Robinhood reiterated $1.9 billion operating expense (OpEx) guidance for the year. This is despite two smaller acquisitions of a crypto exchange and an AI-powered equity research firm during the quarter. Formal guidance is not provided, but it did offer commentary on July and August volumes. July trading volumes were 20% higher than during June. As of its quarterly report, August volume was consistent with July. 

CFO Jason Warnick did tell us that Robinhood has a “strong track record” of 20%+ revenue growth. He’s right, but that’s looking at several years. Brokerages are cyclical, and while market share gains and product diversification do buffer that cyclicality, it’s not immune. For context, revenues fell 25% Y/Y in 2022. The bear/bull debate centers on this issue. Is this just a brokerage with a superior interface? Or a budding financial services super-app bucking the volatility? I see it as a brokerage, but this year has still been very impressive.

Robinhood trades for 19× 2024 earnings. EPS is expected to grow by 94% Y/Y next year and by 1% Y/Y in 2025. Note that the sharp decline in forward p/e shown below is because Robinhood was rapidly inflecting to positive GAAP net income.

Transaction revenue growth was 69% Y/Y, driven by 43% options growth, 161% crypto growth and 60% equity growth. On a sequential basis, this category was flat as crypto volume declines offset growth in other buckets. It added 300,000 funded accounts (2nd largest Q/Q net adds in the last year) to reach 24.2 million. A 1% deposit bonus for Gold members helped drive an added $3 billion in asset transfers, which directly supports overall transaction revenue.

Robinhood sees revenues from margin lending (those leveraging their equities to borrow and buy more) as a massive near-term opportunity. Incumbents collect more revenue from lending than they do from trading. The company previously did not feel like its margin rates or the offering overall were competitive. So? It debuted industry-leading 5.7% margin rates and saw the margin book spike 52% Y/Y. 75% of these accounts have balances over $100,000.

In its pursuit of “winning the active trader” there are a few more product gaps that it needs to close vs. the E*Trades of the world. It has done a lot with futures recently, but does need to enhance index options, fixed income offerings and add products like CDs. That’s the plan. While its active trader product category is the most mature, it sees a lot of product innovation and growth left in the tank. Robinhood will host an October event for these active traders to unveil some of this work. They hinted at a new desktop version coming out with more advanced charting tools, which would be another important product gap to close.

Net interest revenue rose 22% Y/Y to $285 million thanks to growth in interest earning assets, more margin issuing and higher rates. This continued to grow sequentially, as its newly issued industry-leading margin rates helped drive more momentum along with growth in securities lending and carrying larger interest-earning asset balances.

Falling rates will pressure this bucket to a certain extent, but Robinhood has important offsets. Most of its interest-earning liabilities held on the balance sheet are for its 5% cash yield. As rates fall, it will cut that yield to counteract the net interest margin pressure. Still, a quarter point rate cut represents about a $10 million (or 3.5% of total) quarterly revenue headwind.

Other revenue growth was 19% Y/Y due mainly to more Robinhood Gold subscription revenue. Gold members (I need to rewatch Austin Powers) rose 61% Y/Y, while subscription revenue crossed an annualized run rate of $100 million.

Impressively, 8.2% of funded accounts are now gold subscribers vs. 7% Q/Q. That was its largest sequential jump in well over a year. Its 5% cash yield, 3% IRA match, 1% deposit match and now 3% cash back credit card are all driving this success. Robinhood has deeply enhanced the value of this subscription lately, and so was understandably asked about price hikes. That’s not the current focus. It’s pushing to drive more adoption and could revisit price hikes down the road.

The Robinhood Gold credit card and related perks are expected to keep growth here humming… but what about the margin profile attached to this growth. After all, 3% cash back doles out Robinhood’s entire interchange cut. 1% deposit matches and a 5% cash yield without a banking charter are all expensive perks. So how do they expect to make money on this? 

They bank on a large deposit halo effect. Customers are delighted by the value, they move more of their financial service business to Robinhood, and it comes out ahead. That is their point of view. Specifically, Gold members are delivering a 7x revenue lift and ROI gains. So this is working well for now. All I will say is that the brokerage environment will be tougher at some point and credit card losses can mount in a hurry. Recognizing this, they’ve gone very slowly with issuing new credit cards… which I think is the responsible decision. Good for them forgoing near-term revenue to ensure underwriting algorithms and unit economics are favorable. Smart.

GAAP operating expenses (OpEx) rose by just 6% Y/Y thanks to lower stock comp and leverage in other cost categories too. Non-GAAP OpEx rose 14% Y/Y. The team’s incremental EBITDA margin sat at a lofty 77% for this quarter and it’s already at its old 40% EBITDA margin target. Incumbent brokers typically operate at 50% EBITDA margins, and Robinhood sees itself getting there. At the same time, it does not see 50% as the final destination, but a step on its journey. 90% of its costs are fixed and it does not have the same expense base that legacy incumbents have. They should be able to do better than 50%.

The 24-hour market outage from earlier this week wasn’t Robinhood’s fault. Blue Ocean (3rd party vendor) was struggling with the excess demand and so it had to be shut down. Leadership reminded us how new this product is and how it will continue to perfect processes and lead the charge. I think 24-hour markets invite market manipulation and unhealthy moves that push retail to capitulate at the worst times (see last Monday). Still, if this is legal, there’s no reason for Robinhood not to pursue this added volume opportunity. It allows them to stand out from the pack as they can rightfully claim more market access than others.

This was a rock-solid quarter for Robinhood. Tailwinds were in place, and they took advantage. I’m truly fascinated by the polarization of this investment case. I can see a world where growth turns negative amid the next nasty phase of the brokerage cycle when they aren’t enjoying so many successful launches and programs. I can see how the credit card could lead to balance sheet and loss ratio issues down the road. At the same time, these guys are killing it in the market share-taking game. They are standing out wherever they can on the product side and they are delivering great results today. I think bulls should be pleased and bears should be feeling a bit more anxious. Rate cuts are coming, and the volume gains stemming from those cuts should outpace the net interest income headwinds that coincide. Retail mania marches on.

SMCI met revenue estimates & its identical guidance. Its 81% 3-year revenue CAGR compares to 63% Q/Q & 41% 2 quarters ago. $800 million in revenue was delayed to next quarter due to capacity constraints.

SMCI trades for 15x next year’s earnings. EPS is expected to grow by 56% Y/Y next year and by 30% Y/Y the year after. Considering the cyclical nature of this business model, I don’t think EPS growth estimates are all that valuable. They will violently swing as cycles unfold.

The entire call was spent on margin trends and expectations amid the large revenue beat paired with an EPS miss. This stirred up a renewed bull/bear debate, which we will focus on here.

Margin pressure was split evenly between two items: Direct Liquid data center Cooling (DLC) demand and revenue mix shift. Starting with DLC demand, the strong momentum is not surprising. The capacity arms race is not ending, and SMCI hardware is a key piece of AI data centers. Its DLC technology cuts energy usage by 40%, boosts data center performance and uptime, and lowers carbon footprint. Its new data center building block solution (DCBBS) also reduces data center build time from 3 years to 2 years. That’s compelling, as large caps race to build capacity. Simply put, DLC “dramatically improves total cost of ownership” vs. air-cooled systems and should therefore continue to see adoption rates rise.

Demand levels for liquid-cooled racks surprised the team. It responded to this demand by quickly ramping the supply chain and “paid a lot of expedited cost” to support this. That weighed heavily on margins, but should be temporary. Capacity at its new Malaysia factory, expansion projects in the USA and more time to build capacity in an efficient manner are all expected to erode this margin headwind over time. The company sees a clear path back to a 14% - 17% gross margin. 

SMCI sees its DLC tech as industry-leading and it’s choosing to spend more on expediting supply to capture market share gains. Players like Nvidia are trying hard to compete here, so securing demand today does make some sense. This decision fostered a roughly 75% market share of liquid cooling deployments this past quarter. It will ramp rack production capacity from 1,000/month to 3,000/month by 2025. It’s also now extending its easily-deployed DLC rack solutions to next-gen AI data centers, which should simply build on this product’s strong cyclical momentum.

“Now we are further expanding this solution to the entire data center, with rapid deployment of large-scale AI infrastructure.”

Founder/CEO Charles Liang

The other source of the margin compression was a large order from a hyperscaler, which led to revenue mix shift and some overall pricing pressure. This is the potentially more structural and concerning margin headwind. It points to SMCI perhaps losing some pricing power with larger customers, and needing to concede contractual terms to compete.

Simply put, if supply chain bottlenecks are resolved and mix shift stabilizes, margins should improve. I think the bottleneck resolution is inevitable, but the mix shift could continue to weigh on margins. Furthermore, the GenAI data center boom will not be perpetual. If crazy demand normalizes a bit next year, that could weigh further on pricing pressure for this company. If it’s right about leading in DLC, competitive differentiation should diminish the impact of that potential headwind. I’m not sure where I stand here, I don’t focus on this sector, but these are important variables to pay attention to as we move into 2025.

“We expect margins to gradually increase throughout 2025.”

CFO David Weigand

Axon trades for 74× 2024 earnings. EPS is expected to grow by 14% Y/Y this year and by 20% Y/Y next year.

I’ve held this in the past, but I haven’t for a while. I’m excited to dig into the transcript this weekend to see what happened here and if this aggressive pullback is compelling or not. I’ll include that coverage either in the Nu earnings article next week or next Saturday. This is a good company with a dominant share of Fortune 500 brands as customers. Either that’s still true and this could be interesting again, or things have fundamentally soured since I last looked. We shall see.

Q3 guidance also missed estimates across the board.

JFrog trades for 47× 2024 earnings. Earnings are expected to grow by 9% this year and by 18% next year.

ELF trades for 45x forward earnings. Earnings are expected to grow by 16% Y/Y this year and by 17% Y/Y next year.

Two things happened this quarter. Margins were slightly worse than expected, as sales & marketing roughly doubled Y/Y to pursue more growth. This meant the EBITDA raise (and implied EBITDA margin) was a bit disappointing. This will inevitably leave people wondering how much marketing spend is required to keep the growth engine going. Maybe more than previously thought. Next, the world assumed that ELF was aggressively sandbagging guidance. Analysts got a bit ahead of themselves in modeling too large of a raise into their expectations.

Nielsen data is pointing to more (expected) slowing for Celsius. For the 4 weeks ending July 13th, volume rose 18% Y/Y. This compares to 22% Y/Y volume growth for the 4 weeks ending June 29th. As Monster made clear this week, sector growth is very challenged and Celsius is not immune. Pepsi inventory resets could continue, Red Bull launches have been very strong and delayed inventory resets hurt this market share taker more than either of the big boys. I say all of this to acknowledge what we all already know: The energy drink sector is going through a rough patch.

Things will not turnaround overnight. Bank of America downgraded it this week because it sees no meaningful recovery until next summer. Celsius does have marketing programs in place to reignite growth in the back half of the year, but comps do remain difficult through the end of the year. Furthermore, energy drink brands frequently come and go. Celsius has stormed onto the scene more meaningfully than any of the other disruptors, but fizzling out isn’t impossible. Perhaps Pepsi truly did pull forward several years of growth. I don’t think that’s the case (this isn’t telehealth amid a pandemic) but it’s possible. Consumer brands are volatile.

Regardless of this risk, I have built a small stake in the position recently and I have no interest in liquidation. This is falling into meaningful volume support, the next-12 month earnings multiple is around 35x, and EPS growth for the next two years should safely compound at 25%. Market share gains over the last few years have been extremely impressive and international expansion is off to a good start. As the sector backdrop brightens, I think there’s real upside to estimates – just like when the sector was rocking, estimates were far too aggressive. This company continues to take share on a Y/Y basis, address modest monthly share declines due to Red Bull one-off launches and power the entire category's growth. For now, I’ve built this position as large as I want it to be. I like the idea of letting them prove that programs will actually work and letting this grow naturally into a core holding. For now, small is preferred.

After getting through the transcript, nothing has changed about my views towards the company and investment from my mid-week update. For those who missed that piece, I copy and pasted it here.

Progyny trades for 13x 2024 EPS. EPS is expected to grow by 107% Y/Y this year (-55% Y/Y growth last year) and by 15% Y/Y next year.

This was the third bad earnings report in a row for Progyny. It slashed full year revenue guidance by over 5% and by 10% compared to its original 2024 outlook. It cut EBITDA guidance by 7.7% and cut EPS guidance from $1.64 to $1.57.

It called out stabilizing utilization rates, but falling revenue per member, which it doesn’t exactly know the reason for. Leadership thinks the “business remains healthy and well-positioned, based on progress,” and I’m a tad torn on whether or not I agree with them.

On one hand, the guidance revision is awful. The last handful of quarters have featured a hodgepodge of excuses — ranging from med shortages, to treatment mix shift, to abortion news hurting utilization — to explain the disappointment. This quarter, the new excuse was treatment monetization, which the team doesn’t know the true source of. It’s very hard to model both macro and human biology, so I understand how they could struggle. Regardless, investors rightfully demand some level of visibility and this team clearly doesn’t have any. There’s no sugar-coating any of that.

On the other hand, early success in its selling season for 2025 is actually pacing above this past year, and its newest products have been successfully cross-sold to a notable 1 million (16%) of its members. That is undeniably impressive and expands its addressable market materially. It continues to print cash, has no debt and just added the equivalent of another 5% of its market cap in buybacks. It is the clear financial and clinical leader in its space. It saves its members, clients and carrier partners money and vastly uplifts patient outcomes across all categories. It boasts near-100% client retention and trades for about 14× 2024 earnings. Finally, it has an investor day next week. Companies generally don’t schedule these unless they have good things to say to investors.

I can’t buy this dip. I don’t think this team knows how to level-set expectations and I think they are reliant on a normalization in trends that they cannot possibly forecast or bank on. I can’t sell this company because it does create so much value and is deeply profitable. It should compound revenue at 15% during normal times and is oh so very cheap. I’m going to sit on my hands and do nothing. Perhaps I’m being too patient, but holding losers is a far less damaging mistake than selling future winners too early. Either they turn a fundamental corner like every single thing indicates that they should, or this dwindles further and further down as a % of my portfolio.

Amazon is extending ad impression placements to external destinations. Through a partnership with both TikTok and Pinterest, it will place impressions natively on those sites. Consumers will be able to make purchases without bouncing back and forth across the web. Less clicks and pages per checkout mean higher conversion rates. More locations to direct high margin ad impressions will be yet another Amazon margin tailwind. 

Shopify has found great success in partnering across all social channels to place merchant goods and promotions in front of more eyeballs. That has been a large growth driver for this e-commerce rival. That same should be true for King Amazon.

SoFi has been working on establishing excellent service and consistent processes for its very small mortgage business. To enable direct servicing and origination of loans, it purchased Wyndham Capital, and has since integrated the assets. This week, mortgage rates fell to their lowest levels in nearly a year to a still elevated 6.5%, and I would expect mortgage volume this quarter to considerably outperform if that continues.

SoFi has positioned its model to take advantage of more mortgage demand and to create yet another operational tailwind stemming from near future rate cuts. Whether it’s more willingness to originate more unsecured credit with more rate certainty and capital market liquidity, student loan refinancing, better tech platform demand, velocity of money supporting faster financial services growth, or now mortgages, easy monetary policy will be great for this business. 

One may rightfully wonder: How is that the case if this is a bank? While cuts will certainly be a net interest margin headwind, SoFi still has a lot more balance sheet optimization to do. It still has a lot of expensive warehouse debt to swap out with cheaper deposits. That will be a powerful offset to the traditional bank headwind from rate cuts, while all of the other mentioned segments enjoy bluer skies. And one more note: Personal loan variable to fixed refi demand does suffer with rate cuts. There’s intuitively less motivation to fix variable debt when rates are falling. Still, pent-up borrower demand stemming from SoFI’s extreme conservatism and better capital market liquidity will easily overcome that obstacle. That’s management’s view and my own. Cuts are coming. Powell has all but telegraphed a September rate cut, with betting markets now seeing 3-4 cuts in 2024. Obviously, if rate cuts are a response to severe recession, then SoFi and everyone else will suffer. The most likely outcomes, however, are looking like modest recession or soft landing. That scenario would likely bode very well for this company.

Rumors swirled this week that Galileo lost another client in Varo. They haven’t been a Galileo client since 2022. Furthermore, as CEO Anthony Noto explained, Varo committed to regulators to move to a certain processor as part of their bank charter process. That commitment was made before SoFi purchased Galileo and secured its own banking charter. This is old news and will have no impact on 2024 tech platform guidance.

A few weeks ago, Elliott Management disclosed a multi-billion stake in Starbucks. On Friday, another high profile activist firm in Starboard Value disclosed their own investment in the iconic, yet struggling coffee chain.

as another activist involved at Starbucks. I welcome this news. I do think Starbucks is making the right changes. They’re fixing throughput problems, pushing more people to their app and loyalty program, driving better affordability and cutting considerable cost out of the model. The focus on sugar free energy drink growth is also well-placed. Furthermore, they’re right to continue opening stores in China. Cash on cash returns for new stores there are excellent and incremental to overall results… that’s just being clouded by the wildly challenged competitive environment there today. This is still an iconic brand that dominates preference charts across all age cohorts and boasts a still long runway (especially in places like India and Latin America).

With that said, new leadership certainly has been far from perfect. I see it as a large positive to have these two powerfully influential voices in the room to steer decisions in the direction of shareholder favorability and profitable growth. One thing I’d love for them to do is forcefully push Starbucks to embrace the lunch and dinner dayparts more wholeheartedly. They’ve debuted some menu items, but these stores are sitting largely empty for chunks of the day. That’s why Dunkin’ Donuts and Baskin Robbins partner on real estate. I don’t think Starbucks should do that, but I do think the menu needs a lot of work. Stick to your coffee core while also doing a lot more.

SentinelOne has long been a preferred vendor for Google’s Mandiant. Now, the two are expanding this already tight partnership. SentinelOne will integrate its AI Singularity Platform directly into Google’s Cloud’s Threat Intelligence tools. The firms will openly share data in a bid to uplift breach protection and better protect shared customers. SentinelOne also committed to using Gemini models for its Purple AI security assistant and overarching platform.

These are exactly the kind of headlines that I want to see from SentinelOne. Its tech reputation is phenomenal. Its ability to win large customers within its go-to-market is quite poor. Leadership will openly tell you that. That’s why it’s overhauling go-to-market and fixating on large enterprise growth. There’s no product inferiority issue… there’s a selling and communication issue. Deeply partnering with mega-caps like Google can be a powerful tool to overcome this. This will make SentinelOne’s product suite more visible for the search king’s clients and should diminish friction for Google Cloud players to work with SentinelOne.

AWS has been instrumental in powering CrowdStrike's fabulous growth engine. It has already done $1 billion in revenue from that channel alone. SentinelOne needs to build out its large partner and channel network… Alphabet is a fantastic way to do that.

SentinelOne launched a revamped managed detection and response (MDR) tool at this week’s Black Hat security event. This merges SentinelOne’s autonomous security platform with its deep bench of security analyst talent. Like cloud providers let companies enjoy managed infrastructure and data centers without having to build them… MDR allows companies to scale with proper security in place without paying large, expensive teams of analysts.

Snowflake is looking to partner with GenAI darling Cohere to add more large language model (LLM) options for developers on the platform.

Delta is “pursuing legal claims” from CrowdStrike and Microsoft for the $500 million in losses it incurred from the outage. Delta refused help from Microsoft and CrowdStrike, according to the two companies.

Meta raised another $10.5 billion in debt. The deal was done in 5 tranches with maturities ranging from 2029 to 2064 and interest rates ranging from 4.3% to 5.55%.

Disney may have to pay an additional $5 billion for Hulu. Deadpool and Wolverine passed the firm’s best 2023 release (Guardians of the Galaxy Volume 3) in box office sales just 11 days into release.

Alphabet lost its antitrust lawsuit on search. Payments made to Apple to be the default search engine were a focal point of the ruling. It will have its antitrust ad-tech case heard next month.

All of this data continues to point to modest economic slowing. It bodes very well for a soft landing or a modest recession and avoiding severe recession.

My portfolio hasn’t changed since the last update.