• The Pittsburgh Pirates will win the World Series in 2016.
• Warren Buffett will retire in 2020.
• Chrysler will be profitable in 2012.

The first two predictions are made up. The last one is serious.

In documents filed with the bankruptcy court, Chrysler said that it can achieve profitability in 2012. The Wall Street Journal reports (subscription required) that "The company lost $16.8 billion in 2008 and expects to lose $4.7 billion this year. . . An income statement submitted by the auto maker to the court shows the company's forecast for earning about $100 million in 2012 and $1.6 billion in 2013. It sees its net income climbing to $3 billion by 2016."

So not only are they predicting profitability in 2012, but they're actually putting a number on how much they'll make in 2016 (the same year the Pirates will win the World Series).

Isn't 2012 -- nevermind 2016 -- a little far off to be making predictions? Aren't there an infinite number of variables that could -- and likely will -- make that number wildly wrong in either direction? Market-share, raw materials costs, economic conditions, etc. etc. etc. Most companies can't predict their future earnings two years out with reasonable precision, but a company with as much uncertainty as Chrysler making predictions about 2016?

I wouldn't invest any of my money based on those predictions. Oh wait. I already am. We all are. The United States will receive an 8% stake in Chrysler as part of the restructuring.