Patterns of both financial and non-financial data can reveal a wealth of information, and even unearth phantoms.
Every accountant or auditor applies trend analysis to reassure himself as to the reliability and dependability of the financial data he is concerned with. Trends are patterns in data viewed over a span of time. For example, purchases, sales or gross margins examined over a 10-year period might indicate a pattern which could be used for predicting possibilities in the foreseeable future. Trends are usually examined to evaluate ‘goodness of fit’ of the current data in that pattern. If there is any abnormality or deviation, there would be a strong reason to believe something is amiss. To illustrate: if the gross margin has been about 18 to 20 per cent over the last 10 years. a fall of gross margin to 10 percent in a particular year is very likely to indicate some error or fraud in sales, purchase or stock figures. Trends, are indispensable tools to every person concerned with finance, accounts, audit and investigations.
However, in investigations, the usage of trends is even wider. Trend analysis is not restricted only to financial data. A lot of findings emanate from even non-financial data. For example, data relating to number of sales deliveries to a certain customer can reveal whether any nexus exists between the customer and the sales manager. Patterns in over-writings in source documents, number of cancelled cheques, even the time of arrival and departure of employees may offer clues, depending upon the circumstances. The following is one such example. .
In a consumer products’ manufacturing and trading company with 13 factories, 20 regional offices and its head office in Mumbai, the total number or employees exceeded 5,000. The payroll processing was centralized in Mumbai and each employee’s pay slip was generated from is Mumbai. A risk assessment was being carried out and the fraud examiner was examining the payroll and HR related costs. The fraud examine observed that in the last three years, the payroll costs had gone up by about 40 per cent even though the number of additional employees had increased by only 12 per cent. Even allowing for increase in costs on account of inflation and extra bonuses, according to the fraud examiner, the increase should not have been more than 28 per cent. Further, investigation was deemed necessary.
Using his investigation software, he applied several audit tests on the financial data on the basis of which he verified pay slip calculations for correctness. He also checked computation of basic salaries, HRA, bonus, and deductions for professional tax, provident fund, income tax. His findings revealed that the software used by the company was flawless. He now decided to test certain non-financial data to reassure himself that the payroll was correct. He applied ‘abnormality’ tests on the leave and attendance data to examine the relative behaviour of the employees. During this review he came across a strange finding – 42 employees had not taken any leave for the last five years. No sick leave, no leave for any festival, or any other purpose.
This company had electronic digital cards issued to each employee for entry and exit in the premises of the company. A data analysis of the ‘in’ and ‘out’ tame showed that each of these 42 employees came exactly at 9 am and left exactly at 5 pm. They never traveled for the company anywhere out of Mumbai. They never sat late and never claimed any medical allowance nor any conveyance. Full tax was deducted from their salaries, since they never submitted any declaration of any tax saving investment. They had never collected their salary certificates, which were still lying in the files of the company.
On complete investigation, it was found that the total payments made to 42 ‘ghost’ employees during the last three years exceeded Rs 25 lakh. This fraud was conceived of by the HR manager, who felt there would be ‘safety in numbers’. He had doctored the paperwork for recruitment and appointment of these ‘ghost’ employees. He calculated that such a large payroll of the over 5,000 employees would easily shield the 42 ’phantoms’, and even if any of them was verified and audited, the arithmetically accurate pay-slip computation and the supporting paperwork would not raise any doubt. He was almost right – the fraud was not caught for about five years.
Clearly, every risk assessment needs to develop a trend analysis model, which can be used by the company management regularly to detect fraud and minimize leakage. Trends to an auditor (or fraud examiner) are what the heartbeats of a patient are to a doctor.