Method for analyzing car purchase...

[catccc]

We are in the market for a replacement car.  DH and I agreed we'd replace our 2005 toyota matrix when it hits 200K miles.  We are at 199,200 right now, so that will be soon.  The idea is we don't want to drive it until it dies because we do not want to be in a position where we "need" a car right away and can't take our time to find the right deal.  (Unfortunately we both need vehicles; public transport is not an option for our commutes.)  Of course I don't expect the thing to blow up when the mileage rolls over to 200K, but it's about time to start looking, I think.  (Or not?  Argue that, too, if you think we should continue to hold off...)

But what this post is really about...  How I'm analyzing the cost of the car.  I'm looking at cars on the basis of cost per mile, and assuming we drive this next car to 200K.  (Or should I be selling it sooner when there is more residual value?!)  But I'm wondering if that is over simplified because the early miles are likely to be trouble free, and the later miles will require repairs.  So should I be adding on some extra cost for expected repairs and maintenance to the price of the vehicle before calculating the cost per mile?  This may make the used cars look pricier per mile than the new cars.  I also thought we'd save a lot buying a car that is a couple years old, but a lot of the models we are looking at have good resale values and the slightly used versions don't result in any savings per mile over buying new.  I'm not trying to manipulate the numbers so a shiny new car is the "right" decision.  I just want to be as objective as possible.  Any thoughts?

Last note- we'll pay cash unless we can get an APR that makes leaving our $ in a HYS account mathematically better (so a loan interest rate lower than the savings rate after taxes).

[sokoloff]

If your car has been fairly reliable, is in good shape, and doesn't have a rust problem, I'd strongly consider just planning to drive it until it hits 250K miles if you need a "target".

In terms of it dying suddenly and needing a replacement car, rental cars are reasonably available in most markets, and the money you bank by delaying the decision for another 50K miles will more than pay for even a month of car rental if that's needed.

In terms of your analysis, I'd look at discounted cash flow for the car over the expected ownership period (so not a straight cash per mile, but a discounted cashflow per year/mile). This will appropriately penalize a larger upfront expense (that could otherwise stay invested) of a new car. Use a discount rate that's approx the long-run market returns (I'd use 9% if using excel and 10% if doing the math on a napkin.)

[catccc]

Got it, so DCF will account for $ that would have been invested instead of dumped into the car.  That makes sense.

Good point about the possibility of renting a car in case it dies.  But I guess another thing I need to consider here is the value of a vehicle that is running v. the value of a vehicle that is not worth repairing.  At 200K miles & running well, I think I could sell it for $2-$3K.  It is in good shape.  We get the oil changed as recommended, it has had it's share of issues small issues, which we have repaired (replaced O2 sensor, maf sensor, charcoal canister assembly) and some regular things that just wear out on a car that old (clutch replaced and flywheel resurfaced, new shocks and struts).  Not sure about rust.  I am in the mid-atlantic/NE US, so there's that.