"The most valuable commodity I know of is information."
An excerpt from 'Capital returns, Investing Through the Capital Cycle: A Money Manager's Reports 2002-2015', Chapter Two, Value in growth...
Highlighting hidden information and suggesting that money is made in the dark, not the light. This echoes Howard Mark's ideas concerning, 'second line thinking' and my own ideas about 'Wager Value' (focusing on under-used information).
-----------------------------------
"The most valuable commodity I know of is information."
Gordon Gekko
-----------------------------
ESCAPING THE SEMIS’ CYCLE (FEBRUARY 2013)
Niche semiconductor businesses have escaped the ravages of the industry’s
capital cycle
Driven by Moore’s law, the semiconductor sector has achieved sustained
and dramatic performance increases over the last 30 years, greatly benefit-
ing productivity and the overall economy. Unfortunately, investors have
not done so well. Since inception in 1994, the Philadelphia Semiconductor
Index has underperformed the Nasdaq by around 200 percentage points,
and exhibited greater volatility.
The reason for this poor performance is no secret. No part of the tech-
nology world has been more prone to cyclical booms and busts than the
semiconductor industry. In good times, prices pick up, companies increase
capacity, and new entrants appear, generally from different parts of Asia
(Japan in the 1970s, Korea in 1980s, Taiwan in the mid-1990s, and China
more recently). Excess capital entering at cyclical peaks has led to relatively
poor aggregate industry returns.
While the history of the semiconductor business provides a classic
example of the capital cycle, there are companies operating in niches of the
industry which have delivered excellent long-term returns for sharehold-
ers. Two of them are recent additions in our US portfolio: Analog Devices,
based in Norwood, Massachusetts, and Linear Technology, headquartered
in Milpitas, California.
Semiconductors are essential electronic building blocks for electronic
systems and equipment. Analog semiconductors represent around 15 per
cent of the total semiconductor market, with the rest being digital. The func-
tion of an analog semiconductor is to bridge the gap between the real world
and the electronic one – monitoring, amplifying and transforming phenom-
ena such as temperature, sound and pressure. End-markets include mobile
phone handsets (e.g., the digitization of voice), automobiles (e.g., the crash
sensor in an airbag) and the industrial economy (e.g., a temperature sensor
in process automation equipment). This is in contrast to digital semicon-
ductors which operate, predominantly, in the purely digital world of binary
code.
The analog sub-sector has been a notable exception to the low and vola-
tile investment returns of the semiconductor industry. Analog Devices, for
example, has consistently generated high margins over many years, with
robust profits even in stressed environments. On average, between 2000 and
2012, the company’s gross margin was 60 per cent and operating margin was
25 per cent. The level of capital intensity required to achieve these impressive
returns was relatively low. Capex to sales at Analog Devices has averaged 6
per cent since 2000, and has fallen to 4 per cent over the last five years. This
low level of capital intensity has allowed free cash flow conversion at a con-
sistently high level, on average at over 100 per cent of net income.
Linear Technology has displayed even stronger economics. Since the turn
of the century, it has enjoyed an average gross margin of 76 per cent and aver-
age operating margin of around 50 per cent. The ratio of capex to sales has hov-
ered around 5 per cent, with cash conversion again greater than 100 per cent.
In addition to having robust margins, both companies have historically expe-
rienced strong sales growth, driven by the increasing penetration of technol-
ogy into everyday life. Since 1990, Analog Devices’s revenue has compounded
at 8 per cent annually, and Linear’s sales have grown by 14 per cent a year.
How have these companies generated such high returns and to what
extent are these returns sustainable? The answer lies in an understanding
of the supply side of this industry – the specifics of the production process,
market structure, competitive dynamics and pricing power, which together
constitute the essence of capital cycle analysis. Consider first the mechanics
of the analog semiconductor business. As the real world is far more complex
and heterogeneous than the digital one, the product design required to cap-
ture it has to be more complex and heterogeneous. This means that product
differentiation of analog semiconductors is higher and company-specific
intellectual property (whether physical or human capital) more important.
The human capital component is especially hard to replicate because
engineering talent deepens with experience. The design process is much
more trial and error than in other technology disciplines, and less reliant on
computer modelling and simulation. To become an expert in analog semi-
conductor design takes many years – the tenure of the average engineer at
Analog Devices is 20 years. This forms an important barrier to entry. In
addition, each analog company’s process technologies are quite distinct (dig-
ital utilizes a more generic process).
Thus, it is difficult for an engineer to be poached by another analog
company without his productivity being significantly impaired. The supply
of new engineers tends to be constrained for the analog sector – new science
graduates are much more likely to pursue the digital semiconductor route.
This is largely because the learning curve is less steep in digital, and expe-
rience on the job less valued. As a result, research capacity in the world of
analog semis has been, and will likely to continue to be, constrained.
These factors – a differentiated product and company-specific “sticky”
intellectual capital – reduce market contestability. These strategic advantages
are compounded by the fact that analog has a more diverse end market than
digital, with a much wider range of products, numbering in the thousands,
and smaller average volume size. Such market characteristics make it dif-
ficult for a new entrant to compete effectively. Thus pricing power tends to
be robust and market positions relatively stable over long periods. While the
overall market is relatively fragmented – the five firm concentration ratio
is around 50 per cent – it is more consolidated in the various market sub-
segments. Analog Devices, for instance, has over a 40 per cent share in data
converters.
Pricing power is further aided by the fact that an analog semiconductor
chip typically plays a very important role in a product (for example, the air-
bag crash sensor) but represents a very small proportion of the cost of mate-
rials. The average selling price for Linear Technology’s products is under $2.
As a result, competition tends to be less on price and more on product qual-
ity. In addition, once a chip has been designed into an application – a proc-
ess on which the original equipment manufacturer and the analog company
often collaborate, it is costly for the manufacturer to replace it, as the whole
production process has to be revised. Hence switching costs are high, both
improving pricing power over the product lifecycle (often ten years or more)
and the degree to which revenues are recurring.
Finally, and of critical importance, the analog production process is less
standardized than most tech components, and thus far less vulnerable to
obsolescence from the endless march of Moore’s law, significantly reducing
capital intensity. More than a third of sales at Analog Devices come from
products which are more than ten years old. This shelters the sector from
the destructive force of the capital cycle which has wreaked such havoc in the
digital semiconductor industry. Hence there are good reasons to believe that
the high returns historically achieved by these companies can be sustained
into the future.
We are also confident that management will allocate future surplus cash
flow for the benefit of equity investors. Historically, most of the growth of
these businesses has been organic, with excess cash returned to shareholders.
This is a significant achievement for companies in the technology sector, an
area where the temptation to do strategic deals has been strong, often to the
detriment of shareholders. We expect the long-serving management teams
of both companies to continue to allocate capital prudently. Both Analog
Devices and Linear Technology currently offer free cash flow yields of 5 per
cent. With long-term growth in free cash flow likely to be similar to histori-
cal levels, our total annual return expectation is in the low double digits.
---------------------------------
Source
No comments:
Post a Comment