Forums › Ask ACCA Tutor Forums › Ask the Tutor ACCA AA Exams › Cut Off testing
- This topic has 5 replies, 4 voices, and was last updated 8 years ago by
Ken Garrett.
- AuthorPosts
- September 30, 2014 at 3:49 am #202218
How do you decide the sample size for cut off testing?
September 30, 2014 at 7:43 am #202225I don’t know of any scientific approach of selecting a sample size. It is essentially a substantive test, so look at major despatches and receipts just each side of year end. Probably comes to no more than 10 receipts just before/after year end and about 10 despatches just before after year end.
Note that there is no indication that deciding on sample size is on the syllabus.
September 30, 2014 at 8:00 am #202229Anonymous
Inactive- Topics: 0
- Replies: 5
- ☆
First you need to establish whether you have tested the controls over the area you are looking at. So for example if you are doing a sales cutoff you might have done controls testing over the process. If the cotrols have passed then that gives you a bit more confidence and you dont have to do a very extensive substantial testing ( test of detail)
I am not sure how most people do but we do it based on risk and the size and significance of the balance you are testing ( eg sales cutoff). The more you suspect that there will be issues ( based on past experience) the higher the risk so the higher the sample.
Normally the riskiest times are the 3 days before/after the cutoff date ( end of the financial year), then you can exted this to 1 week prior and 1 week post year end, lastly if you really suspect mistatement and you have not tested controls you might be looking at 2-3 weeks either side.
Once you have decided on the peiod you are testing you think about your materiality. Are there thousands of small transactions, or just a few large? If you would consider $1m a material mistatement and you have got thousands of $5 transactions going, then you might use computer assisted testing and might select a random sample or not test at all since the chance that there is a material misstatement is low. For a material misstatement to occur there have to be 200,000 transactions of $5 each posted incorrectly. Now what is the chance of that happening?
If you have got a few large items and a lot small, then you will test the large ones. If the untested balance is below materiality you can say well even if those are posted wrong it wouldnt make a difference.
If you have a wide range of items some high some medium some low you can you MUS and pick a sample based on their monetary amounts.
In many cases computer assisted testing is used where either the computer selects the samples for us and then extrapolates the expected misstatement in the population or the computer can help us as a data analysis tool highlighing unsusual transactions ( for example sales made a midnight when you know the office closes as 5PM)
Hope that helps,
AnnaOctober 3, 2014 at 3:28 am #203291OT Tutor…thanks for your response
Anna1023……thank you very much or such a detailed explanation, you definitely answered my question ! Your answer was very much appreciated. I’m guessing you’re a big 4 auditor who got extremely high marks in this F8 paper?
April 24, 2017 at 2:05 pm #383372Anna 1023…How to proceed in following circumstances:
1) The client has not maintained GRN and GDN records/Listings supporting the process
2) We cannot ask them simply to provide any 10 invoices on the clients discretion prior to and after year end received/issued by the client as it will not serve the purpose because it is highly likely that only invoices recorded in correct accounting periods will be provided by client.Please note the scenario is real and practical. Constructive input from any one will be highly appreciated
April 24, 2017 at 6:38 pm #383462It looks as though you might be jeading towards a modified audit opinion because of lack od,sufficient appropriate audit evidence.
- AuthorPosts
- You must be logged in to reply to this topic.