How New Connected Data Science for Residual Values Will Transform the Automotive Industry

 

By Rahul Sonnad & Haydn Sonnad

Vehicle depreciation is the automotive industry’s economic foundation. While many focus on the sticker price of a car, it is actually depreciation rates that determine monthly payments for leasing or financing. Depreciation also has a major influence on rideshare rates and many areas of the rental car industry, whose economics are highly dependent on the vehicles’ residual value (which is the value of a car after some period of usage).

Estimating depreciation is both an art and science—one that if not performed effectively can result in both small and large scale economic problems. A consumer may lose anywhere between 25% and 40% of what they paid for a new car over their first three years of ownership, and if they buy a car that depreciates quickly, they may not be able to sell it without a huge cash outlay. Large companies can suffer even more significant losses. In November of 2016, Hertz lost $1.5 Billion, 52% of its market value, because “the rental company failed to adequately price the cars in its fleet and depreciate them over time before selling them in used-car auctions.”

Factors of Depreciation

Fundamentally, depreciation is a determination of the market value of the vehicle after some amount of usage. To calculate the value of depreciation, subtract the residual value from the original value.

The value of a used vehicle is determined by a few simple market factors, mainly:

  • Expected lifespan
  • Reliability and operational costs, including maintenance for the used vehicle
  • The feature/brand appeal for the used car buyer, including its specific options

These factors together then overlay onto the basic supply and demand equation for any given type of vehicle.

ICE Depreciation Models

Over the last 30 years, in a world of ICE (internal combustion engines), auto depreciation has been estimated by taking historical data on the market value of cars over their lifespan and finding the closest historical match to a new car today. Different brands tend to depreciate at variable rates. For example, today’s Lincoln vehicles depreciate much faster than Toyota vehicles. Typically, depreciation equations for any given brand and model are dominated by only two factors, mileage and age. Briefly setting aside fuel costs (which makes gas guzzlers less valuable when fuel prices are high), this method has been relatively accurate. However, for any given car, there is a margin of error due to mechanical issues that may have surfaced. Essentially, the more reliable the car, the better accuracy of depreciation calculations as it gets older.

Connected-EV Depreciation Models

Now, the art of depreciation modelling is about to be transformed into more empirical science by a new generation of electric, connected, and increasingly software driven vehicles—embodied today by Tesla vehicles. These connected-EV’s will bring about a fundamental change in the way cars depreciate. Due to the nature of electric drivetrains, battery degradation, continual over-the-air software updates, and rapid technological innovation on both the hardware and software levels, the real-time condition of a vehicle is becoming exponentially easier to dissect.

While the foundational market factors influencing depreciation—appeal, reliability, operational costs, and lifespan—remain unchanged, the characteristics of electric and connected vehicles along these dimensions are radically different.

Let’s examine each of these factors and their potential impacts on vehicle depreciation and valuation.

Lifespan

Tesla-quality electric drivetrains will last over 500,000 miles (about 800,000 kilometers). While some may not believe this, it is easy to extrapolate from the few Tesla vehicles that have reached over 350,000 in mileage (about 550,000 kilometers) that they will easily make it past the 500,000 mark.

Nowadays the chance that most electric cars will ever be driven 500,000 miles is very slim. Even at 20,000 miles per year, driving this distance would take approximately 25 years. This means that typical consumers will never run a modern electric vehicle drivetrain into the ground. Even if someone buys a car with 250,000 miles on it, it’s unlikely they will ever see the end of its utility. Thus, the marginal depreciation caused by miles will trend toward zero.

However, while new electric vehicle drivetrains depreciate minimally with usage, their batteries do not share that same quality. Early electric batteries, such as those of the Nissan Leaf and Fiat 500e, experienced significant issues with capacity retention and thermal management. Newer battery designs, such as those of Tesla vehicles and Chevy Bolts, have not displayed significant problems based on the data emerging from a small but increasing number of high utilization vehicles. While a few of Tesloop’s Tesla vehicles have had battery problems over 200,000 miles, most have not. Meanwhile GM states that they have never had excessive degradation problems on any Bolt batteries (though there may not be any over 200,000 miles). The bottom line is this: the value of electric vehicles depreciates based on a loss of driving range due to battery degradation. Today a Model S loses on average $168 for each mile of degradation (based on internal findings as of February 2019). Fortunately, this loss is unlikely to make the vehicle unusable in a way that a transmission problem in an ICE might. Also, due to the automotive variation of Moore’s law in certain components such as batteries, they can be replaced or reconditioned for a cost that is continually decreasing at approximately 8% year over year.

Tesloop estimation of the depreciation curves for ICE and Electric Vehicles

Reliability and Operation Costs

Beyond lasting a long time, modern connected electric vehicles (EV’s) need very little maintenance. The nature of the electric drivetrain is closer to a laptop hard drive than an internal combustion engine—no oil changes, no timing belts, no gaskets, no transmission, and with the exception of wiper fluid, no fluids to replace—just constant rotation. This means that vehicles with hundreds of thousands of miles on them can provide reliable transportation with little ongoing maintenance other than tires.

Because of low maintenance requirements, operational costs of high mileage electric vehicles are significantly lower than that of ICE vehicles. This translates to high mileage connected-EVs becoming more appealing than vehicles that are more likely to have high-cost unpredictable problems associated with high mileage engines. The operational costs are further reduced by the lower cost of electricity, which in most parts of the world is less than half the price of gas per mile.

Feature Appeal

Perhaps out of all the aforementioned depreciation factors, the one that will change the most is the feature-appeal of used cars. Over the past 30 years, the pace of innovation in new cars was—by today’s tech standards—glacial. Significant new features on cars were infrequent. For example, some big innovations of the past such as remote unlocking, anti-lock braking, airbags, and bluetooth were relatively big innovations, but their slow roll-out and incremental benefit meant they did not levy a significant impact on depreciation rates.

Now, automobiles are entering the digital age through the integration of a wide array of sensors, processing power, cloud connectivity, and software upgrades. Core functionality is being upgraded at rapid pace, and these features have a dramatic impact, not only on how vehicles depreciate, but also on how to measure this depreciation for any given vehicle.

Since 2012, beginning with Model S, Tesla has had the ability to effortlessly deploy software upgrades to cars over a cellular connection that is built into every vehicle. Just like any internet device, the applications running on the vehicle can connect to dynamically updated data sources, such as those for navigation and charging locations. The ability to upgrade software in a vehicle will create greater appeal for used connected vehicles because the features will be new—similar to an old phone running a new app.

It is worth noting that in some cases, features such as Autopilot (Tesla’s driver assistance system) could increase the car’s value over time, compensating for physical depreciation through increased functionality. Today a 2016 Tesla Model S with 40,000 miles on it and an eight camera Enhanced Autopilot system might sell for $55,000, the same price as a new 2018 Mercedes E class with Driver Assist. While some may slightly favor Tesla’s version of Autopilot today, within a few years, the very same vehicle should perform radically better due to software updates. It will likely be possible to upgrade the vehicle’s processor (for a modest cost) to further improve its performance and safety. On the other hand, it is unlikely that the 2018 Mercedes E class will be able to update its driver assistance features, and thus will experience more rapid depreciation as new features are released in new Mercedes cars.

Today, the market for high mileage Tesla vehicles with eight cameras is very limited. As people begin to seek the most affordable and high-quality autopilot hardware, this demand may drive up the value of these cars over time. At the same time, the value of Tesla vehicles without the eight camera Autopilot system may be decreased due to a less demand from consumers.

Sharing Platforms and Cash Flow Potential

Beyond the car’s physical attributes, emerging connected-EV networks will likely impact the value of used vehicles in a manner never before seen. Today’s services like Turo and Getaround enable vehicle owners to list their cars for rent. This creates potential for cash flow, albeit with the labor burden of a side-hustle.

With advancing vehicle connectivity and the evolution of cloud platforms such as Tesloop’s Carmiq Network, it will become increasingly effortless to enlist connected cars into virtual rental fleets that enable hands-off maintenance and customer service. Today, California-based Tesloop rents Model X vehicles that have over 250,000 miles for $1,400 – $1,575 per month in Los Angeles. These cars are appealing to rideshare drivers because they qualify for the “Luxe” and “Black” level services and the mileage on the cars does not impact their earning potential.

Self reported mileage and rental value for high mileage Tesla vehicles.

Thus, there is a huge disparity between the Kelley Blue Book value of a vehicle and its potential cash flow value. As shared rental marketplaces for vehicles become progressively liquid and begin to provide labor and worry-free options to vehicle owners, the cash flow value of a vehicle will begin to have a significant impact on the resale value, and thus the depreciation rate. In today’s car ownership market, most owners would not consider renting out their old vehicle instead of selling it when they want to buy a new one. Rather, they effortlessly trade in their vehicle to the dealer for a significantly lower return than could be earned through a private sale. However, if alternatively the owner can earn 35% of the trade-in value in only one year and a few clicks on an app, this will become a very attractive option for both the initial owner and the prospective used vehicle buyer. As the market becomes more efficient, this disparity between resale value and adjusted cash flow potential will disappear.

Real-time, Vehicle-specific Measurement of Depreciation

Beyond influencing rates of depreciation, connectivity will have another major impact on the automotive industry. Today, depreciation is estimated with inferences, heuristics, and averages. Only when the vehicle is finally sold do the accuracy of these value predictions for the specific vehicle become evident.

However, with connected-EV’s, this averaging methodology will be replaced by vehicle-specific data analysis based on real-time vehicle data feeds. Car value and depreciation will be programmatically measured for each car on a continual basis. Car specific information such as mileage, battery degradation, acceleration stress, as well as details of vehicle options such as which software and hardware packages are installed are now all available minute-by-minute through software interfaces. Individualized vehicle analysis will be much more accurate than on an ICE drivetrain because electric drivetrains do not share the inconsistency of maintenance problems that often occur on ICE—transmissions, gaskets, and timing belts. This can then be correlated with real-time market data for both sales and rentals on sharing platforms to determine sale and cash flow values. The overall data models on services like Carmiq now adjust themselves on a real-time basis as pricing data comes in from sales and rentals.

Measurement of depreciation on a vehicle-by-vehicle basis is especially important for peer to peer sharing scenarios. With a rental car company, any given rental may lose money due to “unlimited free miles”, but as long as the average is positive, the rental car company is happy. But for peer to peer networks, every rental transaction must be demonstrably cash flow positive for the owner. Tesloop’s “Carmiq Value” feature now automatically queries a Tesla Model S and can provide a real-time valuation instantly, before and after the rental transaction.

Sign up your Tesla at Carmiq.net

While the accuracy of vehicle valuation is currently limited by the relatively sparse data available for the models, this method should become the most effective valuation tool as an increasing quantity of listing and sale data is collected in the coming months.

ICE Vehicle and Industry Impacts

While it is not yet fully appreciated due to the relative lack of empirical data, it is likely that modern EV’s due to superior attributes of the drivetrains, will begin to demonstrate increasingly higher residual values. It is interesting to consider how this emerging generation of vehicles will affect the value of the hundreds of millions of ICE vehicles. Adam Jonas, Morgan Stanley’s veteran auto analyst, has cautioned of a point where the appeal of these new, electric, semi-autonomous vehicles starts to materially change the desirability of old unconnected cars, which will increasingly be perceived as less desirable in the market. The impact on the capital markets and industry as this transpires may be severe.

In the next few years, the art and science of depreciation will rapidly evolve into a landscape of connected, electric, and increasingly autonomous vehicles. As these vehicles begin to attached to networks that enable both behavioral analysis and effortless sharing—which creates continual monetization opportunities—a new real-time data science for depreciation will emerge.  And as this new residual math increasingly accounts for characteristics of modern connected-EV’s, it is likely to bring unprecedented economic transformation to the automobile market.