It is a tough practical decision — when do you replace a car versus repairing it?? There isn’t an easy answer, but the intelligent way to approach it is to estimate the present value of the cost of continually repairing the old car versus the cost of buying a new one, net of the sales proceeds of the old car.

It’s a tough decision, with many squishy variables.? Nonetheless, it is the right way to frame the question.? It is also the right paradigm for a number of other questions.

• When will the US give up on bailing out AIG (or any other firm)?
• When will the US Government default, or decide to inflate massively?
• When will China and the Arabs decide that “sunk costs are sunk” and abandon the US and the US Dollar?

After investing in an investment that proves to be bad, the question crops up, “Should I buy more?? If I liked it at 25, don’t I love it at 15?”? The same thing with govenment decisions, except that they are writ larger.

Eventually, there is a tipping point where the investor realizes that things are so bad, that it is better not to invest any more, because it would be throwing good money after bad.? Replace the car, repairing it will cost too much.

In this uncertain environment, that is what we are facing now:

• What will the US do with marginal firms?? AIG or Lehman treatment?…
• When would the US give up on issuing more debt?
• When will foreigners give up on buying US debt?

None of these are necessarily second quarter 2009 issues.? Neither is a car replacement; I can just repair it for now.

My advice is to start thing ahead, and ask when parties that you rely on are likely abandon ship.? Then change your investment processes to avoid the likely path of disaster.? This is messy, but it is the best way to operate.