AA to the rescue

The accruals anomaly is one of the mysteries of investing in company shares that allows investors to make higher returns than they should, ranking close behind the January effect, the small-cap effect and the low-beta effect. In a way, it has been known since the dawn of rigorous investment research. The great Benjamin Graham drew attention to it in Security Analysis, the ground-breaking volume he co-authored with David Dodd, which was published in 1934. Graham wrote about the steps needed to discover what he labelled a company’s ‘earnings power’. This was the underlying profit that a company seemed capable of making through good and bad and, from an investor’s perspective, was well worth estimating.

Finding earnings power, said Graham, involved stripping away the accounting items (usually profits, but costs, too) that accrued to a company in a given accounting period. In that sense, accruals were items that obscured the real picture. They were a distraction that needed to be set aside. In a way, that’s a bit unfair and readers who are not confident about accruals accounting should read our quick guide in the box, above right.

But the point is that the accruals system of accounting involves a higher degree of subjectivity than its counterpart, which is cash accounting. It calls to mind the investment aphorism that ‘profits are an opinion, but cash is fact’; or – to quote from another core investment text from the US, Financial Statements Analysis by Leopold Bernstein – “The higher the ratio of cash flow from operations to net income (ie, accounting profits), the higher the quality of that income. Put another way, a company with a high level of net income and a low cash flow may be using income-recognition or expense-accrual criteria that are suspect”.

So there is an inverse relationship between, on the one hand, accruals and accounting profits and, on the other, cash profits. If accounting profits are high then cash profits are likely to be low and vice versa.

 

On one level, that need not be a problem since the effect of using accruals should be to strip away the ‘noise’ that randomly affects, say, one year’s profits, thus enabling investors to focus on a company’s underlying performance. After all, as the box below explains, all that accruals do is attempt to allocate costs and revenues from unfinished business transactions into discrete accounting periods, which is a sensible aim. But accounting rules become haphazard and sometimes contradictory. Combine this with the opportunities they offer to flatter profits and it is little wonder that accruals accounting can end up producing figures that project a distorted picture of profits.

However – and here is the real concern – although accounting profits are often distorted, they are the numbers that investors fixate upon. If investors did this for the right reason – that accounting profits produce an accurate underlying view – that would be fine. But since accounting profits often don’t provide such a view, investors’ obsession becomes a problem.

This problem – the accruals anomaly – was first rigorously quantified by Richard Sloan, then of the University of Pennsylvania, in a 1996 paper that analysed 30 years’ worth of data – 1962 to 1991 – for US companies, excluding those in financials sectors, that totalled a sample of more than 40,000 company years.

First, Professor Sloan established the rule of thumb that the higher the accruals component in a company’s current profits, the lower the rate of profits growth in the future and – conversely – that the higher the component of cash profits in the figures, the more that profits’ growth would be maintained.

Second, he found that this offered the basis for an investment plan. Investors’ naive fixation on accounting profits meant that share prices tended to trail changes in earnings per share (EPS). So when accruals reversed and EPS growth faltered, share prices suffered, sometimes dramatically. In the long run, therefore, buying shares in those companies with a low accruals component in their profits – ie, those that seemed least likely to serve up nasty shocks – could be a good tactic. At the extreme, buying shares in low-accruals companies and short selling shares in those with a high accruals component could lever up returns.

 

On paper, Professor Sloan’s plan worked out nicely. He divided his sample into 10 groups ranked according to the accruals’ component within profits. For the 30 years under review, the average excess return produced by the group with the lowest level of accruals was 4.9 per cent a year. Simultaneously, the group with the highest level lost 3.5 per cent a year against the same marker. Better still, the so-called ‘hedge’ portfolio, which was long the lowest accruals’ level and short the highest, generated average excess returns of 10.4 per cent.

But Professor Sloan cautioned that the accruals anomaly “does not necessarily imply investor irrationality or the existence of unexploited profit opportunities”. It may be that the time and effort required to find the opportunities was, to use his economists’ jargon, “non-trivial”. In plain English, the game might not be worth the candle because many of the most likely candidates may be illiquid shares.

Perhaps because of that, the anomaly persisted and 10 years on – in 2006 – two New York academics showed that it was alive and well. In addition, the two – Baruch Lev of the Stern School of Business and Doron Nissin from Columbia University – theorised why it would stick around further.

They found that US institutional investors did not trade on the anomaly often enough to arbitrage it away (ie, drive prices up or down so much that prospective risk-free gains vanished). Two connected reasons lay behind this. First, the anomaly tended to dwell in illiquid stocks that were too small for institutions to bother with. Second, such stocks were clearly risky so investing in them might fall foul of so-called ‘prudent-man rules’, which required intermediaries to invest other people’s money as if it were their own.

Simultaneously, the academics suggested three reasons why private investors would not fill the gap left by institutions, starting with the point that the costs (both time and money) of gathering the necessary information were too high since finding a few opportunities involved trawling through the accounts of scores of candidates. Second, dealing costs were often too high because exploiting the anomaly meant holding a big portfolio of stocks and churning its components often. Third – and maybe the killer – the biggest gains came from short selling and, at the best of times, that is difficult for private investors, let alone when they are dealing with small-cap stocks.

Meanwhile, research into the accruals anomaly continues thick and fast. Interestingly, in March Robert Bushman, an accountancy professor at the University of North Carolina, suggested that the negative correlation between companies’ accruals and cash flows – when one rises, the other falls – had faded dramatically. In the 1960s changes in a company’s accruals on average accounted for 90 per cent of the changes in its cash flows. By the 2000s, that ratio had faded to less than 20 per cent. Yet the drop was not so much because accruals had become less significant, said Professor Bushman, but because other factors – especially one-off surprises – had made companies’ cash flows more volatile.

That said, there is so much academic research on this subject that it may be possible to find any result that’s sought. Not just that, but the concept of accruals has flooded every area of a company’s accounts. That has prompted Professor Sloan – now at the University of California at Berkeley – to combine with three other US academics and produce what he hopes will be the definitive guide to measuring a company’s accruals. His team divide accruals into three categories:

■ Working capital accruals, which have been the focus of most research and, arguably, remain the items that most concern investors, as we will explain shortly.

■ Accruals concerning fixed assets, which most likely revolve around asset write-downs – both tangible and intangible. Since these are often big items, their reversal will have a big impact on cash flows but is likely to be quickly absorbed into share prices.

■ Financial accruals, which relate to the changing values of long-term investments and listed securities, but also include the way in which companies amortise the debt they raise.

And the way to tackle the analysis of accruals is either via a company’s balance sheet or – the short-hand way – comparing a company’s income statement (what’s still better known as the profit-and-loss account) with its cash-flow statement. The chief limitation of this short-hand way is that it takes no account of changes in working capital caused by acquisitions or disposals of subsidiaries. Nevertheless, it is an easy way into investigating the accruals anomaly, especially as so many online databases offer basic information on companies’ accounts.

Take Table 1 (below), which juxtaposes accounting profits and cash flows for London-listed companies in the consumer staples sector. Because these companies have similar business models, their ability to generate cash should start from a similar point. Therefore, those that generate more cash in relation to accounting profits may be better businesses. If so, is that reflected in their share price performance?

Click here to enlarge

To expect so much when so many other factors shape a company’s share price – and with just one year’s worth of data – is a big ask. That said, it is interesting that two companies very much in recovery mode – Premier Foods (PFD) and Wm Morrison Supermarkets (MRW) – should show the highest ratio of cash flow to operating profits. This implies that their bosses are straining to squeeze every last drop of cash out of their businesses, but is it reflected in the share price performance – there is quite a difference between the two?

Meanwhile, it may be significant that those companies renowned for steady income generation – the likes of Unilever (ULVR), Diageo (DGE) and Reckitt Benckiser (RB.) – cluster in the middle of the table and have also produced middling share price performances. But when their share prices are rated so highly, that’s all they are expected to do.

And what about those companies producing just a small proportion of cash in their accounting profits – will that be a signal of poor share price performance to come, as the theory says? Certainly, clustered at the top of the table are companies that either continually find the going tough – Dairy Crest (DCG) or J Sainsbury (SBRY) – or which tend to flatter – Tate & Lyle (TATE) and, perhaps, PureCircle (PURE). Have PureCircle’s shares fallen far enough following the company’s recent profits warning? Year on year, Tate’s price performance is good, but over five years the price has done nothing. With Tate, that’s a familiar pattern – will it continue?

Answering such questions requires much more than a close look at accruals versus cash flow. Yet our key point is that understanding and quantifying the effect of accruals on accounting profits is a vital tool in every investor’s box since it wards off losses, which is as good as making profits.

To see how, let’s take Aero Inventory, the ultimate example of the accruals anomaly. This Aim-quoted high flier went like Icarus in 2009, bringing nasty losses to those who forgot to compare its accounting profits with its cash flow.

In the early 2000s, Aero Inventory rose high on the back of acquisitions. Its business model was to aggregate parts control for many companies in the aircraft industry and make profits via the economies of scale it would gather. To expand, it made customers an offer they could not refuse – it would buy and manage each customer’s parts inventory and guarantee to supply parts in future at no greater cost than the customer had previously incurred. Thus customers were freed from the hassle of running vast inventories and of tying up working capital, and they paid nothing for the privilege. All Aero Inventory wanted was a long-term contract – in theory, economies of scale would do the rest.

In terms of revenue building, it worked like a charm. Revenues rose 10-fold to £221m in the four years to 2008. Operating profits responded, rising from £2m to £47m. Yet – you guessed it – cash flow went in the opposite direction. A £20m outflow in 2004 became a £102m outflow in 2008. By the time the company went into administration it was having to find £1.50 of extra working capital to generate an extra £1 of revenue.

Hindsight tells us that Aero Inventory was built on a wing and a prayer. For the past eight years its administrator, KPMG, has been trying to make sense of a company that held 40m aircraft parts across 500,000 lines located in 12 countries. Not surprisingly, it has rescued almost nothing. Unsecured creditors got a one-off payment of 3.6p in the pound last year.

Along the way, Aero Inventory’s finance director has been fined £170,000 and banned from practising accountancy for three years, while its auditor, Deloitte, has been fined £4m. Almost needless to say, shareholders, whose equity had a market value of £140m just six weeks before the company went into administration, will get nothing when the company is finally dissolved in the summer.

Yet some shareholders may have avoided trouble if they had applied the lessons of the accruals anomaly. Table 2 below adopts the method used by Professor Sloan in his 1996 paper to quantify the proportion of accruals in Aero Inventory’s operating profits. The vital rows in the table are for year-on-year changes in current assets and current liabilities, which are italicised. Put very simply, positive changes in current assets put profits into the company but suck out cash, while positive changes to current liabilities do the opposite.

 

Table 2: Aero invention

Year to end June (£m)	2008	2007	2006	2005	2004
Current assets	396.6	242.5	165.4	78.2	43.9
Less: Cash	0.6	0.2	36.4	0.2	0
Change on year	153.8	113.3	51	34.1
Current liabilities	80.8	56.6	31.9	38	11.9
Less: Short-term debt	0	0	0	25.9	8.6
Less: Taxes payable	9.7	8.4	1.9	1.1	0.1
Change on year	22.9	18.3	18.9	7.9
Less: Depreciation	10.1	1.9	0.4	0.4	0.3
=Accruals component of operating profits	120.7	93.1	31.7	25.9
Operating profits	46.8	26.4	11.5	7.9
Accruals/op’g profits (%)	258	353	275	329
Source: Company accounts

To understand in more detail why this is so, let’s switch to Table 3 and do a crash course in accountancy. Common sense says that a company’s gross profit is revenue minus the cost of producing goods or services. For Aero Inventory, which basically managed other people’s inventory, changes in inventory during the year proxied the cost of goods. That’s calculated as the cost of inventory at the start of the 2007-08 year (£195m, which is given in the accounts) plus the cost of inventory purchased and sold during the year (which has to be deduced from the accounts) minus the value of the closing inventory (£346m, also in the accounts). The net of these three amounts is the cost of goods (almost £140m, also disclosed).

 

Table 3: Only make believe

Year to end June (£m)	2008
Revenue	220.7
Opening inventory	-194.9
plus net purchases	-290.7
less closing inventory	346.1
Cost of goods sold:	-139.5
Gross profit	81.2
Operating expenses	-34.4
Operating profit	46.8
Source: Company accounts

The logic behind that arithmetic accords with the accruals concept since it says that the cost of goods is a cash outflow that has been matched with the inflow from the sale of those goods minus a cash cost – the closing value of inventory – that has yet to be matched against sales. However – and this is the vital point – the higher the closing level of inventories, the higher the gross profits. And that can create a perverse incentive for company bosses to inflate the closing level of inventories.

We can’t say that Aero Inventory inflated the closing value of its own inventories. However, it’s clear from Table 2 that, thanks to the soaring value of its current assets (almost all of which was inventory), the accruals component of its operating profits was always huge.

By way of contrast, take Table 4, which does the same exercise for business software provider Sage (SGE). This company was chosen blind, although with the expectation that the accruals component of its operating profits would be small.

 

Table 4: How Sage is that?

Year to end Sep (£m)	2016	2015	2014	2013	2012
Current assets	695	586.3	468.1	414.2	366.9
Less: Cash	264.5	263.4	144.6	100.8	61.6
Change on year	107.6	-0.6	10.1	8.1
Current liabilities	988.3	822.6	927.9	789.3	787.4
Less: Short-term debt	43.3	33.6	125.4	21	8.4
Less: Taxes payable	20.7	31.4	23.7	35.7	29.7
Change on year	166.7	-21.2	46.2	-16.7
Less: Depreciation	30.9	27.2	26.6	30.5	32.2
=Accruals component of operating profits	-90	-6.6	-62.7	-5.7
Operating profits	415.4	361.5	346.1	370.2	349.3
Accruals/op profits (%)	-22	-2	-18	-2
Source: Company accounts

It does not disappoint. In each of the four years 2013 to 2016, Sage’s operating cash flow has been greater than its operating profit – hence the minus figures in the bottom row of the table. This may be a comment on the wonder of a successful software business where marginal costs of products are near zero but marginal revenue remains unchanged. It also reflects a business model where cash is received up front and services delivered later – hence the £536m of unearned revenue (a different sort of accrual) in its latest accounts.

The lesson for investors is to check through their equity portfolio to find which companies perform like Sage and which – Heaven forbid – are more like Aero Inventory. If they find an Aero Inventory look-alike, they should know what to do. Simultaneously, candidates for inclusion should get the once-over.

Online readers can

to access a live Excel file of Table 2. Enter the appropriate data in the yellow cells and the table will calculate whether or not the company in question is in need of a rescue from the AA.