In a prior video we talked about the direct method of testing. The direct method of testing involves testing actual source documents, lining up and comparing them when you have a breakdown in the direct method. Then sales tax auditors will resort to what they call indirect methods. Indirect methods of testing is a fancy way of saying we're going to play guessing games with statistics. So one of the indirect methods of testing is audit path sales. They look at current sales and they'll do statistical comparisons between past sales and current sales. So one of the easiest ones they do is they do an observation test. They'll send an auditor in a business for a couple of days to look at the sales that are being performed, whether the employees are ringing everything up correctly, whether they're charging tax and then the auditor will sit there and literally record every single transaction and they'll compare that against the POS system reports to see if there are any discrepancies. That's called an observation test so if there is an error within that test, then they'll do certain things based on

the error. The other thing they can do is they can take the POS system reports in a current period and do a cash to credit card ratio. So a lot of the times they'll take your credit card ratio which can be verified by your 1099 KS and they'll say okay here's the percentage of credit card sales based on current. Here's the amount of cash sales and we're just going to project that across the board. So the problem that we see with indirect math testing - let's take the classic credit-card ratio for example. The auditor comes in and says I'm going to do two days of testing in a business. I'm going to do a restaurant and I'm going to come in Tuesday at lunch time and I'm going to come in Sunday afternoon. So what the auditor's saying is based on two days of testing. That two days of testing is going to be used for a sample. They're going to take that sample and they're going to apply it to a population (a total data set). They're going to take a two day sample and they're going to apply it to a three-year period of a population so the three year population is over a thousand days. They're going to take two days and make those representative of this entire thousand days of sales. Hopefully you can see the problems with that. Some businesses do more credit card sales on the weekends, some businesses are seasonal and will do sales more frequently at certain times and less frequently on others. With restaurants you have families going out on Saturday nights for dinner - those tend to be larger checks and so you tend to pay larger checks on credit cards unless you're walking around with a fistful of cash. So depending on a few little things, this is why indirect methods of testing are so dangerous. Little hinges swing big doors so these little statistical changes can make a huge result as to the liability. They could have massive impact and we've seen situations like this in the firm. We have a client that the CDTFA came in and they said "look we think that your client has underreported three million dollars in cash sales." I look at the agent and I said if you really think they underreported three million dollars in cash sales,

where's all the cash? They didn't deposit three million dollars in cash in the bank so what you're telling me is they went home, they stack the money under their mattress, they got three million dollars in cash sitting like this giant brick under their mattress and they've just been hiding it. And where have they been going around and spending all this cash all these years? It is not supported by any basis in reality and if you actually thought they were under reporting by any cash you would have sent the guys with the windbreakers in the handcuffs. So that's why indirect methods are so dangerous and I use that as an absurd example because it was an absurd example but that's what happens when you have these statistical things. So you need to be very very careful with indirect methods of testing and one of the reasons I advocate for having somebody represent you in a sales tax audit is because of these situations. We find so many statistical errors. These auditors aren't statisticians so we find so many errors in the way that they're doing their statistical analysis. The tightness of their control, their procedures, the way they're applying samples to populations, I mean it's just a mess so that's the way that indirect methods of testing work. That's frankly why they're so dangerous, so whenever possible you want to avoid indirect methods of testing and focus on direct methods of testing because direct methods of testing are much more reliable than their indirect brother.