What is a 91 NSX worth?

[bud777]

I took data from all over the web and fitted this curve of price based on mileage. It does not take into account condition, modifications or greed. I thought it might be useful as a starting point. It looks like, after 60,000 miles, there is not much penalty for driving it.

Price = 99512 -1.124* Milage +.000005*Mileage^2

 

[keynes]

I took data from all over the web and fitted this curve of price based on mileage. It does not take into account condition, modifications or greed. I thought it might be useful as a starting point. It looks like, after 60,000 miles, there is not much penalty for driving it.

Price = 99512 -1.124* Milage +.000005*Mileage^2

Hi

this post analyses sales by a number of variables. I've not done the maths but I would interested to see how the results from the two equations compare

http://www.nsxprime.com/forum/showthread.php/199875-Gen1-NSX-Current-Market-Analysis

 

[RacerXwing]

Good to see some more data analysis here! It is my passion to understand the pricing of NSXs. To that end, I created the database and the analysis to which keynes links. I currently have 459 1991 NSX data points with mileage all the way up to 340,000. Some points I have are as follows:

1. As you allude, "greed" is a prominent unpredictable factor in NSX pricing because a dealer or individual can list an NSX for whatever whim strikes their fancy. It is valuable to keep your eye on eBay and Bring a Trailer.com results to get a better idea of the "true" market for the NSX. Even then, there are many other factors that effect the price like color, modifications, maintenance, accident history... this is a complicated game.
2. No model will be perfect, but care should be taken in using a quadratic equation (with a "miles^2" term) when modeling price vs. miles. In your equation, the price starts rising again after around 100,000 miles and as much as some of the higher mileage members around here wish this to be true, that just isn't the case in real life. An alternative could be to use a cubic model; but I think the best model for the NSX prices is a logarithmic model. This model has the advantages of quantifying the sharp rise in the prices of lower mileage NSXs and the fact that the prices always decrease at a decreasing rate as mileage increases. This means that the more miles an NSX has to start with, there is less of a drop in value if more miles are added but also less value to begin with.

Lastly, I hope I don't scare you off! I always enjoy posts with NSX data.

Hi

this post analyses sales by a number of variables. I've not done the maths but I would interested to see how the results from the two equations compare

http://www.nsxprime.com/forum/showthread.php/199875-Gen1-NSX-Current-Market-Analysis

Thank you for the referral! I plotted up the two models to see how they compare:


You can see here the difference between a quadratic and logarithmic model. To get a better idea of how the models compare, here is the graph where the range is restricted to the quadratic model's decreasing region:



Looking at the past 365 days of my 1991 data, it can be concluded that the two models start and end at approximately the same prices, at 5,000 and 100,000 miles, but have dramatically different behavior in between. Bud, I am curious to know the R^2 value of your trendline.