Some industries are almost completely dominated by two companies.
Governments are generally wary of monopolies, but sometimes are more open to the existence of duopolies. With a duopoly, two companies share the dominant position of a market, with the third largest competitor and all other competitors being distantly smaller.
I take more than just market share into account when determining whether an industry has a duopoly, however. If two large companies control most of the market, but there is no hard limit against competitors potentially taking market share, then the market is not really "controlled" by those two companies; they just happen to be the biggest players.
For example, Coca Cola (KO) and Pepsico (PEP) share the dominant position in the global non-alcoholic beverage industry. On a global scale, they have the unrivaled distribution networks that give them a solid duopoly. But on any given local scale, the switching costs for buying a smaller brand of beverage are minimal; Coke and Pepsi don't have some untouchable price or taste advantage over their smaller competitors in any given market, and instead rely primarily on their brand strength. So their global duopoly is merely statistical in nature; a smaller competitor can't compete with their whole global distribution networks, but they can compete in their respective local markets, so the dominant positions of these companies are based on the probability that local competitors won't all be successful at once. The barriers to entry aren't particularly high.
Another example would be Verizon (VZ) and AT&T (T). They're the two largest wireless providers in the United States. But they do have two sizable competitors that also own considerable chunks of wireless spectrum rights. If AT&T's T-Mobile acquisition would have succeeded, I would have viewed this industry as a clearer duopoly, but since it did not, I don't view this as a particularly strong duopoly. Certain technologies even promise to improve the efficiency of spectrum usage by orders of magnitude, reducing the scarcity. Much like Coke and Pepsi, they do however have very strong positions, based on their control of spectrum as well as their significant wireless infrastructure and negotiating leverage.
There are several examples of clearer duopolies, like Intel (INTC) and AMD (huge research and development budgets and years of planning are needed to even begin to compete, so smaller competitors would not be viable), or S&P and Moody's (control of the rating market). Those types of businesses are what I view as clearer duopolies, because it's not just a matter of market share; it's a combination of dominant market share and extremely high barriers to entry.
This article focuses on two particular duopolies with high barriers to entry and market share that is consolidated into two main companies in each industry: Domestic Shipping and Trash Collection.
The domestic small package shipping market requires extremely extensive use of the network effect, and there's no way to even begin to compete without enormous capital expenditures and time. Currently, UPS and FedEx have dominant control over this market, with the United States Postal Service being the only other real competitor. Several years ago, Deutsche Post acquired both DHL Express and Airborne Express in an attempt to compete in the U.S. domestic shipping market, but after billions of dollars in losses, they had to pull out. Even a sustained attempt by a large international competitor couldn't penetrate the market with any hope of a good return on investment.
United Parcel Service (UPS)
UPS is the larger of the two competitors, has more integration between ground and air shipping, and therefore has the larger net profit margin of the two. UPS also has the larger dividend yield of 3.01%.
FedEx Corporation (FDX)
FedEx is the smaller competitor, but still competes profitably in this duopoly. It offers the much lower dividend yield of 0.63%, but its stock trades at a lower valuation with a P/E of under 14.
When the third largest trash collector, Republic Services (RSG), acquired the second largest collector, Allied Waste, in 2008, it formed a duopoly with the largest trash collector, Waste Management (WM). The two companies now control over half of the industry. The key asset in this business is ownership of the landfills. It requires considerable cost to open a landfill, but more importantly, it requires allowance from the local government. Landfills face the problem of a "not in my backyard" mentality, so when combined with the large cost, they're not easy assets to develop.
Landfills are sometimes owned by local governments, and sometimes owned by companies like Waste Management or Republic Services. The owner of a landfill can charge tipping fees to smaller trash haulers that don't own their own landfills or transfer stations, and because trash disposal is so necessary, the incoming cash flow for a landfill is almost annuity-like.
Waste Management (WM)
Waste Management is the largest private owner of landfills in the country, pays a considerable dividend yield of 4.07% (though in my latest report I raised the issue of the increasing payout ratio), and maintains a wide moat. The shareholder yield for this business is considerably large.
Republic Services (RSG)
Republic Services is the second largest, and slightly more leveraged competitor. It offers a lower but still decent dividend yield of 3.27%, but also a lower and safer dividend payout ratio.
For these two duopolies, if I were to invest in one of the companies, I'd likely invest in their competitor as well, and therefore invest in the industry as a whole rather than in one individual player. The main enticement for investing in a duopoly is that your investment thesis therefore becomes dependent on the success of the industry rather than on the success of an individual company, and yet only two companies need to be owned in the industry to get that benefit.
Full Disclosure: As of this writing, I own shares of KO and no other companies mentioned.