At what point will an incorrect model outweigh an otherwise excellent audit trail and summary, and thereby cause you to fail? I know this exam focuses on the audit trail and summary but what if your model just isn't anywhere near the correct answer, yet your audit trail and summary for the work you did do were really great?
When I did CA2 we were provided with an example mark scheme on the first day. Although this will naturallly vary from one project to another, you can see the mark allocations in each section, e.g. audit trail, spreadsheet model, etc. so you should be able to get a feel for the answer to your question. In the exam, time pressure will probably dictate that you have one go at getting the model right, without too much debugging time, and then move onto the audit trail and documentation. One final thought: on the first day when I took CA2, the supervisor told a story of someone in a situation of the kind you describe, i.e. excellent written work but fundamental mistake in the model. The supervisor said he just about managed to award a pass, BUT if only the candidate had asked the supervisor during the exam about the issue he slipped up on the whole problem could have been avoided.
Thanks td290 That last paragraph is encouraging. I'll remember to make use of the online tutor as I'm doing the online version..!
It depends on the nature of the mistake in your model. If you have just made some small mistake such as adding a column instead of multiplying or something then it's not likely to fail you. However if you have actually misunderstood the question or your model doesn't satisfactorily answer the question then a good audit trail and summary won't save you. After all, it's no use being able to write pretty reports and document models well if you can't understand what you're being asked.
It's quite some time since I did it, but when I did there were very little marks allocated for the actual spreadsheet and for getting help. So I guess you could make a complete mess of the spreadsheet and still pass. I strongly recommend that you don't let it get that far by checking with the supervisor periodically along the way to ensure you have a sensible approach and answers. However general reasonableness tests are part of the model building / audit process, which should keep you from straying too far off the right path.
