Banks, CVNA, CDLX, Process

It’s been awhile since I’ve written an article, busy news cycle and all that. Figured I’d toss in some updates on pre-earnings CVNA/CDLX thoughts as well as why I’ve been spending time on banks and the general process. Feel free to read each segment as you wish. If you’d like to subscribe we can get the button out of the way here. Paid or free is your choice.

CDLX:

As you may know CDLX has appreciated meaningfully this quarter. After falling below $3 post-SVB collapse we’ve round tripped back to ~$8. It’s currently a ~45% position in my PA and ~27% in my managed account. So what’s up?

First we got a confirmation of JPM launching the new ad server UI in March. I’d estimate this is ~40% rolled out but hard to know exactly. Probably see a full rollout sometime in Q2. As far as I know CDLX has not changed their GTM with advertisers and no official announcement has been made, so we probably won’t see significant revenue impact until maybe Q3. Regardless, the UI looks great and I truly fail to see how it doesn’t cause uplift, which underpins the entire thesis. Hopefully we get some more metrics and commentary here tomorrow or after Q2. I expect an analyst day at some point to outline Karim’s vision of the company and priorities, perhaps even tomorrow. Fingers crossed. Surprisingly this release had basically no impact on the share price, fair perhaps given previous managements lack of follow through.

The stock moves are largely centered around two separate releases. The first one was a guidance update for Q1 released in early April, essentially showing revenue ~10% better than expected. This was absolutely excellent. It showed that CDLX was much more robust than expected, growing 10-15% YoY ex-SBUX in a horrible advertising environment even prior to the new ad server, which only began limited rollouts in the last week of March. Given the FCF neutral/positive goal by Q3 requires ~25% YoY revenue growth, tacking on another 10-15% from the new ad UI seems much more doable than 20-25%. Overall it simply inspires confidence in advertisers view of the model, which was perhaps foggy give niche risk (sorry SNAP and PINS…)

Another big implication here is that if 2023 revenue is 10% above the expected baseline, that’s an additional ~$15m+ of liquidity. Given the earnouts at the time were not resolved and macro was foggy, getting clarification the underlying business was robust was a huge positive. Happy to see the pre-release and curious to see what drove the outperformance given previous guidance was only a month earlier.

The other significantly positive news was that the earnout disputes resolved. Ironically they actually lost the dispute very very hard, and owed almost double the amount they projected! This was due to a calculation error with ARR, etc. The total payment now totals around $200m with $76m paid in cash and something like 3.5m shares, 10% dilution.

Couple theories here as to why, but at the end of the day it was a fuck up, the ARR multiple earnout agreement was a fuck up, and the buybacks in early 2022 were a fuckup. Hence the “resignation” of the prior CFO.

Anyways, the benefit of this fuck up is that the 2nd earnout payment is projected to be $0. Of course given their stellar track record of financial decisions and projections, perhaps that’s worth doubting. The way I see it the independent arbitrator just told them how it’s calculated, so if they proceeded to mess it up again I’d be astonished. This 2nd payment being $0 basically saves us ~$30m in cash and ~15% in equity dilution, but we get an additional $35m on the first payment and an additional 5% dilution on the first payment to 10%. Net effect is maybe -$5m cash, but we save 10% in dilution. Last I checked CDLX was worth a touch more than $50m, so a pretty good deal for us and even better than prior base cases.

The remaining liquidity position is probably ~$40m of cash and $60m of revolver after Q1 with $10m or less of quarterly burn in Q2 and expected FCF positive in Q3. Plenty of runway.

At this point with CDLX, unexpected bankruptcy risk seems off the table with the earnout resolved better than expected. The underlying business is thrumming along nicely and outgrowing most every other ad player (ex-SBUX). The new UI should roll out fully in Q2 on Chase and only further cement this advantage, with probably announcements of other banks by the end of the year. Great steps and what appears to be a positive inflection for the first time in a long time.

At this point my concerns are rather limited, but still quite real. Firstly, we don’t know why Q1 beat. We know T&E is heavily outperforming which may fade if macro deteriorates further. If the beat was due to one offs or non-repeating trends CDLX may struggle to reach FCF breakeven in Q3 despite the new ad server. This would be a liquidity drag and make the story more fragile, but likely still fine given the price. The other concern is that perhaps I’m totally over-estimating the new ad server. I’m guessing it can bring them back to 25-30% annual revenue growth when combined with Bridg and a fresh launch campaign with FI’s, but if this is wrong my return estimates may be far too optimistic. Thankfully at $8 I simply just need “ok” and them to put up something that doesn’t bankrupt them. Do 15% YoY again and we’re alright.

So overall, good stuff from Karim and the team. Disaster avoided for now and numbers moving in the right direction with a very cheap valuation is quite appealing, hence the ~45% position in the PA.

CVNA:

CVNA is being a bit more of a problem child, but less so in a bad way and more so in a “this thesis is going differently but still seems to work way”. The Q1 report will likely be around 77,000 units and is tracking to be around the same in Q2 per AlternativeAlpha. The Q1 guidance update indicated that GPU’s were surprisingly robust despite not selling down their loan backlog from Q4. Additionally the cost cuts are expected to be done a quarter early, clocking in at $420m of SG&A at the mid-point, roughly $5400 per vehicle.

To be frank, I expected costs to be cut at a faster pace and units to be at a higher level. Interestingly CVNA has taken a different pace. Instead of abandoning edge markets and receding to their central USA core to weather the storm, they’ve actually been expanding. New IRC/ADESA recon facilities in CA/OR/MA/NY and sort of generally along the edges of the country have been taking an increasing share of reconditioning. Delivery and shipping fees have cut demand but made purchase intent far higher, lowering returns and creating a higher quality funnel. The overall effect seems to be GPU’s that could potentially be $4500+ structurally.

To back up that assertion, Q2 is expected to be ~$4250 with maybe $250 in one-off finance GPU. This is with an average age of car sold being >100 days versus current sales at 80 days and falling. If we call it ~$2500 in depreciation per year, saving a month equates to $200 higher GPU. If we call it ~20% of vehicles sold in NY/CA/OR/WA/etc and those vehicles are inbounded 500 fewer miles on average, that’s probably $500+ per car, or $100 of GPU. Add in $200 of delivery/shipping fees and we’re at $4500. I’d imagine we can get down to ~40-50 days from 80 and get closer to $4800. Not bad, but somewhat hypothetical until we get the Q1 numbers in our faces.

$4500 of GPU of course doesn’t quite cut it with $600m of interest expense and $5400 of SG&A. Given units are tracking flat QoQ we need some SG&A cuts as we aren’t getting leveraging. There are ongoing cost efficiency decisions being made with market consolidations, reducing 3P logistics, continued layoffs, advertising cuts, and probably more. Hard to quantify but if they can’t get this to $4000 per vehicle I’d be shocked considering they’ve done $3k before. A bit slower and steadier than I would have liked, but such is life.

CVNA’s path is complex and there’s more going on than I’m going to opine about here. I do think it is possible for them to get it done on the unit economics, but math needs to become reality for us to make any money. That said, short term estimates are totally off base and CVNA has plenty of liquidity to blow those out of the water this year which may help with their cost of capital issue. I severely doubt we see any kind of dilutive equity raise as the Garcia’s and management as a whole does not seem interested despite whatever random headlines are tossed out each day. The debt restructuring also probably won’t happen. Status quo and business execution are all that matters.

My only real concern here is units. I can pencil unit econs, I can picture the competitive advantage, but if we can’t translate that into cars sold the math simply doesn’t work and customer perception is being tested. I follow units via alt data as mentioned above. Currently we seem pretty stable at ~76k or so in Q2 and potentially on the upswing, but if degradation continues we may have some issues. I’d imagine they blow guides out of the water on Q2 AEBITDA but volume weakness is a wildcard. CVNA is currently ~20% of my PA and ~12% of my managed account, we’ll see how it goes.

Banks/Process:

If you follow me on Twitter you’ve probably seen an incessant focus on banks, primarily PACW and FRC. A natural question of course is “why are you, the guy writing about investing on a 3-5 year horizon trading bank stocks”. This is a fair question and one I’ve played with perhaps too much, and I figured I’d riff as I find the prevailing common wisdom lacking. Disclaimer is that this was written at 1am and more a stream of consciousness. Feel free to skip. Without further ado.

The bank saga started on March 9th when SVB failed and I had the fun task of checking portfolio exposure given I sometimes invest in what we could call “less than robust” balance sheets. Everything was fine and dandy besides our good friends at CDLX, who bank at PACW and have a LOC stipulating they cannot have significant deposits outside PACW. Given an imminent large earnout and unclear directive on uninsured deposits, this lead me down the PACW rabbit hole.

To be clear, prior to this the only bank I followed was ALLY due to the relationship with CVNA. PACW was my career first “bank deep dive”. My initial checks were that PACW was vaguely liquid and the Fed quickly backstopped SIVB deposits, alleviating any concerns for CDLX (which I still believe is true, I won’t lose sleep if PACW fails).

Interestingly at the same time financials as a sector were seeing broad based panic. Some fund managers I spoke with offered absolutely terrible justifications for that panic and indicated a total lack of fundamental knowledge of what they owned, which was fascinating. Speculations that SCHW or USB or even BAC would fail were seriously entertained for a week or two. We pretended that 50% of deposits leaving in less than 48 hours was a reasonable stress test and all bank balance sheets should be immediately marked to market. Little discussion was focused on NIM’s or loan portfolios or FHLB liquidity. My propensity when something doesn’t make sense is to dig deeper, and the conclusion was vaguely thus:

As usual, the stories certainly weren’t pretty, but it wasn’t financial Armageddon. BKNY was not about to collapse and is actually continuing to spit out dividends. JPM could easily come out of this ahead if they do the FDIC enough favors to avoid stringent litigation. If BAC indeed fails whether I was invested or not probably won’t even be a top 10 concern given the implications. That being said, I do own CVNA which makes me quite credit exposed, so I figured I’d take a look at the worst performer, FRC, and see if I had something to worry about.

FRC was frankly, a shit show. They had some of the worst uninsured deposit ratios, heavy geographic overlap with SIVB, one of the lowest NIM’s, were already getting squeezed on deposit cost, and had a book almost entirely of low yield mortgages originated in 2020/2021. When the WSJ article dropped that they’d seen $70b of deposit outflows, penciling out some pretty basic math showed net income was dangerously approaching $0 and given the loan book duration, would be a very hard hole to get out of. Additionally the company was totally radio silent, hardly confidence inspiring.

PACW on the other hand had much more robust NIM, was proactively managing liquidity, a shorter duration loan portfolio, less deposit concentration, etc. PACW was also proactively giving updates on deposits that clearly penciled to positive net income and if liquidated it was not worth negative tens of billions. Clearly one bank was better than the other, which got me wondering.

At the core my job is to make money. It honestly doesn’t particularly matter the method. If someone offered me 3 to 1 odds on a coinflip I’d be a moron to not take that all day long. The only work there is in actually verifying I’m getting 3:1 odds on a coinflip. Is it a long term compounder? No. Does it have a moat? No. Is it a quality investment? No. But it mathematically makes money.

So if I see two assets where the downside case is $0 and the upside case is 2-4x, and one asset is very obviously better than the other, what should I do? I floated the idea of a 2:1 pair trade with 2 parts long PACW and 1 part long FRC puts in early April.

Historically I have not done pair trades, or even shorted at all as “that’s not what Buffett would do”. Additionally, I hadn’t looked at banks in any seriousness before March 9th and everyone was saying how hard they were and to stay away, etc. Humility is buzzword thrown around a lot to not think you know anything. I’ve been quoted Dalio’s “I’m just a dumbshit who doesn’t know anything” far more than ever necessary. The clear implication was don’t get involved and let your betters figure it out.

A month later when that trade was up 10x on the back of FRC’s failure I came to the sudden realization that I could not care less about looking good and far preferred doing good. In general I’d say markets are quite efficient. Every so often you get a “fat pitch” that looks very stupid to a lot of people, and the price of alpha is taking a swing. Being unwilling to put money down on something blatantly obvious with no clear counter argument based on vague notions of intellectual “correctness” is a mistake I do not intend to make again. Of course I’m not saying go put it all on red, but the failure state of that 2:1 pair trade seemed obscenely unlikely. So what did I do?

Last Tuesday I bought a PACW position pre-earnings and sold for a +15% gain right after as liquidity and NIM concerns were eased. I was not trying to be a “trader” but the two narratives were exceptionally clear. First was that the stock would inevitably move up due to numbers being fine. Second was that FRC was a total failure and probably dead within the month. Booking a 15% gain for an hour of my time was quite worthwhile.

Next I put 10% of my portfolio in FRC puts expiring in late May. Deposit outflows were actually $100b versus $70b as expected and net income was clearly negative. There was a ~$30-$40b hole in the balance sheet that was unsolvable excluding a bailout. To protect against this possibility, I had a small equity position as I figured FRC would easily 5-10x over time if solvency was guaranteed. As of Monday those are worth a touch over 2x the purchase price. Not the 30x+ of the initial pair trade prices, but quite good for a few days.

I say this not to toot my own horn, as it’s entirely possible I get absolutely smoked tomorrow with 3 positions reporting earnings. I say this as a reflection on the dangers of getting too intellectual with the work. Thus far my returns if I simply bought an equal weighted portfolio and held and didn’t transact would be absolutely atrocious. Every single shred of realized gain has been from nutting up. META at $90, CVNA at $4 (hopefully?) CDLX at $3 (hopefully?), shorting FRC, longing PACW, TCEHY at $26, etc. All of those investments besides FRC were totally horrible if not managed as they were. Of course maybe I eat those words tomorrow or later this year, but we’ll see.

The lesson here of course is not dump money into every trade. I’m currently long a 1% position in PACW following announcements of “strategic alternatives” from Bloomberg. That position could very well go to $0 as despite my math showing PACW is fine, reflexivity is a bitch. Ignoring my vested interest, I find it extremely distasteful that we are now at a stage of shoot first ask questions later. The Bloomberg article has potentially caused another bank failure as PACW had $28b of deposits prior to whatever is pending for tomorrow according to an 8-K released tonight. Those strategic alternatives were already underway and communicated. I have no confidence in where we are headed from here, thus a 1% position vs 10% on the FRC short and my overall degrossing of the portfolio as I unwind the FRC trade.

I likely won’t be transacting this frequently again for the foreseeable future nor talking about banks as much as I generally dislike the LT ownership odds and when credit risk is relevant it’s a total pain I don’t wish to deal with. I do plan however on continuing to focus on flexibility in investment vehicles. A circle of competence does not have to be a prison from opportunity and I hope this post-mortem has been of use for those thinking through similar things. At the end of the day returns are based solely on you, and you can either trust yourself or not.

If anyone wishes to chat CDLX/CVNA pre or post earnings I’d be happy to via Twitter @Indrastocks if you aren’t aware. Best of luck and may the fourth be with you as our favorite CEO Lynne Laube would say.