In 1989, my partner Deborah Allen and I wrote a book titled Rust to Riches . The title of the first chapter was "General Who?" In that chapter a professor in the year 2025 asked his class if anyone knew the reason General Motors ( GM :NYSE - commentary - research ) had been forced into bankruptcy. A student raised her hand and answered, "General Who?"

We used that title because people have the illusion that big things can't die. The dinosaurs thought that too. The dinosaurs are the source of the gasoline that we pump into the gas tank of our cars (fewer and fewer GM cars every year).

If my memory serves me correctly, GM's share of the U.S. market was 61% in 1961. Today the market share of the erstwhile Big Three (GM, Ford ( F :NYSE - commentary - research ) and Chrysler ) is less than 60%. 1961 was also the year GM departed from the strategy that had made it so big in the first place. Until then, GM had always chosen a car guy (an engineer) as its CEO, and a money guy (a finance guy) as its chairman. That year, the company decided to appoint a money guy as its CEO as well. Big mistake.

I have always believed it takes two people to build a great company. The CEO should be the guy blazing through the halls preaching his love of the company's mission, its products and its customers. He should know how the product is made and why the product is better than the other guy's product. He should be scouring the earth for every possible high-return investment to grow the business. And he should love to take his meals by eating out of the other dog's bowl.

The chairman is a different cat. He should be the guy who holds the capital purse -- tightly. He should say "Wait a minute, here. Why should I believe this project is going to increase the value of our capital?" He should be old and crusty and no fun at dinner.

It is the creative tension between these two forces that builds value in a business. It is what makes a company able to adapt to change. When it is absent, sooner or later there will be a train wreck.

Well, GM has been run by the bean counters for the last 40 years. The result has been a lot of good financial things along the way (GMAC, vendor outsourcing, etc.), a few excellent adventures (Hughes, EDS), and products that do not measure up to the competitors that have passed them by.

Wednesday's announcement that 2005 won't be so good is just the next step in the process. As a friend told me once, "A big company takes a lot of killing. But it will eventually die."

Besides the obvious sad story about GM, this news should make us think about a few others things as well:

1. Asset markets. Used car, not new car, costs set new car prices. That's because there are so many of them -- more than 230 million, 13.5 years' worth of new cars at 17 million new cars and trucks per year, according to Manheim, the world's largest remarketer of used vehicles and publisher of the Manheim Index of used car prices. When used vehicle prices fall, as they did from 2000-03, the manufacturers of new vehicles have to shave prices by increasing givebacks to the customers. That's not good for margins. So if you want to understand the car companies, you have to watch used car prices.

2. Better, longer-lasting vehicles increase the duration of the car fleet and undermine new-car sales.

3. Rising fleet purchases, as we had in 2004, are bad for car companies. It's like being kidnapped by Kmart . The car companies use fleet sales to stuff unwanted inventory down the market's throat. Think of it as trade-loading. That's why they own car rental companies. When they ship a car they get to count it as a sale and book the standard margin. This helps with earnings, which the car companies must have to feed the twin beasts that are killing them: retired worker pension benefits and retired worker health care costs.

4. Car companies have so many retired workers relative to current workers that the duration of their past liabilities exceeds the duration of their future GDP growth. You can think of their health care and pension contract obligations as fossils representing relative prices and other parameters from the period when they wrote the contracts. I like to call them "shadow prices." For example, the 7% average GDP growth from history implies that GM will have difficulty meeting its obligations whenever revenue and cash flow growth in the future fall short of that number. The deflation scare last year was especially difficult for old-economy companies like GM. The recovery since then has been a blessing, but is it too late? My bet is that we will see 4% real growth and 2% inflation -- i.e., 6% GDP growth -- for some time.

5. Investors should be careful to dig through business plans in order to discover how much of the future free cash flow they are counting on to provide shareholder value is coming from places that are inherently risky. GM, like many other U.S. companies, is making a big bet on China. I believe investors have more China exposure in their portfolios than they realize today.

6. We need to note that the bad news about GM does not represent bad news for the overall economy. Asian growth is straining the world's ability to supply materials (steel), energy (oil) and capital to support the growth. All are becoming more expensive. That means our growth will increasingly be in service companies, which use less of all three. GDP this year will be about 4%. It is driven by productivity growth, not car sales.

Moral of the story: If you want to be a successful car company, you have to make really good cars. The overall economy is growing just fine.