|Shares Out. (in M):||150||P/E||0||0|
|Market Cap (in $M):||4,500||P/FCF||0||0|
|Net Debt (in $M):||0||EBIT||0||0|
· The market doesn’t understand Carvana’s competitive advantage, unit economics, and long-term earnings power.
· A multisided platform is one of the toughest business models to get right as the business needs to overcome the chicken-and-egg problem of creating supply and demand and then having them interact with each other on attractive terms. Carvana’s long-term unit economics are currently being masked by high depreciation costs on their supply of inventory ($10 car depreciation per day, which affects gross profit per unit), large customer acquisition costs ($1,000 of advertising cost per unit sold), and large upfront investments in logistics and R&D.
· We think Carvana will be the lowest cost operator in a fragmented industry. At scale, Carvana has a $1,000 per unit cost advantage vs brick and mortar dealerships due to $250 lower reconditioning costs per unit and $750 lower SG&A per unit.
· The economies of scale in inventory, advertising, logistics, reconditioning, and R&D create enormous barriers to entry that will allow Carvana to be a monopolist and long-term share gainer.
· At scale, Carvana should have steady state unit economics of $1,750 EBIT/car and after extrapolating existing cohort curves, by 2024 we expect them to sell 1MM units and earn $9 EPS.
· Retailers compete along 4 variables in retail: price, selection, service, and convenience. Retailers make tradeoffs among these variables to differentiate themselves and serve different market segments. Reduce service, selection and convenience as low as possible, but offer very low prices you end up with Costco. Provide low selection at high prices in accessible locations with fast service you get convenience stores. Provide great service, good selection, and decent convenience, but at high prices, and you have CarMax. Carvana on the other hand, wins on all 4 variables.
· Carvana is a vertically integrated online platform for buying used cars. They remove the pain points and frictions from the traditional used car buying experience by giving customers what they want:
· Lower and more transparent prices that are +$1,000 below industry averages according to YipitData, which doesn’t include Carvana’s free delivery cost to customers.
· Wider inventory selection of 15,000 cars (and growing) vs <300 on a traditional dealer lot.
· Better service where customers deal with friendly, non-commissioned customer advocates vs the traditional used car salesman the customer doesn’t (or shouldn’t) trust.
· Greater convenience as customers can buy a car in under 15 minutes, have it delivered to their door for free as soon as the next day, and enjoy the comfort of a 7-day test own period where Carvana will pick up the car for free within 7 days if the customer decides to return the car. This convenience compares to multiple hours at a dealer and a test drive that consists of 4 right hand turns around the dealer’s lot.
· Carvana’s website allows customers to complete all phases of a used vehicle purchase transaction:
· Purchase a used vehicle: Carvana currently has a nationally pooled inventory of 15,000 cars (and growing) for sale on their website available to 57% of the U.S. population, where customers can select and purchase a vehicle directly from their desktop or mobile device.
· Finance their purchase: customer can pay for their Carvana vehicle using cash, Carvana’s proprietary loan origination platform, or financing from third parties such as banks or credit unions. Customers who choose to apply for in-house financing fill out a short application form, select from a range of personalized and highly-transparent financing terms (10,000 variations of down payment, monthly payments, and length of loan for each customer), and, if approved, apply the financing to their purchase in the online checkout process. The company finances the whole credit spectrum and approximately 70% of vehicles sold are financed by Carvana.
· Protect their purchase: customers have the option to protect their vehicle with a CarvanaCare-brands vehicle service contract (“VSC”), which is an extended warranty, as part of the online checkout process. VSCs provide customers with insurance against certain mechanical repairs after the expiration of their vehicle’s original manufacturer’s warranty. Carvana also offers GAP waiver coverage to customers, which contractually obligates Carvana to cancel the remaining principal outstanding after insurance proceeds in a total loss event (collision, etc).
· Sell their car: Carvana allows customers to trade-in a vehicle and apply the trade-in value to their purchase, or to sell a vehicle independent of a purchase. Using Carvana’s digital appraisal tool, customers can complete a short appraisal form and receive an offer for their trade-in instantaneously.
· Schedule delivery or pick up of car: using Carvana’s owned first party logistics network, customers can immediately choose available delivery slots for when they want Carvana to deliver or pick up their car.
· To enable this customer experience, Carvana has built a vertically-integrated used vehicle supply chain, supported by proprietary software systems and data:
· Vehicle sourcing and acquisition: Carvana currently acquires approximately 80% of used inventory from wholesale auctions. They also acquire approximately 16% of inventory from consumers and 5% from used vehicle suppliers, including franchise and independent dealers, leasing companies and car rental companies. Using proprietary machine learning algorithms and data from a variety of internal and external sources, they evaluate tens of thousands of cars daily to determine their fit with consumer demand, internal profitability targets and existing inventory mix.
· Inspection and reconditioning: after acquiring a vehicle, it is transported to one of their inspection and reconditioning centers (“IRCs”), where it undergoes a 150-point inspection and is reconditioned to meet “Carvana Certified” standards. This process is supported by a custom used vehicle inventory management system, which tracks vehicles through each stage of the process and is seamlessly integrated with auto parts suppliers to facilitate the procurement of required parts.
· Photography and merchandising: Carvana photographs vehicles using their proprietary photo booths located at each of their IRCs. This allows them to display interactive, 360-degree images of each vehicle on their website. Carvana annotates each vehicle image with a list of features and imperfections to assist customers in their evaluation of each vehicle for purchase.
· Logistics and fulfillment: Carvana transports vehicles purchased by their customers to their local market for home delivery or pick-up. In markets where they have launched operations, delivery to the customer is completed by a Carvana employee through its owned logistics networks in a company branded delivery truck. In certain markets, customers have the option of picking up their car at one of the company’s vending machines. Carvana’s logistics and fulfillment operations are supported by their proprietary vehicle transportation management system, which optimizes the scheduling of transport routes and delivery slots.
· A good video of Ernie Garcia III describing the business can be found here:
Unit Economics Analysis
· Gross profit per unit (“GPU”)
· Inventory turnover: Carvana currently has 65 inventory days to sale, which compares to Carmax and other dealers of approximately 40 days (apples to apples as Carvana has unique definition of inventory days to sale). To improve this metric Carvana simply needs to grow inventory at a slower pace than revenue and inventory turnover increases. We think they can improve inventory days by 15 days over the next several years, which should increase retail GPU by $150 as cars depreciate on average $10 per day.
· Source more cars from customers: Carvana’s competitive advantage in acquiring cars from customers may be greater than their competitive advantage in selling cars to customers. Carvana’s lower cost structure (see SG&A per unit analysis below) allows them to offer a higher price for a consumer’s car than competitors, Carvana’s logistics network provides greater convenience for the consumer because Carvana picks up the car whenever and wherever the consumer wants, and the service is better because the offer process is automated and takes less than 5 minutes (the consumer doesn’t have to drive the car to a dealer and have it inspected for an hour or more). Cars sourced from customers benefits retail GPU and wholesale GPU (where cars are sold to auctions because they don’t meet retail requirements) because they are more profitable than cars sourced from wholesale auctions (no auction fees and less competitive bidding process).
1) Retail GPU: Carvana sourced approximately 16% of retail cars sold from customers in Q3 (vs 6% in Q1), which compares with CarMax at 45%. We estimate the profit difference between an auction car and a car sourced from consumer to be approximately $650 based on CarMax’s wholesale GPU in FY17 of $962 less the auction fee CarMax charges to dealers of $125 and Carvana’s logistics cost of $200 to pick up the car (although this $200 cost is $0 on a trade-in due to free reverse logistics and not included in our forecast). Increasing cars sourced from customers to CarMax penetration levels should increase retail GPU by about $200 ($650 x 30%).
2) Wholesale GPU: Carvana wholesaled approximately 17% of retail unit sales, which compares to CarMax of 57% in FY17. We are not assuming in our analysis that Carvana will run their own wholesale auctions like CarMax (although they likely will). To estimate Carvana’s wholesale GPU we take CarMax’s wholesale GPU of $962 less CarMax dealer auction fee of $125 less the seller auction fee Carvana will pay to Manheim, Adessa, etc of $125 and their transportation cost of $200 to get a wholesale GPU of $500. Increasing penetration of wholesale cars to levels seen at CarMax will increase total GPU by approximately $200.
1) Auto Finance at Traditional Dealerships: Typically, auto financing is arranged through the dealer at the time of purchase and provided through a wide variety of banks, credit unions, and independent finance companies who lend to prime, non-prime, and subprime customers. The dealer is typically compensated by the finance source through a fee based on the spread between the loan offer rate provided by the financial institution and the final loan rate the dealer negotiates with the customer. Most traditional dealers negotiate and manipulate the pricing on every aspect of the car deal. Because salespeople, sales managers and finance managers are paid a commission based on the profit spread in the deal, they have a strong incentive to negotiate the highest profit spread possible on each element of the transaction (total price, monthly payment, interest rate, trade-in value, vehicle service contract, etc) and shift margin between elements of the deal based on customer preference functions. This is why most offers at traditional dealerships are package deals with limited price discovery. The problem this creates for lenders in evaluating a loan is obvious: they don’t know the true price and value of the car, the trade-in or the vehicle service contract, and there is no consistency in pricing from one deal to the next.
2) Carvana Finance Process Advantage: Financing, like all other elements of the sale process, is different at Carvana. Their transparent, no haggle pricing on the car, vehicle service contract, and trade-in offer also extends to any financing offer. This process produces data integrity and provides Carvana with a superior loan origination channel that yields consistent and predictable loan performance. Two principal risks in used auto lending are significantly reduced or eliminated by Carvana: 1) the risk of car collateral is reduced by Carvana’s consistent, 150-point reconditioning process that produces high quality cars. Also, because Carvana sells cars for +$1,000 below industry average, Carvana produces loans with better LTV and payment to income ratios and therefore better performing loans (lower cumulative net losses on like-for-like FICO scores), and 2) the “intermediary” risk is virtually eliminated by Carvana. There are no commission-driven finance or sales managers who may distort the facts on pricing, car quality, or consumer credit information. Thus, Carvana and outside lenders benefit from superior information quality in making financing decisions, which results in lower operating costs and credit costs for Carvana.
3) Carvana Loan Economics: As loan performance history has grown, Carvana is starting to get better loan pricing from its financing partners. Our analysis suggests that Carvana loans have an 8.5% asset yield and are worth approximately $1,500 based on a 3.75% cost of funds, which compares to Carmax’s recent ABS funding costs of 3.15%. This improvement in pricing could add approximately $250-$500 ($700 price increase times either 50% or 100% of loans times 70% finance penetration) of GPU long-term over Q2 2018 GPU (they started getting better pricing in Q3 2018). See Carvana Auto Finance picture below for illustrative loan economics.
· See Carvana Unit Economics picture below for more details on GPU walk.
· SG&A per unit
· Compensation and benefits: compensation and benefits consists of outside customer advocates who drive single car haulers to end customers, 9 car hauler drivers who transport cars from IRCs to local market hubs, HQ customer advocates who handle customer calls and title/registration, and corporate employees in R&D, finance, HR, senior management, etc. We think outside customer advocates can deliver 2 cars per day, which equates to a compensation cost of $80 per car (8 hours per day x $20/hour divided by 2 cars). The 9 car hauler driver we estimate to be about $133 compensation per car (see Carvana Logistics Economics below for more detail on this item). At the December 2018 Analyst Day, Carvana disclosed 600 HQ customer advocates, which equates to approximately $150 per car (this number should come down with PropelAI and other technology developments) based on expected March 2019 volumes and compensation estimates from Glassdoor. We then estimate a corporate cost of approximately $125MM. This corporate cost is based on a letter that Austin Ligon, co-founder and former CEO of Carmax, wrote on July 11, 2003 (published in 8-K) when he detailed CarMax’s economics.
· Advertising expense: A national TV ad campaign costs about $125 million and is a fixed cost unless they want to add more commercials per day. CarMax for example spends about $100 million on their TV ads. As Carvana grows sales volumes these fixed costs decline per unit. We also estimate a variable customer acquisition cost of $75 per unit sold for search and banner ads based on CarMax advertising expenses.
· Logistics: see Carvana Logistics Economics below and Logistics section in Industry Structure Analysis for more detail. Carvana currently has logistics cost per unit of $350, but these costs will decline with scale as capacity utilization increases and IRCs get built closer to end customers thereby reducing freight times. We’re conservatively assuming a slight improvement to $300 per unit long-term.
· Other overhead: this item has similar cost buckets as CarMax, but we’re assuming $100 more per unit to be conservative.
Industry Structure Analysis
· The U.S. used vehicle market is approximately $750 billion in sales, which represents approximately 40MM used vehicle transactions at an average sales price of $19k.
· The used car auto retail industry is highly fragmented. There are approximately 63,000 used car dealerships in North America, comprised of 45,000 independent used car dealerships and nearly 18,000 franchise dealerships. The largest dealership is CarMax with 1.6% of the U.S. market and the top 100 used car auto retailers collectively hold approximately 7% of the U.S. market.
· Carvana benefits from economies of scale that will be extremely difficult for competitors to replicate:
· Positive network externalities in selection: Carvana’s logistics capabilities allow them to offer every car in their nationally pooled inventory to customers across their markets (approximately ¾ of cars have free delivery). As they add markets, it results in increasing demand, which enables them to carry a larger and broader inventory and in turn, improves their offering across markets and increases market shares within markets due to higher conversion rates. The cycle repeats. Based on an extrapolation of Carvana’s cohort curves, we think by 2022 Carvana will have more than +75,000 cars in inventory for customers to choose from, which compares to a traditional dealer lot of less than 300.
· Logistics Route Density: Carvana’s proprietary logistics software and in-house delivery network of 9 and single car haulers allow them to predictably and efficiently transport cars while providing customers a distinctive fulfillment experience. We think Carvana currently ships cars at a cost of approximately 35 cents per mile, which will continue to decline with scale due to low variable costs on unused capacity (adding a 5th car to a 9 car hauler has low incremental cost) and IRCs being closer to end customers (5 IRCs currently and have capacity for 50k cars each). This 35 cents per mile compares to 3rd party providers that charge 75 cents to $1 per mile.
· Reconditioning: Carvana’s IRCs are able to have lower reconditioning costs due to scale economies and flow shop vs job shop manufacturing at a traditional dealership. Carvana’s IRCs have capacity for 50k cars per year, which allows them to have higher labor utilization and they can better match skillset with skill. This allows Carvana to have an average reconditioning cost of $1,000 per car vs competitors like CarMax with $1,250 reconditioning cost per car.
· National and Cumulative Advertising: Carvana has now reached the scale where they can utilize national TV advertising, which costs approximately ½ the cost of regional TV advertising. A national TV ad campaign costs about $125MM and is a fixed cost. As Carvana continues to grow it will reduce their current -$1,000 customer acquisition cost per unit to levels seen at CarMax of approximately -$225 per unit.
· R&D: Carvana has spent the past 6 years improving their website design, vehicle sourcing and acquisition, financing technology, checkout process, logistics technology, etc. The R&D investments continue to grow and continue to increase platform utility and customer experiences. Competitors would have to replicate the cumulative investment Carvana has made in R&D and Carvana could replicate any differentiating features a competitor creates (similar to how Facebook/Instagram incorporated Snapchat-like features on their platform).
· CarMax and Incumbent Dealers Competitive Response
· CarMax recently announced that they will be offering home delivery and the ability to do online only checkout as part of a new omni-channel experience. While this increases convenience and service, CarMax is still at a structural disadvantage to Carvana regarding price and selection. CarMax prices its cars roughly +$1,000 more than Carvana and although they have roughly 70,000 cars in inventory as a company, only 3,000 cars are available in any given market for free shipping. In order to nationally pool their inventory and compete with Carvana’s selection, they will need to add logistics costs of approximately $500 per car. Unfortunately, as seen in the Carvana Unit Economics analysis below, CarMax only makes $1,500 per car and reducing prices $1,000 and adding $500 of logistics costs would blow up their income statement. Incumbent dealers have similar cost structures as CarMax and haven’t been able to improve their service with things as simple as no haggle pricing. Also, it is important to note that while AutoNation, Sonic, Lithia, etc own many stores in some cases, the stores are not centrally managed, but in fact are independently run by local store managers.
· Amazon is likely the only competitor that has the capital to replicate Carvana’s business model. However, we think they are unlikely to do so because 1) it would be a massive incremental leap that doesn’t leverage any of Amazon’s existing assets. It would likely take them at least 3 years to hire people and build out the logistics, reconditioning facilities, and technology. By 2022, Carvana will have +75,000 cars in inventory, and how would Amazon be able to compete on price, selection, convenience, and service, 2) Jeff Bezos has given interviews where he says Amazon always starts small with new product lines (they’d have to go big to not be at a selection and logistics disadvantage relative to Carvana) and the Amazon culture is one of inventiveness where they always fail when they try to compete with their own undifferentiated “me too” type product or service. We think it makes more sense that Amazon will try to buy Carvana. See minute 8 and 47 in first link and minute 9 in second link:
Value is catalyst
|Subject||Tables didnt make it onto idea post; will upload later|
|Entry||12/23/2018 10:47 AM|
Tables didnt make it onto idea post; will upload later
|Subject||Issues raised by CVNA shorts|
|Entry||12/23/2018 02:48 PM|
Thistle, thanks for the idea. I’ve looked at CVNA a bit as a potentially interesting multi-sided platform with network effects, but found enough issues of concern to punt on it. But I would be interested in your thoughts about the issues below raised by some shorts:
Carvana: Looking Under The Hood
Carvana: Disguised As 'Disruption' In The Used Car Sales Business Backed By Unethical Management
Is Carvana “Cooking” Its GPU Metric?
|Subject||KMX and CVNA|
|Entry||12/23/2018 02:57 PM|
Carvana has ~$350 million in inventory as of Q3 2018 in their centralized distribution. Carmax made over $650 million in net income last year. They also already have trucks driving across the country transferring cars from store to store. Seems to me like it would be fairly simple for Carmax to warehouse an addition $300 million in inventory (in addition to their current $2B+) and just add the CVNA sales model to their toolkit
|Subject||Re: Terminal Value|
|Entry||12/24/2018 10:55 AM|
I posted the following bullet points earlier this year in response to the CVNA long post from last year but didn't get a response from the author. Would love to hear your response.
I have spent some time looking at it, and I think it is a short. I would like to hear your counterarguments.
Also, penetration in the Atlanta market might be overstated. There are some posts on Reddit that talk about how if you buy a car from Atlanta, they cover your air ticket and pick you up from the airport.
|Entry||12/28/2018 09:54 AM|
Any thoughts on the financing transaction announced this AM? Any clue as to the background of Sonoran Auto Receivables Trust?
|Subject||Carvana unit economics|
|Entry||01/07/2019 01:43 PM|
|Entry||01/07/2019 02:09 PM|
HAving just bought a used Volvo XC90, and wanting to buy it from Carvana, I am pretty sure this is a slick way that will ultimately get a good shre of the used car market. But I am doubtful that Carvana (or anyone) has a unique advantage. I shopped three channels (Carvana, Vroom and Carmax). Carmax is the most trusted and easy to deal with but as Thistle points out a slight bit pricier, but you do get do drive the car you want, and I will say that their free radius almost always includes the model you are looking for.
I ultimately bought myvehicle from Vroom. The only reason was they had the exact color combo, and accessory package I wanted. I toggled between sites and just waited willing to buy from either one. I had cash in hand, and they fedexed the paperwork to me on a Sat and I had the car at my doorstep 7 days later. They also honored my Carmax appraisal on my tradein. Which I don't think is sustainable. I think this is where Carmax might have the advanatge. If you trade in a car (which happens a lot) then it brings down the cost of your purchase and you shield the sales tax. So I would also model in third part appraisal for tradeins ($150-200 bucks or so, for Carvana). Also to Carmax's advatage it is cumbersome to buy a car out of state. Registration, vin verification financing etc. It took another 6 weeks for me to sort all that out. I would do it again, but I think many folks like having the dealer take care of it all. Vroom had a less slick expereince than Carvana but I can't say that I thought the two were very different, clearly it feels like there is room for a few more players. Cars are such high value purchases, that the conveninece factor is easy to get over since most people do it not very often. I think it will ultimately be any player that is trusted and then followed by the lowest price.
|Subject||Re: Carvana unit economics|
|Entry||01/07/2019 02:35 PM|
Thank you for the Sonoran breakdown. That's really helpful. Your analysis above Sonoran doesn't strike me as totally unrealistic either, despite the insane growth assumed.
My question on Sonoran was more about the background of them and their relationship with Carvana - it looks like it may be Mark Walter based on the address. I don't know if this is an issue or not, but wish they would be more transparent on the non-Ally side.
|Subject||Re: KMX and CVNA|
|Entry||01/07/2019 02:47 PM|
In CarMax 10k:
"We acquire a significant percentage of our retail used vehicle inventory directly from consumers through our appraisal process, as well as through local, regional and online auctions. While in any individual period conditions may vary, over the past 10 fiscal years, 38% to 52% of our retail inventory has been acquired through our appraisal process annually. We also, to a lesser extent, acquire used vehicle inventory from wholesalers, franchised and independent dealers and fleet owners, such as leasing companies and rental companies. The used vehicle inventory we acquire directly from consumers through our appraisal process helps provide an inventory of makes and models that reflects consumer preferences in each market."
Approximately 30-35% of CarMax sales are transfers where car is at one store and transfered to another store. From CarMax IR, approximately 80-85% of those transfers are free because they are within local markets (roughly less than 100 miles away). CarMax also doesnt own its own single car and 9 car haulers; they use 3rd party logistics to transport cars from store to store and utilize the logistics provided by auctions to transport cars from auctions to store.
Therefore, CarMax doesnt have trucks driving across the country transferring cars; they source and sell cars on a local basis (local auctions to save on transportation costs and local consumers) and when they do transport cars they are paying about $0.75-$1.00 cpm to 3 party logistics.
If CarMax were to build out their own logistics network and nationally pool their inventory to compete on selection the transportation cost would at a minimum look like Carvana's on a per unit basis at about $300-350 from store to store (CarMax logistics has extra legs on routes vs CVNA as 9 car haulers would have to stop at multiple stores in a market than a single IRC for CVNA, which is less efficient) and another $200 for last mile to consumer. CarMax would have to eat this +$500 to compete with Carvana since prices already +$1000 more and consumer not willing to pay $1500 premium for the same customer experience as CVNA
|Subject||Re: Re: Carvana unit economics|
|Entry||01/07/2019 03:07 PM|
Spoke with CFO about Mark Walter. If it was Mark Walter or entity associated with Mark Walter it would have to be disclosed as a related party transaction in SEC filings. CFO said it is a party introduced by Mark Walter since Sonoran is financing platform created by Mark Walter and they are still adminstrative agent and payment agent (see quote from proxy below); therefore Sonoran has same address of 227 West Monroe, Suite 4800, Chicago Illinois 60606 as Guggenheim,Delaware Life, etc
"On February 27, 2017, Delaware Life sold its interest in the notes under the note purchase and security agreements to an unrelated third party, but remains the administrative agent and paying agent for the note purchasers."
|Subject||Re: Re: Terminal Value|
|Entry||01/07/2019 03:47 PM|
Ill take your bullets in order
|Subject||Re: Re: Re: Carvana unit economics|
|Entry||01/08/2019 11:15 AM|
Thank you Thistle! However, I don't follow. If Walter created Sonoran. How is it not related?
|Subject||Re: Re: Re: Re: Carvana unit economics|
|Entry||01/08/2019 11:20 AM|
I guess because you don't know who the end purchasers are? Just wondering if it's possible the loans could be being gobbled up by Guggenheim for much higher profits back to CVNA due to the relationship. I mean if they are able to pull that off, great, but just unclear as to the non-ALLY side and why they seem to be so excited about it all of the sudden.
|Subject||Re: Carvana unit economics|
|Entry||01/10/2019 03:53 PM|
FYI, the economics on Ally are based on Carmax Auto Finance and Sonoran are based on combination of Santander Consumer, Exeter, Westlake, ACA
|Subject||Re: Re: Re: Re: Re: Carvana unit economics|
|Entry||01/14/2019 10:25 AM|
I guess I am not alone in asking these questions per the release out this AM. The new information is helpful, but I can't help but be skeptical (maybe I'm mis-reading) given some of the answers (particularly section 4).
|Subject||September 10th, 2018 8-K provides evidence supporting $1,750 EBIT/car economics|
|Entry||01/17/2019 05:54 PM|
September 10th, 2018 8-K provides evidence supporting $1,750 EBIT/car economics. Thought I would add this since it wasnt included in original write-up
• The first calculation in the disclosure (link below) is market contribution, which is equal to GPU less in market advertising (national advertising allocated based on cohort’s share national population, i.e. Atlanta population as % national population; the remainder out of market national advertising where Carvana hasn’t opened is allocated to “corporate allocation”), and last-mile payroll, transaction, logistics (single car hauler), and occupancy expenses, excluding D&A. The next calculation is market contribution less logistics (inter IRC and IRC to hub logistics) and all SG&A (ex D&A) costs not included in market contribution calculation were allocated to “corporate allocation”, which includes HQ customer advocates, out of market advertising not included in market contribution, R&D, and corporate overhead. These costs are allocated to each market based on each market’s share of Carvana’s open markets (if Atlanta has 5.7MM people and Carvana’s open markets in Q2 2018 were 170MM people then Atlanta gets 3.4% of the allocation).
• Per this disclosure, Atlanta made $478 EBITDA per vehicle in Q2 2018, which helps determine total company profitability if every market had the same penetration that Atlanta had in Q2 2018. To get from $478 Atlanta EBITDA/car in Q2 2018 to long-term unit economics and profitability a few things need to be done:
1) Add $1,075 to account for difference between GPU in Q2 2018 of $2,173 and our steady state GPU of $3,250 in unit economics analysis, which results in $1,550 EBITDA/vehicle (3,250 – 2,173 = 1,077 + 478 = ~1,550).
2) Next, the logistics and corporate allocation to Atlanta in Q2 2018 was $810 (1,288-478). Our logistics analysis suggests the inter IRC and IRC to hub logistics cost was about $225/car (including driver expense) in Atlanta vs about $325/car for whole company due to higher capacity utilization on IRC to hub leg (this can be determined by monthly volumes divided by monthly capacity on a 9 car hauler). HQ customer advocates costs are estimated at $150/car. This leaves about $435/car of out of market advertising, R&D, and corporate overhead (810-225-150=435). This $435/car we think is roughly evenly split between out of market advertising, R&D, and corporate. In addition, as of Q2 2018, Carvana’s open markets were only 52% of the population. Therefore, $145 advertising/car needs to be added back and the remaining $290/car is ~2x what it will be in a few years because Carvana will be nationwide. This results in $1,840 EBITDA/car ($1,550 + 145 + 145 = $1,840).
• Carvana had approximately 2% share in Atlanta in Q2 2018. If we extrapolate Atlanta and assume 1) Carvana opens nationwide, 2) gets 2% share in each market, and 3) reaches $3,250 GPU, then Carvana will have EBITDA of $1.47 billion (40MM used vehicles x 2% = 800k cars x $1,840 EBITDA/car = $1.47 billion).
• D&A per car was $233 in Q2 2018, which produces about $1,600 EBIT/car (1,840-233) vs our long term estimate of $1,750 EBIT/car. The difference between these numbers is due to scale in advertising, logistics, D&A, and corporate overhead of 1MM cars in our unit economics analysis vs 800k cars extrapolated.
|Subject||Magic Johnson's Equitrust Life owns Carvana loans|
|Entry||01/24/2019 02:11 PM|
You can see Equitrust Life financial statements on https://insdata.naic.org/; NAIC code 62510
Equitrust Life is unrelated party to Mark Walter since Guggenheim sold Magic Johnson their entire 60% interest in 2015. 3Q 2018 Equitrust Life financials shows equity interest in Carvana auto receivables and debt in CAIH LLC, which is the same title for Sonoran in Carvana’s 10Q, link below
As Carvana said in filing last week, loans not held by Mark Walter or related party nor are they held by Garcia family
“Financial terms weren't disclosed, but Guggeheim no longer owns a stake in the insurance company, said Paul Miller, COO at EquiTrust.”
Sonoran Auto Receivables Trust 2017-1
227 West Monroe, Suite 4800
Chicago, Illinois 60606
|Subject||Re: Re: Magic Johnson's Equitrust Life owns Carvana loans|
|Entry||02/08/2019 10:28 AM|
Agree, what he failed to do was the math showing the Equitrust loans have a 40% ROE
|Subject||Re: Re: Re: Re: Magic Johnson's Equitrust Life owns Carvana loans|
|Entry||02/26/2019 01:44 PM|
First, Carmax only pays -$1000 on deep subprime loans which are about 10.5% of units sold, they pay -$255 on non CAF loans ("CAF" or Carmax Auto Finance is basically prime loans), which are non prime and deep subprime loans. The Spurce Point/Seeking Alpha report makes it seem like all of Sonoran should be paying -$1000, but the proper comparison to Carmax is the -$255 because Sonoran is a mix of non prime and deep subprime like Carmax. This is disclosed in Carmax 10k.
Second, what people are missing on loans is Carmax is not acting as a lender for non prime and deep subprime loans and Carvana is. Lenders make the largest part of the auto loan value pie because as Mark Jenkins put it at investor day "lenders do the most work: they underwrite the loan; the credit score; they price; they verify; in many cases, service the loan, et cetera. And as a result of doing most of the work, lenders also earn the most profits on a given loan". Carmax does none of those things for non prime and deep subprime loans and Carvana does, which is why there is such a big difference in profitability in nonprime and subprime loans between Carvana and Carmax. Deep subprime lenders need to make about $1500 of profit per loan (APR less servicing less losses less cost of funds) to cover their opex of about $750 per loan (SG&A and dealer management costs, which CVNA doesnt have) and earn an adequate ROA of about 2%. To understand what a hypothetic CVNA profit per loan would be and compare to KMX, first you would net out the discount the deep subprime lender pays on a loan (fee KMX pays) and add the profit per loan for the deep subprime lender (Carvana=lender), which results in about $1500 profit per loan for Carvana. CPSS the buy loans at par ($0 discount) and make a profit per loan of $1500; Exter buys loans at 3.5% or -$630 and makes about $2,300 per loan.
Third, Carvana loans should have roughly 25% lower cumulative net losses than subprime peers due to better severity and frequency. Carvana sells cars for roughly $1,100 below market (-4% lower list price or $800, lack of $200 dealer fee, and $100 tax saving from lower price) and they have lower VSC penetration of roughly $600 than average car (CVNA VSC penetration is about 30% vs industry of 60%) so the loss given default will be about $1,700 better or about -19% better severity (avg loan $18,000 and loss given default assuming 50% average recovery is $9,000; -1700/9000=-19%). The $1,700 lower amount financed and Carvana lower interest rate of maybe 1% results in monthly payments that are about 11% lower, which should result in lower frequency of maybe 5%. 19% lower severity and 5% lower frequency result in roughly 25% lower cumulative net losses. Carvana's 25% lower cumulative net losses result in a deep subprime loans being worth about $900 more (18,000 x 20% cumulative net losses x 25% CVNA lower losses= 900). Carvana probably has 1% lower APRs, which then reduces loan value by about -$400 (18,000 x 2.3 WAL x 1%)
So to finalize, Carvana loans are worth $1500 (combined dealer fee and deep subprime lender profit per loan) + 900 from better CNL - $400 lower rates = Carvana subprime loans worth $2000, which is implied by Equitrust financials
Also, another thing worth pointing out, it is very typical for a specialty finance private equity firm (like Guggenheim in the case of Carvana) to help out a business in multiple ways by owning the equity in the business and buying the loans at a decent price. This frees up management to focus on building their business instead of trying to find a buyer of loans every few months when there's no track record of losses. The specialty finance private equity firm makes money on the equity and makes money on the loans because they understand the business. The private equity firm will buy the loans until there is enough scale and track record to securitize the loans, and then the business gets better pricing and profitability on the loans (as Carvana will likely do). There is nothing sinister or uncommon about that relationship unless the private equity firm has a sweetheart deal and overpays for the loans, which is the case of Guggenheim and Carvana, there is no evidence to support that claim
|Subject||Re: Re: Re: Re: Re: Magic Johnson's Equitrust Life owns Carvana loans|
|Entry||02/26/2019 03:05 PM|
Typo, among others
There is nothing sinister or uncommon about that relationship unless the private equity firm has a sweetheart deal and overpays for the loans, which in the case of Guggenheim and Carvana, there is no evidence to support that claim
|Subject||Re: Re: Re: Re: Re: Magic Johnson's Equitrust Life owns Carvana loans|
|Entry||02/26/2019 06:04 PM|
KMX has been doing auto finance a long time and tested a captive subprime lending unit before. Why do you think CVNA will be better? Also, do you disagree with this article on the economics? https://federalautoloan.com/blog/carmax-to-shy-away-from-major-role-in-subprime-lending-space/
|Subject||Re: Re: Re: Re: Re: Re: Magic Johnson's Equitrust Life owns Carvana loans|
|Entry||02/26/2019 09:54 PM|
I suggest you read my comment 22 for bridge from Carmax to Carvana. Maybe I didn't make it clear enoughso here is second attempt
Carmax pays -$1000 to a subprime lender like Exeter as Carmax is not doing the underwriting, pricing, credit scoring, servicing, etc. Exeter makes on average $2,300 per loan (see comment 22 for Exeter loan economics from S&P and S-1). 20% of Exeters loans are Carmax loans btw, see S&P link below
Carvana has prices that are $1,100 below market (lets ignore the fact that Carvana has prices -$1,400 below Carmax since Carmax has prices that are 2% above market according to Yipitdata) and has lower VSC attach of roughly $600 per car. Therefore Carvana loss given default is $1,700 better than peer subprime loans or roughly 19% better severity (typical subprime loan of $18,000 x 50% recovery rate = $9,000 loss given default; Carvana loans have $1,700 better losses therefore severity 19% better because 1700/9000=19%). The $1,700 lower finance amount also likely results in 5% lower frequency because monthly payments about 10% lower; lower monthly payment results in lower likelihood of default. Therefore Carvana cumulative net losses, which are function of severity and frequency, are about 25% better or $900 per loan (18,000 x 20% Exeter cumulative losses on subprime loans x 25% better Carvana loan losses = 900).
Carvana also has probably 1% lower APRs, which reduces loan value by $400 (18,000 x 2.3 WAL x 1% = $400).
Now lets bridge Carmax to Carvana because Carvana acts as lender, has better loan losses, and lower APRs
Carmax pays -1000 to Exeter
Exeter makes +2300 on loan
Carvana loans have 25% lower losses, which make loans +900 more valuable
Carvana loans have 1% lower APR, which make loans -400 less valuable
Therefore, Carvana loans worth $1,800 = -1000 + 2300 + 900 - 400
The difference from $1,800 vs $2,000 in comment 22 is that dealers typically $650 to Exeter vs about $1000 for Carmax. The main difference between Carmax and typical dealer is that Carmax has prices that are on average 2% above market or $400
Also, what the article doesnt mention is that Carmax got out of subprime lending primarily because they worried the high repo rate in subprime would hurt the Carmax brand
|Subject||Re: Re: Re: Re: Re: Re: Re: Magic Johnson's Equitrust Life owns Carvana loans|
|Entry||02/27/2019 05:25 PM|
Spruce Point thesis just got destroyed
|Subject||Re: Re: Re: Re: Re: Re: Re: Re: Magic Johnson's Equitrust Life owns Carvana loans|
|Entry||02/27/2019 05:46 PM|
The stock is up 30-40% since that bozo published - is there really anyone who is short Carvana because of him? Clarification of the lending arm aside, wasn't this a noticeably large miss on everything? I feel like this stock is in one of those twilight zones where any news at all is considered a positive.
I will give them credit, I love the Finance Platform 101 thing and how they handled this. I wish more companies did this as part of their IR efforts, because most slide decks aren't terribly helpful. It's nice to have a company just explain how the business works. Probably saves them time from ansswering fewer of the same questions from us idiots.
|Subject||Re: Re: Re: Re: Re: Re: Re: Re: Re: Magic Johnson's Equitrust Life owns Carvana loans|
|Entry||03/06/2019 03:19 PM|
Equitrust 2018 financials show improving loan performance, consistent with CNL curves Carvana released
|Entry||03/20/2019 09:38 AM|
I have no position in CVNA....but got this email this morning.
|Subject||Gain on sale|
|Entry||08/07/2019 08:31 PM|
it looks like gain on loan sales contributed $918 of $3,175 GPU in the quarter...
1) is that sustainable as the business grows?
2) why / how does that gain originate? (I.e Other auto finance companies don’t have the same gain on sale or nowhere near the size per loan/car).
3) financials mention an agreement to repurchase sold receivables in the future at fair value. Obviously those being repurchased at a future fair value aren’t the same as those being sold for a gain!?
any insights from anyone would be sincerely appreciated.