Asphyxiation is a condition in which the body doesn’t receive enough oxygen. A common side effect is death. Oh, and hallucinations – we start seeing things that don’t exist. When the market is making new highs, valuations are high, and values are scarce, it is common to suffer from value asphyxiation –imagining value when it is not there. Mistakes in this environment predominantly come from the commission (not the omission) of buy decisions. This is why, before you commit your capital, you have to double-check your lucidity and think thrice.

And this is why, when we stumbled on Qualcomm, we could not believe what we saw. Qualcomm should double its earnings over the next four years, and it has an impenetrable moat, great management, a cash-laden balance sheet, infinite incremental return on capital, and is cheap, trading at low-teen multiples. These things are not supposed to happen to companies with a market cap over one hundred billion dollars that are followed by several dozen analysts – unless the street is concerned that the company has caught a deadly venereal disease (in other words, is on its way to becoming obsolete) – and especially not while the market is making new highs.

Qualcomm is a value in plain sight because it is misunderstood by investors – and for good reason: two-thirds of Qualcomm’s revenue comes from the semiconductor segment. Qualcomm designs chips that go into cellphones and tablets. If you read sell-side reports on Qualcomm, most of the ink is spilled about its semiconductor business. This makes sense, because QCOM is covered mostly by semiconductor analysts, and that is what they know – semiconductors. They have an edge in that arena, and so that is what they write about. They can provide many insights about the intricacies of QCOM’s chip-set designs, and how QCOM’s chips are in hundreds of smartphone models while Intel (its largest potential rival) has its chips in less than a handful. They’ll tell you how QCOM combines chips that perform multiple functions into a single chip, and how that gives the company a competitive advantage. Analysts will spend dozens of pages on the semiconductor segment, which has been growing 20-30% a year.

Qualcomm's licensing business gets the least amount of ink; after all, it only accounts for a third of revenues and is a very straightforward, unexciting business that lacks dramatic competitive dynamics – Qualcomm is the only game in town. It collects about $7 (a 3-5% royalty, based on wholesale price) from almost every smartphone sold globally. Of course licensing, despite its lack of excitement and dearth of revenue generation, represents about three-quarters of QCOM’s profit.

Let me explain QCOM’s licensing business through this analogy. There is a continent called Spectrum that for a long time had only one-lane (2G) roads,–which were for the most part built before QCOM arrived on the scene. QCOM employees a lot math geniuses, and its engineers figured out the most efficient way for cars to get from one place to another (they developed complex algorithms for the most efficient use of spectrum) and thus helped to fit more vehicles onto newly built freeways that were based on QCOM’s design. This is very important, because as vehicles get smarter and safer and more and more people want to drive them, traffic jams can result, and so we need ever bigger and better freeways (bandwidth, you see).

QCOM gives away its highway designs to construction firms (makers of hardware the likes of Erickson and Cisco) and highway owners (the likes of AT&T and Vodafone). Qualcomm is like a global tollbooth operator that charges car manufacturers (like Nokia and Apple) a percentage of  phone’s wholesale price. The wireless industry outsources a big chunk of its R&D to QCOM to design the most efficient use of limited, high-demand spectrum. This creates a significant competitive advantage for QCOM, because it gets to spread its massive R&D across a huge number of phones.

Qualcomm makes almost no money on cellphones that work solely on 2G networks. 2G is fine for voice communications but doesn’t do a good job of carrying data. Smartphones require higher-speed networks to function, and this is where 3G and 4G (also known as LTE) come in. These networks were developed in large part based on QCOM’s IP.

 In 2012 there were 3.5 billion global cellphone subscribers. Only about 800 million of them were on 3G/4G networks – the rest are using dumb or feature phones on 2G networks. But 2G gets buried ever deeper with every billion that mobile carriers spend on 3G/4G networks. And they are spending hundreds of billions.

If we assume that the number of global cellphone subscribers will grow 4% a year – a fairly realistic assumption considering global population is at seven billion people and soon dogs will have phones – and that only 5% of subscribers a year will switch from 2G to 3G/4G, the number of smartphones on 3G/4G networks will be growing 22% a year by 2017. Despite that enormous growth rate, only half of users will be on 3G/4G networks by then, so QCOM's growth is unlikely to decline much after 2017. However, the bulk of the smartphone growth is happening in developing countries where cellphones are cheaper and technology tends to get cheaper with time, and so QCOM’s licensing-revenue growth will fall below subscriber growth. The company is guiding for a 2-3% average selling price decline per year, but if we say the decline is 5% a year, then licensing-revenue growth will still be in the teens.

It gets better. Wireless services will advance far beyond cellphones and tablets. They will be embedded in cars and in billions of machine-to-machine devices, from vending machines that accept credit cards and process them wirelessly, to meters that monitor your water consumption and send data to the utility company, to innumerable other uses that we haven't even dreamed up yet. In addition, over the next few years you’ll be hearing a lot more about small cells – your personal cell tower the size of a deck of cards. Small cells will help to solve the wireless bandwidth problem in populous areas. There will be hundreds of millions of them installed, and just as with machine-to-machine devices, QCOM will be receiving its rightful license fee.

Currently the only mobile company that is not paying QCOM is China Mobile – the largest operator in the world. It uses its proprietary 3G standard but is now transitioning to QCOM’s 4G. It is only a matter of time before QCOM starts collecting royalties from smartphones used on China Mobile’s network.

QCOM's competitive advantage is very deep. It derives from its enormous IP portfolio but also from the fact that QCOM spends more money on wireless R&D – $4.7 billion over the last 12 months – than any other company in its space. Half of it goes to develop new chips, but the rest is spent on improving wireless technology. It by far the best way to play on global growth of smartphones.

By our estimates, by 2017 QCOM will earn about $7 of free cash flows, and its debt-free balance sheet will balloon from $17 of net cash per share to about $40. Give it a conservative multiple of 14 and some credit for cash, and in four years it will be a $130 stock.