Thinking about ROIC and Growth

1

• Thinking about ROIC and Growth

2
Empirical Analysis of ROIC

• Through this point, we have examined a general
  model of value creation using economic theory and
  case studies. But how does ROIC and growth
  behave on an aggregate empirical basis?
• To answer this question, McKinsey Company
  analyzed the corporate performance of more than
  5,000 US-based non-financial companies for a
  period of 40 years (1963 2003).

KEY FINDINGS
Median ROIC between 1963 and 2003 9 Percent
of companies between 5 and 15 ROIC 50

• Also, median ROIC differs by industry and growth,
  but not by company size.
• And individual-company ROICs gradually regress
  toward medians over time but are somewhat
  persistent. Fifty percent of companies that
  earned ROICs greater than 20 percent in 1994 were
  still earning at least 20 percent 10 years later.

3
ROIC Over Time Non-Financial Companies

• Since 1986, the ROIC spread across companies has
  gradually widened, driven primarily by companies
  at the top end.
• Companies with strong barriers to entry have
  achieved increased profits from drops in raw
  material prices and labor productivity.
• When measured with goodwill, the spread does
  not widen. This implies that top companies are
  purchasing other top performers yet paying full
  price for the acquired performance.

Annual ROIC without goodwill
Average
15.3
Percent
9.0
5.0
Annual ROIC with goodwill
Average
13.6
Percent
8.3
4.7
Source Compustat, McKinsey Companys
corporate performance database
4
Distribution of ROIC Non-Financial Companies

• 84 of all ROIC observations were below 20.
• If your forecast model requires an ROIC gt 20 to
  generate value, how skeptical should you be?

Annual ROIC without goodwill, 1963-2003
Percent of sample
lt-10.0
-5.0
0.0
2.5
5.0
7.5
10.0
12.5
15.0
17.5
20.0
25.0
30.0
35.0
40.0
gt40.0
22.5
ROIC
Percent of observations below ROIC level
5
7
11
15
25
42
56
66
74
80
87
89
92
94
95
100
84
Approximately 50 of the sample is within ROIC
range of 5-15
Source Compustat, McKinsey Companys
corporate performance database
5
Median ROIC by Industry Group

• ROIC varies by industry, whereas industry
  performance is quite stable. Therefore, industry
  membership can be an important predictor of
  forecasted performance.

Amounts in Percent
The majority of industries had median ROICs
between 9 and 12
Source Compustat McKinsey Companys
corporate performance database
6
ROIC Segmented by Size and Revenue Growth

• ROIC appears to be positively correlated with
  revenue growth, but has no relation to size.
• This does not mean that growth causes strong
  performance, but rather that certain underlying
  factors enable both growth and ROIC (e.g. fast
  growing industries need not compete on price to
  grow revenues).
• Size shows no clear relation with ROIC.
  Efficiency gains from scale may be outweighed by
  bureaucratic inefficiencies or other
  inflexibilities.

Annual ROIC without goodwill, 1963-2003 Percent
Revenues
200-500 M
500-1,000 M
1,000-2,500 M
lt200 M
gt2,500 M
5.2
6.0
6.5
lt0
8.0
7.7
8.0
8.1
9.1
0-5
ROIC increases with higher growth rate
8.9
9.3
9.6
9.5
10.3
3-year real growth rate
5-10
10.8
10.9
11.2
10.9
11.8
10-15
11.9
11.1
11.7
11.5
11.9
15-20
11.9
11.8
11.8
gt20
No clear relation between size and performance
Source Compustat, McKinsey Companys
corporate performance database
7
ROIC Decay Analysis Non-Financial Companies

• ROIC demonstrates a pattern of mean reversion.
  Companies earning high returns tend to gradually
  fall over the next fifteen years and companies
  earning low returns tend to rise over time.
• However, there is a continued persistence of
  superior performance beyond ten years. ROIC
  does not fully regress to the aggregate median of
  9 percent.

Median ROIC of portfolio
ROIC Percent
Percent
gt20
15-20
10-15
5-10
lt5
Number of years following portfolio formation
Source Compustat McKinsey Companys
corporate performance database
8
ROIC Decay Analysis Consumer Staples Industry

• When benchmarking historical decay, it is
  important to segment by industry. For example,
  companies in the consumer staples industry
  regress much more slowly than companies overall.
• In the consumer staples industry, even after 15
  years, the original class of top performers
  outperform the worst performers by more than 13
  percent.

Median ROIC of portfolio
ROIC Percent
Percent
gt20
15-20
10-15
5-10
lt5
Number of years following portfolio formation
Source Compustat McKinsey Companys
corporate performance database
9
ROIC Transition Probability (1994-2003)
Companies with ROICs between 10 and 20 show
little persistence, landing in any group with an
equal probability
43 of companies with ROIC of lt 5 in 1994 still
have an ROIC of lt 5 ten years later
ROIC in 2003
Total
lt5
5-10
10-15
15-20
gt20
50 of companies with ROIC of gt20 in 1994 still
have ROIC of gt20 ten years later
lt5
100
5-10
ROIC in 1994
100
10-15
100
100
15-20
gt20
100
Source Compustat McKinsey Companys corporate
performance database
Transition probability analysis confirms that
ROIC shows considerable persistence, especially
at high and low ROIC performance levels
10
Empirical Analysis of Corporate Growth

• Through this point, we have examined how ROIC
  behaves over time. But how does corporate
  revenue growth behave on an aggregate empirical
  basis?
• To answer this question, McKinsey Company
  analyzed the corporate performance of more than
  5,000 US-based non-financial companies for a
  period of 40 years (1963 2003).

KEY FINDINGS

• Median revenue growth rate between 1963 and 2003
  equals 6.3 in real terms and 10.2 percent in
  nominal terms. Real revenue growth fluctuates
  more than ROIC, ranging from 1.8 in 1975 to
  10.8 in 1998.
• High growth rates decay very quickly. Companies
  growing faster than 20 in real terms typically
  grow at only 8 percent within five years and 5
  percent within ten years.
• Extremely large companies struggle to grow.
  Excluding the first year, companies entering the
  Fortune 50 grow at an average of only 1 percent
  (above inflation) over the following fifteen
  years.

11
Revenue Growth Over Time Non-Financial Companies

• The annualized median (real) revenue growth rates
  between 1963 and 2003 equals 6.3. This is quite
  high, especially when compared to U.S. GDP growth
  of 3.3
• Why the difference? The sample only includes
  public companies, and GDP growth fails to capture
  international growth of domestic companies.

• Median revenue growth demonstrates no trend over
  time.
• Beginning in 1973, ¼ of all companies actually
  shrank in real terms in a given year.

Source Compustat McKinsey Companys
corporate performance database
12
Revenue Growth by Industry Group

• Real revenue growth varies dramatically by
  industry. Unlike ROIC, rankings of industries
  based on growth vary over time.

Source Compustat McKinsey Companys corporate
performance database
13
Revenue Growth Decay Analysis
Median growth of portfolio

• Growth decays very quickly for the typical
  company, high growth is not sustainable.
• By year five, the highest growth portfolio
  outperforms the lowest-growth portfolio by less
  than 5.
• Although ROIC is persistent (high ROIC companies
  often continue to generate high ROIC), growth is
  not.

Revenue growth Percent
Percent
gt20
15-20
10-15
5-10
lt5
Number of years following portfolio formation
Source Compustat McKinsey Companys corporate
performance database
14
Average Revenue Growth Rate for the Fortune 50

• Most large companies struggle to grow once they
  reach a certain size. Consider the real revenue
  growth rate for companies entering the Fortune
  50.
• Although growth is strong before companies enter
  the Fortune 50, growth drops dramatically after
  inclusion. During five of fifteen years after
  inclusion, Fortune 50 companies actually shrink
  (in real terms)!

Source Corporate Executive Board, Stall
Points Barriers to Growth for the Large
Corporate Enterprise, 1998
15
Revenue Growth Transition Probability (1994-2003)

• But are there some companies that can growth
  faster than the norm? In short, the answer is
  no. An analysis of transition probabilities
  shows that most companies growth at less than 5
  ten-years later, regardless of their initial
  growth rate.

Over 50 of companies in each revenue growth
category in 1994 had lt5 revenue growth ten years
later
Only 21 of companies with 20 or greater revenue
in 1994 have at least 15 revenue growth ten
years later
16
Closing Thoughts

• When building a DCF model, we too often become
  caught up in the details of financial statements
  and forget the economic fundamentals a companys
  value is driven by ROIC and revenue growth.
• Therefore, it is critical to benchmark your
  forecasts of ROIC and growth against
  economy-wide, as well as industry-based,
  empirical data.
• A companys ROIC will only exceed WACC for an
  extended period if it has a competitive advantage
  with barriers to entry and imitation. High ROIC
  is typically driven by the ability to charge a
  price premium, low costs, or efficient use of
  capital. Empirically speaking, ROIC over time
  reverts to the mean, but companies can
  persistently achieve high ROICs.
• Conversely, few companies can sustain high growth
  for periods greater than five years. Even
  Fortune 50 companies struggle to maintain revenue
  growth, shrinking in five of the fifteen years
  following entrance into the elite group.