A lot was made last week about $CVNA improving GPU. Investors debate impacts from ASP/turn improvements vs finance receivables. One area not discussed is the reduction of very low volume delivery locations. Carvana had 80+ locations with < 25 deliveries last Q!

I have highlighted to my customers how $CVNA ASP growth has been driven by mix. Here is a YoY comparison of the top 20 selling make/models on Carvana. Top Increases •Jeep Wrangler •Ford F150 SuperCrew •Jeep Cherokee Top Decreases •Ford Fusion •Nissan Altima •Toyota Camry

Or you can just use the list price vs KBB metric I already have highlighting the best value cars in Carvana’s ~40k inventory available for purchase. Obviously want to expand this outside of just Carvana. $CVNA

Interesting thing to track moving forward. I will add some additional data points to this in the near future, but a closer spread between $CVNA retail list price and their internal KBB metric does not seem to be a good thing for margins.

3/ This divergence can be due to many things. However, this internal KBB metric is updated frequently. $CVNA is making improvements this quarter on reducing the age of inventory that may help protect from large market movements, but it remains to be seen how margins are impacted.

2/ Now flashing forward to this past weekend, we can see average list price is up significantly since January (28.1k vs 23.7k). However, the difference between the internal KBB metric and the $CVNA list price has plummeted (1.5% vs 12%)!

1/ Mini-Carvana thread. It is widely known that $CVNA retail list prices have been going up. See the table below from January showing the difference between the average retail list price separated by inventory age and the $CVNA internal KBB metric. Pretty Notable.

Stark difference between the $CVNA core inventory and the third party dealer marketplace inventory. •Core list price ~27.7k vs internal KBB metric of ~26.9k. •Third party list price ~30.1k vs internal KBB metric of 32.6k Third party accounts for ~10% of available inventory.

In a quarter where $CVNA saw retail sales drop significantly, here are the top 10 cars models that saw an increase in units quarter-over-quarter. $TSLA takes two of the top 10 spots.

Random $CVNA things: •Sold ~1/7th the # of cars with a list price over 100k in Q1 compared to Q4 •Tuesday was the slowest weekday retail sales # in over a month •3rd party listings as a % of total listings has been decreasing for ~2 months •Inventory age ⬇️ •Days to sale ⬆️

Here is a breakdown of Carvana's Q1 retail sales by make. The "other" value lumps in the bottom 35 makes (45 total). $TSLA ranks 25th out of 45. $CVNA

Carvana average retail list price end Dec: $24,250 Carvana average retail list price now: $26,900 Projected retail ASP QoQ: ⬇️ $CVNA

TLDR: $CVNA logistics performance varies greatly by each location. Some mid-sized locations that deliver 1/5 the cars as larger location end up taking double the amount of time to get a vehicle to a customer. Largest % of deliveries taking double the amount of time shown below.

3/ Digging deeper, roughly 10% of $CVNA’s retail sales this quarter have taken >14 days to be delivered to a customer, roughly double the average for the quarter. Of the 154 $CVNA locations that have delivered >50 cars this quarter (239 total), here are some of the offenders

6/ It will be interesting to see as sales continue to grow back to recent historical levels, will the pared down head count at $CVNA introduce logistical delays as what have been experienced in the past.

5/ A lot of these locations are only moving about a fifth of the cars as the top few processing locations this quarter. Of the top 10 selling locations, only one location exceeds 10% of sales taking greater than two weeks to delivery.

4/ In the previous table, you will see Adesa taking the top of the list. None of these locations crack the top 50 in overall sales. It is interesting that the mid-tier volume locations are the pain points, while the top selling location has averaged 7.1 days to complete an order.

3/ Digging deeper, roughly 10% of $CVNA’s retail sales this quarter have taken >14 days to be delivered to a customer, roughly double the average for the quarter. Of the 154 $CVNA locations that have delivered >50 cars this quarter (239 total), here are some of the offenders

2/ $CVNA has averaged ~7 days this quarter from when someone buys a car on their site until it is signed, sealed, delivered to the customer (“complete”). This number tends to increase as sales volumes increase, and we are seeing that throughout this quarter.

1/ Carvana delivery logistics thread highlighting the vast differences between some of $CVNA’s most and least efficient order processing locations this quarter.🚗🧵👇

This quarter is shaping up to be a reduced version of Q1 from last year for $CVNA . Sales are down but we are seeing a surge into the end of the quarter like last year, where the surge continued into early Q2.

For anyone who just listened to the $CVNA earnings call, a lot was discussed about needing to turn over inventory in under 90 days. Q4 ended with roughly 42% of available retail inventory older than 90 days. Last weekend we were at 47%, but just under 2500 less total cars.

3/ $CVNA earnings 5. QoQ single digit ASP Declines (Q3 to Q4, Q4 to Q1) 6. QoQ increase in inventory age 7. Stable third party sales as a percentage of overall sales, but continual increase in third party listings as a percentage of total listings

2/ 1. 85,750 $CVNA Q4 Retail Deliveries 2.Q1 tracking to be >10% decline QoQ 3.Early Indications of weekly sales volumes bottoming 4.Very large inventory drawdown, which may be coming to an end to meet predicted sales volume surge from tax return season (like last year)

1/ With $CVNA reporting tomorrow, here are some observations and potential points of conversation for the call based upon my data gathering service. This is info I have provided to my customers many weeks ago and in near-real time throughout the quarter.