is estimating the value of a used car or motorcycle a matter of labor intensive study or of luck?

Factual Questions
21

astro, your claim about inspections not being able to give us any real additional value beyond what we already know may be a valid one. That’s why I specified this case in my OP as one of the options - this is the estimating the roulette case where no matter how we estimate a fundamentally random process we still cannot get any improvement.

Whereas your claim about the prohibitive cost of inspections goes against the spirit of the OP. If inspections for certain key material factors indicative of car quality are costly now, maybe tomorrow we will invent much cheaper ways of doing them. (If ditches are expensive now to dig by hand, maybe tomorrow we will invent bulldozer). But nobody is going to invent anything if we don’t even bother figuring out just what would be useful to have in the first place. Discovering that “inspecting X would have been a wonderful improvement in estimating car reliability, but too bad that it costs $5K per car” is already quite a discovery because once we know that, we can invest in finding cheaper and better ways of doing X.

22

Or we can observe that there’s really not much need for X, and that the likelihood of finding cheaper/better ways of doing it is slim to none, making the appeal of investing in same pretty much nonexistent.

23

Missed the edit window:

ETA: Meaning, if you really believe there is something that can be feasibly achieved in this area, go for it, with my blessing. But note that just desiring it doesn’t mean it’s attainable, and don’t be surprised if others have concluded that it’s not worth pursuing.

24

Gary T, you have said above

then you said that doing that is expensive so that’s why people don’t do that. If we came up with a way to do that for $50, would it be worth doing to improve the car valuation process? What do you think is the “added value” of this procedure?

Are there other such procedures which have not yet been mentioned here and which could give us meaningful added value, if we could magically do them a lot cheaper than what the state of the art is used to?

25

Any invention that reduces time and costs in these areas will just be incorporated by the manufacturers on the production line. So the process on an individual basis, inspection of a used car, will still be substantially more expensive than the original manufacturing costs of that vehicle or subsystem.

The neat invention will drive the cost of new cars down and you will still be stuck with a cost-benefit analysis as to whether inspection and how much is has any value.

26

:dubious: and that would be a bad thing how? So we will get a car in which you can read off the state of wear of the key components from the built-in sensor for free. Great, that would presumably allow for more accurate used car valuation, right?

27

And if it drove the price of a new car up by $1000 it would never get built. How much is it worth to you to pay on a new car when you won’t be the one benefiting from the feature over your ownership of the car?

28

Don’t think so. For starters, most of the things you would be interested in knowing the state of wear of preclude physical sensors being in place. You can’t have a physical device in place to measure the plain bearing journal surfaces on a crankshaft for instances. Could some indirect sensing be used? Perhaps, but then you would get into probability analysis or what the measured result meant. So much chance this problem, so much chance that problem.

And what would you gain with this knowledge? How much do you think it would affect pricing?

29

Hbns, first you tell me that the manufacturer will incorporate the hypothetical invention into the car, ruining my hypothetical evaluation business. So I say, yay! easier to evaluate cars for all. Now you are telling me, nah, wouldn’t happen and cannot be done, so no need to worry about the manufacturer.

Make up your mind, man :confused:

30

Actually, he said:

(Emphasis mine)

That suggests, to me, that he thinks it’d be used as inspection on the production line (to keep “lemons” from making it out the plant door), not as sensors to be placed within each car.

31

maybe. But my impression from this thread so far has been that “lemons” are not sold by the factory but rather created by owners who abuse their car.

Is there in fact also a whole category of fresh from assembly line lemons?

32

Oh, yes. Some cars just aren’t right from the start (though my impression is that this doesn’t happen as often now as it used to). My father, and my father-in-law, both bought cars which were constantly in the repair shop, from the very start. Even if most (if not all) of the repairs are covered by the new-car warranty, it’s still a pain (and the problems will likely continue once it’s out of warranty). I believe that many states now have “lemon laws” which cover cases like these.

But, I think you’re right; most “bad cars” on the used car market are likely the result of what has happened to them since they were new, rather than problems with the initial build.

Even so, his point was that

33

For five bucks, sure, go ahead and do it. For five hundred, nah, not cost-effective. For fifty, maybe but probably not. With a car you really want but have reason to suspect may have that problem, it might be wise and worth it for the peace of mind (or for the avoidance of having to do the repair). As a routine procedure in most cases, I would say no.

A comparable example would be a compression test. (The cost can range from ~50 to a couple hundred or so, depending on the particular engine.) In theory, it sounds like a great thing to do when considering buying a used car. Fifty years ago it was probably a sensible thing to do. In practice with modern cars, though, the reality is that there’s hardly ever (as in I personally know of none) a situation where an engine that runs nicely has a problem that will cost a lot to repair and is only detectable by this test. If it runs right, it’s a 1 in 10,000 chance that a compression test will reveal a hidden problem. There are better ways to spend the money.

34

Let me restate the OPs question:

a) Could a sufficiently intensive mechanical inspection undo the lemons problem, if that inspection was provided for free?

b) Could a sufficiently intensive mechanical inspection undo the lemons problem, assuming this time that the buyer and seller paid for the work somehow and negotiations were perfect and costless?

Have I got that correct?

A preliminary step might be to estimate the magnitude of the lemon’s effect. I would guess that its lower than it was in 1972 (the approximate time of Akerloff’s paper), due to improvements in automotive reliability. In other words, the downside risk of buying a new car is lower, the possibility of owner abuse notwithstanding.

Finally, I’ll note that lemons problem relies on the seller knowing more about the car than the buyer, and the fact that lemons are more likely to be sold than reliability-typical cars. (So if the buyer and seller know each other well, the buyer should be willing to pay above the blue book value assuming that he knows it is not in fact a lemon.)

35

As a few others have pointed out, this question is really sort of imaginary-world-hypothetical, as if the price of a vehicle were somehow intrinsically connected to it’s value.

It isn’t.

Vehicles, like everything else, are worth what people are willing to pay for them. You could have a car that was ready to fall apart, but was beautiful and desirable, and it would be worth far more than my friend’s 92 Honda which will probably run for another 100k miles.

All items up for sale work this way. When I’m in the supermarket, I often try to evaluate how much the cereal or other product is “worth” by weight or caloric “value”. I’m kidding myself. Cereals are priced according to desirability, and if the seller could charge more, they would.

What I do like about this question is that an individual is defining their OWN set of values in order to determine if the art of pricing a vehicle matches those values. My answer is that it doesn’t. Not remotely.

Think about the SUV craze. 20k, 30k, 40k for vehicles that hadn’t been through any of the safety testing normally required for a passenger car, that were basically just big engines bolted to frames. And people loved them, and they were priced according to that love, even on the resale market. Will $35k they spent make the vehicle run for the next twenty years? Nope. SUVs are an extreme example, but the same principle applies to everything we buy.

I share the OPs value, though: to me, a good car serves it’s primary function, and “value” indicates that it will do so with a minimal of maintenance. So we should ignore price, really, and read between the lines.

Best advice? Find out the history of the engine in the car you are interested. Then, find out the “overall history”, such as CR’s reviews of maintenance needs of cars with engines that have the same pedigree.

36

yes, you have captured some of the spirit of this question. But my motivation here is broader than just “evaluate car during a sale”. It could also be “evaluate car when”:

  • writing a warranty on a used car
  • writing an insurance policy for it
  • planning “self-insurance” expenditure on my fleet of rental cars, which really is just isomorphic to the insurance case
  • imposing penalties on the guy who rented a car from me and then messed it up by poor handling

In short, my focus is on evaluation of reality and not on the way this reality may or not be captured in our financial proceedings of various sorts.

So far, it seems that the collective opinion of this thread boils down to “cars today are so great that there will be very little extra benefit from mechanical inspections, since car behavior can be predicted well enough from its make, mileage and brief inspection”. Well, great. Good to know.

37
  • writing a warranty on a used car - lemons problem does not apply, if buyer knows his own car.
  • writing an insurance policy for it - lemons problem does apply, since insurance company has inferior information, relative to the seller.
  • planning “self-insurance” expenditure on my fleet of rental cars, which really is just isomorphic to the insurance case - lemons problem does not apply.
  • imposing penalties on the guy who rented a car from me and then messed it up by poor handling - lemons problem does apply, as does moral hazard.

Either that, or mechanical inspections are so expensive/unreliable that they won’t help.