Bear's Portfolio through 06/2024

Investment Analysis Clubs Saul’s Investing Discussions
22

They are opening new production facilities both this year and in the future, with funds already raised for this, to my understanding. I’ve read this expansion is suppose to give them the capacity for $25B in annual revenue (not sure when all new facilities will be online). If Beth is saying that they STILL are at capacity through some info she has access to, it only means that their revenue is still exploding in the short term.

The positives seem to outweigh the negatives, but I agree that the moment they miss, there will probably be a violent reaction. If that was a driving fear of mine though, I wouldn’t invest in growth stocks.

20 Likes
23

From a June 17th press release, SMCI stated: “Our products are designed and manufactured in-house (in the US, Taiwan, and the Netherlands), leveraging global operations for scale and efficiency and optimized to improve TCO and reduce environmental impact (Green Computing).” Conspicuous in its absence is mention of the Malaysian facility. I don’t know what to make of that. I’m reasonably confident that construction has been completed.

From a November 9, 2023 press release they stated, “Worldwide, Supermicro’s full rack scale delivery capacity is growing from several state-of-the-art integration facilities in the United States, Taiwan, Netherlands, and Malaysia. Future manufacturing expansion and locations are actively being considered to scale to the increasing demand for Supermicro’s rack scale AI and HPC solution portfolio.” The same release continues, ““With our global footprint, we now can deliver 5,000 racks per month to support substantial orders for fully integrated, liquid cooled racks, requiring up to 100kW per rack,” said Charles Liang, president, and CEO of Supermicro.”

So I’m at a loss with regard to Ms. Kindig’s comments. It’s true that SMCI does not employ contract suppliers. Every computer and every rack is assembled in their own facilities. In that the 5,000 racks/mo capacity was asserted late last year I think it’s safe to assume that current capacity is at least that high. I’m guessing that this rate is based on a normal production calendar (one shift, 5 days). If the supply chain can deliver, I speculate that capacity could be increased to a maximum of 3 shifts, 7 days, but that “if” is a very big if.

8 Likes
24

** Liang stated that Supermicro’s Malaysian plant is expected to be operational in the fourth quarter of this year, with a monthly production capacity of up to 5000 cabinets, according to CNA’s report today.**

So this quarter.

8 Likes
25

OK, that article helps . . . I’m still a bit confused about SMCI near term capacity. From the article is appears that the Malaysian factory will add 5,000 racks/mo to current capacity. From other sources it appears that current capacity is already at 5,000 racks/mo. So is near term capacity 10,000 racks/mo? Can the supply chain support that volume? It’s not at all clear to me. But, the target of 15% share is clear. Charles seems very confident that this is both supported and achievable.

4 Likes
26

I’ve seen a lot of angst about SMCI’s ability to scale from other people (not brittlerock). Certainly, SMCI has periodically needed to get expansion funding through issuance of additional stock but the dilution is far less than the resulting advances in revenue. That doesn’t appear to be a problem, at least for me.

But… but… but… “what about suppliers???”. I have two thoughts about that… thoughts I’ve been criticized for unfortunately:

  • We have no insight into the supplier situation and we never will, unless the company says something about it.
  • Given the amazingly good job SMCI has done orchestrating their rapid growth, is it reasonable to think that they are unaware of the supplier situation themselves and that they’ll suddenly be in a situation like “Oh my gosh, boss! I just sent an order to ABC and they can’t increase production any more! I am so surprised! If only we had planned ahead!”? My contention is that it is unreasonable to think they have failed to make plans that include their supply chain.

So… IMO… we can’t have every insight we’d like and we have to rely on management.

Just like every other company we invest in.

Rob
He is no fool who gives what he cannot keep to gain what he cannot lose.

37 Likes
27

Yeah, I’m starting to agree with you, Vince. At first I saw Paycom as a temporarily beleaguered 20%+ revenue grower that would bounce back. Then I realized…maybe they were only growing at 25% in 2021 and 30% in 2022 and 23% in 2023 because of tailwinds that were helping them as the business recovered after growing only 14% in 2020 largely due to Covid.

Hypothetically, this could have been due to:

  • customers re-ramping up with reopening and people coming back to work
  • the clawing back of discounts they might have given hurting customers in 2020
  • pricing increases
    etc etc etc.

What really doesn’t impress me is the underlying customer growth:
2020 end: 31,000 clients
2021 end: 33,900 clients (+2,900 or 9.4% growth)
2022 end: 36,600 clients (+2,700 or 8.0% growth)
2023 end: 36,800 clients (+200 or 0.5% growth)

I haven’t looked far enough to find answers to all these issues, but add in the CEO’s recent rant (https://youtu.be/11YRTZIr--k?si=SwAhmyDHVO5yD_QF) and the recently appointed co-CEO leaving the company, and it’s enough for me to wait until growth picks up as you wisely suggest, or at least until we get some more information.

I’ve reduced it to less than 1% already and I might just dump the rest and watch from the sidelines. Thanks for the nudge to prod at my conviction on this one.

Bear

21 Likes
28

@FallingWallenda I am having trouble taking Beth Kindig’s financial advice seriously when she is recommending buying AMD growing revenue at 2.2% instead of NVDA, and then recommending Dell growing revenue at 6.3% instead of Supermicro.

In both cases she is saying these companies should catch up. AMD because their chip supposedly does better on inference, and Dell because Supermicro is at capacity and Dell will get the overflow orders. These seem like classic cases of putting the horse before the cart in growth investing, where she is not waiting for the company to prove they can grow, but attempting to predict which companies should be growing.

On AMD,

  • She says Nividia owns 98% of the market due to CUDA - so she is proving that Nvidia beats AMD by a ratio of 50:1 in the marketplace
  • There is the claim that AMD’s MI300X is superior to H100, but H100 is already the last generation so it makes sense AMD is producing a chip that is faster than that, but she later states Nvidia is always 6 months ahead of AMD
  • She says Nvidia has a strong toolchain that AMD does not have, and says AMD libraries are “very new and not built out”
  • AMD’s “big problem” as she says is on the software development platform (I thought this was supposed to a bull case for the company?)
  • AMD sells it’s chips for 10-15k and Nvidia for 30-40k as she says, and states that customers will want to save this extra money. But then also says AMD’s margins “will not look as good as Nvidia’s”. I get that AMD has a legacy business too, but I cannot understand preferring the lower margin business that sells it’s product for less
  • Beth says that Nvidia’s pricing power and margin comes from CUDA. Again this seems like a reason to want to invest in NVDA over AMD.
  • AMD is getting in the “niche” for the inference business. If the inference is niche, then why are we investing?
  • AMD is likely to be an a leader in AI PCs she mentions - okay, but AI PC’s don’t exist yet, and she’s once again suggesting to invest based on a prediction. We have no idea who will be first to market with an AI PC. Nvidia is also competing in this space, and what evidence is there AMD will come out ahead of Nvidia in any market?

On Supermicro,

  • Beth said “Supermicro is not running at scale like Dell is”. This is just incorrect information as Supermicro is producing and selling more AI servers than Dell is in their last quarter. Somehow she is lumping the legacy Dell PC business into the same bucket as Supermicro and saying Supermicro cannot scale
  • She says even though Supermicro has “great technology” and is an awesome play in the medium to long term, her firm moved into Dell because Dell is “now the second runner up”. Since when do we want to invest in the runner up businesses over the leader and innovator?
  • She’s saying when Supermicro’s orders get full her “guess” is that those orders flow into Dell. She has no mention of HPE, and she’s proving that Supermicro is the first choice amongst customers.
  • She says Dell has liquid cooling technology. This is the first time I’ve heard someone try to argue that Dell’s DLC can even compete with Supermicro. They do have one system the PowerEdge XE9680L which is DLC, but they literally never talk about it on any earnings call. Meanwhile, Supermicro and HPE are discussing their DLC solutions every chance they get on earnings and conferences. This leads me to believe Supermicro > HPE > Dell on advancements in DLC. Not all DLC is the same, and Supermicro has very advanced innovations in liquid cooling which the others simply do not have yet.

Interestingly enough, there was a short segment in the middle where Beth said we should wait for the moment in software when AI applications are developed and it’s too hard to guess winners. However, for hardware she is guessing on incumbent underdogs to somehow over take the newer innovative companies.

The host sums up the interview the best at the end when he says, it’s “Always fun to root for the underdogs” meaning her recommendations of AMD and Dell. She is literally investing her clients money in the legacy players from the PC revolution and selling the innovative and growing companies. This is why at the start I don’t take her opinion seriously.

47 Likes
29

Bear, I greatly appreciate your comments and kudos as you as you are one of the folks I really look up to on this board.

Vince

18 Likes
30

Great update @PaulWBryant!

What is the new Q NM column, which you have added in this month’s review? I thought it might be the quarterly net margin perhaps? But I checked a few tickers and it doesn’t seem to match. Quick ratio? Net profit margin? :thinking:

31

I was actually hoping someone would ask and you didn’t disappoint! It is indeed the quarterly net (profit) margin, but it’s the adjusted (non-GAAP) margin, for companies that give that. It could be off by 1% or so in either direction, because I calculate it myself based on the shares outstanding I have in my spreadsheet. But if you see one that’s off by more, please let me know!

Why is this important? Well when you’re dealing with “profits” (whether FCF or EPS or EBIT or whatever), you need to know have some idea of how much it can improve. A very quick glance is to see what the margin was in the most recent quarter. But this is just a glance! You have to take it as a single point in time, one data point, and bring to bear all the knowledge you have of the company to give it context. I’ll briefly explain my thinking on each of my companies to give an example.

Axon - 19% is good, but I think they can do a lot better eventually. Their FCF margin is lower at this point indicating substantial CapEx. They’re definitely still in rapid growth mode, so they should eventually be able to spend less and bring more $$ to the bottom line.

Amazon (note, this one is GAAP because they don’t give adjusted) - Just 7%, kind of crazy. Part of that is because their legacy business is super low-margin. Walmart’s GROSS margin is only 25% or so. But Amazon also has AWS, and now Advertising. So their blended Gross margin is close to 50% and improving. I think they’ll drop more and more to the bottom line over time, but it might be a gradual improvement.

Monday - 15% is up from a negative number just over a year ago. But they have lots more to improve IMO. Their FCF this quarter was 40%! I don’t know how close to that net margin can get, but I think it will get close eventually.

Zscaler - 25% is up from about 18% a year ago and steadily improving (but not as rapidly as MNDY’s is improving). CRWD’s also happened to be 25% this past quarter.

MercadoLibre (note, this one is GAAP because they don’t give adjusted) - Like AMZN, theirs is low for structural reasons at 8%, and MELI’s Gross Margin is also similar at arount 50%. But like AMZN it’s also got room for improvement from some of their higher-margin lines of business.

Nu (note, this one is GAAP because they don’t give adjusted) - Like with MNDY, Nu’s margins have shot up rapidly in the last couple years. 14% is a lot better than negative, but I think it will go much higher. It will take time (improvements might be more incremental than they’ve been the last several quarters) as Gross Margin is still only 40-something percent, but that can improve a bit too I think.

ELF - 10% this quarter but it was 16% last Q, 22% the Q before that, and 31% the Q before that. Why? Seasonality! This coming quarter should see a number closer to 30% than to 10%, but don’t get over-excited when it happens.

Twilio - They’re at 13% already after just turning profitable (and FCF positive) a few quarters ago. They can push a good bit higher if they continue to bring OpEx down rapidly as they have started to do. But this is another one where Gross Margin is just over 50%, so don’t expect too much too fast.

Zoom - This quarter’s 37% is pretty wild, but with Gross Margins in the high 70’s they certainly have the ability to eke out lots of profit. The last 3 years they haven’t had much revenue growth, and they tried to buy growth in calendar 2022 which caused profits to drop vs the previous year (30% NM down from 38%). This continued into c2023 but they have started to bring OpEx down and for the full c2023 profits were up (35% NM up from 30%). I think this year they’ll have record profits – perhaps even a 40%+ Net Margin. That’s very high…but when you’re not growing, you should be dumping as much as you can to the bottom line, and when Gross Margin is pushing 80%, you should be able to.

Paycom - 29% is great and they’ve been profitable for a while, but I don’t see a ton of room to improve. I’m souring on this one for other reasons, but interestingly, the Net Margin story here is kinda good but boring.

Docusign - Their 24% is very good, and with a ~80% Gross Margin it could get even better if they bring OpEx down like Twilio and Zoom have. Docusign isn’t really doing that, though, yet.

Supermicro - Interesting profile here. I think 11% is the best they’ve had, but NM has been much higher the last couple years (high single digits) than it used to be (low single digits). They’re just very efficient. Gross Margin is only mid-teens (because they have a lot of what I’d describe as pass-through revenue) but OpEx is minimal and they have been consistently profitable. I don’t know if they can keep NM over 10%, but I definitely don’t think it can get much higher than that, due to GM being only slightly above that. But, hey, this is a rapid revenue growth story, not a margin-improvement story.

Context

In general, when a Net Margin approaches 30% or better, I start to expect that it will top out somewhere. After all, it will never be as high as Gross Margin, because OpEx won’t be 0. But heck, NVDA’s was Net Margin hit 59% adjusted (and I believe 57% GAAP) this past quarter…so, dare to dream!

Again, it’s vital to use all the other info you can (trends, FCF, growth, history…and crucially things like seasonality…etc etc) to add context to this Q NM number, which without context is near-useless.

Hope the above examples for the stocks I hold provides some idea of what I consider relevant contextual nuances. There’s more context to add than I had the time or energy to include. Questions welcome.

Bear

39 Likes
32

Thank you for the detailed explanation! There’s a lot to digest. :pray:

I’m curious about why you prefer a snapshot view as opposed to an approach that might also show trends. Is it because you already have the trend and context in mind, but find the snapshot view offers additional benefits, such as:

  • monitoring how the snapshot evolves to correlate it with other metrics or events or the “story”.
  • quickly comparing across your positions to gauge how this metric tracks across different industries and market conditions.
  • developing a deeper understanding or “feel” for this specific financial indicator, which helps when evaluating new companies. As you’ll have increased experience with it, both across industries and market conditions?

Perhaps other advantages too?

Wouldn’t a higher granularity (e.g. semi or annual) help mitigate seasonal effects? But then the “snapshot” benefits would be diminished too perhaps?

I’m also considering the usefulness of the net profit margin as a potential risk indicator. I imagine scenarios where a significant drop in this metric might signal increased risk, even when other financial metrics appear positive? Do you find that tracking net profit margin in this way helps to identify risks that might not be evident from other indicators alone?

8 Likes
33

I could go with a 4-quarter margin number instead of 1-quarter, which would help account for seasonality, but what about companies that have improved rapidly each quarter without much seasonality, like Monday? Their margin is much better in this quarter than it was 3 or 4 quarters ago. A 4-quarter margin number wouldn’t give you an accurate picture of their progress.

I could include 4 columns or 8 columns with the last several quarters, but I’ve found that you can include 50 or 100 columns if you want, yet you’ll still miss important trends. @SailingDev – I think I saw you have FCF margin on your spreadsheet. That works too – I also track it (on company-specific notes I keep, that can’t be condensed into the summary sheet).

I think the take-away here is, you have to develop a fluency with the numbers for the companies you own. Then, whatever number or numbers you’re looking at for whatever quarters or years, you can fold that into the bigger picture (aka, give it “context”). It takes time to develop fluency with each company, as they all have their own story.

Bear

27 Likes