The Intelligent Investor’s Guide to the Car Market: How to buy a car like a value investor

Most people approach a car dealership with the same enthusiasm they reserve for a root canal. They see a battlefield of asymmetric information, high-pressure sales tactics, and the sinking realization that their net worth is about to take a hit.

As a behavioral scientist and personal finance professor, I see it differently. To me, the used car market is a fascinating "raw" negotiation lab. It is one of the few places where regular people engage in high-stakes financial decisions influenced by cognitive biases, incentives, and the laws of physics. At PFR, we teach our students to stop being "buyers" and start being "value investors" who seek the Depreciation Inflection Point.

Everybody has a different budget constraint. For some, spending $10,000 is a minor drop in the bucket; for others, it is a significant constraint. The point of this article is not whether you can or cannot afford more. The idea is, since cars depreciate, a smart investor minimizes their exposure to declining value assets so they can increase their exposures to investments increasing in value.

The habits of the rich start with living below your means, whatever those means happen to be.

By Jeff Hulett President, Personal Finance Reimagined (PFR) and Personal Finance Professor, James Madison University

The Math of the "Depreciation Sweet Spot"

We are all aware that a car loses value the moment it leaves the lot. However, value doesn’t drop in a straight line. It follows a logarithmic functional shape:

Y = -0.318 ln(x) + 1.0011

(Please Note: This is the functional shape of the curve found in the article's lead image)

First, do not get hung up on the math. This article is about the "so what" of what the math intuition tells us. In this model, Y represents the percentage of value retained, and x represents the age of the vehicle. The critical intuition of non-linear shape is this: During the first three to four years, the curve is steep—the vehicle is losing value at an increasing rate. But around year five or six, the curve begins to flatten. This is the Inflection Point. This is the point at which the decline in value slows.

(The second derivative (f′′(x) tells us the degree to which depreciation is speeding up or slowing down.)

By buying a car after this point, you let the first owner pay for the "new car smell" and the initial massive value drop. We use a utility-focused metric called CPRM (Cost Per Remaining Mile) to validate this. Our goal is to maximize the utility (remaining miles) we get for every dollar spent. A unique CPRM is calculated for each alternative; the idea is to minimize the CPRM.

Finally, an experienced car buyer knows that each make and model has a unique depreciation curve. The depreciation curve shown above is the estimated curve based on data across the most popular cars and found over multiple decades. Our research has determined: 1) The average inflection point is amazingly consistent across makes and models. Also, 2) The PFR car buying decision system accounts for the unique depreciation curves via our bespoke filtering and CPRM approach.

What Is Important To You?

While our system is grounded in math, PFR respects that car buying is deeply personal; we encourage you to determine exactly what is important to you to ensure long-term satisfaction.

(Our car decision system, found in Making Choice, Making Money, includes an app to help you weigh your criteria.)

We anchor our process in 'maximum practical utility'—the baseline need for safe transportation from point A to B—to keep your costs as low as possible. However, we recognize that buyers often have flexibility when weighing preferences like technology, size, or performance. Our approach empowers you to layer these complex desires onto our default criteria once the core utility is secured. Aligning these unique preferences with a solid financial foundation is the key to achieving true long-term value.

The PFR "Fast to No" Filtering Criteria

Before you ever set foot on a lot, you must establish your boundaries. We teach a "Fast to No" mindset: if a car doesn’t meet your pre-determined criteria, you move on immediately.

Based on current market dynamics, here is our initial, standard filter:

• Price: Between $6,000 and $15,000.

• Mileage: Less than 75,000 miles.

• Age: No older than 11 years.

• Distance: 50 miles or less distance from your home

(Most auto-buying marketplaces enable filtering)

Why mileage and age? Mileage is a proxy for mechanical usage. The idea is based on physics (Entropy), and the reality is that mechanical things will wear out. Miles are the true indication of how old a car is, based on how much it has been driven. However, actual age or model year is a proxy for material degradation. Cars are made of rubber and plastics that degrade over time, regardless of how many miles they are driven. Gaskets dry out; bushings crack. This is particularly critical in "salt-belt" regions -- whether coastal or high snow areas -- where salt air or salted roads accelerate corrosion. A car might have low miles, but if it’s 15 years old, the "age tax" eventually comes due.

As such, the true age of a car is grounded in mileage but is constrained by model year.

Leveraging GenAI as Your Research Assistant

A significant challenge when car buying used to be the research—scouring forums and consumer reports to find, for example, "which 2017 models have transmission issues." Today, we’ve codified this into a Tested and Validated GenAI Prompt. You can paste this into your favorite AI tool (ChatGPT, Claude, Gemini, etc.) to generate a shortlist in seconds.

The PFR Prompt:

"I live in the [Insert City] area. I am looking for a used car between 5 and 11 model years old. Please 1) Recommend which car makes have demonstrated the lowest maintenance costs over this period. 2) Within those recommended makes, are there any specific models, engines, or years that should be avoided due to high maintenance, safety issues, or poor fuel efficiency? Then 3) Provide a 'Top 10 List' of models for potential purchase. Requirements for Output: The output is expected to pass a reasonableness check from Jeff Hulett, a Personal Finance Professor and behavioral economist. Prof. H requires that you use high-quality, data-driven sources to inform these recommendations. Please provide no more than 3 citations describing the specific high-quality sources (e.g., Consumer Reports, JD Power, IIHS) used to validate your response."

It is part of a decision system: The prompt is tested. My only qualification is that the prompt works best as part of the broader car-buying system we teach and specify in our book, Making Choices, Making Money.

At this point, add the GenAI-generated makes and models to refine your filter. If the filter results return too few or too numerous car alternatives, filter changes include:

1. Change distance: Closer or farther distance from home: +/- 25 miles

2. Change the size of the list in the GenAI prompt from "Top 10" by +/- 5 models

3. Change the dollar amount from $15,000 by +/- $3,000 (If you can afford it.)

Behavioral Tactics: The "PPI Tell"

Even with abundant data, you are still dealing with human incentives. Those selling cars want to get the highest reasonable price for the car. Some used car dealers play "fast and loose" with the data because they want to lure you into the showroom. At this stage in the buying process, your job is data validation. The Pre-Purchase Inspection (PPI) is your greatest weapon. Yes, the PPI costs money, and yes - it is the best money you will ever spend. It is a necessary investment to optimize your car's investment value.

When the data passes your filters and CPRM minimum, I use a unique "fast to no" strategy, called the "PPI Tell." I find a reputable mechanic near the dealership and then say to the salesperson:

"I am a serious buyer, but I want to validate the car's condition before I drive all the way there. I have already authorized payment for a PPI at [Mechanic Name] nearby. If you can drop the car off there for an hour, and it checks out, we have a deal."

Here is the behavioral secret: If the dealer refuses to take the car, that is your "tell." Since you are paying for the inspection and it only costs the dealer a few minutes of time, a refusal suggests they are not confident the car will pass. This small amount of time for the dealer to take the car to a mechanic is not unreasonable. However, this approach transfers just enough skin-in-the-game to tell you whether they are confident the car will pass. If they say "no," you say "next."

By the way, I do not actually pay for the PPI until after the PPI is completed. Also, I do not always execute the PPI Tell strategy. Sometimes I just go to the dealership and drive the car to the mechanic for the PPI.

Once the PPI is obtained, it has 2 purposes:

1) Determine if there is a "deal killer" like a bent frame from an undisclosed accident or some other serious defect. (Yes, the CarFax can get gamed too!) To validate, I usually ask the mechanic, "Would you recommend one of your family members to buy this car if they were in the market for a similar one?" If the mechanic's answer is "no," this is your deal-killer tell. If Step 1) is passed, then

2) The PPI contains your negotiation data. The PPI will itemize the maintenance plus the costs for the mechanic to repair it.

Keep in mind, this is a used car. It will have maintenance needs. You cannot expect a used car to be completely clean. Alternatively, if the dealer does present only clean used cars, the dealer will likely extract the value. This is like the CarMax model.

Closing the Deal

I have a profound appreciation for ethical used car dealers. When you find one who welcomes a PPI and provides transparent data, buy from them and refer everyone you know.

Remember, car buying is not about the car—it’s about using a car buying decision system. By using math to find the inflection point, GenAI to filter the makes, and behavioral tactics to screen the sellers, you transform a stressful expense into a high-value investment.

Finally, when negotiating and assuming no deal killers, I provide the PPI to the seller. Remember, it is a used car. You can expect some maintenance. Since I am certainly not a car professional, I do not know:

1. How much is reasonable maintenance I should pay vs.

2. How much is unreasonable deferred maintenance, the seller should pay.

As a result, I go with a "split the pie" approach. My negotiation message to the seller is: "To be reasonable, I will split the difference with you. Let's take the total maintenance and divide it in half. I will do the deal if you discount the price by half the PPI total." I have a high negotiation success rate with this split-the-pie approach.

A concluding thought...

At this point, some readers may be questioning, "Wow, this feels like a lot of work! Maybe I would be better off buying new or from CarMax?"

To that, I answer, this work is a high-value investment in a consistent, repeatable decision process, which you will use throughout your life. Once you practice this decision system, it gets easier and more rewarding.

IT IS WORTH THE INVESTMENT!

My biggest takeaways are:

1) Using a car-buying decision system is your best friend. We presented a part of it in this article. The book "Making Choices, Making Money" provides the full system, plus our decision technology are very useful for implementing the decision system.

2) Appreciate the incentives and constraints of car dealers. Data is not always clean, and it should not be surprising, given the dealers' incentives and constraints. Think of a PPI as the best way to generate your own data. It is the best money you will spend to support a high-value car purchase.

Good luck!