Industry attractiveness
The standard reason given to conduct an industry analysis is to understand whether an industry is attractive or not. This is a rather silly reason. If all you are interested in is attractiveness, the best approach is to look at industry average profitability. Finance Professor Aswath Damodaran has a very nice website from which I extracted the following information. The table shows the ten most profitable and ten lest profitable industries in four geographic markets (return on invested capital, banking is excluded, data from financial year 2011, unweighted average). Ideally, I would like weighted ten year average, but that can be calculated relatively easy.| United States | Europe | Japan | China | |||||||
| Industry Name | ROC | Industry Name | ROIC | Industry Name | ROIC | Industry Name | ROIC | |||
| Computer Software | 43.31% | Advertising | 79.15% | Retail (Internet) | 275.22% | Retail (Internet) | 187.50% | |||
| Educational Services | 42.33% | Computers/Peripherals | 66.93% | Internet software and services | 66.99% | Beverage (Alcoholic) | 72.65% | |||
| Computers/Peripherals | 34.66% | Computer Software | 57.14% | Tobacco | 44.57% | Beverage | 42.44% | |||
| Tobacco | 31.97% | Information Services | 56.40% | Insurance (Life) | 32.62% | Internet software and services | 33.29% | |||
| Retail (Softlines) | 30.03% | Computer Services | 54.74% | Telecom (Wireless) | 14.80% | Broadcasting | 32.42% | |||
| Telecom. Equipment | 26.25% | Healthcare Products | 30.35% | Oil/Gas (Production and Exploration) | 14.66% | Office Equipment & Services | 31.85% | |||
| Metals & Mining (Div.) | 25.19% | Retail (Special Lines) | 28.15% | Heathcare Information and Technology | 14.31% | Healthcare Products | 29.84% | |||
| Semiconductor | 22.88% | Tobacco | 25.95% | Heavy Construction | 13.38% | Furn/Home Furnishings | 28.52% | |||
| IT Services | 21.77% | Food Wholesalers | 25.67% | Hotel/Gaming | 13.32% | Healthcare Facilities | 28.20% | |||
| Internet | 21.54% | Healthcare Services | 25.62% | Healthcare Facilities | 12.50% | Retail (Automotive) | 24.23% | |||
| Water Utility | 5.89% | Oil/Gas (Production and Exploration) | 3.91% | Insurance (Prop/Cas.) | 1.25% | Retail (Building Supply) | 1.76% | |||
| Investment Companies | 5.87% | Oil/Gas Distribution | 3.87% | Financial Svcs. | 1.24% | Utility (General) | 1.72% | |||
| Precision Instrument | 5.78% | Steel | 3.64% | Steel | 1.12% | Shipbuilding & Marine | 1.68% | |||
| Automotive | 5.77% | Educational Services | 3.52% | Shipbuilding & Marine | 0.22% | Food Wholesalers | 1.04% | |||
| E-Commerce | 5.46% | Real Estate (Development) | 2.34% | Investment Co. | 0.20% | Semiconductor | 0.79% | |||
| Foreign Electronics | 5.44% | Financial Svcs. (Non-bank & Insurance) | 1.96% | Financial Svcs. (Non-bank & Insurance) | 0.19% | Steel | 0.40% | |||
| Property Management | 4.72% | Biotechnology | 1.59% | Brokerage & Investment Banking | -0.02% | Thrift | 0.00% | |||
| Maritime | 3.54% | Telecom. Equipment | 0.66% | Semiconductor | -2.79% | Brokerage & Investment Banking | -0.62% | |||
| Building Materials | 2.98% | Semiconductor | 0.46% | Power | -2.82% | Financial Svcs. (Non-bank & Insurance) | -1.64% | |||
| Power | 2.62% | Financial Svcs. | 0.08% | Biotechnology | -19.91% | Trucking | -4.55% | |||
Why bother with a full industry analysis? It is only worth-wile to conduct a full industry analysis if you want further knowledge than can be gained by the averages. Here are some valid reasons:
- You are new to the industry and want to understand the key structural mechanisms in the industry. This is typically the case for students as well as private equity professionals or equity analysts.
- You want to understand how the attractiveness of an industry is changing in the future. First, there is a fair amount of continuity in industry attractiveness, but the structural characteristics do change with time. By looking at each important characteristic a better forecast of future industry attractiveness can be generated. Second, in some cases change occurs quickly. This could be due to legal changes, technological, or demographic changes. Whether a particular innovation will make an industry more or less attractive is best studied by how that innovation influences the structural characteristics of the industry. This is typically the case for companies or equity analysts.
- You want to explore new strategies in an industry. To craft an original strategy you must be very familiar with the the structural (and behavioural) characteristics of the industry. If you are not familiar with the industry structure, it is likely that your "original" strategy will be a copy of an existing company's strategy. Only by seeking out new knowledge can you craft a new strategy. This is typically the case for companies.
Industry definition
This is housekeeping, but your definition of the industry should correspond to the underlying reality. In some cases, it is a straight forward description of the industry (e.g. tobacco, steel, beverages). In other cases, the description is a bit more complicated (e.g. metals and mining, financial services non-bank, office equipment and services, biotechnology) requiring a bit more thought. In yet other cases you might judge an extant industry to be defined wrongly (e.g. retail special lines, internet software and services, hotel/gaming).
There are a lot of choices that needs to be made, so judgment is required. However, if you miss to include some products in the industry definition, you will have a second chance when you study substitutes to cover them. So the choices you make on industry definition will turn out not to be critical for the analysis. It is however, important to present a clear definition so that everyone involved in the analysis are using the same vocabulary. The following section will describe three important aspects of industry definition:
Substitutes. At the core, the industry definition should capture a set of related companies selling similar kind of products. The economists describe this as a high cross-elasticity of demand; when prices of one product goes up the demand for other products goes up. When the price of BMW goes up, the demand for Lexus increases. However, the price of BMW is not likely to influence the demand for Toyota. Based on this, BMW and Lexus belong to one industry (luxury cars) and Toyota to another industry (standard cars). While theoretically correct, this way of defining an industry is often not practical. Normally we do not have access to cross-elasticity of demand for all products. This is where judgement plays a role in defining the industry. For practical reasons we might settle with an industry definition that includes all cars. However, we would not include motorcycles or taxi services in the automotive industry.
Often it makes sense to follow existing industry definitions even if they are not perfect, but in other cases the existing definitions will lead you in the wrong direction. Internet retail is often classified as a separate industry, but it is more correctly classified into existing retail channels (e.g. book retailing, clothing retailing, travel agencies). Treating Internet retail as a separate industry when it has largely killed a lot of high-street shops just does not make sense. Hotel/gaming is classified as one industry in Japan, but it is very unlikely that most hotels have gaming facilities, and vice verso.
Geography. If the customers are local (i.e. they only buy from from companies present locally) the industry should be defined on a local level and if the customers are global (i.e. they could buy from anywhere in the world), the industry must be defined correspondingly. The market for telecom equipment is global. Companies like Huawei and Ericsson are from different continents but they compete in one global market. Huawei might be strong in China and Ericsson strong in Sweden, but what is important is that the buyer both in China and Sweden would consider both companies as suppliers. In contrast the market for water utilities is regional or national. Even if there would be one multinational water utility company, buyers would not buy water globally. Water is just too heavy to transport. The market for emergency healthcare facilities is local.
For practical purposes it is often reasonable to study a national market instead of multiple local markets. However, if the industry is global it does not make sense to study a national market independently.
Geography. If the customers are local (i.e. they only buy from from companies present locally) the industry should be defined on a local level and if the customers are global (i.e. they could buy from anywhere in the world), the industry must be defined correspondingly. The market for telecom equipment is global. Companies like Huawei and Ericsson are from different continents but they compete in one global market. Huawei might be strong in China and Ericsson strong in Sweden, but what is important is that the buyer both in China and Sweden would consider both companies as suppliers. In contrast the market for water utilities is regional or national. Even if there would be one multinational water utility company, buyers would not buy water globally. Water is just too heavy to transport. The market for emergency healthcare facilities is local.
For practical purposes it is often reasonable to study a national market instead of multiple local markets. However, if the industry is global it does not make sense to study a national market independently.
Ownership. The output from an industry is products/services sold in a marketplace. So the existence of markets is a good to delineate industries. In the US market for softdrinks there used to be a small number of concentrate manufacturer and a large number of bottlers. The bottlers bought concentrate on the market. After 2010, most of the bottlers had been acquired by the concentrate manufacturers. Prior to 2010 it made sense to consider concentrate manufacturing and bottling as two separate industries as there was a market between them. Currently, there is no such market so it would be artificial (and practically) difficult to consider them as two separate industries.
Very often the situation is less clear cut. Historically, there was a point where around 50% of the concentrate was shipped to fully integrated bottlers and 50% to independent bottlers. As long as there is a substantial market, it is probably best to consider two industries. 50% is a substantial part of a market, 20% is not a substantial part. The real difficult would be if around 70% was done internally and 30% on the market. In such cases it might also be useful to consider the trends. If internal transactions are increasing then the industries could be considered an one industry.
Very often the situation is less clear cut. Historically, there was a point where around 50% of the concentrate was shipped to fully integrated bottlers and 50% to independent bottlers. As long as there is a substantial market, it is probably best to consider two industries. 50% is a substantial part of a market, 20% is not a substantial part. The real difficult would be if around 70% was done internally and 30% on the market. In such cases it might also be useful to consider the trends. If internal transactions are increasing then the industries could be considered an one industry.
No comments:
Post a Comment