to the forums 🙂
    EMM gives excellent advice (as usual) and it’s always good to plan specifics on sampling criteria with Audit. As some of this might be open to interpretation, this can be more of ‘art than science’ 😉
    If you search our forums, with the keyword of ‘sample sizes’ you’ll get at least 80 hits. Below are some of posts which might help in planning sample size requirements:
    Posts discussing sample sizes based on control frequencies
    http://www.sarbanes-oxley-forum.com/modules.php?name=Forums-and-file=viewtopic-and-t=1780
    http://www.sarbanes-oxley-forum.com/modules.php?name=Forums-and-file=viewtopic-and-t=1811
    http://www.sarbanes-oxley-forum.com/modules.php?name=Forums-and-file=viewtopic-and-t=1640



  • EMM gives excellent advice (as usual) and it’s always good to plan specifics on sampling criteria with Audit. As some of this might be open to interpretation, this can be more of ‘art than science’ 😉
    You would think that interpretation would be provided by the SEC and the PCAOB for matters such as this, and not left up to the EA’s (and management) to define for themselves. Hopefully, the releases/updates due out from the SEC and PCAOB relating to guidance for management on section 404 and the new version of AS2 will better define issues like this.
    Thanks again for the information.


Log in to reply