What I've Been Working On: Including Free Alt Data for $CVNA and $CDLX

You may have seen my Twitter thread earlier today here and this post will basically be rehashing providing some more context in a way that isn’t character capped. I’ll also do an article on CDLX probably next week as I’m just slammed through Monday with what hopefully ends up being exciting news. Let’s begin.

The tweet is a great summary of what I’ve been tinkering with and how it fits together, so we’ll copy paste:

1. Democratizing access to alt data - Free coverage of $CVNA and $CDLX (https://clarity-markets.com/datasets)
2. Making >$50m on CVNA
3. Empowering X analysts (and Substack!)
4. My humble goals to reshape equity research

For this article we’ll focus on point 1, and relegate the others to when I have better updates and more time!

Democratizing Access to Alt Data

Why:

I’m a huge proponent of alt data for investing, just not necessarily in the industry typical way. The standard is to check if Yipit/MScience/some CC data provider have a data set on a ticker you are interested in. This has 2 major problems

1. The vendor is incentivized to obfuscate methodologies

   1. Obfuscation makes the vendor sound much smarter and thus justifies fees

   2. Obfuscation makes products seem standardized, when they definitely aren’t

2. The vendor is a vendor, not an analyst

   1. Extremely valuable data points may be missed/overlooked

   2. Coverage must be “consistent” and “measurable”

For obfuscation the results can be pretty messy. A great example is Yipit reporting SEA Limited GMV. I forget the exact numbers, but essentially the expectations based on alt data going into the report were much higher than the actual reported values. This led to the stock being cut by something like 30-40% as a plethora of firms who thought they knew what they were walking into got rugged.

Many in the industry don’t fully grasp how coverage is unique per company. Something like CVNA is uniquely alt data enabled in a way that is vastly more accurate than SE. Vendors however typically want to charge the same amount per data set and treat them as equal from a consumer facing perspective. This is an obvious inefficiency in my eyes. Imagine if an ER heart surgeon charged the same as a nurse practitioner because it’s both “healthcare”. That’s the scale of the silliness obfuscation of “data” and “software” enables.

For the vendor being a vendor, it’s just a natural thing that the closer you are to the work, the better it is. Would you rather read the 10-Q of a company you’re following or only have access to the Motley Fool summary? Vendors try their best, but given they want to obfuscate, there’s bound to be some drops. Additionally coverage needs to be consistent and productized typically. This causes some large gaps where something that isn’t necessarily a clean weekly snapshot can still have HUGE benefits. Vendors don’t want the data equivalents of call options, they want the data equivalents of bonds. Call options can sometimes really outperform bonds.

That’s the basic gist of my issue with the alt data industry as is. As someone really into long term equilibriums, I don’t view the current environment as stable. Due to rapidly decreasing software development costs, rapidly increasing data availability, and a progressively more technologically enabled finance work force, I believe the industry will be quite a bit different a decade from now. I’d like to be a part of that and here’s how.

How:

Let’s discuss CVNA and CDLX

CVNA:

CVNA is THE alt data stock I’m not sure I can emphasize this enough. Like yeah base rates of -99% to back again are terrible, but the stock simply should never have been -99%. It is not a normal business in terms of opacity. There’s a lot of companies where you get a guidance that’s maybe useful, take some wild guesses on assumptions, hope an insider leaks info, and build out a vague model.

CVNA you just know what they’re going to report 2 months before they do. The margins become mathematics. It’s honestly remarkable. I haven’t done the math but I’d bet if you bought CVNA each Monday close and sold each Tuesday close, essentially getting Yipit report day exposure, you’d have some absurd Sharpe as the numbers just kept coming in better and better. The fun thing is, you could know what Yipit would say before they reported it, and I did.

I build my own alt data tool for CVNA and as mentioned in the intro, I’m providing free access (to a good chunk of it).

You also don’t need to do much for access.

1. Simply go to https://www.clarity-markets.com/datasets/Carvana

2. Make an account

3. Download daily files on a 2 week delay

I’m not full on Robinhood though, not that rich. Live data/commentary are premium locked and quarterly sheets are on a 1 quarter delay, but it’s I believe the first instance of free Yipit level alt data. (We’ll see how that goes, I love you CVNA guys, don’t kill my project thanks).

It’s early so we don’t yet have the full suite of dashboards and email toggles blah blah. WIP and hope it helps. Feel free to send any feedback to indra@clarity-markets.com or my Twitter DM’s. If you’d like to subscribe fully and can’t swing $500, also feel free to reach out.

CDLX:

CDLX is the call option supreme of data coverage, and

Austin Swanson

is pure concentrated alpha in that arena.

Like I mentioned most providers want the “bonds” of data coverage. Stable, consistent, you can point to it on an 8-K. CDLX (and I assume many other smallcaps) will just randomly leak a ton of great info that you can methodically capture!

Probably the best example is when Starbucks ads came back onto the platform in a limited basis earlier in 2024. Austin noticed via his data collection there was a Starbucks ad. Starbucks used to be >10% of CDLX billings but hadn’t been on the platform in months. In hindsight of course it didn’t end up making a huge impact, but we have to consider the moment in time:

1. CDLX was not pricing in growth

2. CDLX is a super opaque business

3. CDLX is a fixed cost business, so layering on volume is a big deal

Combine those 3 facts and you get a very silly intraday move when someone notices a Starbucks ad. I believe it ended up being >10-15% as the information disseminated. That one little tip from Austin if you had a time machine was probably worth >$10m, all for a measly couple hundred in subscription fees.

As a result the plan is to work with Austin to provide similar freemium access to his CDLX reporting. You’ll get some of his data on a delay that helps you with your research, and if you’d like to subscribe for real time invaluable insights, that’s also available.

Goal is to have that live in an early state over the next couple days. Same deal, https://www.clarity-markets.com/datasets it will pop up soon.

The March Continues:

These two companies, while being the majority of my portfolio and posts, are not the only things we plan to cover. We will try to maintain the freemium model to the best of our ability (while also remaining solvent). There will also be a variety of other exciting things I hopefully get to share over the coming months. Maybe it works out, maybe it doesn’t. If you are a reader that has some kind of unique research, or insight, or quantitative metric you’d like to share (for money even), feel free to reach out. Let’s make equity research a bit more efficient one step at a time.