Global Industry Overview

Global
06 January 2009

Our Approach to Global Equity Valuation,

Accounting, and Quality of Earnings
Global Valuation and
Analytics Research
Karl Debenham
„ Highlights
„In this report, we update our basis for company financial valuation and
Karl_debenham@ml.com
(44) 20 7996 2386
analysis. We continue to refine our systematic approach to maintaining
consistency in a global context – what we call iQmethod SM, part of our suite of
iQanalyticsSM offerings. The key features of iQmethod are:
Prepared with the assistance of
„ A consistently structured, detailed, and transparent methodology, with Jon Duchac, Ph.D., Merrill Lynch
data sourced directly from our analysts’ models. Professor of Accounting at Wake
Forest University.
„ Guidelines to maximize the effectiveness of the comparative valuation
process, and to identify some common pitfalls.

„ The iQmethod framework focuses on standard measures under three broad

headings:
For educational purposes only. Not to be
„ Business Performance – judging management effectiveness. used as the sole basis for any investment
„ Quality of Earnings – assessing the sustainability of business decisions.
performance, and the attendant risks.

„ Valuation – making the connection between business performance and

market value. We look at the common metrics in fine detail to maximize
consistency and highlight possible false signals.

„ Our global iQdatabaseSM facilitates this analysis:

„ The iQdatabase is designed to provide both breadth and depth of data;

and to effectively obtain, organize, and present consistent, detailed, and
explicit financial information, in a manner most useful to investor
investment priorities.

„ More than 100 detailed line items for each of the 2,500+ stocks in our
global equity research coverage universe are available.

„ A suite of products is available to access and manipulate the rich content of

the iQanalytics platform. These include:

„ The iQtoolkitSM, which presents sector and regional analysis via our MLX
web portal, Bloomberg, and coming soon, Reuters Knowledge for
Investment Management platforms, leveraging the power of the
database, including the metrics presented in this work.

„ iQworksSM software, which further exploits the power of the iQdatabase,

enabling us to create custom reports and charts including user-defined
ratios and calculations based on the detailed data in the iQdatabase.

Merrill Lynch does and seeks to do business with companies covered in its research reports. As a result, investors should be aware that the firm may
have a conflict of interest that could affect the objectivity of this report. Investors should consider this report as only a single factor in making their
investment decision.
Refer to important disclosures on page 56.
 Globa l
0 6 Jan ua ry 2 009

Quick Reference
We summarize below the definitions of our standard measures under three broad
headings: Business Performance, Quality of Earnings, and Valuation. A more
detailed explanation of how these metrics can be used (and misused) is given in
the body of this report.

Business Performance
These measures essentially provide a checklist to judge management; Return on
Capital Employed tests the effective deployment of resources across the
enterprise in general, Return on Equity considers the rate of return to equity in
particular. The Operating Margin indicates pricing power, while EPS Growth and
Free Cash Flow give a feel for future value.

Table 1: Business Performance

Standard Measure Numerator Denominator Units Page
Return on Capital Employed NOPAT = (EBIT + Interest Income) * (1 - Tax Rate) + Goodwill Total Assets – Current Liabilities + ST Debt + Accumulated % 5
Amortization Goodwill Amortization
Return on Equity Net Income Shareholders’ Equity % 12
Operating Margin Operating Profit Sales % 13
EPS Growth Expected 5-Year CAGR From Latest Actual N/A % 14
Free Cash Flow Cash Flow From Operations – Total Capex N/A Mn 17
Source: Merrill Lynch

Quality of Earnings
Not all earnings are equal! “High-quality” earnings should be viewed with greater
confidence and, by implication, others treated with more caution. These metrics
are intended to test the underlying security of the business performance.

Table 2: Quality of Earnings

Standard Measure Numerator Denominator Units Page
Cash Realization Ratio Cash Flow From Operations Net Income X 19
Asset Replacement Ratio Capex Depreciation X 20
Tax Rate Tax Charge Pre-Tax Income % 21
Net Debt/Equity Ratio Net Debt = Total Debt, Less Cash & Equivalents Total Equity % 22
Interest Cover EBIT Interest Expense X 24
Source: Merrill Lynch

Valuation
These measures connect the economic performance of the business to its market
value in various ways. The Price/Earnings Ratio is a payback indicator; the
Price/Book Ratio reflects the valuation of the company’s equity. The Dividend Yield
and Free Cash Flow Yield reflect tangible and potential monetary rates of return,
respectively. Enterprise Value/EBITDA is a general “structure-neutral” cash-
generation multiple, and Enterprise Value/Sales indicates “volume leverage.”

Table 3: Valuation
Standard Measure Numerator Denominator Units Page
Price/Earnings Ratio Current Share Price Diluted Earnings Per Share (Basis As Specified) X 26
Price/Book Value Current Share Price Shareholders’ Equity/Current Actual Shares X 35
Dividend Yield Annualized Declared Cash Dividend per Share Current Share Price % 35
Free Cash Flow Yield Free Cash Flow Market Cap. = Current Share Price * Current Actual Shares % 37
Enterprise Value/EBITDA Enterprise Value = Market Capitalization + EBIT + Depreciation + Amortization X 38
Minority Equity + Net Debt + Other LT Liabilities1
Enterprise Value/Sales Enterprise Value Sales X 39
1 Pension liabilities, deferred taxes, capitalized lease, and other post-retirement benefits are explicitly included in other non-current liabilities. However, this item is not limited to these factors, as new
and inventive forms of funding are always emerging and need to be captured.
Source: Merrill Lynch

2
 Globa l
0 6 Jan ua ry 2 009

CONTENTS
Section Page
Quick Reference Standard Measure Definitions 2
At a Glance

First Principles 1. Steps in Measuring the Investment Process 4

Business Performance
Quality of Earnings
Valuation

Business Performance 2. A Checklist for Management 5

Return on Capital Employed
Return on Equity
Operating Margin
EPS Growth
Free Cash Flow

Quality of Earnings 3. Measures to Test Resilience of Business Performance 19

Cash Realization Ratio
Asset Replacement Ratio
Tax Rate
Net Debt/Equity Ratio
Interest Cover

Valuation 4. Connecting Market Value With Business Performance 26

Price/Earnings Ratio
Price/Book Value
Dividend Yield
Free Cash Flow Yield
Enterprise Value/EBITDA
Enterprise Value/Sales

The Good DCF Guide 5. A Best Practice Guide to Discounted Cash Flow and Its Limitations 41
What (and How) to Discount
Different Approaches to the Residual

Appendix The iQdatabase Global Template 49

3
 Globa l
0 6 Jan ua ry 2 009

1. First Principles
The common thread between all equity investments is that investors commit
capital to enterprise management in the expectation that this will produce
leveraged returns. That leverage acts in two stages – first: that the enterprise
earns a rate of return on its capital in excess of its cost; second: that its market
value reflects that economic return.

In updating our iQmethod, we have revisited (but not changed) a set of measures
that in our view can best serve as the basis (and only the basis) for comparative
assessment for both of these stages. Prudence requires that we test the
confidence of our assessments, so we include measures to provide an indication
of “quality of earnings.”

Criteria for Selection

There is a particularly long list of metrics that could be included in our standard
list, and almost as many ways of defining each metric. We continue to use the
following criteria in to determine our standard measures:

Relevance: our key aim is to gather information from all directions; the metric
should add incremental information, rather than complement an existing signal.

Simplicity: the concept should be explicitly understandable by all users.

Applicability: the metric should be meaningful across the entire spectrum of

commercial/industrial sectors.

Forecastability: some definitions, while intellectually rigorous, are of limited

practical use owing to the difficulty of forecasting some of their inputs. Where
necessary, we have modified the “business school” definition in favor of a more
pragmatic approach.

Our Sixteen Standard Measures

Business Performance
„ Return on Capital Employed
„ Return on Equity
„ Operating Margin
„ EPS Growth
„ Free Cash Flow
Quality of Earnings
„ Cash Realization Ratio
„ Asset Replacement Ratio
„ Tax Rate
„ Net Debt/Equity
„ Interest Cover
Valuation
„ Price/Earnings Ratio
„ Price/Book Value
„ Dividend Yield
„ Free Cash Flow Yield
„ Enterprise Value/EBITDA
„ Enterprise Value/Sales

4
 Globa l
0 6 Jan ua ry 2 009

2. Business Performance
These measures essentially provide a checklist to judge management: Return on
Capital Employed tests the effective deployment of resources across the
enterprise in general, Return on Equity considers the rate of return to equity in
particular. The Operating Margin indicates pricing power and the ability to control
production costs, while Earnings Growth and Free Cash Flow give a feel for
potential value.

Return on Capital Employed

Concept
ROCE is an indicator of how effectively resources are used within the enterprise.
If the rate of return on capital exceeds its cost, management is producing positive
economic returns for shareholders.

The expression capital employed (the denominator) refers to all resources used in
the business on which someone is expecting a return.

The return (the numerator) represents the total fund from which all required
returns, dividends, interest paid, etc., will be met.

Use after-tax approach for This ratio is sometimes presented at the pre-tax level. We think this approach
valuation severely limits the value of the measure for two reasons. Tax is a critical element
of the cost of capital, so pre-tax ROCE cannot be used as a value indicator. In
addition, international comparison is rendered meaningless because of the wide
range of tax rates across the globe. Accordingly, we use an after-tax approach.

Critical to consider write-offs As it is possible to inflate the ratio merely by writing off resources, usually as
goodwill, we add back the cumulative total of all such write-offs, amortization, etc.,
to the denominator, and the annual charge to the numerator.

And also operating leases Because significant inconsistencies can arise where a company makes extensive
use of assets that are leased, as opposed to owned, where applicable, total
assets should be adjusted upward to reflect the notional value of assets provided
under operating leases. Because there are many payment schedule permutations
and, as yet, no definitive set of rules to determine such values, the calculation can
be on a simplified basis only – perhaps grossing up the annual cost of lease
payments by a yield factor based on interest rates.

Definition
Return on Capital Employed Numerator: Net Operating Profit After Tax (NOPAT). This is calculated as: EBIT
(consolidated operating profit, plus fixed-asset income such as associates), plus
NOPAT = (EBIT + Interest Income) *
(1 - Tax Rate) + Goodwill interest income, less adjusted tax, plus goodwill amortization charge.
Amortization
_____________________ Interest income is included, being part of the fund available to provide returns, but
Average of opening and closing interest payments are not deducted, being a claim to be met from that fund.
Capital Employed = Total Assets –
Current Liabilities + ST Debt + To maintain consistency, the tax charge is adjusted to reflect the shield effect of
Accumulated Goodwill Amortization interest paid; this is best achieved by applying the appropriate tax rate (actual, or
where anomalous, long term) to the sum of EBIT and interest income.

Goodwill write-offs arising from impairment, or amortization (where it is still charged),

are then added back, ignoring the (normally minimal) tax impact.

Denominator: total assets, less non-interest-bearing current liabilities, plus

cumulative goodwill amortized and/or written off. Non-interest-bearing current
liabilities being calculated as Total Current Liabilities less Short-Term Debt. We

5
 Globa l
0 6 Jan ua ry 2 009

recommend this approach, rather than the classic exercise of adding up the
relevant components of either side of the balance sheet. It is simpler, and focuses
on the concept of capital employed, and more practically, it is considered more
robust in that all components, including items such as special provisions, are
included unless they are specifically excluded.

The asset base almost always changes between the opening and closing of the
financial year; which data should be used?

In strict economic terms, it is the opening (previous) capital on which the return is
earned. In measuring business performance, however, because the numerator is
essentially a flow over a period of time, the denominator should represent the
average capital over that period. It is, however, definitely inappropriate to use the
year-end capital alone; we use the average of the year-end and previous
year-end values.

Use the average of opening and The most rational approach, which we adopt as a standard, is to take the average
closing balance sheet values of the opening and closing balance sheet values, excluding the first year. It is also
sometimes appropriate to weight these points if, for example, a significant
transaction occurs early or late in the year.

The expressions “Return on Capital Employed” and “Return on Invested Capital”

are almost synonymous, but strictly speaking, ROIC refers to the opening capital
rather than the average.

Interpretation
A test of management Sustaining a high level of ROCE is a broad test of management. Because the
ratio can be decomposed into the product of margin (profit/revenue) and asset
turnover (revenue/assets) it can be viewed as a composite indicator of both
pricing power/cost control (the margin) and management’s ability to position the
enterprise and assign capacity (the asset turnover). Ideally, both of these should
be maximized, but where margins are low or declining, management should
demonstrate that asset turnover is either high, or at least rising.

And a basis for valuation ROCE is also the starting point for many other valuation concepts and techniques
in wide use today. Two in particular are worth noting:

Economic Return/Economic Value Added/Market Value Added

ROCE measures the extent of a company’s accounting rate of return on capital,
but the value of that return also depends on the cost of that capital (which is
defined and discussed in the next section on page 10).

The difference between a company’s ROCE and its weighted average cost of
capital (WACC) is known as the Economic Return. This can, of course, be
positive or negative. Multiplying the economic return by the capital employed
indicates the total amount of value that has been generated for shareholders,
known as the Economic Value Added (EVA)®.

This can then be compared to the premium or discount that the market applies to
the value of the company’s capital; the Market Value Added (MVA), which is
defined as enterprise value/capital employed. Enterprise value is the market value
of the company’s capital: market capitalization plus net debt and other long-term
liabilities and minority equity. This relationship is usually illustrated by plotting the
economic spread against the MVA across a peer group, as in Chart 1.

6
 Globa l
0 6 Jan ua ry 2 009

Chart 1: Valuing Economic Returns by Comparing to MVA

5.0

Buy

ROCE – WACC (Or “Economic Return”)

4.0
Outliers of Interest
„ Buys – with high Economic
3.0 (High economic returns
Returns and low MVAs (where and low MVAs)
mean reversion would imply a
higher market value in future). 2.0

„ Sells – with low Economic (Low economic returns

1.0 and high MVAs)
Returns and high MVAs (where
below-average returns point to Sell
possible drops in market value). 0.0

-1.0
0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3 1.4 1.5
Enterprise Value / Capital Employed
(or “Market Value Added”)

Source: Merrill Lynch

Two types of outliers are of potential interest: those with high Economic Returns and
low MVAs (where mean reversion would imply a higher future market value), and the
opposite, where below-average returns point to possible drops in Market Value.

Another way of illustrating this effect in a single number is the Rating of Economic
Profit ratio (REP). The REP is defined as: (Enterprise Value/Capital Employed)
divided by (ROCE/WACC).

In simple terms, a REP greater than unity implies a premium valuation,

suggesting that either the stock is overpriced, or that there will be an offsetting
increase in ROCE relative to WACC. Symmetrically, a REP under one indicates
either potential price appreciation, or declining future Economic Returns.

Cash Flow-Based Measures

Limitations of ROCE: One key limitation of basic ROCE analysis is that it is effectively limited to
Fails to reflect the age of assets accounting returns, rather than economic returns or cash flows.

In particular, ROCE fails to reflect the age of the assets involved, which can lead
to distortions – flattering the rate of return for companies with an older, long-
depreciated plant, for example. Although such businesses may produce
acceptable returns at their current level of activity, incremental returns may be
significantly lower, reducing the value of future growth.

CROCE attempts to The concept of Cash Return on Capital Employed (CROCE) attempts to account
account for this by using for this; in effect, trying to estimate the “replacement cost” return on capital in
replacement cost terms of cash flow. In general terms, this involves replacing fixed (non-current)
assets in the capital employed calculation as follows:
First: estimate the average age of the assets in years as:
asset life = (gross fixed assets – net fixed assets)/annual depreciation
Then: obtain the replacement cost fixed assets at an assumed x% inflation rate as:
repcost fixed assets = gross fixed assets * (1+x%) ^ asset life
where gross refers to the historical cost, and net to the book value

7
 Globa l
0 6 Jan ua ry 2 009

The numerator in the ROCE calculation must also be amended, with EBIT
replaced by replacement cost EBITDA (EBIT plus the implied depreciation
associated with assets at replacement cost).

We should be aware of two proprietary techniques that attempt to refine the basic
concepts of ROCE:

Cash Flow Return on Investment Developed by Holt Value Associates (now owned by Credit Suisse First Boston),
(CFROI) CFROI seeks to connect accounting and economic returns by treating the firm as
a single economic project, comprising a series of outward investments and inward
cash flows, and calculates the Implied Internal Rate of Return (IRR).

CFROI assumes that over a certain period (depending on the nature of the
business) returns on investment will inevitably trend back to the firm’s cost of
capital. The value of this approach is that it avoids having to make assumptions in
perpetuity for the residual cash flows in the IRR calculation. CFROI also attempts
to indicate a real, rather than nominal, return that requires assumptions for both
asset life and replacement cost.

Cash Return on Capital Invested CROCI is Deutsche Bank’s equivalent valuation technique. As with Holt, it
(CROCI) evaluates a cash flow return on capital, adjusting both assets and depreciation
charges for inflation in different ways.

8
 Globa l
0 6 Jan ua ry 2 009

Soup-to-Nuts Example
Table 4 exhibits the key inputs and Table 5 is a step-by-step worked example to
evaluate the Return on Capital Employed, both on a plain vanilla historical cost
basis, and also on a cash basis. In this (hypothetical) example, the historical cost
ROCE is good, but the cash returns less impressive, highlighting that capex is no
substitute for improving margins or asset turnover.

Table 4: Summary Financial Data

The starting point for the numerator is Company A 2002 2003
EBIT, plus interest and investment Revenue 4,050 4,350
income, but not interest expense. We
- Cash operating costs -3,000 -3,150
also need the tax rate and – where
applicable – goodwill amortization. = EBITDA 1,050 1,200
- Depreciation -1,200 -1,350
= EBITA 900 1,050
To evaluate the cash-based returns, - Goodwill amortization -100 -100
we need details of the historical cost = EBIT 800 950
non-current assets, and the -Interest expense -200 -220
depreciation charge. + Interest income 30 25
+ Investment income 20 20
= Profit before taxation 650 775
Conceptually, the denominator - Tax expense -228 -271
represents “all the resources on which
the provider is expecting a return.” The
Non-current assets at cost 12,000 13,500
numerator is “the flow available to
provide those returns.” Property, plant & equipment 9,000 10,500
Goodwill 2,000 2,000
Other non-current assets 1,000 1,000
Returns are normally calculated on a
book-value basis, but a full approach Non-current assets at book value 7,500 7,550
would also consider cash returns on Property, plant & equipment 6,000 6,150
historical cost assets, or on their Goodwill 900 800
estimated replacement cost. Other non-current assets 600 600

Current assets 666 715

Inventory 222 238
Receivables 333 358
Cash 111 119

Current liabilities 921 975

Short-term debt 333 358
Payables 388 417
Other current liabilities 200 200
Source: Merrill Lynch Imagination

9
 Globa l
0 6 Jan ua ry 2 009

Table 5: The Spreadsheet

Replacement
Book Value Historical Cost Cost
2002 2003 2002 2003 2002 2003
Add up the various components of Denominator: Average Capital Employed
non-current assets – the values may PPE 6,000 6,150 9,000 10,500 10,168 12,288
alter according to the basis – add the Goodwill 900 800 2,000 2,000 2,000 2,000
current assets, deduct total current Other intangibles 600 600 1,000 1,000 1,000 1,000
liabilities but add back short-term Current assets 666 715 666 715 666 715
debt. This may seem convoluted, but Total assets 8,166 8,265 12,666 14,215 13,833 16,003
does ensure full capture. Also add - Current liabilities -921 -975 -921 -975 -921 -975
the cumulative goodwill amortization + Short-term debt 333 358 333 358 333 358
to get the year-end capital employed. + cumulative goodwill amortization (see below) 1,100 1,200
This is the “capital invested” (the = Year-end capital employed 8,677 8,848 12,077 13,598 13,245 15,386
opening balance) for the next year.
“Capital employed” is the average of Capital invested (opening balance) 8,677 12,077 13,245
the opening and closing levels. Capital employed (average) 8,763 12,838 14,315

Cumulative goodwill amortization calculation

- Goodwill at book value -900 -800
+ Goodwill at cost 2,000 2,000
= Cumulative goodwill amortization 1,100 1,200

Replacement cost calculations

Historical cost PPE 9,000 10,500
Replacement cost is estimated by - Book value PPE 6,000 6,150
first establishing the average age of = Accumulated depreciation 3,000 4,350
the assets (accumulated depreciation
/ Depreciation charge 1,200 1,350
as a multiple of the annual charge)
= Average age (years) 2.5 3.2
and compounding the historical cost
PPE by the assumed inflation rate. ^ (1+Inflation) 5% 5%
= Replacement cost escalator (applied to 113% 117%
historical cost )

The numerator, NOPAT, applies Numerator: NOPAT

nominal tax to the sum of EBIT, EBIT 800 950 800 950 800 950
interest, and investment income, + Interest income 30 25 30 25 30 25
adding back the annual goodwill + Investment income 20 20 20 20 20 20
amortization charge where applicable Subtotal 850 995 850 995 850 995
(not generally being tax deductible, it Tax rate (%) 35 35 35 35 35 35
- Nominal tax -298 -348 -298 -348 -298 -348
is recredited in full).
+ Adjust for goodwill amortization 100 100 100 100 100 100
NOPAT 653 747 653 747 653 747
Cash NOPAT adds back
+ Depreciation 1,200 1,350 1,200 1,350
depreciation as well, but as this is tax
- Nominal tax on depreciation -420 -473 -420 -473
deductible, it is taxed in this
calculation, to ensure that only the Cash NOPAT 1,433 1,624 1,433 1,624
net equivalent is added back.
Return on capital employed (%) 8.5 5.8 5.2
Source: Merrill Lynch Imagination

10
 Globa l
0 6 Jan ua ry 2 009

Watch Out for...

Thoroughly check the definition – in particular, if the return is pre- or post-taxation
and whether the effect of goodwill has been reflected.

Give the WACC a thorough “reality check” – especially if the argument is driven
by a “low” Beta.

“Artificial” balance sheet valuations. There are some situations in which balance
sheet items, typically asset valuations, while fully compliant with GAAP, could
vary significantly from their “open market” levels. This reflects the hybrid nature of
balance sheet valuations: some items are recorded at historical cost (property
and equipment); some at market value (marketable securities held for sale); some
at lower of cost or market, amortized cost, or some other valuation method
(inventory, equity investments); and others not at all (internally developed
patents). In addition, assets acquired through a business combination are written
up to market value at the date of acquisition. Examples of companies whose
balance sheets may not reflect current open market values very well include
regulated utilities, real estate companies, and enterprises created as a result of
corporate reconstruction. The effect, usually, is to overstate ROCE.

Defining the Cost of Capital

Defining the WACC A company’s Cost of Capital is defined as the weighted sum of its cost of equity
and debt. The weights are determined in line with the definition of Enterprise
Value on page 34, in that the equity component includes the minority interest in
equity, and the debt component includes other non-current liabilities, such as any
shortfall between projected assets and liabilities of funded pension schemes, or
the explicit liabilities of unfunded schemes (especially in Germany).

Cost of Debt: In simple terms, the cost of debt is determined by the effective
interest rate, but in practice, this can be difficult to establish.

Both of the widely used determinants: using the yield on quoted bonds, or dividing
the average debt by the interest paid, present difficulties in practice; we strongly
suggest that all assumptions be rigorously tested. One useful approach is to
compare the company’s peer group, by ranking both cost of debt and financial
leverage. There should be strong correlation between the two hierarchies.

Because interest is tax deductible, the effective cost of debt is lower than the
interest rate (net cost = interest rate * [1 – tax rate]). It is also very important that
unrealistic tax rates are not inadvertently applied to the tax shield. A “standard”
rate, based on the average of local marginal rates, should be used where the
actual rate is distorted by anomalous factors.

Cost of Equity: The cost of equity involves three components: the Risk-Free Rate
of Return, the “Equity-Risk Premium,” and the systematic risk (usually known as
Beta).

(1) The Risk-Free Rate is usually – but not necessarily – taken as the yield on
long-term paper issued by the relevant central bank.
Three Components of Cost of
Equity: (2) The Equity-Risk Premium is the difference between the expected returns on
„ Risk-Free Rate of Return equities as an asset class and the risk-free rate. The return on equities is
determined as the internal rate of return generated from the aggregate expected
„ Equity-Risk Premium dividends – which is a formidable calculation undertaken regularly by the Global
Valuation and Analytics Research Team.
„ Beta

11
 Globa l
0 6 Jan ua ry 2 009

A company’s cost of equity is calculated as the Risk-Free Return plus the product
of the Equity-Risk Premium and the stock’s specific risk, or Beta.

(3) The Beta is a concept that is easy to misapply; there is a very real risk of
generating a flatteringly positive value where it appears low. The Beta is the
systematic risk of investing in a particular security; it is usually estimated by
calculating the slope of the regression line obtained by plotting price movements
of the stock against those of the reference index.

The misunderstanding arises because the slope of the regression line is only an
estimate, and even for liquid stocks, the lack of precision (measured by the
standard error) is usually far greater than appreciated. For example, if the slope of
the regression line is 0.9, and the standard error 0.3, there is a one-in-ten chance
that the actual Beta is greater than 1.5 (0.9 + 2 * 0.3).

Return on Equity
Return on Equity Concept
The accounting rate of return on shareholders’ contributed capital.
Net Income
_____________________
Definition
Shareholders’ Equity Net income as a percentage of shareholders’ equity.

Shareholders’ equity is issued and fully paid capital, plus other paid in
capital/share premium account/reserves, plus retained earnings and preferred
stock. This represents the base of equity attributable to the company’s
shareholders; minority interests in equity are not included.

As with the definition of Return on Capital Employed, we take the average of the
opening and closing balance sheet values in the denominator.

Interpretation
The link between the P/E Just as the Return on Capital Employed measures the rate of return on the
and P/BV enterprise in general, the Return on Equity indicates the accounting returns to the
shareholders in particular.

The connecting factors are the level of financial leverage and the cost of debt. If
the Return on Capital exceeds the cost of debt, the Return on Equity will rise with
increasing leverage.

In valuation terms, the ROE is useful as a crosscheck between the Price/Earnings

Ratio and the Price/Book Value. These three metrics are closely connected, in that
the Price/Book is equal to the product of the ROE and the P/E. For example, where
the Price/Book is average but the P/E is relatively high, so must be the ROE.

Watch Out for…

Inadequate equity. Where shareholders’ funds have been depressed by unusual
factors, such as write-offs, this ratio can send a false-positive signal. Ensure that
the P/E, Price/Book, and ROE are internally consistent.

12
 Globa l
0 6 Jan ua ry 2 009

Operating Margin
Operating Margin Concept
Operating Profit This is a composite indicator of management’s pricing power and cost control, of
_____________________ which the former is more important; successful businesses supply goods or
Net Sales Revenue services that their customers want, and they sell them on their own terms.
Successful managers build a portfolio of such businesses. Low-margin activities
can produce positive economic returns (if asset turnover can be sustained), but
their lower visibility generally depresses valuation.

Pricing power is critical Definition

Operating profit as a percentage of net sales (after returns, allowances, and
discounts). Note: strictly, operating profit rather than EBIT, and sales rather than
total revenues.

Although it is often used as such, operating profit is not a synonym for EBIT. This
can include the group’s share of profits from associates, but these are not
reflected in consolidated sales. In practice, the impact is rarely significant, but it is
not impossible for distortions to arise where associates, joint ventures, etc.
represent a significant portion of the business. Where this arises, it is worthwhile
comparing EBIT to proportionate sales (consolidated sales plus the attributable
portion of associates, etc.) as a cross-check.

Similarly, the distinction between sales (which are revenues from trading
activities) and total revenue (which includes other revenue) may seem arcane, but
our essential focus is on core trading activities here, so sales is the more
appropriate measure.

Interpretation
Peer group hierarchy The operating margin is generally used as a relative indicator, building up a
hierarchy across a peer group. Higher margin businesses demonstrate a possible
combination of good positioning, pricing power, and cost control – and the key
word is demonstrate. Good margins are highly visible, and valuations reflect this.

Watch Out for...

False Positives – Where high gross margins arise from (possibly transient) market
power, a strong operating margin can disguise weak cost control.

Anomalies – These are usually credits, such as gains on fixed-asset sales

included in operating profit. There is some debate about whether these should be
stripped out of the calculation. Are they exceptional in character, and so
excluded, or do they arise purely because of prudent book values, and so should
be included as actual profits? Similar arguments apply to restructuring costs,
which have a habit of acquiring perennial status in some cases. There is no
correct answer to this question; the best practice is to evaluate both the reported
and the adjusted figures, and explain any significant variation explicitly.

Inconsistent Depreciation Rates – Occasionally, variations across a peer group

arise because of different depreciation/amortization policies; this issue can be
resolved by cross-checking the equivalent EBITDA margins.

Changes to the Business Mix – Another question is the extent to which margin
change is driven by the business mix. It is positive that management adjusts the
portfolio to maximize returns, but gains achieved by eliminating problems are by
definition unsustainable, and so less valuable than like-for-like progress.

13
 Globa l
0 6 Jan ua ry 2 009

EPS Growth EPS Growth

The expected five-year CAGR from the Concept
last actual EPS The trendline growth rate expected for the business over the medium term.
Or, if this is not meaningful,
This is connected to, but not necessarily the same as, the sustainable growth rate that
The sustainable trendline growth rate = the business, in its mature, steady state, is expected to deliver over the long term.
ROE * (1 – DPS / EPS)
Definition
Five-year CAGR The expected five-year Compound Average Growth Rate (CAGR) in EPS from
the last actual year, where meaningful.

Because the brackets have to be exact in Excel, it is worth recapping the formula:

CAGR = (((EPSY5/EPSY0) ^ (1/5)) – 1) * 100

There are many ways of defining EPS. Whichever is used in calculating the P/E
Ratio should be used again here.

“Sustainable” growth Where the five-year CAGR is not useful (where it is negative, or at cyclical turning
points) we should alternatively use the analyst’s assessment of the sustainable
growth of the business. For a mature business, in steady state, this can be
determined objectively by multiplying the Return on Equity (where appropriate,
averaged over a cycle) and multiplying by the Earnings Retention Ratio
(1 – dividend/earnings).

Interpretation
This is a critical driver for a wide range of valuation metrics and techniques.

Key determinants of DuPont Analysis

EPS growth The components of EPS growth were definitively decomposed in the 1920s by the
then-CFO of the General Motors Corporation (which at the time owned DuPont).
In very simple terms, it can be shown that the product of the asset turnover and
the EBIT margin generates the ROCE. Grossing this up by the Net Debt/Equity
Ratio leads to the ROE, and the growth in EPS is the ROE multiplied by the
Earnings Retention Ratio.

So the key drivers of EPS growth are:

„ Capex (investment in productive assets)

„ Asset Turnover
„ EBIT Margin
„ Net Debt/Equity Ratio
„ Dividend Payout Ratio
Note that EPS growth is always an increasing function of the first three drivers,
and always a decreasing function of the Dividend Payout Ratio. The Net
Debt/Equity Ratio is usually an increasing function, unless the ROCE is lower
than the cost of debt!

Watch Out for...

Meaningless Answers
The calculated CAGR may not be meaningful, either because earnings move
from negative to positive (or vice versa), or because the base earnings are close
to zero. Equally, the CAGR may be arithmetically valid but misleading if the
opening or closing periods are near turning points in the earnings cycle (trough to
peak effects, or vice versa). In all of these cases, we should specify an estimate
of the underlying projected growth, rather than the calculated CAGR.

14
 Globa l
0 6 Jan ua ry 2 009

How to calculate the CAGR: This can be evaluated in various ways:

Logarithmic regression. This is the most rigorous approach, taking full account
of the impact of compounding.

In Excel, the growth rate in percent (the slope of the regression line) is given by
the expression:

= (INDEX ( LOGEST (Range) , 1) – 1) * 100

Where range is the range of cells holding the EPS values.

It suffers from the drawback, commonplace in practice, that any negative or zero
EPS value will render the calculation impossible (because logarithms are not
defined for negative numbers, or zero).

Linear regression. This is less valuable, being equivalent to the average of the
annual growth rates, but ignoring compounding. It does have the advantage that it
will always return a meaningful number, and so can be used where the
logarithmic approach fails because of a negative or zero value.

The Excel formula is:

= INDEX (LINEST (Range), 1)/AVERAGE (Range) * 100

Again, where range refers to the cells holding the EPS values.

Weighted averages. For those with long memories, the Wells Fargo (Sharpe)
method used a good protocol to estimate sustainable growth. This was a weighted
average of: the last three years’ actual growth (50%); the next two expected (30%);
and the Return on Equity multiplied by the Earnings Retention Rate (20%).

Over-optimism
Long-term growth is rare In reality, exponential growth is very hard to sustain. Very few companies have
actually maintained high double-digit growth for a decade. Even where growth is
well established, there are ceilings in every market place – and as that ceiling is
approached, growth becomes harder to sustain and eventually fades.

This logistic growth pattern is intuitively consistent with many real world situations.
In botany, it can be applied to how tall plants can grow, and in economics, to
market shares. The issue for us, in terms of valuation, is that we sometimes fail to
appreciate just how quickly this effect can come into play, especially where initial
rates of growth are impressive.

To illustrate this, Chart 2 plots a hypothetical series (for example, sales of a new
must-have consumer product) against time. Initial sales are 50 units, growing at
20% a year, and research suggests this could ultimately rise to 1,000 units. The
difference between exponential growth (constant 20%) and logistic growth (which
slows as the ceiling is approached) is dramatic; if our valuation model assumed
sales of 600 units, the exponential assumption would hit that level at year 13, but
logistic growth not until year 17. At higher assumed growth levels, the effect is
even more spectacular.

15
 Globa l
0 6 Jan ua ry 2 009

Chart 2: Exponential Versus Logistic Growth

1,000

Exponential growth
800
The difference between exponential growth
and logistic growth:
Logistic growth
If our valuation model assumed sales of 600 600
units, the exponential assumption would hit
that level at year 13, but logistic not until 17
400

200

10

12

14

16

18

20

22

24

26

28

30
0

8
Time (years)

Source: Merrill Lynch

An excellent reality check is to consider the absolute level of earnings implied by

extending our growth forecast out as far as the P/E in years…

Quality of Growth
Compare the CAGR and ROE The CAGR in EPS can be distorted by arithmetic, by cyclicality and by
overoptimism; one test that can screen for all three effects is to compare the
expected CAGR to the “sustainable” growth rate: ROE * (1 – EPS/DPS). Where
this ratio is equal to unity, all of the expected growth is explainable on
fundamentals. Where it is less than one, it suggests either that positive cyclical or
recovery effects dominate the medium-term prospects, or that high expectations
may need more supporting evidence. Equivalently, the opposite holds true where
the ratio exceeds unity.

Visible Is Valuable!
Growth coming from recovery factors (EBIT restructuring, interest savings, tax-
loss carryforwards) is usually ephemeral. Similarly, margin growth from
costcutting is harder to sustain than revenue expansion, and organic (or internal)
growth is more visible than hoped-for acquisitions.

Where revenue growth is concerned, price increases are much more valuable
than volume gains because they involve no increase in costs, and have a much
greater impact on profits.

Growth at Any Price?

Growth is not a goal in its own right; there is a major difference between achieving
growth efficiently, in terms of investing in high-return projects, or diluting the firms’
overall economic returns by inefficient investment. Where expansion dilutes the
ROCE, it is less valuable.

16
 Globa l
0 6 Jan ua ry 2 009

Free Cash Flow Free Cash Flow

Cash flow from operations, less all Concept
capital expenditure The extent to which the enterprise generates cash in excess of its essential
spending. Equivalently, the cash that could be withdrawn from the business
without damaging its structure or prospects.

Definition
Strictly defined as cash flow from operations less maintenance capex.

Cash flow from operations is defined as:

Net income
plus: depreciation, amortization, and deferred taxes
plus/less: any other non-cash charges/credits to the income statement
plus: minority interests, less dividends to minorities
less: the difference between dividends received from associates and the profit
from them included in net income
plus: the net change in provisions
less: the net change in (non-cash) working capital.
The problem comes in defining maintenance capex in an objective manner. The
textbook definition is the minimum capital investment necessary to maintain the
business in its current condition. It is virtually impossible, however, to distinguish
between the essential capex necessary to sustain the business in its current
state, and that which is directed toward future expansion. Moreover, our valuation
often assumes that the structure will change; in this respect, planned growth is
essential to the argument, so all the forecast capex is essential.

On these grounds, our definition of Free Cash Flow is cash flow from operations
less total capex.

Dividends: in or out? The issue of dividends often crops up in debate, and our view is clear: if it is
possible for management to avoid payment, it is not essential spending, and does
not represent a reduction in free cash flow. Thus dividends paid to minority
holders, which are authorized by the subsidiary board, and preference dividends,
which are a fixed commitment, are both deducted in the calculation.

On the other hand, dividends paid to ordinary shareholders – regardless of

precedents or expectations – can be cut or cancelled by management if
necessary, so ordinary dividends do represent discretionary spending, and – as
such – are not chargeable against free cash flow.

Interpretation
How much to take out Free Cash Flow to a business is analogous to an individual’s net salary, less
committed expenditure.

Because all essential spending has been deducted, Free Cash Flow indicates the
discretionary spending power of management; in investment terms, it represents
option value, allowing management to either reward shareholders with dividends
or buybacks, expand the business by raising capex, or alternatively, reduce risk
and interest costs by repaying debt.

Symmetrically, free cash flow is also a good measure of financial flexibility, which
has multiple strategic implications.

17
 Globa l
0 6 Jan ua ry 2 009

Watch Out for...

Traps
Although intuitively appealing, this measure can send a dangerous false-positive
signal when a company is stagnating or contracting. It is essential to use this
measure in conjunction with the Asset Replacement Ratio (page 18) to ensure that
Free Cash Flow is sustainable, and not merely flattered by declining capex.

Equally, the option value of positive free cash flow may be illusionary if debt
reduction is an over-riding priority.

As with many other metrics, a high free-cash-flow yield is an indicator of potential

value. In practice, a catalyst (such as a change of management) is needed to
release that value.

Quantification Issues
Hard to quantify Although historical figures based on actual data are extremely useful, forecasts of
Free Cash Flow are subject to an even wider margin of error than other
accounting projections. The reason is that they depend on changes in working
capital, which are notoriously difficult to forecast. This is a structural issue, arising
from the fact that working capital essentially represents the difference between
two large numbers. Even if the analyst were successful in forecasting the
components with reasonable precision, the potential error in the difference will
always be significantly higher.

Some suggest that the change in working capital should arbitrarily be set at zero,
others that it be omitted altogether. We disagree: the problem is difficult, but not
insurmountable. It is, however, critical to bear in mind that forecasts are far less
valuable.

18
 Globa l
0 6 Jan ua ry 2 009

3. Quality of Earnings
Not all earnings are equal! We believe that “high-quality” earnings should be viewed
with greater confidence, and by implication, others treated with more caution. These
metrics are intended to test the underlying security of the business performance.

Defining High-Quality Earnings

Changing perceptions of quality When the bull market was raging, investors viewed high quality rather differently
than today. The priority then was earnings momentum; investors prized stocks
that delivered regular, visible growth in EPS or EBITDA, and in general, were
happy to focus solely on that issue, as long as expectations were met or
exceeded. Times have indeed changed, and a broader range of issues needs to
be addressed in the current climate. We suggest that high-quality earnings be
defined as those that meet the following criteria:

„ Substantially realized in cash

„ Sustainable, in that the productive asset base is maintained by adequate
investment
„ Do not arise from potentially transitory tax benefits
„ Are not prejudiced by over-aggressive, financial leverage
„ Interest payments are adequately covered

Cash Realization Ratio

Cash Realization Ratio
Concept
The proportion of profit that is realized in cash.
Cash Flow from Operating Activities
_____________________ Definition
Net Income Cash Flow from Operating Activities divided by Net Income, where Net Income is
positive.

Interpretation
Profitability measures firm performance over a period, and differs from cash flow
in that it measures the completion of the transaction cycle when products are
delivered (or services are performed), not necessarily when cash is received.
Cash Flow measures the collection of cash. Thus, Cash Flow involves a longer
cycle than profitability because it typically takes longer to collect cash. Net Income
can be thought of as the sum of Cash Flow and Accruals. The former is intuitively
more valuable, and this ratio effectively measures the balance between the
tangible and intangible aspects of economic performance.

Cash Flow from Operating Activities represents the Cash Flow from trading
activities, as opposed to investing and financing. In broad terms, this amounts to
Net Income, plus Depreciation and Amortization, Deferred Taxation, minus
Increases in Working Capital, so this ratio is essentially driven by the balance of
these factors. A positive (high value) would be consistent with conservative
depreciation policies, high reinvestment (hence deferred tax), and effective
working capital management. Equivalently, a number below unity could indicate
the contrary.

Note the qualified language here: consistent with does not mean implies. There
may be perfectly valid reasons for cash and profit to be misaligned in the short
term, but the trend should iron these out. The relation between Net Income and
CFFO is also dependent on the current stage of the company’s life cycle.

19
 Globa l
0 6 Jan ua ry 2 009

Watch Out for…

False positives arising when the calculated ratio is boosted purely because net
income has collapsed; low net income is not positive!

Asset Replacement Ratio

Concept
The rate at which productive assets are replaced.

Definition
Asset Replacement Ratio
The Asset Replacement Ratio is equal to Capital Expenditure divided by
Capital Expenditure Depreciation.
_____________________
Depreciation In this context, both “capital expenditure” and “depreciation” refer exclusively to
investments in productive, tangible fixed assets (property, plant & equipment).

Interpretation
Bottom line: Is the asset base If the value of this ratio is less than unity, the asset base is shrinking. This is
growing or shrinking? essentially a reality check on growth expectations and Free Cash Flow. Although
it is possible for a business to expand while simultaneously shrinking its asset
base (by increasing asset turnover), and equally that Free Cash Flow can be lifted
by underinvestment, both of these effects have clear limits, and are intrinsically
unsustainable. As with so many ratios, it is worth comparing current levels to
previous trends, as well as across peer groups.

Watch Out for…

In practice, depreciation rates do not always reflect the actual life of underlying
assets. The Asset Replacement Ratio could send a false-negative signal if the
depreciation rate is conservative. Equally, an inadequate charge will flatter the
result.

Disparities between assets’ historical costs could significantly affect this relation.
Acquisitions that require the acquired entity to be consolidated at the market
value of assets and liabilities on the date of acquisition could also affect the
comparability and interpretation of this number.

The classic test is to divide the historical cost (or gross) book value by the
depreciation charge; this multiple is the effective average write-off period, and should
be compared to an intuitive assessment of the assets’ expected economic life.

Another clue can sometimes be found in the notes to the accounts under other
operating income, this often includes gains or losses arising from disposal of fixed
assets. A pattern of regular gains suggests that assets were written down to
levels below their market value, that depreciation was higher than necessary, or
vice versa.

Cross-check the age of assets Where the ratio is sending a strong signal, say under 0.5 or above 1.5, and for
peer group comparison, it is well worth cross-checking on the average age of the
assets. This is calculated by dividing the accumulated depreciation (historical cost
minus book value) by the annual depreciation charge. Where assets are
approaching the end of their lives, a high replacement ratio should carry far less
weight, as it represents catching up rather than expansionary investment.

These two effects are highlighted in Table 6 for three similar companies. Stocks A
and B have similar asset replacement ratios (both in the positive zone), but stock
B’s assets are significantly younger than A’s. In reality, B could have invested

20
 Globa l
0 6 Jan ua ry 2 009

rather less than A and still have had a comparable asset base. On the other
hand, stock C has a low ARR, but solely because it depreciates over a shorter
period. Again, its asset base could be in perfectly adequate condition, despite its
low nominal replacement rate.

Table 6: The Impact of Asset Life and Depreciation Policy

Fixed Assets Depreciation Write-off Average Age/Wo Cap
Gross Net Accum Charge Period Age P% Ex ARR
A 3,000 1,000 2,000 300 10.0 6.7 67 350 1.2
B 3,000 2,000 1,000 300 10.0 3.3 33 350 1.2
C 3,000 2,000 1,000 400 7.5 2.5 33 350 0.9
Source: Merrill Lynch

Tax Rate
Concept
Tax Rate At first inspection: “The proportion of profit claimed by the state.” More subtly:
Tax Charge “How much of a company’s profit are its shareholders able to keep?”
_____________________
Pre-Tax Income Definition
The tax charge in the income statement, as a percentage of the pre-tax income,
where pre-tax income is positive.

Interpretation
Defining a low tax rate? Corporate tax rates vary considerably around the world. In particular, some
countries (such as Ireland and the UK) have relatively low taxes, while others
such as Germany and Japan impose significantly higher rates.

For multinationals, the effective rate can be thought of as the average official rate
(weighted by the proportion of its profits in each fiscal center), less the extent to
which it is able to take advantage of concessions. A “low” tax rate in this context is
where the effective rate is significantly lower than the average official figure.

We need to draw the distinction between companies that are lightly taxed because
they are based in favorable fiscal regimes, and those that depend upon manipulation
to save tax. In practice, where this ratio sends a warning signal, we must consider
the geographical mix of earnings before assuming that there really is risk.

The risks of a low tax rate The idea of minimizing tax payments is intuitively appealing, but there are valid
reasons why apparently low tax rates should ring warning bells about valuation.

Underlying profit may be lower than the reported figures suggest.

Many businesses, quite properly, take advantage of financial engineering to
minimize the tax payable on their underlying activities. In one sense, reported
profit can be thought of as the sum of real earnings and artificial tax maneuvers.
Where a tax rate appears low, this is consistent with a higher proportion of
artificial profits, and underlying profits that fall short of the headline number.

Tax-driven benefits can prove transitory.

Tax benefits rarely last forever. Many tax benefits are timing differences that
simply shift the payment of taxes into later periods, rather than permanent
GAAP/tax law differences. Loss carryforwards become exhausted, and fiscal
loopholes are eventually closed. Even if new tax-sparing opportunities
subsequently arise, a significant element of volatility is unavoidable when an
existing concession is withdrawn (or threatened), for example, US proposals to
limit Tax Inversion earlier this year.

21
 Globa l
0 6 Jan ua ry 2 009

Watch Out for…

This is an asymmetric indicator, in that even though a low tax rate can be grounds
for caution, a tax rate above the expected rate does not necessarily imply
prudence, or high quality.

As with many other quality of earnings measures, a single year’s data may be out
of line for valid reasons, but a consistent trend of apparently low rates is a strong
danger signal.

Net Debt/Equity Ratio

Concept
Net Debt/Equity Ratio The proportion of capital provided by lenders, compared to that provided by
Debt less Cash and Equivalents owners; internal capital compared to external.
_____________________
Definition
Total (Shareholders + Minority) Equity
Net Debt expressed as a percentage of equity capital.

In this context, debt is defined as all interest-bearing liabilities to banks, credit

institutions and capital markets; finance leases; and all non-equity capital
instruments such as redeemable Preferred shares, or equity-linked instruments
(such as convertibles), including those issued by subsidiaries.

Similarly Net Debt is debt less cash and equivalents. Cash equivalents are
marketable securities, or claims that are readily convertible into cash at three
months or less.

Include all equity Equity consists of issued and fully paid capital, plus all reserves, retained
earnings, and additional paid-in capital. The consolidated figure (shareholders’
plus minority) is taken because, on the debt side, that proportion of borrowings
attributable to minorities is usually neither explicitly identifiable, nor stripped out.

Interpretation
How long is the lever? The Net Debt/Equity Ratio is the most commonly used measure of financial
leverage. It indicates both financial risk and capital efficiency (in that the cost of
debt is generally lower than the cost of equity). In the current, subdued climate,
the former predominates. High levels of debt or low levels of equity lead, at
worst, to bankruptcy and, at best, to the curtailment of strategic ambitions. In
more buoyant economic environments, capital efficiency – using cheap debt
rather than expensive equity – is also indicated by this ratio, with a higher figure
suggesting a lower cost of capital. Whatever the merits of using financial leverage
to optimize returns, it is certainly the case that low debt levels have an implicit
option value relative to high leverage.

Finally, this measure must always be placed in context: we must have data on
Interest Coverage, Free Cash Flow, and ROCE to pass judgment on whether a
given level of financial leverage is appropriate.

Watch Out for...

Significant holdings of cash
Although Net Debt/Equity is a generally valid measure, we must be cautious
where net borrowings differ materially from the gross. There are four key reasons:

„ Debt and cash are seldom offsettable at par. It is rarely true that a dollar
of cash can extinguish a dollar of debt, because most loans have a fixed
component and a term structure, and redemption will invoke cost penalties
(especially when rates are falling).

22
 Globa l
0 6 Jan ua ry 2 009

„ It might seem obvious, but most businesses hold cash in order to spend
it. Just because the last balance sheet showed cash at high levels, it does not
necessarily follow that those reserves are still available. That cash might have
been the proceeds from an IPO or bond offering that have been put to work
since the balance sheet date.

„ Some businesses need to maintain high levels of liquidity to deal with

lumpy cash flows. The effect is to flatter underlying net debt, because that
cash is essential to the operation of the business; it is not actually available
to reduce debt.

„ Certain sectors (contracting is the classic example) receive significant

proportions of their turnover in advance. Although the money is legally
theirs, these deposits really represent an interest-free loan from the
customers. Again, not all the money is available to reduce debt, as it is
required to meet future commitments or costs. It may well be that similar
deposits are received to maintain the pipeline, but if the flow of new orders
dries up, liquidity will dry up as well.

Prepayments from customers are Soup-to-Nuts Example

actually debt Table 7 exhibits the key inputs, Table 8 is a step-by-step worked example to
evaluate the Net Debt/Equity Ratio for this (hypothetical) example, and
demonstrates two ways in which a misleading situation can arise.
Cash equivalents: marketable securities,
or other claims that can be readily Table 7: Summary Financial Data
converted into cash at three months or Company B Year to April (EUR mn) 2002 2003
less Current assets 2,120 2,380
Inventory 620 650
Receivables 450 480
Debt: all interest-bearing liabilities to Cash and equivalents 1,050 1,250
banks, credit institutions and capital
markets, and finance leases. Current liabilities 1,785 1,857
Includes non-equity capital instruments Short-term debt 330 360
(redeemable preference) and equity- Payables 405 417
linked instruments (convertibles) Payments in advance 850 880
Other current liabilities 200 200

Equity: the total of all classes of capital + Non-current liabilities 1,900 1,915
reserves + retained earnings = common; Long-term debt 1,200 1,200
common + Preference = shareholders’; Provisions for risks and charges 100 105
shareholders + minority interest = total Other long-term liabilities 600 610

Equity 6,150 6,505

Payments in advance: customers’ Contributed capital 1,000 1,000
payments for goods and services not yet Reserves 3,200 3,400
provided. Belongs to the company Retained earnings 1,250 1,400
legally, but not economically. Preference stock 500 500
Minority interest 200 205

Interest expense -200 -220

Source: Merrill Lynch Imagination

Note that, at first inspection, financial leverage appears modest; the numbers show
the latest reported Net Debt/Equity Ratio at 43%. But our attention was drawn to the
high level of cash, relative to debt, and a closer look at the accounts reveals that most
of this cash appears to flow from customer payments in advance, and is not – in
practice – available to repay debt. Taking this into account, the adjusted Net
Debt/Equity Ratio rises from 43% to 57%.

23
 Globa l
0 6 Jan ua ry 2 009

Our suspicions aroused, we decided to conduct further tests. Averaging the

opening and closing total debt levels, and matching this to the interest expense
incurred over the latest year, it seems that the company has been paying interest
at an effective rate of 14%. This compares to a typical borrowing cost of, say, 7%.
It is possible, of course, that the company has made a bad bargain with its
bankers. It is more likely that actual borrowings over the year were much higher
than at the year-end, which in this hypothetical exercise was clearly selected to
coincide with the peak of the annual cash cycle.

If we gross up the annual interest expense by a more realistic interest rate, the
impact on the implied level of financial leverage is substantial: 43% rises to 68%,
and allowing for the payments in advance, 82%.

Table 8: The Spreadsheet

Leverage looks reasonable, but note Raw Basis Adjusted Basis
that cash is two-thirds of debt; why
should this be? That cash is actually Year to April 2002 2003 2002 2003
Calculation of net debt/equity
inflated by high levels of customer
payments in advance, which are not, Short-term debt 330 360 330 360
in reality, available to offset debt. + Long-term debt 1,200 1,200 1,200 1,200
= Total debt 1,530 1,560 1,530 1,560
Adjusting for this shows the effective
- Cash & equivalents 1,050 1,250 1,050 1,250
leverage to be higher.
+ Payments in advance 850 880
= Net debt 2,580 2,810 3,430 3,690
/ Total equity 6,150 6,505 6,150 6,505
= Net debt/equity 42% 43% 56% 57%

Calculation of implied interest rate

The company appears to be paying Interest expense 220 220
14% interest; why? Clue is the year-
/ Average total debt 1,545 1,545
end, picked to coincide with peak of
= Implied interest rate 14% 14%
trading cash cycle for this industry.
Gross up the interest paid by an Calculation of implied average net debt/equity
estimated market rate to get a Interest expense 220 220
measure of the average debt over / Estimated interest rate 7% 7%
the year; implied leverage soars. = Implied average total debt 3,143 3,143
- Average cash & equivalents 1,150 1,150
+ Average payments in advance 865
= Implied average net debt 4,293 5,158
/ average total equity 6,328 6,328
= Implied average net debt/equity 68% 82%
Source: Merrill Lynch Imagination

Interest Cover Interest Cover

Concept
EBIT
_____________________ The ratio of operating profit to interest cost. As with all cover metrics, it essentially
compares the size of a claim on resources to the funds available to meet that claim.
Interest Expense
Definition
Net interest not enough EBIT divided by interest paid. Please note: we should not use “net” interest
(interest expense less interest income), for two reasons. First: a reduction in the
denominator of a ratio has a highly leveraged effect compared to an increase in
the numerator; deducting interest received from the bottom line flatters the ratio
excessively. Where interest income is significant, and ongoing, the logical
approach would be to add the ongoing interest income to EBIT in the numerator.

Our second reason is more telling: Interest Cover is essentially a measure of risk,
and when stocks appear on the risk horizon, the fact that they earned interest

24
 Globa l
0 6 Jan ua ry 2 009

income last year is usually academic; the cash has gone, but the debt remains.
Therefore, we adopt the more stringent approach, and ignore interest income
altogether.

EBITDA in the numerator Although the equity industry practice is to base the numerator on EBIT, there is
considerable merit in replacing that with EDITDA. Such a move would bring us
closer to general practice on the credit side, and is also logically appealing –
given that interest has to be paid in cash, which EBITDA comes closer to
measuring than EBIT.

Interpretation
This is principally an indicator of security, becoming relevant in the context of
weak free cash flow or high leverage. Alternatively, it is used to estimate how
much additional borrowing capacity remains before risk becomes excessive.

Watch Out for...

Use of net interest rather than interest paid.

25
 Globa l
0 6 Jan ua ry 2 009

4. Valuation
These measures connect the economic performance of the business to its market
value in various ways. The Price/Earnings Ratio is a payback indicator; the
Price/Book Ratio reflects the valuation of the company’s equity. The Dividend
Yield and Free Cash Flow Yield reflect tangible and potential monetary rates of
return, respectively.

Enterprise Value/EBITDA is a general structure-neutral cash generation multiple,

and Enterprise Value/Sales indicates volume leverage.

Price/Earnings Ratio
Price/Earnings Ratio Concept
Current Share Price Essentially, this is the payback period in terms of earnings. How many years’
_____________________ earnings per share does it take to recover the share price?
Diluted EPS
(both Reported and Adjusted) Definition
The share price divided by earnings per share, where EPS is positive.

Earnings per share

Conceptually, attributable net earnings per unit of capital.

Basic and diluted EPS Our definition of basic EPS is:

Net income before abnormal items, less preference dividends, divided by

the time-weighted average of dividend-franchised shares in issue.

But because many companies have current obligations that could require them to
increase the number of actual shares, such as options, convertibles, etc., the basic
EPS often overstate the underlying earnings per unit of capital. A prudent approach
demands that valuation should always be done on the basis of diluted EPS.

The definition of diluted EPS adds the total of new shares that would have to be
issued if those options, convertibles, etc. were to be exercised to the
denominator; it now reflects the potential, rather than the actual, shares in issue.
In addition, the basic EPS numerator can be adjusted to reflect any potential post-
tax impact on earnings that would arise from that exercise. There are many
permutations and combinations of possible adjustments, depending on the
instruments that create the dilution, and some of these are detailed on page 25.

Please note: the denominator is not necessarily a constant. It is essential to ensure

that historical average shares are adjusted whenever there is a stock dividend,
bonus issue or split, or a capital increase at a discount to the ex-rights price.

Adjusting EPS for consistency in This detailed, but apparently straightforward, definition masks a key issue in that there
valuation is no unique definition of earnings per share that manages to transcend the
boundaries of local accounting standards and investor practice. The situation has
improved in recent years, and further convergence is expected, but substantial
differences still exist in three areas: goodwill amortization, deferred tax, and pensions.

It is generally accepted that where net income is affected by unusual factors,

these should be stripped out of the P/E calculation – because our goal is to value
the shares on an ongoing basis. Similarly, in seeking to compare stocks reporting
under different conventions, it is essential to eliminate inconsistencies arising
from differing accounting practices – for example, by adding back goodwill
amortization.

26
 Globa l
0 6 Jan ua ry 2 009

The heart of the problem is defining an unusual factor; there has always been a
temptation to stretch the concept in order to flatter EPS, and in the current
climate, arguments based on adjusted (or in the US, pro forma) earnings often
stand in a poor light.

Getting our story straight Comparative P/Es are often distorted by semantic differences across different
cultures; in particular, we have the problem of aliases where various expressions
are used to describe EPS that have been adjusted in similar, but not always
identical, ways. How, for example, could we distinguish between clean and
underlying EPS? Equally, why – outside the UK – can an investor be expected to
be familiar with the esoteric distinction between exceptional and extraordinary?

In order for our global definitions to be consistent, we must start by defining

unambiguous, explicit terminology. There are two discrete reference points:
reported earnings, presented by the company under local GAAP, and adjusted
earnings, used to calculate the P/E. We should use these expressions, rather
than synonyms or aliases, in all global comparisons. We then need to map the
route from reported to adjusted.

Setting the rules A Systematic Protocol to Define Adjusted EPS

The common starting point is local GAAP-reported Net Income attributable to
ordinary shareholders (after minority interests and preferred dividends). There are
three distinct types of adjustments that are then applied (where appropriate) to
lead us to Adjusted Net Income (and so to adjusted EPS):

„ Adjustments allowable under local GAAP – already included in reported EPS

„ Analysts’ adjustments to reverse unusual factors
„ Analysts’ adjustments to reverse inconsistencies among local GAAPs
(1) Adjustments Allowable Under Local GAAP
Most GAAPs calculate EPS by adding back selected unusual items to Net
Income, and so these will already have been included under reported EPS. The
adjustment must not be double counted!

Under US GAAP (and generally, elsewhere) these are limited to:

Discontinued Operations
For this purpose, a discontinued operation refers to a separately identifiable
segment of a company, a major line of business, or class of customer. It usually
takes the form of a clearly identifiable subsidiary or division whose operations and
financials can be clearly distinguished from the rest of the company. If a disposal
qualifies, the gain or loss on sale and the results of its operation until the sale is
complete are reported separately from the company’s income from continuing
operations. All prior periods presented in a financial statement should show
separately the results of operation of the segment to be disposed.

The fact that the results of the portion of the business that is being sold cannot be
separately identified argues against treatment as a discontinued operation. For
example, the sale of an entire business unit may qualify as a discontinued
operation, whereas the sale of one of the three plants and half of the sales offices
that make up the unit would not.

Changes in Accounting Principle

This represents the one-time effect of a change from one accounting principle to
another. It is generally occasioned by the introduction of new accounting standards
and the adjustment made to transition from the old standard to the new standard.

27
 Globa l
0 6 Jan ua ry 2 009

Extraordinary Items
Extraordinary items are those which are both unusual and infrequent. The test for
whether an item not specifically identified by US GAAP is extraordinary is very
demanding. Infrequent means that it is not likely to recur in the foreseeable future,
given the company’s environment.

US GAAP specifies certain items that are always extraordinary and others that
are virtually never to be considered extraordinary, and this constitutes a very
useful model for other GAAPs:

Examples of items required to be considered extraordinary:

„ Gains from negative goodwill on an acquisition

(in the rare case where the value of the assets exceeds the purchase price)
„ Most expropriations of property
„ Losses resulting from prohibition under a newly enacted law
(2) Analysts’ Adjustments to Reverse Unusual Factors
These are items generally deemed to be not extraordinary under GAAP, but
which exert a distorting influence in assessing the future trend, and so should be
properly reversed out of net income in calculating adjusted EPS.

We must stress that unless these items have been stated on an after-tax basis,
nominal tax must be deducted from the gross amount to reverse the impact on
Net Income. The key here is that these are items that impact only the current
year’s earnings, and do not impact future earnings. Investors and analysts should
be cautious here, as companies often try to encourage financial statement users
to treat certain recurring expenses as one-time items (e.g., pro forma items).

„ Write-downs/write-offs of receivables, inventory, equipment leased out,

deferred R&D costs, or other intangibles
„ Foreign exchange and translation gains/losses
„ Gain/loss on disposal of a portion of a business unit (as opposed to the entire
unit)
„ Other gain/loss from sale of property, plant & equipment used in the business
„ Effects of a strike
„ Adjustment of accruals on long-term contracts
This list is, of course, by no means exhaustive.

(3) Analysts’ Adjustments to Reverse Inconsistencies among GAAPs

Although there has been substantial convergence between national GAAPs and
IAS, difference still remain. In particular, major differences exist in three key areas:

„ Goodwill amortization
„ Unfunded pension liabilities
„ Dilution
Reverse goodwill amortization We regard it as appropriate to add back the cost of goodwill amortization in the
calculation of adjusted EPS, on the grounds of comparability. The decision by the
FASB to end the requirement for US companies to amortize goodwill marked a
sea change in this ongoing controversy, and we expect the process of
convergence to support our position.

28
 Globa l
0 6 Jan ua ry 2 009

Because, outside the USA, goodwill amortization is rarely a tax-deductible

expense per se, it is probably justifiable to add it back in full.

Pension costs: no adjustment The issue of pension liabilities is more problematic. There are two points:
underfunded liabilities and unfunded liabilities. We think that the quantified costs
of restoring underfunded liabilities, where these arise, should not be stripped out
of the EPS calculation – even if it was the case that other GAAPs would not
require them. This is because, unlike with goodwill amortization, there is a very
wide range of possible treatments, rather than a straightforward yes/no approach.

On the issue of unfunded (pay-as-you-go) pensions, as in Germany, we feel that,

as there is no satisfactory adjustment possible to render comparability with
funded schemes, this source of accounting inconsistency is best ignored!

Dilution: different conventions We must also take account of the different conventions on dilution. In the US,
EPS is always diluted to reflect the impact of stock options, as well as equity-
linked capital instruments. In Britain and Europe, basic EPS is used unless the
impact of dilution is material (over 5%) when the lower of basic or diluted is taken.
The divide is driven by the much greater prevalence of options in the US; these
are rarely material in other countries, where dilution generally arises through
warrants and convertibles.

The rules for calculation of diluted EPS are convoluted in the case of stock
options, but straightforward for other instruments. For convertibles, the
denominator is increased by the number of potential new shares, and the
numerator also increased by the estimated post-tax interest saving. For warrants,
the denominator is first increased by the number of potential shares, then reduced
by the number of shares that could be bought back in the market with the
exercise proceeds.

Interpretation
The P/E Ratio is the most widely used and most valuable metric in our iQmethod,
and is applicable to virtually all stock and sector situations.

It is vital to stress that a low P/E in itself is rarely a sufficient argument to consider
a stock undervalued (or vice versa). Given our expectations of earnings, we need
evidence to suggest why the rating should change before we can draw a valid
inference. This usually takes the form of comparing the stock’s P/E against either,
or both, of two benchmarks:

„ Peer Group Comparison

There is typically a valuation hierarchy within a sector, with stocks trading on
higher P/Es perceived as having better prospects or lower risk than their peers.
Such hierarchies may have little objective basis, and investment opportunities
often arise when the case for change can be made – for example, that a stock’s
discount (or premium) to the sector is at the limits of its rational range.

„ Cyclical Trends
P/Es often show strong cyclical trends, driven by the overlay of
macroeconomics and investor sentiment. Current P/E levels are often
usefully compared to the limits of their cyclical range, or to levels reached “at
this point in the last cycle….”

Connecting the P/E to growth In absolute terms, however, the P/E is conceptually linked to growth in earnings
per share (as defined earlier).

29
 Globa l
0 6 Jan ua ry 2 009

Most investors expect their stocks to have positive EPS growth, at least over the
medium term, and it is rational for stocks with higher expected growth rates to
command a premium P/E. The problem is in finding a useful connection between
the P/E and the expected growth rate.

PEG: a first step The simplest measure is the PEG Ratio, obtained by dividing the P/E by the
expected growth rate. Note that to maximize the objectivity, this should use the
historical (known) EPS and the expected trendline growth rate. Stocks with a low
PEG Ratio are seen as better value, even if their P/E is higher, by virtue of the
implied value of future earnings.

Although its simplicity makes it widely used, the PEG Ratio has a serious flaw in
that the relationship between growth and value is non-linear. This has the effect of
biasing the metric of higher levels of assumed future growth. For example, in
simple terms, 20% growth for one year is actually less than 10% for two years.

As well as flattering high growth, this non-linearity also makes the PEG Ratio
excessively sensitive to changes in the assumed growth rate.

This bias toward growth is exacerbated by the fact that the PEG substantially
ignores the value of current operations relative to future expectations; it is
intrinsically risky.

In order to overcome these issues, some analysts modify the basic ratio in two
ways:

PEG: modifications „ Dividing the P/E by the logarithm of the growth rate, rather than the actual
rate, to restore the linear relationship.
„ Adding the current Dividend Yield to the growth rate in the denominator to
acknowledge the value of current operations (the PEGY Ratio).

Lateral thinking: the payback The payback period is probably the most widely used valuation tool in commerce,
period being easy to calculate, universally applicable, and intuitively rational.

We introduced the concept of the P/E as a simple static payback period, and it is
worthwhile to extend this idea to take account of earnings growth. Chart 3 plots
EPS against time for a hypothetical stock. Just to simplify the calculations, the
price is 100, the historical EPS is 10, and earnings are expected to grow at 10%
yearly. The historical P/E can be thought of as the point along the Time (x) axis at
which the area under the historical EPS line equals the price: 10 years.

Equivalently, the payback period can be identified as the point on the Time (x)
axis at which the area under the expected EPS curve equals the price, in this
example about seven years. The area under the curve is effectively the historical
P/E adjusted for the logarithm of the growth rate.

This measure avoids the bias and excessive sensitivity of the PEG Ratio. It has
the advantage of sharing the look and feel of a P/E; its drawback is the complexity
of the calculation!

30
 Globa l
0 6 Jan ua ry 2 009

Chart 3: The Earnings Payback Period

Dynamic Payback 30
At the payback point, the area under the Expected EPS
curve of expected EPS is equal to the share
price. The higher the growth, the earlier this 25
is achieved. Payback period
20

Earnings per share

15

Historic P/E
10 Historic EPS

5 Price = 100

0
0 1 2 3 4 5 6 7 8 9 10
Time in years

Source: Merrill Lynch

Watch Out for…

Adjustments for pre-tax Errors in Adjusting Reported Earnings
adjustments must consider It is not unknown for the gross amount of pre-tax unusual items to be added back to
nominal tax net income, significantly flattering the adjusted figure. If the abnormal cost had not
been incurred, pre-tax profit would have been higher, but that gain would still have
been taxable and the rise in net income lower.

Cyclical Lows
Where earnings are volatile, P/Es can often drop to extremely low levels, and
mean reversion can make these look attractive. The problem is that low P/Es can
revert to the mean in either of two ways, higher P or lower E. The latter can (and
do) drop faster and farther than we often expect.

Sensitivity to Expectations
High-growth, high-P/E stocks are always at risk from expectations of higher discount
rates, let alone loss of earnings momentum. Visibility of growth becomes critical.

Weighted averages can be Representative Relatives

distorted Ensure that relatives are calculated on an appropriate basis. It is well understood that
simple averages can be misleading, but it is perhaps not so widely appreciated that
even weighted averages can introduce significant distortion. This typically arises
when a small stock has earnings close to zero, and hence a very high P/E. Even
though its weight is tiny, its contribution to the average is exaggerated by its
extremely unrepresentative P/E.

The only truly representative process is aggregation. This involves multiplying

both the price in the numerator and the EPS in the denominator by the number of
shares for each stock, then adding these across the group to obtain the total
market capitalization and total earnings, which then give the aggregate P/E.

This approach avoids the distortion caused by both near-zero and negative EPS.
It is equivalent to consolidating all the stocks in the peer group as if they had
merged, and is applicable to all similar multiples.

31
 Globa l
0 6 Jan ua ry 2 009

Table 9 illustrates this point:

Table 9: P/E Averages – How Distortion Can Creep in
Risk of Flattery
Stock Price Shares EPS P/E Mkt Cap Weight (%) Net Income
The weighted average P/E for these four
A 52.0 96.1 1.40 37.1 4,997 35.6 134.5
stocks is 42.0, but the aggregate P/E is 38.8.
B 48.0 100.2 1.23 39.1 4,810 34.3 123.0
Thus, using the weighted average can flatter C 46.0 86.5 1.20 38.3 3,979 28.3 103.8
stocks trading at a discount. (Stock A is D 5.0 51.0 0.02 250.0 255 1.8 1.0
trading at more of a discount to the weighted Total 14,041 100.0 362.3
average number.) Source: Merrill Lynch Imagination

The weighted average P/E for the four stocks above is 42.0, but the aggregate
P/E is 38.8, which is about 8% lower. This has a dangerously flattering effect on
stocks trading at a discount. For example, Stock A is trading at a 12% discount to
the weighted average P/E, but this drops to only 4% when compared to the
aggregate multiple.

Use the median as an alternative If this process is impractical, perhaps because of the number of stocks involved,
we would strongly suggest using the median P/E as the most representative
benchmark. In the example above, the median is 38.7, which is reasonably close
to the aggregate.
Soup-to-Nuts Example
Table 10 exhibits the key inputs, Table 11-14 outline a step-by-step worked
example to evaluate EPS and P/E on both reported and adjusted bases.
Company C has been stuck in time, but the incredible stability of its trading and
financial performance makes it ideal to illustrate the various factors that can impact
EPS, and how we can fully reflect all of these in our valuation.

Although its operations have been Table 10: Summary Financial Data
remarkably steady over the years, Company C, Year to December 2000A 2001A 2002A 2003A 2004E
there have been some changes under EBIT before goodwill amortization 1,210 1,210 1,210 1,210 1,513
the surface. - Goodwill amortization -200 -200 -200
= EBIT before exceptional items 1,010 1,010 1,010 1,210 1,513
In 2001, Company C was forced to +/- Exceptional item (pre-tax) -200
make a one-off $200 mn reduction in = EBIT 1,010 810 1,010 1,210 1,513
the value of its inventory. +/- Financial income (expense) -120 -120 -120 -120 -120
= Pre-tax profit 890 690 890 1,090 1,393
In 2002, it sold off a complete
- Tax expense -327 -267 -327 -327 -418
subsidiary (that had failed to generate => Tax rate 37% 39% 37% 30% 30%
any profit). = Profit after taxation 563 423 563 763 975
In 2003, it stopped amortizing goodwill. +/- Minority interest in profit 22 22 22 22 22
= Ordinary net income 585 445 585 785 997
In 2004, it announced a June capital +/- Extraordinary items -130
increase to acquire another business, = Net income 585 445 455 785 997
which (we hope) will increase profits. - Preference dividends -20 -20 -20 -20 -20
= Net income to common 565 425 435 765 977
Other disclosures
Marginal tax rate is 30%
Shares in issue at year-end 1,000 1,000 1,000 1,000 1,500
Potential shares from options 100 100 100 100 100
Potential shares from 5% convertible bond 120 120 120 120 120
Source: Merrill Lynch Imagination

Year-end basic shares need two adjustments to reflect the impact of the mid-year
discounted capital increase. First, for the time-weighted average, and second, for
the effect of the discount. In essence, shares issued for free are assumed to have
existed for all time (think of a split), otherwise EPS growth would be understated.
The proportion of new shares thus covered depends on the issue terms, and the
ex-rights share price.

32
 Globa l
0 6 Jan ua ry 2 009

EOP shares are first averaged on a Table 11: Shares in Issue

time-weighted basis (in this case at 2000A 2001A 2002A 2003A 2004E
50% since the issue was at the end of End-of-period shares in issue 1,000 1,000 1,000 1,000 1,500
June). Time - weighted average shares in issue 1,000 1,000 1,000 1,000 1,250
/ historical shares adjustment factor (see below) 0.917 0.917 0.917 0.917 1.000
We then adjust for cumulative historical = Basic shares 1,091 1,091 1,091 1,091 1,250
factors, arising from discounted issues + potential shares (options + converts) 220 220 220 220 220
(of which a stock split is the ultimate = Diluted shares 1,311 1,311 1,311 1,311 1,470
example).
Capital Increase Discount Calculation
Diluted shares are obtained by adding
on the potential shares that would arise 1 new for 2 old shares issued 30th June 2004 @ 6, cf: ex rights price of 8
2 old @ 8 + 1 new @ 6 22
from exercise of options and
Divided by 3 implies: theoretical ex-rights price 7.3
conversion of the bond.
Implies: Discount to actual (1-733/800) 8.3%
(Note: these would normally be Implies: Historical shares adjustment factor (733/800) 0.917
historically adjusted only for splits.) Source: Merrill Lynch Imagination

Reported (GAAP) EPS are based on ordinary net income to common shares, and
the calculation is straightforward. In this (illustrative) example, everything stays
the same except for:

„ 2001: $200 mn charge to EBIT arising from nonrecurring, but ordinary event
(inventory write-down). The net impact of $130 mn passes straight through to
EPS.
„ 2002: $130 mn post tax (equivalent to same $200 mn pre tax) charge to net
income. As this arose from an event recognized as extraordinary under
GAAP (loss on disposal of a discrete business unit), the effect is reversed at
the ordinary net income level.
„ 2003: Groundhog ceases amortizing goodwill. Effect fully recognized in
reported EPS.
„ 2004: mid-year capital increase lifts shares in issue; historical years also
impacted.
The starting point for reported, or GAAP, Table 12: Reported EPS
earnings is net income from ordinary, 2000A 2001A 2002A 2003A 2004E
continuing operations. Preference Net income 585 445 455 785 997
dividends are then subtracted to give -/+ Extraordinary items (reversal) 0 0 130 0 0
“ordinary net income to common shares” = Ordinary net income 585 445 585 785 997
– the numerator. - Preference dividends -20 -20 -20 -20 -20
For diluted shares, the numerator is = Ordinary net income to common 565 425 565 765 977
increased to reflect the economic benefit / Basic shares 1,091 1,091 1,091 1,091 1,250
of the new shares – in this case, the = Basic earnings per share 0.52 0.39 0.52 0.70 0.78
interest saving on the convertible bond.
Ordinary net income to common (above) 565 425 565 765 977
The denominator is basic or diluted + Interest saving on converts 20 20 20 20 20
shares, respectively. - Nominal tax on interest -7 -8 -7 -6 -6
= Ordinary net income to diluted 578 437 578 779 991
/ Diluted shares 1,311 1,311 1,311 1,311 1,470
= Diluted earnings per share 0.44 0.33 0.44 0.59 0.67
Source: Merrill Lynch Imagination

The use of adjusted EPS must be limited to the elimination of distortions arising
either from:

„ Non-recurring, or exceptional elements that do not qualify as extraordinary

under GAAP, but should be reflected in valuation. Even though they
represent a fair deduction from (or, occasionally, an increase in) ordinary

33
 Globa l
0 6 Jan ua ry 2 009

earnings, their unusual nature distorts the trend, and they are reversed (on a
net basis) to give a clearer picture; or

„ Inconsistent local GAAPs. For example, whether goodwill is amortized.

Table 14 shows how the distorting effects of these events are reversed out to give
a consistent picture of the company’s prospects. Note that the impact of taxation
must be fully reflected. In particular, it is critical to ensure that the pre-tax
exceptionals are subjected to nominal tax before adjusting.

One further practical caveat: make sure that the effects of extraordinary events
are only counted once. With this in mind, it is probably safer to start from net
income than ordinary net income, especially if there are both exceptional and
extraordinary elements involved.

Note that the impacts of all three events: extraordinary, exceptional, and goodwill
are ironed out in the calculation of adjusted EPS, which are unchanged over the
relevant period.

Table 13: Adjusted EPS

Analysts’ adjustments involve reversal
of distorting effects in order to identify 2000A 2001A 2002A 2003A 2004E
the trend in EPS, rather than their strict Net income 585 445 455 785 997
value. - Preference dividends -20 -20 -20 -20 -20
= Ordinary net income 565 425 435 765 977
-/+ Extraordinary items (reversal) 0 0 130 0 0
Ordinary net income to common 565 425 565 765 977
+ Adjustments to reverse unusual factors
Exceptional items (reversal) 0 200 0 0 0
- Tax impact 0 -60 0 0 0
+ Adjustments to reverse inconsistencies
Goodwill amortization (reversal) 200 200 200 0 0
= Adjusted net income to common 765 765 765 765 977
/ Basic shares 1,091 1,091 1,091 1,091 1,250
= Basic earnings per share 0.70 0.70 0.70 0.70 0.78

Adjusted net income to common (above) 765 765 765 765 977
+ Interest saving on converts 20 20 20 20 20
- Nominal tax on interest -7 -8 -7 -6 -6
= Ordinary net income to diluted 778 777 778 779 991
/ Diluted shares 1,311 1,311 1,311 1,311 1,470
= Diluted Earnings per share 0.59 0.59 0.59 0.59 0.67
Source: Merrill Lynch Imagination

And finally, the P/E is obtained by dividing the current share price by EPS. Having
It is worth calculating all four definitions of calculated all four permutations of reported and adjusted, basic and diluted EPS,
EPS to see if material differences arise.
we decide that the most useful measure is the adjusted diluted, and so base our
For valuation purposes, the most useful
measure is adjusted diluted, but where valuation on the highlighted measure.
adjustments are significant, the reported
P/E should be referenced as well. Table 14: P/E Ratio at the Current Price of $9.00
2000A 2001A 2002A 2003A 2004E
Reported basic 17.4 23.1 17.4 12.8 11.5
Reported diluted 20.4 27.0 20.4 15.1 13.4
Adjusted basic 12.8 12.8 12.8 12.8 11.5
Adjusted diluted 15.2 15.2 15.2 15.1 13.4
Source: Merrill Lynch Imagination

34
 Globa l
0 6 Jan ua ry 2 009

Price/Book Value
Concept
The valuation multiple the market applies to the company’s equity.

Definition
Price/Book Ratio The current share price divided by Book Value per Share (common shareholders’
equity per share.)
Current Share Price
_____________________
Common shareholders’ equity is issued and fully paid capital, plus other paid-in
Common Shareholders’ Equity/ capital/share premium account/reserves plus retained earnings. This represents
Current Number of Shares the base of equity attributable to ordinary/common shares; minority interests in
equity are not included, neither is preferred stock.

Note that the number of shares in this ratio is the current actual number; it is not
diluted, or averaged.

Interpretation
With the ROE, a check on the P/E Essentially, a high Price/Book Ratio implies high, expected rates of return on
Ratio equity and vice versa.

The Price/Book Ratio is essentially a complement to the Price/Earnings Ratio. It is

equal to the product of the P/E and the Return on Equity, and as such acts as a
confirming indicator, given the P/E and expected ROE.

Watch Out for…

Inconsistencies among the P/E, Price/Book, and ROE.

Remember that book value is a historical cost number that can differ significantly
from market value, and may not be comparable across firms. This is especially
important in analyzing firms that have assets whose market values can be fairly
readily estimated (such as real estate companies) but according to GAAP are
recorded at historical cost.

Dividend Yield
Concept
The tangible rate of return to shareholders from company cash. While the Free
Cash Flow Yield (see page 33) represents the potential cash return to
shareholders, the Dividend Yield represents actual cash payments.

Dividend Yield Definition

The annualized declared cash ordinary dividend, expressed as a percentage of
Annualized Cash Dividend per Share
the share price.
_____________________
Share Price There are two points of detail here:

How are dividends “annualized?”

According to regional convention:

Regional variations outlined In Continental Europe, dividends are generally paid once a year in arrears; the
dividend is declared and paid in the year following that in which it was earned. We
annualize in terms of the year in which the dividend was earned, rather than paid.

In the US, UK, and occasionally Europe, interim dividends are often paid on
account, with the final dividend paid in arrears. In the UK, the practice is to combine
the (single) interim and final for the year in which they were earned.

35
 Globa l
0 6 Jan ua ry 2 009

In the US, where dividends are usually declared and paid quarterly, the basis of
annualization is rather different. Investors generally calculate the yield based on the
“current indicated annual rate,” which is the latest declared quarterly dividend,
multiplied by four. Depending on where we are in the calendar, this is not always the
same as the cumulative dividend for the fiscal year. Although this implies the risk of
inconsistency when benchmarking against European practice, the practical impact is
rarely significant.

„ What is a declared dividend?

Avoid net and gross That declared by the company, before any tax considerations.

Dividends have long been described as net or gross according to their tax status.
In strictly local terms, these expressions are explicit, but even in a regional
context (let alone global), they can become highly ambiguous. The reason is that
there are two separate tax issues involved:

„ Credits sometimes added to declared dividends in respect of the fact that

they have been paid out of the issuer’s post-tax profits, and/or
„ Withholding tax debited in respect of the imputed fiscal liabilities of the
recipient.
The expression net usually refers to the declared dividend, before tax credits, but
can also mean the declared dividend after deduction of withholding tax, with
equivalent ambiguity attached to gross: after-tax credit or before-tax deduction?
Ideally, the dividend used in valuation should be the amount effectively received
by the shareholder, but this is impractical because of the number of permutations
and combinations between the fiscal status of the issuer and holder. We think the
most rational approach is to base the yield on the dividend explicitly declared by
the issuer, ignoring any associated tax impact, which will depend entirely on the
fiscal status of the particular holder.
Where the declared dividend attracts a tax credit, this will sometimes lead to an
understatement of the effective yield; for example, French recipients of French
dividends are entitled to a significant tax credit. But this cannot be generalized: in
this example, holders outside France cannot usually realize this benefit in full.
Equally, it is also possible that the dividend will be declared before mandatory
withholding tax (common in Scandinavia), in which case the declared yield will
overstate the effective rate of return.

Interpretation
Components of total return Investors expect a total return, which is the combination of Dividend Yield and
Capital Appreciation. The relative attraction of these components will be primarily
determined by the holder’s attitude to risk; tangible immediate income versus
potential future gains, overlaid with fiscal issues, which, independent of the
probabilities involved, affect the value of the expected income and appreciation.

It is sometimes argued – generally in bull markets – that dividends are

unnecessary, in that income requirements can also be met by realizing capital
gains. Undeniably true, provided these are not ephemeral; capital gains can be
reversed, but a dividend can always be spent. Similarly, it is fallacious to ignore the
value of potential growth. Where a company’s expected returns justify it,
shareholders will benefit more from reinvestment than distribution.

Theoretical arguments aside, the Dividend Yield is a critical determinant of value

when comparing stocks in mature industries. It can also point to possible recovery
situations; when high-yielders maintain their dividends, the implied risk premium
can unwind rapidly.

36
 Globa l
0 6 Jan ua ry 2 009

Dividend discount models The Gordon Model

The dividend can also be used to derive an absolute valuation for stocks by
dividing the current dividend by the difference between the company’s WACC and
its expected growth. This simple model works fairly well at low growth rates
(obviously, they must be below the WACC), but does suffer from two drawbacks.
The Gordon approach is excessively sensitive to changes in the assumed growth
rate, a typical issue when the determining factor is in the denominator. In addition,
it tends to undervalue future growth, working best for high-yielding stocks. For
these reasons, its applicability is limited to value stories.
It is possible to extend the Gordon approach to more general stock situations using
two-, or even three-stage models. Although these exercises can be stimulating, they
suffer from the problem that the numbers of assumptions, and thus the degrees of
freedom, rise rapidly, increasing the uncertainty of the conclusion.
Internal Rate of Return
IRR: a useful benchmark It can be shown that the expected total return from holding a dividend-paying
security is equal to the yield plus the expected growth. This simple measure
makes an excellent benchmark for comparing stocks in many mature sectors, and
can easily be calculated from our standard measures.

Watch Out for…

One-time special payments Specials
Simply screening a range of stocks for dividend yield can fail to spot when a
do occur
dividend represents a one-time payment. Such “specials” can often be paid after
a major disposal, or after an exceptionally good year (at the peak of a cycle, for
example). Sometimes boards just feel generous; many German companies pay
an “anniversary” dividend to commemorate corporate milestones.
Security
The conventional measure of dividend security is EPS/DPS. This cover ratio
completely ignores the fact that the dividend has to be paid in cash; we feel that
the best indicator of cover is the ratio of free cash flow to the cost of dividend.
False Security
Be cautious of dividend yields that are inflated by significant reductions in share
price. It may well be a signal that the market believes that the most recently paid
dividend is unsustainable and ripe for a cut or outright elimination. Remember that
the shareholder buying the stock today is entitled to the next dividend, not the last
one already paid.

Free Cash Flow Yield

Free Cash Flow Yield Concept
The potential rate of cash return that could be obtained; an upper bound to the
Free Cash Flow
dividend yield.
_____________________
Share Price x Current Shares Definition
Free cash flow divided by market capitalization.

Interpretation
We have traditionally used cash flow-based metrics in the form of a multiple, but
this approach has significant limitations:

The generic expression price/cash flow is completely ambiguous, and requires

qualification; what exactly is cash flow in this context, and what should it be?

The generally accepted definition is net income, plus total depreciation and
amortization, plus deferred taxation. The value of this multiple lies in its simplicity,
being easy to calculate and consistently applicable across a wide range of sectors.

37
 Globa l
0 6 Jan ua ry 2 009

Distinguish between cash However, its simplicity is its key limitation, in that it is essentially incomplete: it
generation and availability gives a useful signal on a business’s ability to generate cash, but says nothing at
all about the claims on that cash flow that must be met before shareholders can
benefit. It is equivalent to an individual’s confusing his salary with his effective
spending power. To carry the analogy further, while credit may well be granted
on the former basis, it must be repaid from the latter.

On the other hand, price/free cash flow is a more rigorous valuation measure,
despite being difficult to forecast, because free cash flow, being struck after all
non-discretionary expenses, indicates the value that could either be taken out of
the business by shareholders, used to fund expansion, or used to reduce debt.
We focus on free cash flow because it is the most relevant cash-flow metric in the
context of valuation.

Alternatively, if the multiple is inverted and expressed as a percentage, it

represents the yield in terms of the cash that could be withdrawn from the
business. We feel that this is the optimum approach because it represents a
valuation floor in that where the Free Cash Flow Yield is above an investor’s cost
of capital, the total economic return will be positive, regardless of the share price.

Finding a floor…
In other words, this indicates the point of equilibrium when it is worth taking the
enterprise private.

…and a ceiling The free cash flow yield also represents a ceiling to the dividend yield; where
there is substantial room between these, there could be a story.

Watch Out for...

The option value of positive Free Cash Flow may be illusionary if debt reduction is
an overriding priority.

The intrinsic difficulty in forecasting Free Cash Flow (please see page 15).

Enterprise Value/EBITDA
Concept
A “structure-neutral” metric This is a structure-neutral metric; it offers a broad indication of the cash-
generation potential of an enterprise as if it were debt free.

EBITDA is the cash generation available to meet first interest payments, then
taxation.

Enterprise Value is the market value of the company’s capital employed at this
time and in its current structure. In addition to equity and net debt components,
pension liabilities, deferred taxes, capitalized lease, and other post-retirement
benefits are explicitly included in other non-current liabilities. However, this item is
not limited to these factors, as new and inventive forms of funding are always
emerging and need to be captured.

Definition
Enterprise Value/EBITDA
Enterprise Value divided by EBITDA, where EBITDA is positive.
Enterprise Value = Market Capitalization
+ Net debt + Minority Equity EBITDA is EBIT plus depreciation and amortization.
+ Other Non-Current Liabilities Where nonconsolidated earnings (associates, joint ventures, etc.) are significant,
_____________________
the nominal depreciation and amortization of these contributions should be
EBITDA = EBIT + Depreciation & estimated, and included with consolidated depreciation and amortization.
Amortization
Enterprise Value is the sum of: market capitalization, net debt, minority interest in
equity, and other non-current liabilities.

38
 Globa l
0 6 Jan ua ry 2 009

Market capitalization covers all classes of shares. Net debt should be the analyst’s
estimate of the level at this point in time, rather than the historical actual or next
forecast level. Moreover, the debt should be at market value where this is
determinable, especially where the difference is material, as with some convertibles.

We feel that the minority interest in equity should be taken at book value, rather
than scaled up by the Price/Book Ratio, because that measure takes full account
of the fact that the minorities exist. The most important elements of other non-
current liabilities are usually unfunded pension or post-retirement benefits,
deferred taxes, and capitalized leases, but others could include environmental
decommissioning provisions and all long-term claims on the business.

Consolidation Issues
Nonconsolidated entities: A limitation of this measure is that the contribution of nonconsolidated entities’
depreciation and amortization is not explicitly available, and often ignored,
Adjust EBITDA, not EV
understating EBITDA. To compensate for this, we sometimes see EV reduced by
the value of these non-contributing entities. We do not feel that this is
appropriate, because, although this approach appears more rigorous, that
valuation is only based on part of the enterprise. If, for example, the associates
were dilutive (which is quite likely) this would not be picked up. Thus, there is a
real danger of swinging from understatement to overstatement.

We think that estimating the missing D&A contribution of nonconsolidated items,

and including this in EBITDA, provides a better assessment of this metric that fully
covers the enterprise.

Interpretation
When use of cash is critical The classic application of this metric is in mature sectors in which cash flow is
highly visible – Building Materials, for example. The key determinant of quality is
not so much how cash flow is created, but how it is deployed. Good companies
know when to take on debt to acquire strategic targets, and how much to pay.
EV/EBITDA, being structure neutral, is a useful starting point to evaluate
acquisition targets.

Another argument runs that for any given present level of Equity Value, positive
cash flow will reduce debt over time, implying that the equity value should
increase to compensate. This argument is fine, provided the cash flow actually
materializes.

Watch Out for...

Net debt at book value, when market value is significantly higher (convertibles).

Capital liabilities (such as unfunded pension exposure) omitted if they are not
recognized on the balance sheet.

Enterprise Value/Sales
Concept
Enterprise Value/Sales
This is a measure of volume leverage: how much is the enterprise worth in
Enterprise Value = Market Capitalization relation to its turnover?
+ Net Debt + Minority Equity + Other
Non-Current Liabilities Definition
_____________________
Enterprise Value divided by Sales.
Net Sales
Interpretation
This measure is fairly narrow and should never be relied on in isolation, but it is
useful in two particular situations –

39
 Globa l
0 6 Jan ua ry 2 009

Benchmark for deals… In the context of acquisitions:

The purchaser generally starts from the view that he or she will be able to run the
business better than the vendor (raise margins). So current earnings multiples are
less important than sales, because revenue better represents the level of
opportunity.

…and an indicator of volume When valuation is highly dependent on macro factors:

sensitivity Where valuations are driven by external factors, such as the price of oil or the level
of consumer expenditure, it is useful to determine the exposure of the business to
that macro factor. Sales is the best general indicator of volume sensitivity.

Watch Out for...

Understatement of enterprise value (as before).

40
 Globa l
0 6 Jan ua ry 2 009

5. The Good DCF Guide

The valuation metrics covered above are generally used in a relative sense; the
most common form of absolute valuation is Discounted Cash Flow, based around
the Capital Asset Pricing Model (CAPM). Many of the concepts involved have
been referred to already; this section seeks to tie those threads together and
outline our best-practice approach.

The basic concept is disarmingly simple: the absolute value of any capital asset is
the present value of the expected cash flows, discounted at the cost of that
capital. In practice, however, several key issues underlie that statement.

What Is the Asset – Enterprise or Equity?

There are two possible approaches. We can either:

„ value the equity directly, by discounting the free cash flow (which is fully
attributable to equity, because interest has already been deducted); or
„ value the enterprise as a whole, by discounting free cash flow before interest
(adjusting for the tax shield), and then deducting current debt to value the equity.
Direct equity valuation looks The first approach looks to be the simpler option, as it enables us to ignore the
simpler… cost of debt (which is difficult to calculate rigorously) because this is fully reflected
in the post-interest cash flow. It also feels closer to real-world situations where
DCF is widely used to value specific projects.

…but ignores the value of Although the equity can be viewed as a stream of projects, it differs crucially in
leverage that it has a distinct capital structure independent of its constituent projects.
Essentially, we cannot ignore the significant impact of leverage on equity
valuation. This is positive, provided that the return on capital employed is greater
than the cost of debt. This being generally the case, normal practice should be to
discount the enterprise rather than the equity.

A similar issue arises with minority equity. Should we treat the minorities the
same way as debt, deducting the current balance sheet figure from the Enterprise
Value to leave the attributable equity valuation? Or should we factor in the
minority by deducting the P&L charge from the cash flow? Neither approach is
completely rigorous; for the former we strictly need to deduct the market value of
the minority, rather than the book. In the latter, we tacitly assume that cash flow
equals profit. Fortunately, the impact of minorities is rarely significant, and the
inconsistency can usually – but not always – be ignored.

We suggest that the best approach is to deduct the minority profit from the cash
flows, rather than the minority equity from the EV. The reason is that, whereas
debt remains essentially a fixed element of the structure, the minority moves in
line with the cash flow. To ignore the future growth in the minority interest is to
overstate the attributable equity value.

41
 Globa l
0 6 Jan ua ry 2 009

What Are Expected Cash Flows?

A definition for DCF We suggest the following definition, which is a compromise between pragmatism
and rigor.

EBIT
Minus: nominal tax
Plus: depreciation and amortization
Plus: net change in provisions
Minus: net change in working capital
Minus: capital expenditure
Minus: minority interest in profit

Reverse the tax shield on interest The tax charge is applied to EBIT, rather than pre-tax profit, to reflect the tax shield
of the interest charge. Strictly speaking, it should be applied to EBIT before
deduction of any goodwill amortization (which is generally not tax deductible) and
this should be done where goodwill is material. The nominal tax rate should be
based on the analyst’s assessment of the long-term trend rather than any
anomalous immediate rates. Dealing with the change in provisions as a whole
enables us to fully reflect the timing differences between the accrual and cash
accounts – including deferred tax.

Working capital (inventory plus receivables less payables) is extremely difficult to

forecast – to the extent that some prefer to omit this item altogether. However,
because working capital usually rises over time, this would effectively overstate
value. We would suggest that recent trends in working capital be extrapolated in
line with expected sales growth.

As discussed on Page 15 under Free Cash Flow, capital expenditure should be

total, not maintenance.

Financial assets and equity DCF is most effective for modeling operating, as opposed to holding companies; but
accounting in practice, many enterprises have more complex structures, and these must be
factored in where appropriate. In particular, we must take account of the value of
any financial assets that have not contributed to the free cash flow as calculated,
and where significant, adjust for elements of EBIT that are not fully received in cash.

Dealing with financial assets is straightforward; the value (at market if known, if
not at book) of purely financial assets should be added to the implied EV in
calculating the equity value. Note this does not apply to an operating asset that is
currently not contributing!

Dealing with equity-accounted entities (associate companies and joint ventures) is

more complex, for two reasons. Although their operating contributions are usually
included in EBIT, this is not always the case (they can be included at the net
income level, for instance), and it may be necessary to estimate the implied EBIT
contribution. In addition, the EBIT contribution may be at significant variance with
the effective cash generation. This can cut two ways: on one hand, the share of
the associate’s depreciation is not included, but equally, the parent’s actual
influence over the associate’s cash flow may well be limited to the dividend
received, and it may be important to adjust accordingly.

42
 Globa l
0 6 Jan ua ry 2 009

Over What Period Should We Discount?

A distinguished professor was acting as a consultant to a chemical company,
assessing its DCF-based project appraisals. “Why,” she asked the technical
director, “are all your DCFs based over different periods?”

“Well,” the director replied, “sometimes we have to go farther out to get a positive
number.”

The problem: balancing the Our task is rather different. DCF models of projects can usually be calculated over
residual their entire expected lives, but – conceptually at least – equities have no finite
lifespan. This means dealing with a series of explicit periodic forecasts together
with a residual, representing the ongoing value. But how many periods should we
forecast? Most models are built on the basis of a detailed assessment of the
current and upcoming situation, then overlaid with our view of key trends.
Inevitably, the further ahead we extrapolate, the greater the uncertainty; but
equally, if the forecasting horizon is too short, our process is little more than
valuing the residual. How can we strike a balance?

The law of diminishing returns dictates that over time, returns on capital inevitably
revert to the cost of capital. Successful enterprises will make higher returns, but
their “comparative advantage period” (CAP) is finite, and characterized by
Porter’s Five Forces model that defines industry structure and profitability.

The Porter Principles and the „ Strength of current competition

comparative advantage period „ Threats from new entrants
„ Threats from substitutes
„ Bargaining power of suppliers
„ Bargaining power of customers
Although these cannot be objectively determined, it is useful to use an intuitive
assessment as the basis for our DCF: say five years for a technology business,
10-15 for most mature businesses, and perhaps 20 for a utility.

Tapered extrapolation Our models will rarely cover more than five explicit forecast years, but it is not
difficult to extrapolate the trend in free cash flow (just the bottom line, rather than
its input assumptions) to reflect the CAP. As a refinement, we would suggest
tapering the trend in growth at the end of the explicit modeling period down to 4%
(expected nominal GDP growth) over our assumed CAP.

For example, if FCF had been growing at 14% p.a. over the five years of our
detailed forecast, and the CAP was 10 years (five to extrapolate) in our DCF, we
would assume FCF growth of 12% in year six, 10% in year seven, etc. – bringing
growth down to nominal GDP by the end of the CAP.

Watch Out for…

In discounting annual cash flows at their year-ends, we must avoid the
discontinuity that can arise as our valuation date approaches the year-end month.
The best way to deal with this in Excel is to record the explicit year-end of each
period-end, and discount on the basis of partial years:

PV (Cash flown) = Cash flown/(1 + WACC) ^ ((year endn – valuation date)/365)

43
 Globa l
0 6 Jan ua ry 2 009

How Should We Deal with the Residual Cash Flows?

Coming to terms with infinity Growth in Perpetuity
The residual is commonly calculated by assuming that after the last explicit period,
cash flow grows at an assumed constant rate, then evaluating the sum of the
resulting geometric series as: cash flow/(discount rate – growth rate). The present
value of this is then added to that of the explicit annual cash flows to estimate the
enterprise value.
Although this is mathematically correct (provided that the growth rate is below the
discount rate) as a modeling technique, it is rarely satisfactory, and often yields
an implied value far higher than the market price. The problem is that in any
model where the input variable (the growth rate in this situation) is in the
denominator, the output value can be extremely sensitive to changes in the input.
Moreover, for small numbers, the sum of the geometric series converges very
slowly, so the residual can often dominate the valuation. Again, although this may
be mathematically correct, it is intuitively problematic; should an investor buy an
asset on the basis of a valuation primarily dependent on cash flows that are
receivable several decades ahead?
For example, a hypothetical initial cash flow of 100 is forecast to grow by 10%
each year for the 10 years of our explicit model. Discounting this at a cost of
capital of 8% yields a present value (PV) of 1,208. The analyst then assumes a
conservative 5% terminal growth rate (only half of the trend rate) and adds the PV
of this residual to that of the annual cash flows to reach a value of 4,916. That
may or may not represent fair value, but the key point is that the residual
accounts for 75% of the total valuation. Looking at it another way, if we had not
used the geometric sum formula, but had continued to forecast year by year, it
would take 93 years to reach that figure!
Moreover, that estimate is extremely sensitive to the figure we take for terminal
growth, as Chart 4 illustrates. If, alternatively, we had taken a figure of 6% instead
of 5%, the estimated value rises to 6,770, an increase of 38%. If we had gone the
other way, and used 4%, the PV would have dropped to 3,989, a fall of 19%.
If growth in perpetuity models are used, the terminal rate should always be below
the expected growth in nominal GDP – say 4%.
Chart 4: Growth in Perpetuity Model, Sensitivity to Terminal Growth Rate

14,000

12,000

Sensitivity to Growth Rate:

If we had taken a figure of 6% instead of 5%, 10,000
the estimated value would have risen to
Present Value @ 8%

6,770, an increase of 38%. 8,000

If we had gone the other way, and used 4%, PV (periodic cashflows + residual)
the PV would have dropped to 3,989, a fall of 6,000
19%.
4,000

2,000 PV (residual)

0
0% 1% 2% 3% 4% 5% 6% 7%
Terminal growth rate

Source: Merrill Lynch

44
 Globa l
0 6 Jan ua ry 2 009

Residual Yield
An alternative – and simpler – approach to handling the residual is to assume that
the equity is sold at market value (based on a standard benchmark) at the end of
its comparative advantage period.

As well as avoiding the necessity to make an assumption about perpetual growth,

this methodology sits comfortably with investment theory. The essential reason
for buying an equity is growth, and when that growth has run its course, the equity
will then essentially trade as a (undated) fixed income security, and should be
valued on its hypothetical yield.

We would suggest taking the current market average free cash flow yield, or a
suitable proxy: twice the average dividend yield, perhaps. The residual is then
evaluated by grossing up the final cash flow by the yield, and discounting to
obtain the PV.

A more conservative approach This approach generally gives a much lower value for the residual (because we
are dividing the final cash flow by a larger number). In the example above, if we
use a free cash flow yield of 6%, the DCF values the residual at 1,854, which
represents 61% of the total, as shown in Table 15.

Table 15: Residual Methodologies Compared

Growth in Residual Yield
Present Value (WACC = 8%) Perpetuity (5%) (6%)
Forecast cash flows 1,208 1,208
Residual 3,708 1,854
Total 4,916 3,062
Residual/Total 75% 61%
Source: Merrill Lynch

Reduced input sensitivity Although we still have to make an assumption about the yield, this approach is far
less sensitive than the growth-in-perpetuity model. If we had taken a 4% yield
instead of 5%, the valuation would have risen by 16%, less than half the impact of
the equivalent input sensitivity under growth-in-perpetuity.

Chart 5: Residual Yield Model – Sensitivity to Yield

6,000

5,000

4,000
PV (periodic cashflows + residual)

3,000

2,000

1,000
PV (residual)

0
3% 4% 5% 6% 7% 8% 9% 10%
Market Average FCF Yield

Source: Merrill Lynch

45
 Globa l
0 6 Jan ua ry 2 009

What Is the Cost of Capital?

We discussed the issues involved in calculating the Cost of Capital in detail in
section 2, page 10. Some additional issues, however, need to be addressed in
applying this concept to our DCF. These arise, in the main, because of the
extended time horizon involved; it will usually be appropriate to use long-term
estimates for many of the inputs, rather than the immediate values.

How Should Leverage Be Treated?

Use long-term estimates where Actual debt levels will almost certainly vary – possibly considerably – over the
current levels could change DCF term, and this can impact the WACC significantly. This arises both through
changing weights of equity and debt, and also the effect on the stock Beta.

Where debt levels have been, or are expected to be volatile, it is worthwhile

basing the weights in the WACC on a representative or long-term target level of
gearing, rather than the current actual. Note, however, that it is still the actual
level to be deducted from the estimated EV when the discounting is complete!

Risk-Free Rates
The yield on Government bonds (aggregated across the G7 countries) has been
falling steadily since 1990: the current levels (4.1% in September 2004) may well
prove to be too low longer term.

Chart 6: G7 Aggregate Government Bond Yield (%)

When the latest figure differs from the trend
On average, G7 government bond yields
have been steadily declining for years. In our
very long-term DCF assumptions, are we
justified in assuming this trend could never
reverse?

Source: Merrill Lynch

The Equity/Risk Premium

The Implied ERP can be calculated as the difference between the Implied Return
on Equities (obtained from an aggregated dividend discount model) and the risk-
free return. Chart 7 shows that, although the ERP has spiked occasionally, it has
essentially remained within a single standard deviation of its average of about
3.5% since 1990. The current level (5.1%) appears likely to be high for long-term
forecasting.

Please note: updated information on this data is available from the Global
Valuation and Analytics Team upon request.

46
 Globa l
0 6 Jan ua ry 2 009

Chart 7: Implied Equity Risk Premium (%)

Source: Merrill Lynch

What Beta to Use?

It is well worth reiterating the basic principles: the Beta is the systematic risk
(volatility) of an asset within a portfolio. In this context, it is the relative variability
of a particular stock to the market as a whole. The equity risk premium reflects the
risk of holding equities as a class; multiplying the ERP by the Beta indicates the
risk of holding that specific stock. The Beta is the link between macro and micro in
equity valuation.

The problem lies in the uncertainty involved in measuring the Beta. This is
usually measured as the slope of the regression line between changes in the
stock and market values. This is only an estimate, however, and is subject to a
considerable margin of error. Are we really justified in projecting this into the
distant future?

Aggregate Betas can be better There is a strong argument, especially for less liquid stocks, new issues, and
situations where the capital structure has changed, to use the aggregate Beta for
the relevant industry/sector as the basis of a proxy for the stock’s own specific risk.

Table 16 lists aggregate Betas for various industries. The unlevered Beta is the
observed Beta of the stock adjusted for the impact of its leverage. Please note
that the unlevered is always less than or equal to the levered: removing the debt
reduces the risk.

Table 16: Aggregate Betas for Various Industries

5Y Avg Current Current
Sector Levered Levered Unlevered
Oil & Gas 0.63 0.67 0.57
Construction & Building Materials 0.60 0.60 0.46
Forestry & Paper 0.87 0.79 0.56
Steel & Other Metals 0.72 0.74 0.58
Aerospace & Defense 0.68 0.92 0.78
Diversified Industrials 0.89 0.64 0.46
Electronic & Electrical Equipment 1.09 1.17 0.85
Engineering & Machinery 0.81 0.86 0.65
Automobiles 0.91 0.95 0.68
Household Goods & Textiles 0.77 0.76 0.71

47
 Globa l
0 6 Jan ua ry 2 009

Table 16: Aggregate Betas for Various Industries

5Y Avg Current Current
Sector Levered Levered Unlevered
Beverages 0.41 0.39 0.32
Food Producers & Processors 0.34 0.39 0.32
Health 0.63 0.61 0.53
Personal Care & Household Products 0.49 0.39 0.31
Pharmaceuticals 0.73 0.77 0.72
Retailers, General 0.92 0.78 0.74
Leisure, Entertainment & Hotels 0.82 0.70 0.60
Support Services 0.84 0.98 0.88
Transport 0.59 0.63 0.48
Food & Drug Retailers 0.57 0.76 0.58
Telecom Services 1.23 1.20 0.81
Electricity 0.24 0.67 0.47
Banks 0.92 1.01 0.58
Insurance 0.98 1.28 1.10
Life Assurance 1.01 1.32 1.12
Investment Companies 0.63 0.42 1.09
Real Estate 0.53 0.42 0.28
Specialty & Other Finance 1.15 1.18 0.99
Information Technology Hardware 1.75 1.55 1.45
Software & Computer Services 1.48 1.29 1.29
Source: Merrill Lynch estimates, September 2004. NB: Calculations based on 104 weekly geometric observations against
the MSCI World Index in US Dollars.

Levered and unlevered Betas Strictly speaking, we cannot, however, simply substitute the industry Beta into our
calculation. The reason is that the measured, actual Beta is significantly
influenced by the weights of debt and equity, and if our particular stock’s structure
varies from the industry as a whole (which is quite likely), an adjustment is
necessary.

In order to use an industry/sector Beta, we must start with the unleveraged

industry Beta, and apply the following formula:

⎛ D⎞
β LeveragedStock = βUnleveragedSector * ⎜1 + (1 − t )* ⎟
⎝ E⎠
Where:

t = Tax rate
D = Debt value
E = Equity value

Summary
The weighted average cost of capital is:

WACC = Cost of debt * D/(D + E) + Cost of equity * E/(D + E)

Cost of debt = Effective interest rate * (1 – t)
Cost of equity = Risk-free rate + Equity risk premium * Beta

48
 Globa l
0 6 Jan ua ry 2 009

Appendix: The iQdatabase Global

Template
The tables in the appendix provide an outline of the standard industrial
accounting layout used by Merrill Lynch analysts to populate our iQdatabase.

Several industries also have industry-specific measures, both financial and

operational. We employ specialized layouts to meet the needs of the finance
sector: banks, insurance, financial services, and real estate investment trusts.

Explicit links between reported The key objective is to enable the analyst to include both reported (local GAAP)
and adjusted figures and adjusted (standardized) data in an explicit and consistent manner. We show
only the general measures used in the commercial/industrial sectors, our analysts
can, and do, add their own industry-specific measures in addition to these. Note
that the number of adjusted metrics is kept to a minimum; in the Income
Statement, we consider adjustments to EBIT and Net Income; in the Balance
Sheet, we provide adjusted fields for both Equity and Debt.

We also incorporate the costs are negative convention, so that profits are
calculated additively, and adjustments (reversals) subtracted. Although a little
counterintuitive, this approach does avoid possible ambiguities.

49
 Globa l
0 6 Jan ua ry 2 009

Table 17: Income Statement

Measure Definition
Sales revenue Trading revenue from consolidated operations, net of returns and discounts
Other revenue Non-trading revenue
Total revenue Sales + other revenue
Cost of goods sold Cost of goods sold
Gross profit Total revenue + COGS
Depreciation Total depreciation (excluding Amortization)
Amortization Amortization of all intangible assets (inc. goodwill & other intangibles)
Total depreciation & amortization Total depreciation + total amortization
o/w: depreciation PPE Depreciation of PPE
o/w: depreciation Other NCA Depreciation of PPE
o/w: amortization of intangibles ex. Goodwill Amortization intangibles other than goodwill
o/w: amortization Goodwill Amortization minus the above
Share of associates Dep & Amort (est) Explicit or estimated (adjustment to EBITDA)
EBIT (reported) EBIT as reported under local GAAP
Included contribution from associates etc The part of EBIT (reported) from non-consol entities (if any)
EBIT excluding associates’ contribution EBIT – inc contrib assocs (“consolidated EBIT”)
Interest expense Total interest expense (not “net” unless only that is given, if so, set IntInc=0)
Interest income Total interest income (not “net” unless only that is given, if so, set IntExp=0)
Other financial income/expense Financial items not covered by the above
Financial income/expense Sum of the above three
Profit/loss before tax EBIT (reported) + financial inc/exp total
Tax expense/benefit (total) Total tax (expense/benefit) for the year
o/w: deferred The part of Tax which is deferred
o/w: on exceptionals The part of Tax which relates to exceptionals
Post-tax contribution from associates etc. Where not included in EBIT
Post-tax goodwill amortization Where not included in EBIT
Profit/loss after tax PBT + total tax + post-tax assocs + post-tax goodwill
Minority interest in profit/loss Minority profits negative, loss profit. Give the net figure.
Net income before exceptionals PAT + Minority interest in profit/loss
Effect of discontinued operations As per US GAAP
Effect of accounting changes As per US GAAP
Effect of extraordinary items As per US GAAP
Other exceptional Equivalents under other GAAP
Exceptional items post-tax (total) Sum of the above four
Net income (reported) Net Income before exceptionals + exceptionals
Adjustments to EBIT
EBIT (reported) EBIT as reported under local GAAP (as per P&L above)
Add: associates EBIT inferred from post tax Post-tax assocs * (1 - standard tax rate) or actual if no standard
Add: amortization inferred from post-tax where goodwill is amortized at the post tax -level
Financial income treated as operational Financial income relating to operations
Financial expense treated as operational Financial expenses relating to operations
Add: total financials treated as operational Sum of the above two
Restructuring costs Reversal: adjustment for exceptional restructuring charges
Asset write-downs ditto: asset write-downs
Gain/loss on sale of assets ditto: gain/loss on asset sales
Other pre-tax exceptionals Other adjustments to EBIT
Less: total pre-tax exceptionals Total pre-tax exceptionals
EBIT adjusted EBIT reported + post-tax assocs, amort + op fins - pre-tax excepts
Adjustments to net income
Net income (reported) Net Income reported under local GAAP (as per P&L above)
Exceptionals pre-tax (est post-tax equiv) Reversal: est. post-tax equiv of pre-tax exceptionals (@std tax)
Exceptionals post-tax Reversal: pre-tax exceptionals
Goodwill amortization Reversal: goodwill amortization
Other profit adjustments Any other reversals of reported net income under local GAAP
Less: Total adjustments to net income Total post-tax exceptionals
Net income (adjusted) Net Income reported - total post-tax exceptionals
Source: Merrill Lynch

50
 Globa l
0 6 Jan ua ry 2 009

Table 18: Cash Flow Statement

Measure Definition
Net income Net income (per P&L)
Total depreciation & amortization (contra) Add back total depreciation & amortization
Deferred taxation charge (contra) Add back deferred tax
Cash earnings Sum of the above three
Minority interest (contra) Add back minority interest
Difference between assocs’ divs & sh of profits Because assocs cash contribution is only the dividend
Change in provisions etc. Includes equivalent other non-current liabilities
Change in working capital To capture cash effects
Other operating cash items Balances op cash flow to cash earnings + the above 4
Cash flow from operations As reported under local GAAP. Cash earnings plus the above 5
Capital expenditure Total capex, not “maintenance” (considered indeterminable)
Proceeds from sale of non-current assets As published
Acquisitions/disposals of investments As published
Other investing cash items As published
Cash flow from investing Sum of the above four
Shares issue/repurchase As published
Cost of dividends paid As published
Change in debt As published
Other sources/uses of cash Balances the above three to the total below
Cash flow from financing Balances chg cash (total cash flow) to cash flow op + cash flow inv
Total cash flow (change in cash & equivs) As published. Equates to sum of op, inv & fin cash flows
Free cash flow Operating cash flow plus total capex (which is defined negative!)
Change in net debt Change in debt + change in cash
Source: Merrill Lynch

Table 19: Supplementary Measures

Measure Definition
Restricted stock Restricted stock compensation
Stock options Stock options compensation
Total equity-based compensation Sum of restricted stock and stock options
Expense options at fair value? Indicates (yes/no) whether options are expensed
Proportional revenue Consolidated revenue plus pro-rata associates, etc.
Revenue from discont. ops As per US GAAP
Profit before tax - adjusted Profit before tax adjusted for exceptionals: EBIT adjusted + financial income/(exp)
Tax on non-recurring items Tax on exceptionals
Proportional EBITDA Consolidated EBITDA plus pro-rata assocs, etc.
Headcount Year-end full-time equivalent employees
Personnel costs Total wages, salaries, social security, pension & post-retirement benefits charged to EBIT
Capitalized interest As published
Standard tax rate % Analyst's assessment of the "normalized" or sustainable tax rate
Cash tax rate % Same as tax rate, but excluding deferred tax
Tax rate on ordinary activities % Tax rate excluding exceptionals. NM if negative
Cash taxes paid Cash taxes actually paid
Source: Merrill Lynch

Table 20: Balance Sheet

Measure Definition
Property, plant & equipment Property, plant & equipment (“Tangible FA”)
Goodwill Exclude other intangibles
Other Intangible assets Patents, trademarks, etc.
Investments Financial fixed assets
Other non-current assets Balances total below to the above four.
Non-current assets Book value after accumulated dep/amort
Inventory At book value (after B&D debts)
Trade receivables Accounts receivable (debtors) arising from sales
Other receivables & prepayments Other accounts receivable

51
 Globa l
0 6 Jan ua ry 2 009

Table 20: Balance Sheet

Measure Definition
Receivables (total) Sum of the above two
Cash & equivalents “Equivalents”= items promptly convertible into cash at certain value
Other current assets (balances) Balances total CA to inventory, cash & receivables
Current assets As published
Total assets CA plus NCA
Trade payables Accounts payable (creditors) arising from purchases for sales
Other payables & accruals Other accounts payable
Payables (total) Sum of the above two
Short-term debt All maturities under 12m, includes current portion of LT debt
Other current liabilities Balances total CL to payables & ST debt
Current liabilities As published
Long-term debt Interest-bearing liabilities maturing over 12m
Provision for deferred tax The balance of deferred tax payable, net of credits
Provision for pensions/PRB Accruals for pension and PRB liabilities as published
Other provisions All other accruals
Provisions Sum of the above three
Other non-current liabilities Balances total NCL to LT debt and provisions
Non-current liabilities Enter total as published
Total liabilities Current liabs + non-current liabs
Contributed capital As published
Reserves As published NB: this field can be negative!
Cumulative retained earnings As published NB: this field can be negative!
Ordinary shareholder equity Sum of the above three
Preferred stock As published
Minority equity As published NB: this field can be negative!
Total equity Total of ordinary, pref & minority equity
Adjustments to debt
Total debt As reported under local GAAP (as per BS above)
Off balance sheet borrowings Leases, securitized working capital, etc.
Other obligations Other obligations equivalent to debt for valuation purposes
Less: Total adjustments to reported debt Total post-tax exceptionals
Total debt (adjusted) Total debt + adjustments
Adjustments to equity
Ordinary shareholder equity As reported under local GAAP (as per BS above)
Unrealised capital gains After tax
Other Untaxed reserves, provisions with equity character, etc.
Less: Total adjustments to reported equity Total post-tax exceptionals
Ordinary shareholder equity (adjusted) Ordinary equity + adjustments
Source: Merrill Lynch

Copyright 2005 Merrill Lynch, Pierce, Fenner & Smith Incorporated (MLPF&S). All rights reserved. Any unauthorized use or disclosure is prohibited. This report has been
prepared and issued by MLPF&S and/or one of its affiliates and has been approved for publication in the United Kingdom by Merrill Lynch, Pierce, Fenner & Smith Limited,
which is authorized and regulated by the Financial Services Authority; has been considered and distributed in Australia by Merrill Lynch Equities (Australia) Limited (ABN 65
006 276 795), licensed under the Australian Corporations Act, AFSL No 235132; has been considered and distributed in Japan by Merrill Lynch Japan Securities Co, Ltd, a
registered securities dealer under the Securities and Exchange Law in Japan; is distributed in Hong Kong by Merrill Lynch (Asia Pacific) Ltd, which is regulated by the Hong
Kong SFC; and is distributed in Singapore by Merrill Lynch International Bank Ltd (Merchant Bank) and Merrill Lynch (Singapore) Pte Ltd (Company Registration No.
198602883D), which are regulated by the Monetary Authority of Singapore. The information herein was obtained from various sources; we do not guarantee its accuracy or
completeness.

52
iQanalyticsSM Product Suite
Featuring:
iQanalyticsSM – Entire suite of proprietary iQ valuation and metrics products, populated by Merrill
Lynch’s iQdatabaseSM.

iQdatabase – Globally integrated real-time research database sourced directly from Merrill Lynch
equity analysts’ earnings models. Includes forecasted as well as historical data for income statements, balance sheets and cash flow
statements for companies covered by Merrill Lynch.

iQmethodSM – Merrill Lynch’s unique approach to valuation, accounting and quality of earnings. Includes detailed descriptions of our
accounting calculations for 16 key standard measures that assess business performance, the quality of earnings and valuation metrics.

iQtoolkitSM – Electronic portal to a range of Microsoft Excel based applications, including the iQvirtual analyst, which leverage the
power of the iQdatabase and iQmethod.

iQvirtual analyst – Interactive Microsoft Excel based valuation applications for analyzing stocks on a cross-regional and cross-sector
basis.

iQworksSM – Software applications that provide Merrill Lynch’s institutional salespeople with direct
real-time access to the iQdatabase and the ability to prepare customized reports and analyses for institutional investor clients.

iQcustomSM – Customized data projects and reports tailor-made for clients by the Global Valuation & Analytics group. This service is
available to investor clients upon request to Institutional Sales representatives.

iQanalytics, iQmethod, iQdatabase, iQtoolkit, iQworks and iQcustom are service marks of Merrill Lynch & Co., Inc.

53
 Globa l
0 6 Jan ua ry 2 009

54
 Globa l
0 6 Jan ua ry 2 009

Important Disclosures
FUNDAMENTAL EQUITY OPINION KEY: Opinions include a Volatility Risk Rating, an Investment Rating and an Income Rating. VOLATILITY RISK
RATINGS, indicators of potential price fluctuation, are: A - Low, B - Medium and C - High. INVESTMENT RATINGS reflect the analyst’s assessment of a
stock’s: (i) absolute total return potential and (ii) attractiveness for investment relative to other stocks within its Coverage Cluster (defined below). There
are three investment ratings: 1 - Buy stocks are expected to have a total return of at least 10% and are the most attractive stocks in the coverage cluster;
2 - Neutral stocks are expected to remain flat or increase in value and are less attractive than Buy rated stocks and 3 - Underperform stocks are the least
attractive stocks in a coverage cluster. Analysts assign investment ratings considering, among other things, the 0-12 month total return expectation for a
stock and the firm’s guidelines for ratings dispersions (shown in the table below). The current price objective for a stock should be referenced to better
understand the total return expectation at any given time. The price objective reflects the analyst’s view of the potential price appreciation (depreciation).
Investment rating Total return expectation (within 12-month period of date of initial rating) Ratings dispersion guidelines for coverage cluster*
Buy ≥ 10% ≤ 70%
Neutral ≥ 0% ≤ 30%
Underperform N/A ≥ 20%
* Ratings dispersions may vary from time to time where Merrill Lynch Research believes it better reflects the investment prospects of stocks in a Coverage Cluster.
INCOME RATINGS, indicators of potential cash dividends, are: 7 - same/higher (dividend considered to be secure), 8 - same/lower (dividend not considered
to be secure) and 9 - pays no cash dividend. Coverage Cluster is comprised of stocks covered by a single analyst or two or more analysts sharing a common
industry, sector, region or other classification(s). A stock’s coverage cluster is included in the most recent Merrill Lynch Comment referencing the stock.

55
 Globa l
0 6 Jan ua ry 2 009

Other Important Disclosures

Information relating to Non-U.S. affiliates of Merrill Lynch, Pierce, Fenner & Smith Incorporated (MLPF&S):
MLPF&S distributes research reports of the following non-US affiliates in the US (short name: legal name): Merrill Lynch (France): Merrill Lynch Capital Markets
(France) SAS; Merrill Lynch (Frankfurt): Merrill Lynch International Bank Ltd, Frankfurt Branch; Merrill Lynch (South Africa): Merrill Lynch South Africa (Pty) Ltd;
Merrill Lynch (Milan): Merrill Lynch International Bank Limited; MLPF&S (UK): Merrill Lynch, Pierce, Fenner & Smith Limited; Merrill Lynch (Australia): Merrill Lynch
Equities (Australia) Limited; Merrill Lynch (Hong Kong): Merrill Lynch (Asia Pacific) Limited; Merrill Lynch (Singapore): Merrill Lynch (Singapore) Pte Ltd; Merrill
Lynch (Canada): Merrill Lynch Canada Inc; Merrill Lynch (Mexico): Merrill Lynch Mexico, SA de CV, Casa de Bolsa; Merrill Lynch (Argentina): Merrill Lynch
Argentina SA; Merrill Lynch (Japan): Merrill Lynch Japan Securities Co, Ltd; Merrill Lynch (Seoul): Merrill Lynch International Incorporated (Seoul Branch); Merrill
Lynch (Taiwan): Merrill Lynch Securities (Taiwan) Ltd.; DSP Merrill Lynch (India): DSP Merrill Lynch Limited; PT Merrill Lynch (Indonesia): PT Merrill Lynch
Indonesia; Merrill Lynch (KL) Sdn. Bhd.: Merrill Lynch (Malaysia); Merrill Lynch (Israel): Merrill Lynch Israel Limited; Merrill Lynch (Russia): Merrill Lynch CIS Limited,
Moscow; Merrill Lynch (Turkey): Merrill Lynch Yatirim Bankasi A.S.; Merrill Lynch (Dubai): Merrill Lynch International Bank Ltd, Dubai Branch; MLPF&S (Zürich rep.
office): MLPF&S Incorporated Zürich representative office.
This research report has been prepared and issued by MLPF&S and/or one or more of its non-U.S. affiliates. MLPF&S is the distributor of this research report in
the U.S. and accepts full responsibility for research reports of its non-U.S. affiliates distributed in the U.S. Any U.S. person receiving this research report and wishing
to effect any transaction in any security discussed in the report should do so through MLPF&S and not such foreign affiliates.
This research report has been approved for publication in the United Kingdom by Merrill Lynch, Pierce, Fenner & Smith Limited, which is authorized and
regulated by the Financial Services Authority; has been considered and distributed in Japan by Merrill Lynch Japan Securities Co, Ltd, a registered securities dealer
under the Financial Instruments and Exchange Law in Japan; is distributed in Hong Kong by Merrill Lynch (Asia Pacific) Limited, which is regulated by the Hong
Kong SFC; is issued and distributed in Taiwan by Merrill Lynch Securities (Taiwan) Ltd.; is issued and distributed in Malaysia by Merrill Lynch (KL) Sdn. Bhd., a
licensed investment adviser regulated by the Malaysian Securities Commission; is issued and distributed in India by DSP Merrill Lynch Limited; and is issued and
distributed in Singapore by Merrill Lynch International Bank Limited (Merchant Bank) and Merrill Lynch (Singapore) Pte Ltd (Company Registration No.’s F 06872E
and 198602883D respectively). Merrill Lynch International Bank Limited (Merchant Bank) and Merrill Lynch (Singapore) Pte Ltd. are regulated by the Monetary
Authority of Singapore. Merrill Lynch Equities (Australia) Limited, (ABN 65 006 276 795), AFS License 235132, provides this report in Australia. No approval is
required for publication or distribution of this report in Brazil.
Merrill Lynch (Frankfurt) distributes this report in Germany. Merrill Lynch (Frankfurt) is regulated by BaFin.
Copyright, User Agreement and other general information related to this report:
Copyright 2008 Merrill Lynch, Pierce, Fenner & Smith Incorporated. All rights reserved. This research report is prepared for the use of Merrill Lynch clients and
may not be redistributed, retransmitted or disclosed, in whole or in part, or in any form or manner, without the express written consent of Merrill Lynch. Merrill Lynch
research reports are distributed simultaneously to internal and client websites eligible to receive such research prior to any public dissemination by Merrill Lynch of
the research report or information or opinion contained therein. Any unauthorized use or disclosure is prohibited. Receipt and review of this research report
constitutes your agreement not to redistribute, retransmit, or disclose to others the contents, opinions, conclusion, or information contained in this report (including
any investment recommendations, estimates or price targets) prior to Merrill Lynch's public disclosure of such information. The information herein (other than
disclosure information relating to Merrill Lynch and its affiliates) was obtained from various sources and we do not guarantee its accuracy. Merrill Lynch makes no
representations or warranties whatsoever as to the data and information provided in any third party referenced website and shall have no liability or responsibility
arising out of or in connection with any such referenced website.
This research report provides general information only. Neither the information nor any opinion expressed constitutes an offer or an invitation to make an offer,
to buy or sell any securities or other investment or any options, futures or derivatives related to such securities or investments. It is not intended to provide personal
investment advice and it does not take into account the specific investment objectives, financial situation and the particular needs of any specific person who may
receive this report. Investors should seek financial advice regarding the appropriateness of investing in any securities, other investment or investment strategies
discussed or recommended in this report and should understand that statements regarding future prospects may not be realized. Investors should note that income
from such securities or other investments, if any, may fluctuate and that price or value of such securities and investments may rise or fall. Accordingly, investors may
receive back less than originally invested. Past performance is not necessarily a guide to future performance. Any information relating to the tax status of financial
instruments discussed herein is not intended to provide tax advice or to be used by anyone to provide tax advice. Investors are urged to seek tax advice based on
their particular circumstances from an independent tax professional.
Foreign currency rates of exchange may adversely affect the value, price or income of any security or related investment mentioned in this report. In addition,
investors in securities such as ADRs, whose values are influenced by the currency of the underlying security, effectively assume currency risk.
Officers of MLPF&S or one or more of its affiliates (other than research analysts) may have a financial interest in securities of the issuer(s) or in related
investments.
Merrill Lynch Research policies relating to conflicts of interest are described at http://www.ml.com/media/43347.pdf.
Fundamental equity reports are produced on a regular basis as necessary to keep the investment recommendation current.

56