How we misprice used cars | Hacker News Reader

How we misprice used cars

priceonomics.com

In Robert Cialdini's Influence, he describes a scheme his brother used to pay his way through college. Basically, he'd buy a car at the bottom of its fair blue book range, clean it up, and then place an ad for it at the top of its price range, with a well-worded ad, asking people to call for an appointment. The sneaky bit is that he'd then schedule all of the interested buyers for exactly the same time. They'd arrive minutes apart, but he'd only let them look at it on a first-come-first-serve basis, making the late-comers stand and wait. The pressure created by the competitive atmosphere would always ensure a sale.

If you combine this with the information in this article, you could potentially never pay for a car (or even major maintenance) again, and even make a profit in the long term.

1: Find a popular and reliable model of car that doesn't require any standard maintenance in the 25k-45k mile range other than oil changes.

2: Write some generic ad copy for this car. Make sure to research this and make sure you're advertising as effectively as possible.

4: Buy this car used, but in good condition, near the bottom of its price range and just over 30k miles.

5: When you hit 39k miles (if you're driving an average of 60 miles/day, this will be about 5 months), sell it using the scheme above. You've already got the ad copy; just tweak it to fit the specific car you have.

6: Goto 4.

This scheme will probably work, and make you some money. In general, there's a word for doing work for money, it's called a job.

That's true, but for the average person, it would be worthwhile, even with zero profit.

According to my cursory internet research, the average US household spends 1.5% of their annual income on car repairs[0]. The average adult works 1,300 hours per year[1]. The average household has 2.1 adults[2]. So in a five month period, the average household puts about 17 hours into car maintenance.

So if you live in an average household, you'll come out on top with this scheme as long as you spend less than 17 hours on the process each time.

[0] http://auto.howstuffworks.com/under-the-hood/cost-of-car-own...

[1] http://www.zerohedge.com/news/2013-05-13/how-average-america...

[2] http://www.popcouncil.us/pdfs/wp/144.pdf

the only problem is - tax, which eats into yours, already small, profits; and if you continue to do this through the year - you definitely need to get a used car license with all prerequisites (i.e. parking lot, bond and bunch of other stuff which do not comes free)

above assuming you want to stay legal.

This seems to vary by state. In Colorado, for example, a dealer is someone who sells more than three cars per year with profit in mind. So as long as you're keeping each car for more than 4 months, it seems you'd be fine in Colorado.

Georgia, on the other hand, defines it simply as selling for profit. So to comply with the law, you wouldn't be able to mark up the cars at all.

No one sells cars at 30k anymore. They're either coming off lease, typically at 36k, or they're financed for 5 years or longer which means that they're more likely to be at least 60k before the owner even thinks of selling.

First of all, that's clearly refuted by the article itself. If nobody sold cars at 30k, they wouldn't have data for that mileage range.

Secondly, you can see that it is untrue by going to any used car site and filtering by mileage. There are tons of cars available right at 30k miles.

On maintenance intervals: If all cars really did get scheduled maintenance done on time, then cars that have just over X0,000 miles should be worth more because they just had this expensive maintenance done.

That is a car with 51,000 miles will just have had its 50,000 mile maintenance done while a car with 49,000 miles will not have.

So scheduled maintenance cannot explain these jumps, since it would predict jumps in the opposite direction.

The bias discussed in the original article is pretty well documented. It also has nothing to do with the other inefficiencies in used car pricing that we've been discussing here :-) Of course, there's also nothing automatic about getting schedules maintenance done when the odometer rolls over to some mileage that ends in 000.

I'm aware of the original articles point, my post was of the form "even when you account for X, their analysis is still valid". And neither did I mean to imply that people actually get maintenance done exactly on schedule.

A more relevant question would be how much dead weight loss this bias results in. I'm guessing not that much.

Cars often have specific expensive maintenance at specific intervals. My car is worth maybe $2k more if I HAVE completed the timing belt replacement at 115k than it was at 114k before completing it. Warranty expiration also might influence value -- I'd far prefer a car with 1000 miles left on a comprehensive warranty, to know I could deal with undisclosed issues, than just expired from warranty.

In general, there are just a lot of information asymmetries associated with buying used cars that make it difficult for a seller whose high mileage car has been well-maintained and is running well to get what is perhaps objectively the "right" value for a car. In your example, you'd be crazy to do your scheduled maintenance at 115K if you were planning on selling your car in the near future. You'd never recoup that $2K you spent.

The reality is that this well-maintained car that doesn't have problems (and/or which has cosmetic problems that you've learned to accept) is probably going to be worth more to you than to a random buyer who just sees "high-mileage car" (that they're only looking to buy because they need a car and don't have much money).

Some of the information asymmetries can be mitigated by buying from a dealer with an add-on warranty but then the prices are such that they often don't look like a particularly good deal relative to a new car.

It would seem a pretty rare situation that a new car would be financially a "good deal" compared to a used car at a dealer, regardless of warranty or financing differences.

Most new cars are bought for psychological reasons, not financial reasons, IMO. (Nothing wrong with that, as long as you aren't confused into thinking you're "saving money" buying new.)

Need a "new" car and want to save money? Buy a late model (3-5 years old) car with higher than average miles on it. It will be cheaper than the average used car, because people are inordinately "afraid" of miles on cars. Late model, as it will still "look kind of new" (so scratching that itch), will likely have recent safety advances, and in areas of the country that salt the roads, it will have fewer years of salt exposure than a similar miles at an average rate used car. High miles in 3-5 years probably got a lot of highway use and surely got a lot of regular use. Those miles aren't as hard on cars as the "creampuff", low miles, sat in traffic for hours each day cars.

Once you've bought it, find a good independent mechanic for that brand (or a general independent). Will be way cheaper than the dealership, and in my experience, just as good. I do all my own maintenance (ex-tires and ex-exhaust), but a good indy will save you a bunch of money as well.

My car ('98 Mercedes E300) just flipped over 210,000 miles on the way into work today. It will be retired in 1-3 more years due to rust, not mechanical malady. I paid about $6000 about 4.5 years ago, and I've got an additional $650-ish worth of parts into it over the 4 years. Brakes will be good for the duration, and if it lasts 3 more years, I'll need tires, but if I retire it next year, it won't. That'll mean more than 5 years of service for $5000 in depreciation and out-of-pocket maintenance expenses, as I'll be able to sell it for $1500 or so running but rusty.

That's a little bit of an extreme case as I do all my own work, so the $650 number would be more than double that (glowplugs, brakes all around, and two window regulators would be the labor needed), but I've driven a pretty decent-looking, mid-sized diesel Mercedes for 4.5 years and over 30,000 miles for under $100 per month in purchase and maintenance costs. I paid almost as much in diesel as I paid for the car over its life. (I get just under 30 mpg, and diesel is a hair under $4/gal.)

Buying the new equivalent of my car? $51,400 base MSRP. http://www.mbusa.com/mercedes/vehicles/model/class-E/model-E...

Drive that 4 years and 30,000 miles and it's likely worth $30K, or a per-month cost of around $500/mo, five times as much. (Granted, that's for a brand-new car, which has some appeal, even to me, but it's not worth 5x as much, IMO.)

Well, much depends on exactly what sort of car you're buying and when you're buying it, too.

I bought my first car in 2006. It was a GM product, about a year old, and with about 12,000 miles on it. I paid just over half of what it would have cost to buy it new. That was a good deal.

I bought a new car this past summer from Toyota. I would have saved a couple of thousand dollars going used, which just wasn't worth it to me compared to getting something with a known provenance. The price difference wasn't very big, in any case.

Between the brand (GMs depreciate like mad, Toyotas not so much) and the economic climate (I get the impression that Cash for Clunkers took a lot of used cars off the road, and demand for used cars is higher in a weaker economy anyway), the difference between the two cases was massive.

I bought a Lincoln LS in Dubai, UAE for $22100 from the ford dealer with 300 miles on it in 2007 (2006 MY); it was $45k list price. There were 8 in country and it had just been cancelled. Extra bonus is I found a $1k satellite phone under the seat.

Cash for Clunkers wasn't so big a program (regardless of its merits), and the mpg requirement means that it didn't hit sedans and small cars that much.

I found the same thing with Toyotas when I was shopping for a new SUV a few years back. I was open to buying used but the savings I saw just weren't worth it to me relative to buying new--especially given that the dealer came down to pretty much invoice.

Ref: parent post. The situation if you do your own maintenance and/or have a good independent mechanic for your brand can certainly change the equation. For example, I've been told that high-end car (Mercedes, BMW, etc.) maintenance, including scheduled maintenance, gets really expensive after the warranty period so you probably don't want to buy a used one if your idea of dealing with a car is dropping it off at the dealer as needed. It's also a question of how tolerant you are of dealing with car repairs generally as opposed to just owning something that gets you reliably and comfortably from Point A to B with a minimum of service overhead.

VW diesels are another good example - diesels in general hold up a lot better.

A Golf TDI off the lot at $25k brand new..and the dealers ask for and get 20k for ones with 25-40k on them.

Agreed 100% on the diesel longevity and overall lower maintenance. I'd happily buy a 25k or 40k mile 2 year-old diesel for 20% off vs a new one for MSRP. Even moreso if 6+ months of factory warranty was left. That way, you can get anything broken fixed by the dealer, and you're only only the hook for wear items.

Even though it seems small, that $5K saving is real and would pay me for all the maintenance the car would need for the next 100k+ miles. Even if one went to an Indy mechanic, that would cover all the maintenance for a long time.

If you finance a car, you can often get a much better rate on a new car than on a used car, which may be a factor toward getting a new car.

Definitely true, but it's a small and not enough to make up for a 20% reduction in purchase price.

I'm seeing 0.49% on new cars and 1.49% on used cars from USAA, and several other banks/CUs are all showing 1.99% or better rates. Even if you got 0% on a new car as an incentive and paid 2% on a used car loan, over a 5-6 year loan, you'd pay about 7% more with the loan vs the "free" financing on the new car. (Nevermind that a 5-6 year car loan is insanely long, IMO; buy a car you can afford for cash or on a 3-4 year note. With money at 2%, that's less important than it used to be, of course.)

There are also programs where a captive (or semi-captive) credit agency will finance people who couldn't qualify for a used car loan. It frustrates me when I hear car ads touting exactly this, which I view as borderline predatory.

"With my credit, I couldn't afford a car, but FooBar brand got me into a new car for only $95 a week!" I believe that such practices should be legal (because you don't generally make someone better off by legislating away their choice), but it's not helpful in the majority of cases. If it's the difference in being able to keep your job vs not, it's helpful. If it's the difference between being able to brag to your friends that you're rolling a 2014 Camry instead of a 2005 Corolla that you could actually afford, well neither they nor you are doing you any favors...

I'm ok buying a Tesla on as long a note as I can get; I expect to keep it in-service for 10 years, and operating cost savings (and tax credits, etc.) will make it worth it ($100k for a car is a stretch for me right now, so 6y vs. 4y would be a win).

Depreciation still makes my current car a better deal, but if my car suddenly died, I'd probably replace with a Model S (maybe after a period of time on zipcar/drivenow). Otherwise waiting on the rumored S AWD in 2015.

Yep, that's true and made it worth it for me (plus VW throws in the first 30k of maintaince, which saves another $2k or so)

It's somewhat luck of the draw of course. But with a new car (especially these days) you're getting a pretty well-defined quantity with a good warranty. The one fairly low mileage used car I bought from a dealer ended up having more than its fair share of expensive maintenance; I should probably have sold it sooner than I did as well. And last time I bought a new vehicle, I looked at used but nothing seemed like much of a deal.

The fact that I drive vehicles a relatively modest number of miles a year and keep them until they're basically not reliable transportation any longer may color my experiences as well. I just want to buy something and have it run with the minimum fuss and bother for as long as possible.

except that there is NO discontinuity at the 115K mark nor any X5K mark.

It's also a matter of scheduled maintenance: If you buy a car right before a bunch of critical components are replaced, expect a repair bill of $500-1K at the appropriate time. Although a lot of cars at say 60,100 miles will probably not have their 60K done - at which point you can use that as a bargaining point. If it's a high-activity auction, it might be hard to get the exact details on each car for sale

I have found this can be pronounced in RVs as well. Totally trashed RV with < 10K miles selling for much more than a expertly maintained one with 50K miles.

That's not the point. What the article is talking about only applies to the difference between vehicles at the bottom of a 10k mile bracket, and vehicles at the top of the same bracket. It's independent of maintenance, and doesn't have anything to do with cars differing by 40k miles.

I would love the priceonomics guys to do the same survey on RVs. My experience shopping for RVs is that RVs have a price co-efficient that is strongly correlated with "mileage" and poorly correlated with "condition." This was exactly their point on cars when a 79,900 mile car was $210 more expensive than a 80,100 mile car. I read the article to say that people use mileage rather than condition to price cars.

So far the closest I've seen on that topic is this entry http://blog.priceonomics.com/post/32944888191/living-in-a-va... which talked about the Westfalia RVs.

Reply on news.ycombinator.com