Most people rely on expert opinion to make investment decisions. But the essence of being a good investor involves thinking through things carefully and making your own assessments. After taking a close look at Apple  (AAPL) and its presumptive investments in the automotive field, this Bloomberg article regarding the confluence of Tech and Autos caught my eye.

The article’s centerpiece is an infographic showing investments, partnerships, and personnel shifts between Tech firms and Automakers and provides a good high-level perspective on the evolving landscape.

While the infographic was good, my eye was caught on this quote discussing a consultant’s assessment of the future of the auto industry.

McKinsey estimates that rideshare and onboard-data services could generate an additional $1.5 trillion of annual automotive revenue by 2030, adding to the $5.2 trillion from traditional car sales and services.

The people that work at McKinsey are geniuses. It would be easy to read their assessment and take it for the gospel truth. But to me, the idea that automakers would gain revenue from a shift to rideshare is laughable.

By my back-of-the envelope calculations, automaker revenues will fall by roughly a half over the time mentioned in this report, thanks to a paradigm shift in the business of getting from point A to point B. This shift will be characterized by the gradual replacement of autos with atus /ah-tooz/ – “automatic transport units.”

Being able to assess the demand environment for a company is the first essential step in making an intelligent valuation of a firm. Let’s walk through our assessment step-by-step and see if our argument rings true with you.

Auto Spending is Unsustainable

Operating a personal car can scarcely be justified on economic grounds. Another quote from the Bloomberg article admirably underscores this contention:

It costs an average of $8,558 per year to own a car in the U.S., but each vehicle is used just 4 percent of the time.

Let’s do a quick, back-of-the-envelope calculation meant to give a sense for what this $8,558 figure means to Americans.

The U.S. Census Bureau estimates that the average household income in this country is around $52,000 per year. Note that this is a pre-tax figure, and assuming an adjusted tax rate of 15%, that implies post-tax, take-home pay of $44,200. Further assuming that one-third of pre-tax salary is spent on housing and is non-discretionary, the average household has around $27,000 left over after paying taxes and housing payments.

Using this logic, on a percentage basis, $8,558 per year works out to around 32% of discretionary spending. In other words, just under one-third of most people’s discretionary income stream is directed at an asset that is used only 4% of the time!

If you ran a manufacturing company, would you commit to sinking one-third of your after-tax profits into a factory that you anticipated would have an average capacity utilization of 4%? If you did commit to this arrangement, how long would it take before the first class-action shareholder lawsuit was filed?

Where are we going?

In the not-too-distant future, automakers (atumakers) will no longer sell to individuals through dealership networks. Instead, they will mainly sell directly to companies like Lyft and Uber, which will put these atus into service under their own brand names. They will also likely sell to “transportation capacity brokers” a kind of company that does not yet exist, but which will rent out peak-time transportation capacity to Lyft and Uber. A few atus will be ordered directly off the Internet (somewhere out there, the Michael Dell of the atu world may or may not be toddling around in diapers today), but the big-volume sales will be through the first two channels.

How many atus will be needed in this scenario?

An initial, off-the-cuff estimate would be that total atu units would be about half that of total auto units. Right now, Americans own two autos on average per household; we’ll see this drop to around 1.0 atu per household within forty years, even without taking into consideration changes in custom and habit. (We already take an Uber ride with another person – the driver – so sharing a ride with a non-driver is not such a huge shift. Uber and Lyft will charge us less for shared rides and give us the option to share or not when we call for a pick-up.)

High-speed Internet service and a generation of people used to online social interactions will make commuting to most white collar jobs much less necessary, a social change that will ease peak atu demand during “rush hours.” My 18-year old son meets his friends in person one time for every ten times he meets them online. Once he and his cohort are making managerial decisions, the pointless time expenditure of 45 minutes one-way to attend a one-hour meeting will be avoided in about the same proportion.

Over the last twelve months, Ford (F) has generated $153 billion in revenues and generated $18 billion of Owners’ Cash Profits. In the atu world we have envisioned here, it is easy to imagine the company generating $5 billion of OCP on $77 billion in revenues. Revenues will drop due to falling unit volumes, profitability will drop as maker-customer negotiating dynamics change and inter-atu differentiation drops.

Presumably, Ford will need to continue spending on capital projects in the atu-future, and it will do so out of its lowered profitability. The cash left over after making these expenditures is its “free cash flow” and the value of a firm is the sum of its free cash flows over time.

So if we are right about an atu future for automakers like Ford, the value of these companies is vastly overstated at present market prices.

Don’t take my word for it – work through the steps for yourself. There are only a few drivers of value and these drivers are visible to anyone who takes the time to look at them.