Discover more from Stock Market Nerd
News of the Week (May 22 - 26)
SoFi; Nvidia; Snowflake; PayPal; JFrog; Match; Uber; Amazon; Meta; Nanox; Upstart; Palo Alto; Cresco Labs; Earnings Round-Up; Macro; Portfolio
Welcome to the hundreds of new subscribers who have joined us this week. We’re delighted to have you and determined to provide as much value as possible.
1. SoFi Technologies (SOFI) -- Leadership Interview
We published a defense of the SoFi bull case after Wedbush aggressively downgraded the name for reasons we disagreed with. Well? This week, CEO Anthony Noto sat down with JP Morgan for a conference. Most of the same topics were covered and his answers were nearly identical to the explanation we provided.
The strong deposit trends SoFi enjoyed in Q1 continued into May. CEO Anthony Noto reiterated plans to add at least $2 billion in deposits this quarter. He was expecting just 35% of new accounts to utilize direct deposit but 50% of its accounts are doing so. The company has plans to continue rapidly growing its deposit base to fund originations thanks to the “very high quality” of these deposits, its bank charter, and its conservative leverage ratios. Interestingly, these deposits are not coming from the Silicon Valley Bank debacle, they’re coming from large money center banks. That points to this trend being durable and not a temporary reaction to a systemic shock.
SoFi continues to reject 70% of applicants and continues to tighten its credit parameters. It never got too liberal with credit bands during the pandemic boom. In other words, the rapid origination growth is not coming at the expense of credit quality.
It continues to be able to pass on higher rates to customers. This works wonders in maintaining loan valuations as well. Its average interest rate is rising faster than benchmark yields, demonstrating its pricing power.
Its life-of-loan loss rate continues to be FAR below its risk tolerance. It’s 4.5% vs. a 7%-8% loss rate tolerance.
Loss curves continue to outperform as its newest loans season.
SoFi’s guide continues to assume -2.5% GDP growth. That’s likely too pessimistic and a better result would feed its growth.
Fair Value Accounting vs. At Cost Accounting:
Noto called fair value accounting the best method for loan valuation. Why? Because under fair value, loans are marked-to-market every 3 months with those impacts being reflected on the income statement and balance sheet that same quarter. That’s not the case with at-cost accounting at other banks. Under that method, banks set current expected credit losses (CECL) upfront and adjust occasionally over time based on performance. Noto pointed out that at cost accounting frees companies to report healthy equity balances one day and be taken into receivership the next as they hadn’t recorded losses on financial statements to accurately portray capital ratios.
He corrected the Wedbush analyst who downgraded SoFi partially due to concerns of loan valuation falling because of changing loan classification from held-for-sale to held-for-investment. That change has zero impact on valuation methodology.
Finally, Noto spoke on fair value adjustment assumptions from last quarter. SoFi was criticized for raising the fair value of its loans amid soaring rates and rising risks. Two things. First, this is related to its strategy to hedge away all interest rate risk. It locks in cash flow to remove downside and forgoes some potential upside in doing so. That’s why the balance sheet loan balance impact has been so small. And again, credit quality is why the losses have been so tiny. Secondly, he told investors that its 2.97% loss rate for Q1 was assumed to rise above 4.5% based on economic outlook. That likely pessimistic forecast was part of the fair value adjustments… which again are set by a third-party firm and not SoFi directly.
On No Loan Sales Last Quarter:
SoFi’s singular goal in the holding or selling loans decision is ROE optimization. Recently, it has been able to generate a 6% net yield from holding vs. 4% for selling. With its excess liquidity and conservative leverage ratios, it has the freedom to select the best yield… so that’s what it’s doing. It did the same thing 5 years ago when rates were rising amid the 2018 taper tantrum. When conditions improved and selling yields became more attractive, it resumed loans sales once more.
Noto is fully confident in capital market demand being there to sell loans at favorable gain on sale margins. He talked about the Q1 ABS deal, exploding net interest income, low charge-offs, comparable deals and discussions with the demand side as evidence.
On the Tech Platform (Galileo + Technisys):
Total contract value from the current “robust pipeline” is “multiples higher” than at any point during Noto’s tenure. The pipeline consists of far larger customers than what the tech segment had been pursuing and servicing in the past.
The revenue ramp from these wins will not be immediate. It will start with moving new customers to the updated tech stack and then eventually migrating the client’s pre-existing user bases to the platform.
The firm is confident in growth here accelerating towards 20% by end of year.
Real time asset and liability management in light of the Silicon Valley bank issues has been a compelling selling point. Its ability to track deposits in real time and by customer segment is a small but tangible edge.
Noto told us to expect travel to simply be the first non-financial service offering to allow members to save on purchases and to expand SoFi use cases. He teased auto and entertainment ticketing products. He expects this vector, financial services and the tech platform to lead to more of its growth coming from non-lending segments over time.
SoFi in 5 Years:
Noto thinks the pace of quarterly member adds (currently around 400,000) could at least double by 2024. He wants SoFi to have over 20 million members 5 years from now and 40 million total products. His goal is for SoFi to be a top 10 financial institution in the USA with plans to expand globally from there (first in Latin America). It’s in Hong Kong currently, but not with the full product suite or anything close to it.
2. Nvidia (NVDA) -- FY Q1 2024 Earnings Review
This was the best large cap beat & raise I’ve ever covered. I do not own any shares. I’m just wildly impressed.
Beat revenue estimate by 10.3% & beat guide by 10.6%.
Beat EBIT estimate by 19.1% & beat guide by 20%.
Beat net income estimate by 19.4% & beat guide by 20.4%.
$11 billion revenue guidance beat estimates by 55%.
gross margin (GPM) beat by 330 bps; EBIT beat by 93%.
c) Balance Sheet
Stock based compensation = 10.2% of sales vs. 12.2% Q/Q & 7% Y/Y.
No buybacks this quarter vs. $2 billion Y/Y.
Dividend payments were flat Y/Y.
Inventory rose 45.9% Y/Y but fell 10.7% Q/Q as levels appear to have peaked.
$15.3 billion in cash & equivalents; $11 billion in debt (with $1.3 billion current).
d) Call & Release Highlights
32.7% 3-yr revenue CAGR vs. 24.8% last Q & 25.4% 2 Qs ago.
Data center revenue powered the outperformance here as AI use cases and compute needs explode higher.
Auto revenue was flat Q/Q and more than doubled Y/Y on a very small base of $138 million.
Professional visualization revenue rose 30.5% Q/Q and fell over 50% Y/Y on a small base of $622 million.
GPM was 66.8% vs. 66.1% Q/Q & 67.1% Y/Y.
GAAP operating expenses fell 3% Y/Y despite this massive demand outperformance. Wonderful combination.
Gaming, Auto & Professional Visualization Platforms:
Both segments are “emerging from inventory corrections.”
Gaming demand demonstrated “resilience amid a challenging consumer backdrop.”
Its auto design pipeline is now $14 billion over the next 6 years vs. $11 billion Y/Y.
Data Center – the AI-powered standout of the quarter:
The explosion of generative AI and language learning model (LLM) interest is vastly bolstering Nvidia’s growth. Leadership says that it has the most “versatile, energy efficiency and lowest cost approach to train and deploy AI.” The demand here is broad based. Cloud titans are all “racing” to deploy its graphics processing unit architecture in Hopper and Ampere. Jensen Huang called generative AI an “iPhone moment” and Nvidia was ready to pounce before the ChatGPT demand spike began. Its experience with building 5 of its own data centers and full stack approach also makes it an ideal partner to hold client hands and steer them to evolve in the most optimal manner possible.
Despite players like Amazon hard at work on building its own training and inference chips, it’s still relying heavily on Nvidia.
Meta also recently added Nvidia’s H100 GPUs built on Hopper to power its supercomputers.
BloombergGPT and AT&T are also utilizing it along with more large enterprises than for any competitor out there.
Its next generation of H100 (called DGX) is now being shipped. This extends use cases to on-premise environments and offers firms pre-trained templates to use to fortify models with 1st party data. ServiceNow was called out as an early DGX adopter.
Beyond AI model training, model inference is another Nvidia strength facilitating this historically strong print. It sees its inference capabilities as “orders of magnitude ahead” of others with “unmatched versatility across different workload types.” Importantly, data center revenue should keep growing sequentially in Q3 and Q4. This is not just a Q2 anomaly but the beginning of a more durable demand trend. It also expects to secure a lot more needed supply to meet pent up demand for H100 and its Hopper/Ampere infrastructure by Q3.
3. Snowflake (SNOW) -- FY Q1 2024 Earnings Review
Beat product revenue guide by 3.4%; beat revenue estimate by 2.4%.
Crushed $4.5M EBIT estimate by nearly $30M; crushed EBIT guide.
Beat $0.05 EPS estimate by $0.12.
Beat free cash flow (FCF) estimate by 54%.
b) Full Year Guidance
Lowered product revenue guide by 4%.
Lowered EBIT margin guide from 6% to 5%.
Reiterated 76% gross margin.
Raised adjusted FCF margin guide from 25% to 26%.
Next quarter EBIT margin guidance implies a material miss vs. estimates.
Despite this disappointment, it reiterated its schedule and path to $10 billion in annual recurring revenue.
c) Balance Sheet
Stock comp was 42.4% of sales vs. 42.5% Q/Q. This is quite elevated to say the least, but it did offset about 72% of that compensation with stock repurchases.
Stock comp = 108% of total FCF generation.
Roughly $4 billion in cash & equivalents. No debt.
d) Call & Release Highlights
“We are operating in an unsettled demand environment and we see this reflected in consumption patterns across the board. While enthusiasm for our products is high, enterprises are preoccupied with costs in response to their own uncertainties… This may continue near term.” -- CEO Frank Slootman
Consumption was strong through March but has been weak from Easter up until now. Week over week (W/W) growth ground to a halt. Its guidance assumes W/W growth resumes but doesn’t fully recover. Snowflake attributed this slowing to uncertain macro leading customers to seek cost optimization. They’re doing so by deleting stale data. Less storage means less revenue based on Snowflake’s consumption revenue model. Some of its largest customers switched from 5-year data storage to 3 years. Data query requests are still rapidly growing at 55% Y/Y… They’re just costing customers less due to less storage and so quicker fetching.
Its consumption-based model means it’s easier for customers to add or subtract costs vs. a software as a service-based (SaaS-based) approach with more pre-set commitments. This will mean slightly more downside amid poor macro and potentially more upside when things are good.
Importantly, leadership told us multiple times that this has nothing to do with competition. Its win rates are stable. This is a matter of weaker bookings trends as customers are “hesitant to sign large deals.” Still, when a firm carries one of the most expensive valuations in the market, misses are not easily forgiven.
“Productivity is not where we want it to be… we are not satisfied with our results.” -- CFO Michael Scarpelli
As an important aside, Snowflake and AWS have significant customer overlap. So? AWS slowing to 11% Y/Y growth in April can be taken as a decent hint of demand softening for Snowflake. SNOW investors should pay close attention to AWS Q2 growth rates to gauge how pessimistic Snowflake’s guidance could turn out to be.
79% 3-yr revenue CAGR vs. 88.7% last Q & 96.9% 2 Qs ago.
373 customers with over $1M in product rev vs. 330 QoQ & 206 YoY. It calls 590 of the Forbes Global 2000 its customers vs. 580 Q/Q and 513 Y/Y.
NRR was 151% vs. 158% QoQ and 174% YoY. 151% is still elite.
Remaining performance obligations (RPO) of $3.4 billion rose 31% Y/Y. 57% of its RPO is current.
Margins and cost cutting:
While poor macro persists, Snowflake will intentionally cut back on growth marketing and will slow the pace of hiring.
Snowpark is a newer Snowflake tool which frees developers to write and build within Snowflake and in a wide range of source code languages. This merges Snowflake’s data storage, querying, organization and leveraging prowess with an actionable ability to utilize this insight to create applications all in 1 place. This cuts down data transference costs while fostering cohesive simplicity and productivity. It also means developers can tap into data cleansing and quality issues in real-time while they build.
30% of all Snowflake customers are already using Snowpark weekly vs. 20% Q/Q. Consumption of this service rose 70% Q/Q. Its cost advantages over alternatives like Spark were cited as reasons for this strength.
AI relies on vast quantities of data to train models and make them valuable. How do customers combine and access all of their fragmented data silos in one place? With firms like Snowflake. That, to leadership, will make it a prominent piece of the current generative AI wave. This quarter marked the beginning with Snowflake delivering 91% Y/Y growth in AI/ML customers.
Snowflake’s recent acquisition of Applica (which creates language models to “solve real business challenges and understand unstructured data”) will be a key piece of its participation. It will also acquire Neeva for its “next generation search capabilities” to enhance Snow customers’ ability to build “rich, search-enabled and conversational consumer experiences” -- similar to Bard and ChatGPT. Neeva allows this to happen with significantly less technical expertise to lower the barrier to model building. To help ease usage further, it released new standard query language (SQL) tools to power data anomaly search without “mastering data science.”
Launched a manufacturing data cloud for supply chain management. Blue Yonder (a software supply chain company) is the first to re-platform onto Snowflake for this tool.
Gross margin strength was via more favorable cloud service provider contracts and the marketing pull-back helped EBIT margin further.
FCF was boosted by early May receivables collections.
There was no change in the firm’s approach to guidance. It’s not being more pessimistic than it has been in past quarters.
Azure revenue growth is now outpacing AWS and Google Cloud for Snowflake following Microsoft naming Snowflake a tier 1 partner. It appears this relationship is becoming more & more friendly.
4. PayPal Holdings (PYPL) -- Investor Conference
On Macro Trends:
The year continues to be off to a better start than expected with e-commerce growth assumptions rising from about 0% to low single digit % growth. PayPal’s growth is tightly correlated to this as a predominately online discretionary checkout player.
Discretionary spending is “returning a bit.”
Reiterated expectations for high single digit revenue growth for the year vs. original guidance calling for mid-single digit growth.
Newer custom cohorts continue to “perform extremely well” and engage more frequently with the brand.
E-commerce growth re-accelerating will feed PayPal growth and allow it to extract more value from its now leaner fixed cost base.
Cross-border volume (especially in China) is beginning to stage a comeback. This is higher margin transaction revenue for PayPal. The U.K. is even starting to show signs of a rebound which has been its weakest market recently due to the structure of their mortgage laws and spiking rates.
The unbranded segment moving downstream with PayPal Commerce Platform (PPCP), expanding globally and adding other high margin services. This has been a key theme of PayPal’s conferences in recent weeks and its last earnings call. It has a clear line of sight to unbranded margin expansion. It expects expansion to begin later this year.
Accelerating branded growth which is expected to come throughout the year.
Unbranded & Branded Teamwork:
Unbranded continues to perform extremely well for PayPal with a “really strong pipeline” and existing customers allocating more and more volume to Braintree. PayPal will take all the unbranded growth it can get despite the lower margin vs. branded revenue. Why? Because Braintree and PPCP come with the latest branded integrations for PayPal and Venmo branded checkout which bolsters its branded share.
PayPal’s long history means a maze of legacy integrations that need updating. Winning Braintree business gives them an easy touchpoint to implement these modern checkout flows. It has 1 out of 3 of its largest merchants on its latest flows which will reach 50% by year’s end. Its new PPCP tool for smaller businesses should expedite reaching 100% as it sells through channel partners. This mean all merchants under that single partner will be automatically upgraded to the latest and greatest PayPal/Venmo process.
On PPCP Prospects:
Schulman alluded to “massive pent-up demand for unbranded products” geared to smaller businesses. That’s what this is. He repeated sentiment around the old version of this (PayPal Pro) being “completely insufficient to address market needs.” PPCP, conversely, is “state of the art.” He expects this to be “one of the most successful product launches it has ever had” and compared his excitement to the BNPL roll out which rapidly rose to top 3 in the world despite a large head start for other players.
Teen accounts on Venmo are now live. This adds millions of potential users to its addressable market with a large chunk of those potential customers already having parents with a Venmo account.
PayPal saw its checkout preference and adoption rise from 2021 to today at roughly the same rate as Apple Pay. It outpaced Apple Pay in 2022 but is slightly lagging behind so far in 2023 according to Schulman. This performance would shock investors if they paid close attention to the growing FinTwit narrative of PayPal being a dinosaur. Not only are Venmo and Braintree not dinosaurs, but legacy PayPal also isn’t either. This preference is being powered by its omni-channel ubiquity and the fact that the world is split between iOS and Android. PayPal offers a unifying layer for both. It offers more checkout options than anyone else and sees no quality gap between its latest checkout integrations vs. any other products on the market. Not Shop Pay, not Google Pay, not Apple Pay… nobody.
Still, PayPal’s main source of competition isn’t other digital wallets, it’s other antiquated forms of checkout. 65% of checkout volume still uses manual data entry or card on file by single merchant. Digital wallets are more convenient and utilitarian vs. these other options. Digital wallets will keep taking share and branded PayPal is positioned to benefit.
Next Generation Checkout:
PayPal’s password-less, 1 click next-gen checkout is now rolling out to a select group of merchants. Its massive consumer scale means massive data scale and a massive card vault to store data and remove the need for customers to enter any data manually. All they need to do is verify their identity with duo authentication. Schulman said the team showcased this at a closed sales conference with “huge excitement.”
The 10% decline in non-transaction operating expenses this year was called the “beginning of what we need to do” by Schulman. It will get back to its 5-yr OpEx CAGR this year and wants to fall below it in 2024. He sees recent AI advancements as significantly cutting other costs for PayPal like legal and customer support. PayPal’s massive data scale gives it an inherent edge over the competition on seasoning built or externally utilized models.
5. JFrog (FROG) -- Investor Conference & a Study
JFrog Advanced Security (new product) is now live with 15 clients. The revenue ramp should start in 2024. 50% of its customers use its original security product called X-ray which offers software composition analysis. Advanced Security adds end to end use cases like container security and JFrog thinks the entire X-ray customer base is ripe for upselling to the newer product. This new tool routinely consolidates 10+ point vendor solutions to lower total cost of ownership for clients… ideal in today’s environment.
JFrog paid for a Forrester study to quantify the savings of its automated software delivery architecture. According to the findings, clients save $13.5 million in costs over 3 years with JFrog vs. alternatives. They save $6.7 million via automation of vulnerability protection and regulatory compliance activities in open-source environments. Customers then save $3.4 million more via better productivity. All of this leads to a 393% 6 month return on investment. JFrog saves customers time, money and headaches. That’s always valuable; It’s even more valuable in 2023.
Leadership continues to see optimization headwinds subside.
Products like Microsoft Copilot make source code creation far quicker and seamless. This enhances the need to store more binary-based software packages. That supports JFrog Artifactory growth. So? The OpenAI boom is a tailwind for this business.
6. Match Group (MTCH) -- Leadership Interview
New CEO Bernard Kim and CFO Gary Swidler expressed more optimism in Tinder’s performance than we’ve heard since Kim started. Kim shared that while it took him a few months to learn the business and build conviction over the correct path, the last few months have shown considerable promise. In the last few weeks, product velocity has “increased dramatically.” Its pricing optimizations are working as planned in the U.S. and it is set to roll them out in 5 more countries soon. This will lead to negative Q/Q payer growth in Q2 and Q3 (but a positive revenue impact) with payer growth returning to positive territory in Q4. Tinder fell behind the competition on price hikes over the last few years. It’s now “playing catchup.”
The weekly subscription take rate is meeting expectations while the a la carte revenue contribution from this is exceeding expectations. The subscription revenue boost is less dramatic as a lot of weekly subs go with that option instead of monthly subs -- it’s not all incremental. Vitally important is that the encouraging revenue trends Tinder saw in April continued through the conference this week. This gives it enhanced confidence in Tinder exiting 2023 with revenue growth of at least 10%. That’ll work.
“We’ve had some recent wins and think that will continue.” -- CEO Bernard Kim
Finally, Tinder’s new brand marketing campaign is working as hoped. Its user intent and consideration has been rising every day that it has been live. This boost is strongest with GenZ women, which is ideal. That’s a massive key to improving the gender balance on Tinder. It’s working on “The Vault” to enhance the female experience as well with more granular profile matching.
Margins & Capital Allocation:
The team was asked about getting back over 40% EBITDA margins eventually. It refrained from committing to this but added there’s significant leverage in the model to enjoy. Tinder’s 50%+ EBITDA margin means recovering growth there would be a large margin driver. App store fee reform would greatly help too as that’s its largest cost of revenue bucket.
Swidler hinted at recently becoming more interested in smaller acquisitions in the field of dating specifically. It will not buy another Hyperconnect which Kim bluntly stated the old team “overpaid for.” He’s right. It seems like Match is growing less excited about the prospects of Hyperconnect’s two apps but more excited about the AI/ML talent bench it came with to merge with the rest of its apps. Success there could make the $2 billion price tag look less ridiculous over time.
Kim is “excited about recent performance” in Asia with new CEO Malgosia Green leading the way.
New CTO Will Wu is rolling out AI-infused tools to remove friction from profile creation, enhance matching algorithms and debut coaching tools to help users talk to their crushes.
7. Uber Technologies (UBER) -- A Better Approach
Two important Uber developments his week. First, it will cut down its headquarter footprint in San Francisco to save on fixed costs. With all the vacancies in that real estate market today, we’re not sure if it will find an easy way out of the lease or a sublet. We shall see.
Secondly, Uber is partnering with Google’s Waymo across its business for autonomous technology. It was already partnered here on freight, but extended that to the rest of the business. The program will begin in Phoenix for both rides and eats.
Under old leadership, the company had been trying to build this autonomous capability internally. Now, it’s going about it in a far more asset light and rational way. Uber has an unmatched network of riders and eaters eager to consume its services. That is its strength. Investing billions to build autonomous tech was an expensive distraction with low probability of success considering all of the mega caps chasing the same goal. The current team is more reasonable and focused on delivering profitable growth. And it has done that while yielding share gains across all of its markets.
Google deciding to partner with Uber is a wonderful vote of confidence in the value of Uber’s massive network (and all the data it will be able to bring to the table). Rather than creating its own network and a directly competing product, Google saw the value in partnering with Uber. That’s encouraging.
8. Amazon (AMZN) -- Back in Style
Amazon caught two encouraging notes this week. The first was from Mizuho which discussed channel checks pointing to accelerating demand for AWS’s generative AI tools. The “ease of transition and product differentiation” was credited for the traction. It shared case studies on AWS winning vs. competition due to privacy and data security while reminding us that Generative AI workload business for hyper-scalers is high margin. It sees AWS growth bottoming in Q2 while that feeds positive sentiment and erodes the erroneous idea that Amazon can’t compete in AI. Maybe they read our introduction. :)
As an aside, Bill Gates made a bold prediction this week that AI will replace shopping search functions and marketplaces like Amazon. I agree, but also think Amazon’s upcoming ChatGPT-like search release will be its first step in being a large, large part of that future. It has the seller scale and logistics network.
Mizuho thinks growth will be 10% next Q (in line with consensus), ramp up to 14% by Q4 and rise to 20% in 2024. If that 20% comes to fruition, our Amazon model presented last week would be even more overly pessimistic than we thought it was.
Citi also released new data on Amazon’s retail growth in Brazil pointing to meaningful share gains. Traffic for Amazon there was up 35% in Q1 while that traffic fell 6% Y/Y for the space as a whole. Perhaps its gaining share vs. the formidable Mercado Libre?
Final Amazon notes:
Prime Air has now reached 100 drone deliveries in a small but important milestone.
Amazon is expanding its office space capacity in the UK through WeWork.
Dish Network will sell wireless phone plans through Amazon.
9. Meta Platforms (META) -- Miscellaneous
Rumors are building that Meta will fire more people.
The EU fined Meta $1.3 billion over unlawful data transference from the EU to the USA. It’s appealing.
Meta is working on a new intellectual property licensing partnership with an augmented reality firm called Magic Leap.
WhatsApp is adding an edit message feature.
TikTok is unsurprisingly suing Montana over the statewide ban.
Meta will sell Giphy to Shutterstock to remove FTC concerns and guarantee continued access to the content.
10. Nano-X Imaging (NNOX) -- Earnings Review
As a reminder, revenue today is from its teleradiology and AI segments, not the X-ray hardware (Nanox ARC). It generated $2.4 million in revenue vs. $1.8 million Y/Y (almost all from teleradiology). Peanuts. The entire investment case is based on ARC. Everything but the balance sheet today is irrelevant. It has $91 million in cash & equivalents. Share count is stable.
Its net loss of $11.8 million compares to $21.7 million Y/Y as it paid out less stock comp and earn out liabilities. Losses were impacted by an $8 million legal fee from a shareholder lawsuit.
b) The Path to ARC Commercialization
The headline of the quarter was FDA clearance of the ARC system (and Nanox.Cloud) model with enough power to perform musculoskeletal scans. This clearance is for Nanox.ARC “as a stationary system to produce tomographic images of the human musculoskeletal system for adults.” It will now apply for clearance on more use cases like mammography. With clearance secured, it’s standing up a demo center in Ft. Lauderdale this year with the goal of beginning U.S. ARC deployment THIS year vs. previously hoping for 2024. The team fully expects FDA clearance to serve as a vote of confidence and domino effect to accelerate clearance in other nations.
It signed a new pre-sale agreement to deploy 270 ARC units in Morocco over 3 years “subject to first unit acceptance and clearance.” The first unit has already been sent. Deployment efforts in Ghana continue with that expected to be its first commercial market to contribute meaningful revenue this year. Nigeria will follow soon after. It thinks it has the production capacity in place right now to service these contracts thanks to its Dagesh production facility which is now live.
Received ISO 13485 certification for its X-ray tube in Korea. This covers design, manufacturing and sales of X-ray tubes in what could mark a key step in building its X-ray tube licensing business. These tubes can be used for manufacturing quality control and security use cases.
It continues to collect more clinical data in Israel and just passed an Israeli Ministry of Health evaluation which green lights it to expand trials to strengthen regulatory apps for incremental use cases.
c) Final Notes
Nanox and USARAD settled on cash and stock payouts in connection with meeting performance benchmarks. This is now fully settled.
Its Nanox AI solutions have been installed in 4 of the 5 planned NHS clinics in the U.K. and 10 other clinics in the U.S.
Continues to seek out 1 to 2 more supply chain partners for its ARC components (chip, tube and system).
d) My Take
FDA clearance was wonderful news. Rampant claims of fraud should now be put to bed by this stamp of approval. Still, the investment is wildly speculative and still I’m not ready to lean into aggressive share purchases. I want to see ARC revenue start to flow in and want to see what kind of gross margin this revenue carries. All we know is that it’ll be at least 51%.
This information will give me the needed details to know the true potential of this investment. I’m fine with missing some upside while I wait as there is still a long way to go for this investment to work. This quarter was an objectively massive step in the right direction. There are more steps.
11. Upstart (UPST) -- Funding Supply
After beginning a partnership last February, CME Federal Credit Union is deepening its relationship with Upstart. It will join the budding referral network rapidly pushing towards 100 total partners. The credit union caters to first responders and is rather small at $407 million in total assets. Two things are true here. First, every single piece of added funding supply is good news. Secondly, the $4+ billion in announced committed funding is much more significant news than this is. That was the reason I started adding to my position once more.
12. Palo Alto Networks -- (PANW) Earnings Review
As a company operating in both endpoint (and network) security, Palo Alto Networks is a decent gauge for CrowdStrike’s (& SentinelOne’s) upcoming report.
Met revenue estimate & beat guide by 0.6%; beat billings guide by 1.3%.
Beat $0.93 EPS estimate by $0.17 & beat guide by $0.18.
Beat $0.14 GAAP EPS estimate by $0.17; Beat EBIT estimate by 19%.
Beat FCF estimate by 21.2%; beat GPM estimate by 140 bps.
b) Full Year Guidance Updates
Raised billings guide by 0.6%; slightly raised revenue guide & slightly missed revenue estimate.
Raised EPS guide from $4 to $4.27 & beat estimate by $0.28.
Raised adjusted FCF margin guide from 37% to 38%.
Next quarter guidance was slightly ahead on revenue and comfortably ahead on net income.
c) Balance Sheet
~$6.7 billion in cash & equivalents; $3.7B in convertible notes.
Stock-based compensation = 17% of sales vs. 17.9% Q/Q & 19.2% Y/Y.
No buybacks this quarter.
d) Call & Release Highlights
Cautious spending, added deal scrutiny and sales cycle elongation all continued to be headwinds. Palo Alto’s strong results amid putrid macro are byproducts of two major factors. First, cybersecurity is the least discretionary spend bucket in IT. There’s “no impending recession in threats” as CEO Nikesh Arora stated. Second, it fosters significant vendor consolidation to cut costs for clients with its 3 platforms. Cost savings is at the top of every company’s mind in 2023.
25.6% 3-yr revenue CAGR vs. 26.5% last Q & 26.5% 2 Qs ago.
Bookings growth for $1 million+ customers, $5 million+ customers and $10 million+ customers rose 29%, 62% and 136% Y/Y respectively.
Lifetime value of its top 200 customers rose 32% Y/Y to $63 million.
53% of the global 2000 has purchased its Strata platform (network security), Prisma platform (cloud security) and Cortex platform (endpoint security) vs. 33% 3 years ago.
Next generation ARR rose 60% Y/Y to reach $2.57 billion.
$1.10 in EPS compares to $0.60 Y/Y; trailing 12-month adjusted FCF is up 68% Y/Y.
Gross margin was helped by a larger software mix and easing supply chain issues. Less hiring helped EBIT margin on top of these gross margin boosts as well (remember EBIT = gross profit - operating expenses).
The company is 300 bps ahead of where it was supposed to be for EBIT margin based on its 2021 investor day targets.
Enjoyed 90% Y/Y growth in 4+ cloud module customers.
Will launch a new cloud module with its Cider acquisition this year.
First cloud vendor to reach FedRAMP High Ready classification.
Cloud credits rose 44% Y/Y.
Palo Alto is the only vendor named a Gartner leader in Security Service Edge (SSE) and Service Defined Wide Area Network (SD-WAN). SD-WAN allows for automated uncovering of best traffic routes and emulates a virtual network to lower costs and raise bandwidth.
Its the share leader in virtual firewalls and a Gartner leader in network firewalls. This gives it a set of capabilities it calls “the clear leader across Zero Trust network access.”
It landed 5 different 8 figure SASE deals this quarter with ARR here growing 50% Y/Y.
It sees network revenue as being a bit soft on the hardware side. Conversely, demand for software and cloud-based products remains robust.
Cortex reached $1 billion in bookings. Its newer extended security intelligence and automation management (XSIAM) is tracking to be its fastest new product to reach $100 million in bookings. This is similar to extended detection and response (XDR) which infuses 3rd party data sources into EDR products to move threat protection beyond the endpoint. XSIAM adds incremental layers of AI-fostered automation into the XDR fold.
13. Cresco Labs (CRLBF) -- Q1 2023 Earnings Review
Beat revenue estimates by 0.9% and met its vague guidance of a modest sequential decline.
Missed EBITDA estimates by 9.5%.
Missed -$0.02 GAAP EPS estimates by a penny.
All states grew Y/Y besides Illinois.
Cresco held share in every state with available data and jumped into a top 5 market share position in Michigan for the first time. It expects share gains throughout the year and Cresco remains #1 in Illinois, Massachusetts and Pennsylvania.
It now expects total revenue to be flat Q/Q in Q2 vs. previous guidance of sequential growth. Not good.
Remains the #1 branded wholesaler in Cannabis.
Margins were hit by revenue softness in its high margin Illinois market. This softness came via Missouri legalization going live and diminishing demand at dispensaries along the border.
Last quarter, GAAP margins were hit by a $141 million impairment charge from exiting its 3rd party California business. There was another $10 million inventory impairment and intangible asset valuation charge. GAAP EBIT margin for Q4 2022 would have been 4.2% with GAAP net income margin -14.5% with gross margin being around 50%.
Despite earning just $3.6 million in GAAP operating income, it paid out $16.8 million in cash taxes during the quarter. It cannot legally deduct ordinary business expenses from its tax bill and needs 280E reform to happen to do so. This is a when not if. But when it happens is anyone’s best guess.
Cost benefits from exiting some California and Arizona margins will begin to boost margins in Q2. It expects to reach 50% GPM and a 20%+ EBITDA margin for the second half of 2023.
SG&A fell modestly despite opening 8 new stores. Those openings added $14 million to its $21 million in CapEx which will fall sharply through the rest of the year. It reiterated its CapEx guide for the year.
b) Balance Sheet
$90 million in cash & equivalents vs. $122 million Y/Y. $286 million in current assets.
$382 million in debt.
Stock compensation was 3% of revenue.
c) Call & Release Highlights
2023 continues to be the “year of the core business and cash.” It continues to pull out of low margin and under-utilized operations in Arizona and California to bolster margins while it awaits new markets coming on line and federal reform. Pennsylvania, Ohio and Florida are all supposed to compound over 25% for the next 3 years as rec sales begin there. That’s where its focus will be.
It told us to expect more of these “footprint rationalization efforts” going forward. It is also leaning into its premium FloraCal brand by launching it in 3 new states. This premium brand is showing more pricing durability to complement its High Supply value brand. Enthusiasts are eager to pay up for the highest quality flower, solventless wax and live resin (not from distillate) edibles. That has not changed.
Despite rapid new store openings in Illinois, its absolute retail market share was maintained as its stores outperformed the competition.
Its Sunnyside.shop e-commerce platform crossed a billion in sales and is contributing insights and data to guide promotional activity.
Its new loyalty program added 100,000 members during the quarter, reaching 160,000 total. These users are spending 12% more than non-members.
Units sold rose 7% Y/Y and transactions rose 3.5% Y/Y despite cutting its retail employees by 10%. The decline in revenue is all pricing pressure related and will not last forever. Unit volume growth of 32% Y/Y offers explicit evidence of this being the cause.
Leadership expressed more caution about closing this deal than it has since it was announced. It all but told us that it will not close unless the required divestitures in Ohio and Florida (where it’s struggling to find a buyer) happen at favorable enough deal terms. If this doesn’t happen, it will likely look at other M&A opportunities to establish a presence in states like New Jersey.
Regulatory handcuffs continue to weigh on this industry. In terms of what Cresco can control, its retail operations remain the most efficient among large players, its wholesale business is the largest and its share across states is stable or improving. When will regulation finally create a fair environment for conducting normal business? No clue. But when that happens, this company and its (in our view) best-in-class team should be poised to dominate.
For now, we must continue to play the waiting game. These cannot be viable businesses when laws prevent them from accessing basic banking services, prevent them from listing on real exchanges and force them to pay 10x their operating income in taxes. That will not be permanent. A wildly frustrating investment thus far… but we are remaining patient.
14. Earnings Roundup
There are unfortunately not enough hours in the week to cover every single report of interest during earnings season. This section functions to (very) briefly summarize the results of firms we did not have time to cover.
a) Intuit (INTU)
$6 billion in revenue missed estimates by 1.5% & missed its guide by 1.8%.
Beat $6.82 GAAP EPS estimate by $0.56 or 8.2% and beat its $6.85 guide by $0.53.
Beat $8.49 GAAP EPS estimates by $0.43 or 5.1% and beat its $8.45 guide by $0.47.
Beat non-GAAP EBIT estimates by 5.3%.
Annual Guidance Update:
Raised revenue guidance by 1.1% to $14.3 billion which represents 12.5% Y/Y growth. This was also 0.9% ahead of analyst estimates.
Raised Small business and self-employed growth (includes QuickBooks which grew 25% Y/Y) guide from 19.5% to 24% Y/Y.
Lowered consumer group growth guide from 9.5% to 5.5%.
Raised Credit Karma growth guide from -12.5% to -11%. The segment is struggling with loan volume demand.
Raised GAAP EBIT guidance by 6.2%.
Raised EBIT guidance by 1.7% and beat estimates by 2.6%.
Raised $7.07 GAAP EPS guidance by $0.73 and beat estimates by $0.46. This represents 7.5% Y/Y growth.
Raised $13.74 EPS guidance by $0.49 and beat estimates by $0.41.
Next quarter guidance of 9.5% Y/Y revenue growth compares to 1.5% Y/Y growth estimates. GAAP EPS was a bit light, EPS was a bit strong.
$4.3 billion in cash and equivalents.
$6.6 billion in debt.
Repurchased $483 million in stock with $2 billion left on the program.
Dividends rose 15% Y/Y.
b) Autodesk (ADSK)
Met $1.27 billion revenue estimates and its revenue guide.
Missed EBIT estimates by 3.5%.
Missed $1.56 EPS estimates by a penny but beat its guide by 2 pennies.
Net revenue retention was “between 100% and 110%.”
Annual Guidance Update:
Reiterated $5.1 billion in bookings for -12% Y/Y growth.
Reiterated $5.4 billion in revenue for 8% Y/Y growth. This met estimates.
Reiterated $1.93 billion in EBIT for a 35.7% margin. This met estimates.
Reiterated $1.06 billion in GAAP EBIT for a 19.7% margin.
Raised the old $3.80 GAAP EPS guide to $3.85. This missed estimates by 2 pennies.
Raised the old $7.15 EPS guide to $7.24. This beat estimates by a penny.
Reiterated $1.2 billion in FCF.
Next quarter guidance was a bit light on demand and profits. This is a matter of timing as the full year guide was maintained or slightly raised in a few places.
$1.9 billion in cash and equivalents.
$2.28 billion in long term notes payable (none current).
Stock compensation was 13% of sales vs. 14% Y/Y. It bought back $512 million in stock vs. $457 million Y/Y. Buybacks represent over 3x total stock compensation.
c) Workday (WDAY)
Beat revenue estimates by 0.6% and its guide by 0.6%.
Also beat its subscription revenue guide by 0.7%.
Beat EBIT estimates by 9.8% & beat its margin guide by 200 bps. Also beat its GAAP EBIT margin guide by 330 bps.
Beat ($0.28) GAAP EPS estimates by $0.28.
Beat $1.12 EPS estimates by $0.20.
Sharply missed OCF estimates.
Annual Guidance Update:
Slightly raised its revenue guidance and slightly beat estimates.
Also slightly raised its subscription revenue guidance.
Maintained EBIT margin guidance which implies a small EBIT dollar raise based on the revenue raise. This is barely ahead of estimates. Reiterated expectations for a 1% GAAP EBIT margin.
Its 23% EBIT margin guide compares to 19.5% in 2022.
Lowered its FCF guidance from $1.71 billion to $1.65 billion. This missed estimates by 2.4%.
Next quarter guidance was just ahead on revenue and comfortably ahead on EBIT by 2.5%.
$6.33 billion in cash & equivalents.
$2.9 billion in long term debt with none current.
Stock comp is elevated at 22% of sales vs. 27.1% Y/Y. Basic share count rose 2.8% Y/Y and diluted share count rose 3.8% Y/Y. For a company beginning to enter a more mature phase of its growth curve, this is too high.
Named Zane Rowe as its new CFO.
Launched a new AWS integration with Workday Extend.
Won Dollar Tree and Nissan business during the quarter.
75.2% GAAP GPM vs. 71.9% Y/Y
Gross revenue retention remained over 95%. Anything in the high 90% range is good.
15. Final market headlines:
Block told investors that volume growth accelerated in May at the JP Morgan conference.
Microsoft debuted a ChatGPT-powered virtual assistant as part of its product release event this week. Unsurprisingly, this event centered around new AI-infused capabilities.
a) Key Data from the Week
Building Permits came in at 1.15 million vs. 1.42 million expected and 1.43 million last report.
April new home sales of 683,000 beat expectations of 663,000.
Pending home sale growth was flat vs. 0.5% expected and -5.2% last month.
Manufacturing Purchasing Managers Index (PMI) for May was 48.5 vs. 50 expected and 50.2 last month.
S&P Global Composite PMI for May was 54.5 vs. 50 expected and 53.4 last month.
Services PMI for May was 55.1 vs. 52.6 expected and 53.6 last month.
GDP for Q1 rose 1.3% vs. 1.1% expected and 2.6% last month.
Employment & Consumer Health:
229,000 initial jobless claims vs. 250,000 expected.
Michigan Consumer Expectations reading was 55.4 vs. 53.4 expected and 60.5 last month.
Michigan Consumer Sentiment reading was 59.2 vs. 57.9 expected and 63.5 last month.
Personal Spending M/M for April was 0.8% vs. 0.4% expected and 0.1% last month.
Core durable goods M/M for April was -0.2% vs. 0% expected and 0.3% last month.
Durable Goods Orders M/M for April grew 1.1% vs. -1% expected and 3.3% last month.
Core Personal Consumption Expenditures (PCE) index Y/Y for April was 4.7% vs. 4.6% expected and 4.6% last month.
Core PCE index M/M for April was 0.4% vs. 0.3% expected and 0.3% last month.
PCE Y/Y for April was 4.4% vs. 3.9% expected and 4.2% last month.
PCE M/M for April was 0.4% vs. 0.4% expected and 0.1% last month.
5-year breakeven inflation continues to hover around 2%:
High Yield Option Adjusted Corporate Credit Spreads:
b) Level-Setting the Data
Our base case remains a mild recession in 2023, a rate hike pause beginning at the next meeting and an end to quantitative tightening by Q1 2024. The recipe of slower economic growth and more favorable monetary policy with falling discount rates is greatly preferred for long duration growth asset vs. runaway inflation and endlessly rising rates. The data we’ve gotten in 2023 has consistently eroded the risk of runaway inflation. Real-time inflation indicators point to continued disinflation. Supply chains are normalizing, rent inflation is cooling, labor supply is just starting to improve to curb wage inflation and energy continues to trade well off of its cycle peak.
So? I continue to gradually trim into excess multiple expansion and add into contraction with a slight bias towards accumulation and cash deployment. Real structural growers will find demand growth in any environment. They’ll soon likely do so while enjoying less intense future cash flow discounting as rates ease. This will bolster valuations as revenue and profit growth troughs across some sectors.