oalvarez

^ yep yep...that's exactly what we're after here. what is your locale? thanks for the post!

uberspeed3

Got it, thanks. Seems like relatively solid extrapolation then.

911seeker

I am trying to test the values and have some questions / comments.


There are several values to which you could apply the depreciation percentages, and you would reach different conclusions (I am going to use my own car - 2014 C4S - data to illustrate, hope it does not seem that I am trying to hijack the thread):

• The original MRSP (132k - before fees and taxes)
• The actual purchasing cost when new (I would assume in jan 2014 you could get a 5-7% discount on this car, so say 124k.
• The price at which you could TODAY buy a similar specd car NEW (and if you find it from the previous model year, you have two values here, say 20% discount over the 2015 increased price if you find a 2015 - or 12% discount on the 2016 twice increased price if you just find a 2016). I am not going to consider the 2017 because of the change of model, if not, a third value could be in play. So, 107 (2015 MY new) or 118k (2016 MY new).

So, to which of this values do we apply the depreciation %? I bought the car 2.25 years after the guarantee start, so the residual values (using the table and increasing the values 4% for the mileage of 7,1k miles / year) would be around 67.5% at that point, which gives us several values:

• 89.3k if we use MRSP
• 83.4 if we use the MSRP minus 6%
• 80.1 if we use the price of a MY16 today
• 72.1 if we use the price of a MY15 today (which is getting more difficult to find each month)

Just for the record, they were asking 89 (coming down from 96 I believe) and I paid around 84-85 (depending on how we interpret the market value of my trade in), and this was in Florida, which is not supposed to be one of the "cheapest" places to buy preowned porsches.


I assume the one to use is MRSP, but wanted to clarify...

oalvarez

Correct....the residuals are based off of MSRP. most importantly, only you (and the dealer) know what you paid for the car.

911seeker

A few comments:

• Imagine the exact same car could be purchased new with a 25% discount. Would it make sense to purchase a 2 year used car for 5% additional discount? Surely not. So, I understand that it si more convenient to use the residuals on MSRP, but the price new has to remain "somewhat stable" for that to work.
• Should residuals be different for the base price and the options? I think so, for some options the depreciation will be even greater than for the base car itself. I don't know if there are options that don't lose value. In any case, as a rule of thumb, could we assume that the options depreciate at twice the rate of the base car?

This is probably making it too complex, but there are web-based resources that help approximating that value, I think cargurus values are somewhat strange, but there are other on-line appraisal websites that help... and you can always ask in this forum for an expert opinion on the specific build out and save you the work...

dd356

The difficult variable in the new/used scenarios is always the warranty situation. Clear there is a value to it, but some cars due to the cpo's etc have it, some do not. Obviously, if you never use it, it's a neutral factor. However, nice to have, and you generally always pay for it.

911seeker

The remaining guarantee or CPO period should be easy to value... the cost of the CPO divided by 730 days ($3,5/day) and multiplied by the number of days of remaining guarantee could be a proxy... A car with 4 years left of guarantee would be worth 5k more than the same car w/o any guarantee left, does that make sense?

ace37

While the details you're hitting at are certainly things that will have a real impact on the resale value of the car, I think you're getting into the noise if you want to do that manually. Creating such a model has the combination of high complexity and sensitivity to changing market demands that makes the effort highly unrewarding for most purposes - the model will take extensive development and then a great deal of time to properly maintain for very little extra precision. And it doesn't matter at all to leasers as PFS gives you linear credit for all options up to the MRM regardless of 'true' market value. PFS is writing checks for these cars at the end and they don't charge us any differently excepting a past ding for PDK transmissions - that should tell you something about how hard it is to get a better answer.

If you're undeterred and you do want to build such a model, to do it properly I would recommend you perform an ANOVA (statistical analysis). Grab a large number of VINs and get Manheim auction prices. Flesh out all the options as variables to see how each really matters. Throw in overall "equipped" levels, a few options combinations, and the traditional value-influencing variables like mileage, season of the year, and age of the model within the cycle. You'll learn a great deal and I'd love to see the results of that.

That methodology could produce a model good enough to sell to a commercial entity. I'd laugh if it's better than some of the mainstream car evaluation service models.

911seeker

I am certainly not going to do that, although it is definitely an interesting suggestion!


I think it should be possible to "keep it simple" yet be a little more accurate without going crazy and trying to apply actual science to it... But at the moment I already have my car and can't spend the time to try...


For the sake of the argument, I would just apply some depreciation percentages to the base model price (either the ones oalvarez used or the ones that Edmunds uses, for example), higher depreciation to the options, and add some value back if it has additional guarantee / CPO, and that should do the trick with a 95% reliability... and then I would ask here as well to test the result!


Best

oalvarez

i absolutely LOVE the discussion, but in truth, was hoping to create a "place" where folks could either come to learn how to value either their own car or car they're contemplating buying, or to ask the question of "is this a good price?" and have it answered within some acceptable range. again, love the discussion and thought!

Austin-Guy

Ok so how'd I do? JUst bought this CPO w/15k miles, paid less than 80k

911seeker

I think you did fine... Do the math with oalvarez's depreciation factors and you should have paid mid 80s

oalvarez

you know what's fun about the example above? instead of someone asking "is this a good price" we finally have someone that was honest enough to come out and said "hey, here's what i paid."

at the end of the day, and what is most important to me, is the fact that even if someone "overpaid" he/she is still happy with their purchase. that's all that matters, not what is supposedly the best deal ever.

Speedster91

Nice topic.

Just trying to run through the math. 2014, would make it 2 years old and a total depreciation of about 34%. But at 15K (7500 / year) you'd take away about 8% (residual of 74%)

.74 x 128K is about 95,000.

Is that right?

By these numbers, this gentleman found himself a very nice deal.

blschaefer1

I like where this thread is going but I believe a key variable in the math is being overlooked. Often, lease residuals set by the manufacturers do not correlate to true market value at the end of the lease term. Most leases are subsidized to some extent by the manufacturer. Porsche less so, BMW more so, for example.

My 2014 C2S had an MSRP of 128k, MRM of 123K, and at lease end this November the residual is set at 88k. Come November, the true market value of the car will be certainly less than that. Porsche will eat the difference, which is the subsidized amount. Clearly I won't be buying the car from PFS.

OP - not sure how to capture this in your 'formula' but this part of the equation is critical for your tool to be accurate.