Skip to content

← Accounting

The Auditor

Chapter 27 · Keeping Everyone Honest

In this chapter 4 sections
  1. What “true and fair” promises
  2. Who may be the auditor
  3. A visit to the shop
  4. Four ways to sign

In 1896 a court in London had to decide what an auditor is for. The Kingston Cotton Mill Company had gone bust, and when the people winding it up looked at its books they found that for years its managing director had been writing down more cotton and yarn in the mills than were ever there. With the stock puffed up, the mill had seemed to make profits it never made, and dividends had been paid out of them. Every year the auditors had copied his stock figures into the balance sheet with the words “per manager’s certificate” and signed. The company’s liquidators said they should pay the money back. A judge agreed with them. The Court of Appeal didn’t. It was no part of an auditor’s duty to count the stock, the judges said, and the manager was a trusted man with nothing about the figures to arouse suspicion. One of them, Lord Justice Lopes, put it in a line that has outlived the case: an auditor “is a watch-dog, but not a bloodhound”.

The case is remembered for the line, and for how much has changed since. The report of the case notes that a simple sum, the stock at the start of the year plus what was bought less what was sold, would have shown the figure was so large as to call for an explanation. Today the auditing standards make an auditor attend the stock count wherever the stock matters, unless that is impracticable. The watchdog has been asked to sniff a good deal more.

An auditor, लेखापरीक्षक, is an outsider with the right training who examines a business’s books and the evidence behind them, and then says in writing whether its accounts show a true and fair view. The managers, or the owner, make the accounts. The auditor checks them and gives an opinion. That opinion, and the care behind it, is the whole product: an auditor must bring “that skill, care, and caution which a reasonably competent, careful, and cautious auditor would use”, in Lopes’s words, and one who doesn’t can be made to pay for the loss, as the liquidators tried to make Kingston’s auditors pay. It is why a bank lends against audited accounts and why shareholders who never see the shop floor can trust what the directors tell them.

What “true and fair” promises

The people who own a business are often not the people who run it. Shareholders put in money and leave its running to directors; a sleeping partner leaves it to the one at the counter. Those who run it are stewards of other people’s money, and once a year they owe them an account of what they did with it. The trouble is that the account is written by the very people it reports on. An audit is the owners’ way of having someone check it who has no reason to make it look good.

The phrase the check is made against is old. Britain’s Companies Act of 1948 required every company’s balance sheet to “give a true and fair view of the state of affairs of the company” and its profit and loss account to do the same for its profit. Nepal’s Companies Act, 2063, asks the same in its own words: a company’s books must show the true state of its business clearly (section 108). Britain’s accounting regulator says plainly that its law has never defined the phrase. What they have come to mean is roughly this: not exact to the rupee, which no set of books can be, but free of anything big enough to mislead someone who relies on them.

That idea is materiality, which Chapter 5 met among the bookkeeper’s rules; for an auditor it sets the size of the job. The International Standards on Auditing, written by an international board and used in Nepal as the Nepal Standards on Auditing, call a mistake material if, alone or added to others, it “could reasonably be expected to influence the economic decisions” of the people using the accounts. A wrong figure of Rs. 90 in a paint shop’s year doesn’t change anyone’s mind. A wrong figure of Rs. 90,000 might. Where the line falls is the auditor’s judgment, and the standards give an example: five percent of the profit before tax for a business that trades for profit, though higher or lower can be right.

The standards also say what an audit promises. It gives reasonable assurance that the accounts are free of material misstatement, “whether due to fraud or error”. Reasonable assurance is “a high level of assurance, but is not a guarantee”. An auditor looks at a sample, not at everything, and much of the evidence persuades rather than proves. And it is the managers, not the auditor, who are responsible for the accounts and for the controls that keep them right. The auditor finds mistakes, fraud among them, but the audit is built to catch what is big enough to matter, not to find every rupee that has gone missing.

So an audit has three objects: to find the honest mistakes, an entry left out, a sum posted to the wrong account, an expense recorded as an asset, two mistakes that cancel each other, an entry made twice; to find the dishonest ones, figures falsified and money taken; and, by being there, to make people at the counter less likely to try either. Accounting makes the books; auditing tests them. It needs everything accounting knows, and more: the law, the business, and the habit of asking for the paper behind each figure.

Not every audit is of this kind. Internal audit, आन्तरिक लेखापरीक्षण, is done by the business’s own staff or by people it hires, all year, for the managers, to see whether the controls work and the money is spent as it should be. Final audit, अन्तिम लेखापरीक्षण, is done once the year is over by an independent outsider the law provides for, and its report goes to the shareholders, or, for public money, which in Nepal the Auditor General audits, to the President, who has it laid before Parliament. Internal auditors work for the managers and can be told what to look at; final auditors are chosen by the owners and can’t be. Audits are also sorted by what they look at: the financial accounts, the work done for the money spent, the managers’ decisions, whether laws and rules were followed, the computer systems, taxes; and by when: continuously, in visits at set times (interim audit), or once at the year’s end, on part of the books or the whole.

Who may be the auditor

An auditor’s whole worth lies in being independent. One who owes the business money, owns a large share of it, keeps its books or is married to its manager is checking himself, or his family. So the law draws lines. Nepal’s Companies Act, 2063, says every company shall have its accounts audited (section 110), by an auditor chosen from those who hold a license to audit, normally by the shareholders at the annual general meeting (section 111). It lists who may not be one (section 112), among them a director, an adviser or an employee of the company, or their partners and close relatives; anyone who owes the company money; anyone holding one percent or more of its paid-up capital; and anyone convicted of an offense relating to audit in the last three years, or of corruption or fraud in the last five. Before taking the job the auditor must say in writing that none of these applies, and an audit by someone who was barred doesn’t count.

The same Act gives the auditor rights to do the work. The company must show its books whenever the auditor asks during office hours, and its directors and staff must answer the auditor’s questions truthfully (section 114). The report goes to the shareholders and must say whether the auditor got all the information and explanations needed, whether proper books were kept, whether the accounts follow the accounting standards and agree with the books, whether they fairly show the company’s position, profit and cash flows, and whether anyone has acted against the law, misused its property or committed an accounting fraud (section 115). The auditor signs and dates it personally (section 116). And an auditor who has found something awkward can’t simply be sacked: until the year’s audit is done, removal is allowed only for misconduct or acting against the company or the law, after telling the Institute of Chartered Accountants of Nepal, with the regulator’s or the Company Registrar’s approval, and with a fair chance to answer (section 119).

Who holds a license is settled by the Nepal Chartered Accountants Act, 2053, which replaced the Auditors Act of 2031 and set up the Institute of Chartered Accountants of Nepal, ICAN, to train, examine and govern the profession. It has two kinds of member (section 16). Chartered accountants have passed ICAN’s examinations, or a recognized foreign body’s, with practical training. Registered auditors, दर्तावाला लेखापरीक्षक, are those who were licensed under the old Act, now in classes B, C and D. Nobody may offer audit services without ICAN’s certificate of practice (section 29). A chartered accountant may audit a business of any size. Each class of registered auditor has a ceiling, in rupees, on the audits it may take, which ICAN’s council revises every three years by the central bank’s price index: from साउन 1, वि.सं. 2078, up to Rs. 1,200 million for class B, Rs. 300 million for class C and Rs. 60 million for class D. A revision since may have moved them.

The Act also sets a code of conduct (section 34), and it reads like a list of ways an auditor could stop being independent. An auditor may not share fees with anyone who isn’t a member of ICAN, or take a fee that depends on the profit shown; must keep the client’s affairs confidential except where the law asks; may not certify statements that the auditor’s own firm didn’t examine; must disclose any interest in the business; must report every material fact known and every misstatement found; may not audit an organization they left as an employee within the last three years. Behind the rules are the qualities a good auditor needs, honesty, skill and care, knowing the law and the trade, tact, restraint and fairness, and the principles behind every audit: independence, competence, confidentiality, legality, materiality, fair presentation, due care and ethics.

A visit to the shop

A paint shop is not a company, and the Companies Act’s audit sections don’t reach it. No law found while writing this book makes a sole trader or a partnership like Yasmin’s have its accounts audited. Yasmin and Gaurav did it anyway. When Gaurav joined, the partners agreed that the shop’s accounts would be audited each year: he paints houses all week, sees the books on Saturdays, and wanted someone from outside to say they were right. That is stewardship again, in a shop. The bank, which had lent Rs. 200,000, liked the idea too. The auditor they chose for their second year was Anisha, a chartered accountant with a small practice in the town, who had never kept their books, owed them nothing and was no relation. Everything about her audit, like every figure from the shop, is made up for this book.

She started before the year ended, by asking how the shop worked. Who takes the cash, who orders, who signs cheques, who keeps the bin cards. A business with good controls needs fewer checks, and one without them needs many, so learning the controls first tells an auditor where to look hardest. Then on असार 31, the last day of the year, she came to the stock count. She watched the helper and Yasmin count, and counted some racks herself, a few lines from the count sheets to the racks, and a few from the racks back to the sheets, so that tins neither added nor left out could hide. Behind the enamel she found twelve tins of a shade the maker no longer made, counted at their cost of Rs. 900 each. Sujan wouldn’t take them back and they would fetch, at best, Rs. 450 each. Stock is valued at cost or at what it will fetch, whichever is lower, so the tins were worth Rs. 5,400 less than the sheet said. On its own that was well under what would matter, but it was wrong, and the partners put it right: the draft profit of Rs. 285,440 became Rs. 280,040.

Five percent of that profit is about Rs. 14,000, and Anisha took it as her line: a misstatement, or several added together, of more than that would matter to the partners and the bank. The year’s payments had left 961 vouchers in the file, the bill or receipt behind each payment, Rs. 4,728,050 in all. Checking a voucher, called vouching, means taking the entry in the books and finding the paper behind it: a bill in the shop’s name, for the shop’s business, of the same amount, approved by the person who should approve it, and paid when the books say. She checked all 24 of the big ones, every voucher of Rs. 50,000 or more. They were Sujan’s bills, Rs. 2,855,200 between them, and each agreed with his invoice and the helper’s note of what arrived.

That left 937 small vouchers worth Rs. 1,872,850. Checking them all would take days and cost more than it found. Instead she took a sample: every 23rd voucher, starting from the 4th, a number picked at random so that she couldn’t choose which ones to see. That gave her 41 vouchers worth Rs. 83,550. Picking at a fixed interval from a random start is one of the ways the standards allow; picking at random throughout is another. What matters is that every voucher has a chance of being chosen. A sample has a risk the whole doesn’t, that it happens to miss what is there or show what isn’t, and the standards ask the auditor to keep that risk acceptably low.

Forty of the forty-one were right. One was a tinter bill for Rs. 3,780, entered in the books as Rs. 3,870, the two middle digits swapped. The error was Rs. 90, about a tenth of one percent of the sample’s value. An auditor doesn’t stop at the error found. She projects it onto the vouchers she didn’t check: Rs. 90 is 0.108 percent of the Rs. 83,550 she looked at, and the same share of the Rs. 1,872,850 that all the small vouchers came to is about Rs. 2,017. Projecting by value is a common way; the standards require a projection but leave the arithmetic to the auditor, and the answer is a rough estimate, not a count. Rs. 2,017 was about a seventh of her line, well inside the smaller margin the standards have an auditor keep for each kind of item, so that errors in several places can’t add up unseen; and it was a slip, not a theft, so the sample needed no more than a note. Had it come to more than the line, she would have checked more vouchers, or asked the partners to go through all of that kind and correct them.

Vouching has a blind spot, and Anisha knew it. It starts from the books and looks for the paper, so it can find an entry with no good paper behind it, but never a sale or a bill that was never entered at all. For those an auditor runs the test the other way: from Sujan’s delivery notes to the books, from the bill book’s numbers to the sales, from the year-end statements to the ledger. She wrote to the bank and to Sujan herself, asking each to confirm, directly to her, the balance the shop owed or held at असार 31, and both replies agreed. She checked the bank reconciliation, the depreciation on the scooter and the racks, the debts owed by builders and whether any looked unlikely to be paid. Each check went into her working papers, what she looked at, found and concluded, so that another auditor could follow her work.

At the end she wrote her report. The standards say what a report contains. A title saying it is an independent auditor’s report, the people it is addressed to, then the opinion first, under its own heading. Then the basis for it: that the audit was done by the standards and that the auditor is independent. Then what the managers are responsible for, and what the auditor is, and the signature, place and date. Anisha’s was addressed to the two partners. The report is the whole document; the opinion is the one paragraph that matters most in it, the auditor’s conclusion in a sentence. Hers was clean.

Four ways to sign

A clean opinion, called unmodified in the standards and often unqualified, says in the words of the standard that the accounts “present fairly, in all material respects” the business’s position and results, or, in words the standards treat as the same, “give a true and fair view”. Often it comes after the managers have fixed what the auditor found, as Yasmin and Gaurav fixed the old tins. When something can’t be put right, there are three other opinions, and which one comes depends on two questions: what is wrong, and how much of the accounts it touches.

Figure 27.1 Which opinion an auditor gives. Across: how much the problem matters. Down: whether the auditor found the accounts wrong, or couldn’t get the evidence to know. The wording is from the International Standards on Auditing, ISA 700 and ISA 705.

Suppose the partners had refused to write down the twelve tins, and suppose the tins had been worth much more, enough to matter, but nothing else in the accounts was wrong. Anisha would give a qualified opinion: the accounts present fairly “except for the effects of the matter” she describes. A problem that is material but confined to one part of the accounts earns a qualification. If the error ran right through the accounts, so much of the stock invented, say, that the profit and the balance sheet were both meaningless, the opinion would be adverse: the accounts “do not present fairly” at all. The standards call a problem like that pervasive: not confined to one part, or a large share of the whole, or something a reader must know to understand anything else.

The other row is for when the auditor can’t find out. Suppose the stock had been counted without her and the count sheets lost. If only the stock was in doubt, and it mattered, she would qualify again, this time for the “possible effects” of what she couldn’t check. If the doubt reached everything, the whole bill book lost, the bank refusing to reply, she would give a disclaimer: the auditor does not express an opinion, because she couldn’t get enough evidence to have one. A disclaimer isn’t a gentler word for bad news: it tells a bank that nobody can say what the accounts are worth.

A modified opinion is meant to hurt. A bank reads a qualification before it reads the profit. That is what the watchdog is for. It doesn’t catch every fraud, and Kingston Cotton Mill’s auditors didn’t catch theirs, but because it is outside the business, can’t be easily dismissed and must say in public what it found, the people who run a business know their account of themselves will be read by someone with no reason to believe it. Most of the time that is enough to make the account true.

Found a mistake? Please tell us.