Friday, 8 October 2010

Bob’s Famous COP ratio

‘One ratio, one ratio to rule them all’. It sound like a line from the Lord of the Rings but this ratio’s real claim to fame is that it gives a quick means of testing the reliability of a set of accounts. With it we can start to identify those who have moved to the dark-side of the financial force. So as a precursor to financial analysis I always take a careful look at the COP ratio.

Creative accounting is the name we give to the dark art of fiddling the accounts – or to give it its polite name: ‘earnings management’. Earnings management is where the creators of financial information employ the simple devices accounting permits to overstate or understate earnings relative to a true and fair view. Whether earnings management has occurred will be a matter of judgement because so far no one has come up with an absolute definition of what true earnings should be. But of one thing we can be sure: if there is not an explicable relationship between earnings and the underlying cash flows of the business then mischief is afoot.

Accounting standard setters are not helpful in that the rules set for the measurement of earnings are severely constrained by the asset valuation and capital maintenance concepts employed. Make no mistake, most earnings management occurs under the bonnet – deep in the engine of the business. It is here that judgements about appropriate accruals are made, it is here that the year end is adjusted at departmental levels, stocks are valued, revalued, written off; receivables counted in or counted out and payables counted for or ignored according to choice. It is at this level that real mischief occurs.

Ah-ha you say: ‘what about the auditors?’ Well, what about them? Transactions auditing, grubbing around in the bought and sales ledgers, counting stock, all that stuff is done by the lowest of the low. First year audit juniors, not the brilliant brains that are the audit partners and managers, are the ones sent to the basement to ‘sample’ the systems.

It is well recognised by those of us who research these things that the plainest signal of accounting mischief is what we refer to as ‘abnormal accruals’. That is: the to-ing and fro-ing of adjustments converting the ‘relatively’ objective cash flow into the very, very subjective statement of operating income. There can be grand fiddling below the EBITDA line, over or understated capital charges, and other lines of account but they are more difficult to manipulate and more difficult to hide. So our focus of attention is on the EBITDA and how it relates to cash flow. It is this relationship: EBITDA to operating cash flow which reveals the story. It was this relationship between earnings and cash flow that Loren Fox of Fortune Magazine puzzled over and led her to question that ultimate pack of financial cards – Enron’s accounts.

‘The flow of cash in and out of Enron told a less spectacular story than its soaring revenue, In 1998, Enron actually had a negative cash flow of $59 million….In 1999, Enron had a positive cash flows of $177 million – but that came from a rise in short term borrowing, while the net cash generated by operating activities fell by 25 per cent from 1998 levels….Investors might have started to wonder if they had paid closer attention to these subtler measures of efficiency and profitability rather than to earnings, or to ‘operating earnings’.’
Extract from: Loren Fox, Enron – the rise and fall, pp 175-176


COP is the ratio of EBITDA and operating cash flow. So far so good – it is not too difficult to measure: take the operating profit from the income statement and add back depreciation, amortization, impairments, net income expense and any other ‘profits’ on non-operating activities. The trick is to match the EBITDA as closely as possible to the construction of the operating cash flows. It’s measurement takes just a few moments perusal of the accounts and the cash flow statement (where depreciation etc is shown as part of the reconciliation of net income and operating cash flow).

We now need to set parameters of ‘normality’ on the COP. Obviously in an unchanging world EBITDA and OCF will be the same and the COP will equal one. Say that again? If a company neither expands nor contracts its business and if it holds its net working capital constant then opening and closing working capital balances cancel out and EBITDA will necessarily equal cash flow. This will be the case with cash based businesses where changes in sales revenue and costs are matched closely by changes in cash flow. The typical retailer holding constant levels of inventory, who pays its suppliers quickly and makes most of its sales for cash should also have a COP ratio close to one. At the other extreme we have the contract based businesses where cash flow is very sluggish. In this case the other limiting extreme where the COP equals one plus the percentage change in EBITDA over the year can occur.

So the range of COP is one (where cash flow maps EBITDA exactly) to one plus the percentage change in EBITDA over the year (where cash flow is constant with respect to the change in EBITDA). The next job is to predict the change in EBITDA which should occur with a given change in activity. To do this look first at the change in the revenue of the business as the primary signal of change in activity level. Let us say it has risen by 10%. In that event, the EBITDA should rise by at least 10% - possibly much more if the business’s cost structure is highly geared. Working out the operating gearing of a business is virtually impossible from a single year’s accounts. Accountants probably do not know, and certainly cannot disclose the fixed costs of the business.


The way I sort out the cost structure is to look at how EBITDA and revenues relate statistically. In the figure above there appears to be a good relationship between EBITDA and Revenue using 12 years of data. With an R2 of 0.8 the correlation is +0.89 which suggests the relationship is dependable. So with the function of the best fit line we can turn this year’s revenue into an EBITDA figure taking account of the cost structure of the firm. For those who are interested the slope coefficient is equal to one minus the variable cost percentage (expressed as a percentage of revenue).

So we now know what the EBITDA should be if it’s on trend. Take the change in EBITDA from the previous year’s actual and this (plus one of course) represents the outer limit of our COP ratio – up or down depending on the change in direction of revenue.
So now we have it. Our COP should sit between one and the figure you have just estimated. Anything outside this range and the rat gets smellier the further away you go. Once you leave this range the accruals process is pushing the EBITDA into territory where it cannot be supported by the cash flow of the firm. Loren Fox had spotted that Enron's COP ratio was 19! Why did those high rolling equity analysts working for the great finance houses of the US not spot the problem?

I know it looks a lot of work but once you get the hang of it the COP ratio is a very powerful tool in your hands. The prediction of the acceptable range has another use too: it helps you convert revenue forecasts into EBITDA forecasts and that is vital if you ever want to value a firm. But that's another story.

10 comments:

Investor88 said...

This is a very interesting technique. So, you assess the change between this years theoretical EBIDTA and last years actual, use this as a 'confidence interval' as such, and compare it to the ratio of this years op.cashflow to actual EBITDA? This method does however hinge on a respectable R-squared between EBITDA and Revenue - to calculate the theoretical EBITDA which is needed to assess the %change. Is an R-squared of 0.8 even realistic?

Anonymous said...

dear prof,

can u explain how this method can convert revenue forecast to ebitda forecast, and valuing the firm as well.

Thanks

Professor Bob Ryan said...

What the OLS regression does is to give you a relationship EBITDA = a + b x revenue. The OLS tells you the value of a and b in the formula. So, if you have got a good revenue forecast (say) over 5 years then this formula will let you convert that revenue to an EBITDA number taking into account the fixed and variable elements in your cost structure. What you then need to do is deduct your forecast for depreciation and amortization, your forecast of finance charges and tax and you have an earnings forecast. From there doing stepped valuation using NOPAT, dividends even FCF is within your grasp. Obviously, in preparing the historical record for analysis you do need to clean the data as far as possible.

Professor Bob Ryan said...

Investor88:

You have the method spot on. Obviously, much does hinge on the R2 - I normally look for and clean up 10-12 years of back data if using annual figures. If the actual COP is lying close to the edges of the range and particularly if I am not expecting it to be there (eg a retailer) then I would look hard. If the R2 is too low then I would check to see why first of all, then I would try to isolate fixed costs in the firm's recent accounts to see if I could get any clear sense of the operating gearing of the business (not easy). Most times I have used however the method as described works fine.

Anonymous said...

dear prof,

what does COP stands for actually?

Anonymous said...

Hi Bob,

You said that
"The prediction of the acceptable range has another use too: it helps you convert revenue forecasts into EBITDA forecasts"
In what way does it helps and how do you apply it?

Thanks

Professor Bob Ryan said...

COP stands for cash to operating profit ratio.

On the problem of converting revenue to EBITDA: the conventional approach is to use either the latest or the average of recent gross and operating profit margins. However, the use of margin ratios when applied to forecasted revenue is that you are assuming, implicitly, that all costs are variable. What the regression does is to give a relationship between revenue and EBITDA which recognises the leverage effect of fixed costs in the cost structure. So, once revenue is forecast, just use the regression factors (a) and (b) to convert the forecast into EBITDA for each year.

Darrel Freisinger said...

What does EBITDA stand for?

Darrel Freisinger said...

What does EBITDA stand for?

Professor Bob Ryan said...

Darrel: EBITDA is 'earnings before tax, depreciation and amortization'. We estimate it by adding back depreciation, amortization (and impairments) to operating profit. It measures the operating capacity of a business before charges for the consumption of capital assets during the year in question.