Incredible research OP. Thank you for sharing that with the community.
-
Protip: Profile posts are public! Use Conversations to message other members privately. Everyone can see the content of a profile post.
***STICKY*** Gen1 NSX Current Market Analysis
check out Haggarty's evaluation tool - even higher values!
https://www.hagerty.com/apps/valuationtools/1996-Acura-NSX?id=116228
https://www.hagerty.com/apps/valuationtools/1996-Acura-NSX?id=116228
Part 2: Appreciation
First, I must post my appreciation for everyone here; thanks again for your continued support. Now we're going to look at a different kind of appreciation...
I mentioned in the original post that I wanted to improve my analysis on the effect of time on NSX prices. In that post, I looked at the average and median prices of all NSXs vs time. Those two charts are recreated here, with fancy new 6-month moving average trend lines in red.
All that can be concluded is that yes, "prices are going up". But everyone already knows that, so I wanted to find some new information.
Some limitations of this first analysis are:
So, here's what I did to improve it:
This new analysis gives a much more powerful look at the effect of time on prices. An important note is that NA2 Coupes and Zanardis were excluded from the analysis, as the scarcity of data points would not result in any meaningful trends.
Without further ado, here are the charts:
I'll talk you through the first one (NA1 Coupe). These are all graphs of Price vs Time, with the different colors of data coming from the different mileage bins. In the NA1 Coupe graph, these are <15000 miles, 15000 - 30000 miles, 30000 - 45000 miles, 45000 - 65000 miles, 65000 - 90000 miles, 90000 - 120000 miles, and >120000 miles. My choice in mileage bins was somewhat arbitrary; my general goal was to have enough data points in a bin to show a meaningful trend and there was no exact science to it. Each mileage bin has a linear trendline fit; this assumes the prices are increasing at a constant amount per day; the slope of the line. This slope is the number next to the * x in the equation, i.e. NA1 Manual Coupes with <15000 miles are appreciating at a rate of $17.88 per day. I will call the slope the "appreciation rate". You can see for yourself the variety of appreciation rates.
Some limitations and disclaimers before the conclusions:
Conclusions over the period of 6/1/2013 to present:
As always, I hope you find this information valuable and I am happy to answer any questions.
First, I must post my appreciation for everyone here; thanks again for your continued support. Now we're going to look at a different kind of appreciation...
I mentioned in the original post that I wanted to improve my analysis on the effect of time on NSX prices. In that post, I looked at the average and median prices of all NSXs vs time. Those two charts are recreated here, with fancy new 6-month moving average trend lines in red.
All that can be concluded is that yes, "prices are going up". But everyone already knows that, so I wanted to find some new information.
Some limitations of this first analysis are:
- Mileage is not considered
- Type (NA1 or NA2, Targa or Coupe...) is not considered
- Clean / salvage title is not considered
- Degree of modification is not considered
So, here's what I did to improve it:
- Break the NSXs down by type (only manual transmission NSXs considered)
- Break each NSX type into "mileage bins", i.e. all cars with under 20k miles
- Remove all special factors; only consider clean title NSXs
- Remove all highly modified NSXs, only consider stock and lightly modded NSXs
This new analysis gives a much more powerful look at the effect of time on prices. An important note is that NA2 Coupes and Zanardis were excluded from the analysis, as the scarcity of data points would not result in any meaningful trends.
Without further ado, here are the charts:
I'll talk you through the first one (NA1 Coupe). These are all graphs of Price vs Time, with the different colors of data coming from the different mileage bins. In the NA1 Coupe graph, these are <15000 miles, 15000 - 30000 miles, 30000 - 45000 miles, 45000 - 65000 miles, 65000 - 90000 miles, 90000 - 120000 miles, and >120000 miles. My choice in mileage bins was somewhat arbitrary; my general goal was to have enough data points in a bin to show a meaningful trend and there was no exact science to it. Each mileage bin has a linear trendline fit; this assumes the prices are increasing at a constant amount per day; the slope of the line. This slope is the number next to the * x in the equation, i.e. NA1 Manual Coupes with <15000 miles are appreciating at a rate of $17.88 per day. I will call the slope the "appreciation rate". You can see for yourself the variety of appreciation rates.
Some limitations and disclaimers before the conclusions:
- List prices: since the majority of prices here are list prices, the trends here may not be representative of actual market behavior.
- For example, two 02+ NSXs, each with nearly 50k miles, were listed in the past 7 days for over $100k. This is not in line with the pricing I would expect and as such cars like these may skew the data.
- The shear amount of data points, especially for the NA1 Coupes, is hopefully enough to mediate the damaging effect of outliers like these two NSXs.
- Assumption that prices are increasing linearly each day is not representative of actual market behavior.
- This analysis is simply a report of trends from my dataset and may be completely inaccurate compared to the actual market behavior.
Conclusions over the period of 6/1/2013 to present:
- Within an NSX type, as miles decreases, the appreciation rate generally increases.
- The NSXs with the fewest miles generally have the highest appreciation rate.
- NA2 Targa (Facelift) NSXs have the highest appreciation rate.
- NA1 Targa NSXs have the lowest appreciation rate.
- All clean title, low modification NSXs have positive appreciation rates, regardless of miles and type.
As always, I hope you find this information valuable and I am happy to answer any questions.
Last edited:
Amazing work, thank you! I'm in the market now and this is extremely helpful. I recently sold my S2000 and used the one on s2ki as a pricing guide as well.
2017 Summary
I have not given an update in a while, but I have not been sitting idle. It is hard to believe, but the number of NSXs in my database has more than doubled since I first made the original post here (current NSX count is nearly 2200, compared to nearly 900 originally). This was made possible through diligently recording new listings as well as traversing further back in time through Prime and an eBay sale archive to the beginning of 2012.
Here are some changes I made since the original post:
.
New conclusions:
What happened in 2017??
.
Summary:
I have not given an update in a while, but I have not been sitting idle. It is hard to believe, but the number of NSXs in my database has more than doubled since I first made the original post here (current NSX count is nearly 2200, compared to nearly 900 originally). This was made possible through diligently recording new listings as well as traversing further back in time through Prime and an eBay sale archive to the beginning of 2012.
Here are some changes I made since the original post:
.
Structure change to the price models:
.
.
Improved the way time effects the price equation, now a new equation is created for each NSX group (1991, 1992, 1993-4, 1995-6, 1997-9, 2000-1, 2002-5) each year.
Where possible, created separate models with only listing prices and only actual sale prices. Much of the original post is devoted to the struggle and reservations of making conclusions based on listing prices rather than actual sale prices. Creating model(s) that are based solely on actual sale prices removes those reservations.
.
.
Improved calculated effects of special factors:
.
Now special factors are based on actual sale prices
.
Now special factors are based on actual sale prices
.
Added maintenance as a factor:
.
<tbody>
</tbody>
| -2 | Major maintenance is needed, ex. timing belt & clutch & ABS & coolant hoses |
| -1 | Some maintenance is due, ex. timing belt or clutch or ABS or coolant hoses |
| -- | Unknown or minimal maintenance |
| 1 | Some mention of maintenance, one major item complete, ex. timing belt |
| 2 | No or minimal maintenance needs, multiple major items complete |
| 3 | Major/regular maintenance by prestigious NSX service name (the Whisperer, etc.) |
<tbody>
</tbody>
.
Only about 100 NSXs have this information, so conclusions are forthcoming. So far, both -2 and 3 are both showing strong effects on listing prices.
.
.
New conclusions:
.
The rule of thumb for actual sale prices.
Across all years, types, and colors of NSXs, through out all of time, I observed a trend. When compared to listing prices, the actual sale price of NSXs averaged out to be 90% of the predicted listing price. This means if you want a rough idea of what a listed NSX is worth, take its listed price and multiply by 0.9.
.
Definitive effects on price:
Based on only actual sale data. Each effect was calculated by filtering the data to show only actual sale prices of one factor at a time (i.e. when calculating the effect of an automatic transmission, I removed the cars with both automatic and a salvage title)
.
<tbody>
</tbody>
| Factor | Value Effect |
| Automatic Transmission | -10% |
| Salvage Title | -21% |
| Accident history (but no salvage title) | -4% |
<tbody>
</tbody>
What happened in 2017??
.
I saw a fair amount of fearful comments throughout the year about the state of the NSX market . To the effect that values are going down, Senna forbid, and the sky is falling. In some instances, these comments were not entirely without merit. In particular, BringaTrailer.com saw some lower-than-expected sale prices and several NSXs there did not meet reserve. Well, I'm here to provide a voice of reason... with data! We're going to look at the actual sale prices for each class of clean title, low modified, manual transmission NSX. These graphs show price vs. miles; the different colors indicate the different years of data. 2012 is the blue line and points, 2013 is red, 2014 is yellow, 2015 is green, and the two years we'll be looking at, 2016 and 2017, are purple and aqua. A higher trendline indicates that NSXs with the same mileage are selling for more money, i.e. values are increasing.
.
I have the most data points for NA1 Coupes so they present the strongest case; we'll start there
.
NA1 Coupe 1991-1994
.
NA1 Coupe 1991-1994
.
The trendlines, Actual Price (Value) vs. Miles, for 2016 and 2017 are nearly right on top of each other. This data suggests that no significant change in NA1 Coupe values occurred from 2016 to 2017 across all mileage.
.
While all other NSX types have fewer data points, the trend continues:
.
NA1 Targa 1995-1996
While all other NSX types have fewer data points, the trend continues:
.
NA1 Targa 1995-1996
.
As miles increases, the 2017 trendline does seem lower than the 2016 trendline, albeit with considerably fewer data points than the NA1 Coupe data. This data suggests that NA1 Targa values went down slightly at higher mileage.
.
NA2 Targa 1997-2001
.
Due to lack of high mileage sale data, the 2016 trendline was removed. The 2017 data points seem to follow a similar pattern to the 2016 data, though again a lack of sale data hinders the conclusion. This suggests that NA2 Targa values did not significantly change. A notable sale at $82,000 came from Bringatrailer.com, a strong price for a 2000 Silverstone.
NA2 Targa (Facelift) 2002-2005
.
Due to lack of high mileage sale data, the 2017 trendline was removed. The 2017 data points seem to fall in line with the 2016 trendline, with a similar lack of data. This suggests that NA2 Targa (Facelift) values did not significantly change. A noteable sale of a purported $135,000 came from an 2005 Imola Orange / Orange.
.
As miles increases, the 2017 trendline does seem lower than the 2016 trendline, albeit with considerably fewer data points than the NA1 Coupe data. This data suggests that NA1 Targa values went down slightly at higher mileage.
.
NA2 Targa 1997-2001
.
Due to lack of high mileage sale data, the 2016 trendline was removed. The 2017 data points seem to follow a similar pattern to the 2016 data, though again a lack of sale data hinders the conclusion. This suggests that NA2 Targa values did not significantly change. A notable sale at $82,000 came from Bringatrailer.com, a strong price for a 2000 Silverstone.
NA2 Targa (Facelift) 2002-2005
.
Due to lack of high mileage sale data, the 2017 trendline was removed. The 2017 data points seem to fall in line with the 2016 trendline, with a similar lack of data. This suggests that NA2 Targa (Facelift) values did not significantly change. A noteable sale of a purported $135,000 came from an 2005 Imola Orange / Orange.
.
An important note is that these graphs only show completed sales, NSXs that did not meet reserve are not included. One reason for an NSX to not meet reserve is that the seller has an unrealistically high reserve. A more common reason is the right bidders are not present in the auction to drive up the bids to be in line with the car's predicted value. Such is the fickle nature of auctions, but their data is the best source for determining the raw value of an NSX and they show values are holding steady.
.
.
.
Summary:
- To roughly estimate an NSX's value, multiply the listing price by 0.9.
- The decreases in value caused by Automatic transmissions, Salvage titles, and Accident histories are properly quantified.
- The NSX market generally held constant from 2016 to 2017 and values are not going down.
.
As always, feel free to chime in with questions. I'm looking forward to seeing what 2018 brings!
Last edited:
SW2 Heritage Cars
Guest
- Joined
- 6 April 2016
- Messages
- 179
Awesome data as always. In my own dealings, I have seen coupes hold fairly strong this year, but targas (and especially '02+) fall somewhat in the last quarter of 2017. It has not been a big correction, but it's there IMO.
just as amazing as before, thank you for the update!
Hello RacerXwing
Your dedication to data collection and presentation is honorable. Most of the time, after purchasing our cars, we probably stop looking at prices and simply enjoyed our purchase. I am glad that you are continuing your study and watching the trend from year to year.
When I was doing my study, I noticed a slight trend based on region. Are you able to see that in your analysis?
In reviewing your recent "NA1 Coupe 1991-1994" chart, it is interesting to see the 2016 (purple) and 2017 (blue) trend lines cross at approximately the 70k miles mark. This suggest that higher mileage cars were selling at higher prices in 2017 than 2016. From my point of view, perhaps the lower amount of data points for higher mileage cars could have shifted the trend line. What are your thoughts?
Overall, I agree with your conclusion. I am just curious about the trend lines crossing and if you saw any significant changes based on regions/states.
Your dedication to data collection and presentation is honorable. Most of the time, after purchasing our cars, we probably stop looking at prices and simply enjoyed our purchase. I am glad that you are continuing your study and watching the trend from year to year.
When I was doing my study, I noticed a slight trend based on region. Are you able to see that in your analysis?
In reviewing your recent "NA1 Coupe 1991-1994" chart, it is interesting to see the 2016 (purple) and 2017 (blue) trend lines cross at approximately the 70k miles mark. This suggest that higher mileage cars were selling at higher prices in 2017 than 2016. From my point of view, perhaps the lower amount of data points for higher mileage cars could have shifted the trend line. What are your thoughts?
Overall, I agree with your conclusion. I am just curious about the trend lines crossing and if you saw any significant changes based on regions/states.
Thank you all for your support.
Thank you for providing your insight; it is good to add another source of data to this discussion. I am frustrated by the lack of data to draw a stronger conclusion. It almost makes me wish I had chosen to something more common like the S2000 instead... almost. There are a fair bit of NSXs coming up in January auction season (including an NA2 Coupe at Barrett Jackson!), so I will be watching those with a keen eye.
Thank you for your kind words natcc. This kind of data analysis is a passion of mine, so my fun is simply doubled now in that I can enjoy both an NSX and NSX data at the same time.
Region is one factor that I have not examined. I do notice a large portion of listings come from California or Florida, but I haven't been tracking that effect. It would require manually recording the location of each NSX.
Your interpretation of the trend lines is correct, but I would hesitate to say the effect is strong enough to make that conclusion. The maximum difference between the two lines in the low-miles region you point out is less than $5000, but there is still a difference, so I investigated potential reasons why.
Even though the NA1 Coupes have the most sale data, the trendlines are still vunlerable to outliers. This uncertainty was enough for me to reach the conclusion that they were essentially equal. I hope that makes sense.
Thank you for providing your insight; it is good to add another source of data to this discussion. I am frustrated by the lack of data to draw a stronger conclusion. It almost makes me wish I had chosen to something more common like the S2000 instead... almost. There are a fair bit of NSXs coming up in January auction season (including an NA2 Coupe at Barrett Jackson!), so I will be watching those with a keen eye.
Thank you for your kind words natcc. This kind of data analysis is a passion of mine, so my fun is simply doubled now in that I can enjoy both an NSX and NSX data at the same time.
Region is one factor that I have not examined. I do notice a large portion of listings come from California or Florida, but I haven't been tracking that effect. It would require manually recording the location of each NSX.
Your interpretation of the trend lines is correct, but I would hesitate to say the effect is strong enough to make that conclusion. The maximum difference between the two lines in the low-miles region you point out is less than $5000, but there is still a difference, so I investigated potential reasons why.
- I noticed that the NA1 Coupe chart excluded cars with >200k miles, so I have fixed that. In fact, there is a low sale price ($35k) in 2016 with just over 200k miles, lending credence to your idea. The is no corresponding high mile NSX sale in 2017.
- Another reason came from this surprising sale in 2017, a 1991 coupe with 9800 miles sold for only $58,850 (https://www.ebay.com/itm/1991-Acura...ash=item440c3ed030:g:FRsAAOSwi~5ZwrC~&vxp=mtr). This outlier brings down the 2017 trendline in that area.
Even though the NA1 Coupes have the most sale data, the trendlines are still vunlerable to outliers. This uncertainty was enough for me to reach the conclusion that they were essentially equal. I hope that makes sense.
SW2 Heritage Cars
Guest
- Joined
- 6 April 2016
- Messages
- 179
Of course! If you would like the confirmed sales numbers I have from my past year, e-mail me at [email protected]. Happy to share if it helps your data grow.
Nice work! The data are compelling
Wow, thank you for this kind offer Sterling. I am happy to keep all the individual data points private if you wish. I'll get in touch. Thanks again!
Bought two NSX's in 2017. I'll get you the purchase prices (inflated by $20 K each so we can pump them up :tongue: ).
Outstanding, keep up the stellar work young man!
This data is invaluable. Thank you again for keeping up with it. It’s at least as good as, and probably better than, the Hagerty data!
Of course, how else will we make sure they go up? :tongue:
Thanks again to everyone commenting here. All these appreciative comments are what keep me working on this project.
I am aware that a static forum post is not the best way to display my insights. They are updated daily as I get new data, but the leading post in the thread has not been updated since last year. As such, much of what is in that post has now changed due to the increased power of insight of more data. I apologize and intend to update it soon. More importantly, please find a link to a live-updated selection of insights below. These charts will update any time new data meets the criteria. Save it to your bookmarks!
https://docs.google.com/spreadsheet...-WUwbcHb7L/pubhtml?gid=1777763775&single=true
https://docs.google.com/spreadsheet...-WUwbcHb7L/pubhtml?gid=1777763775&single=true
Last edited:
Awesome graphing of the data. Thank you! I'm going to use this to figure out my insurance coverage needs.
wow... early A.M. JUly 4.. on first cup of coffee... my quick research says...... with all due respect... THAT YOU HAVE A LOT OF TIME ON YOUR HANDS!!!
I would show this to my wife... who loves mining data...and understands it..... but then my 4th of July would be blown with her looking at this "data porn"
thank you for going where no one has gone before!!
I would show this to my wife... who loves mining data...and understands it..... but then my 4th of July would be blown with her looking at this "data porn"
thank you for going where no one has gone before!!
Would be interesting if we knew if the outliers for a year/type are the result of vehicle condition, originality and history or something else. Haggerty adds approx 20% value between Concours, Excellent, Good and Fair condition cars. Even though that data is not readily available in sales data alone, and is even somewhat subjective, it could be inferred from the deltas in outliers.
Thanks! Glad you are finding it useful.
Hah, thanks. Because this analysis is my passion, it feels like no time at all. It is very much data porn for me too!
Now that I have more factors available since the first post, such as the title status, maintenance level, and mod level, I've found low outliers are usually salvage title NSXs with due maintenance, while high outliers are usually overly ambitious dealers / sellers. However, a higher than expected sale price can also result from comprehensive, complete maintenance.
Thanks everyone! I am currently working on adding 2011 data. I wish the first post could be updated or deleted, as it is no longer relevant or correct anymore.
Similar threads
