Whew!! Finished August BARELY positive. Was much more positive earlier in the month, getting my YTD to a negative number that started with a 3 (not rounded but still). Then the FED decided to do some more talking. Overall though, feeling good about August.
-47.5% YTD
August July June May
Snowflake 23% 25% 25% 23%
ZScaler 15% 21% 22% 22%
MongoDB 13% 13% 13% 11%
Data Dog 12% 15% 16% 20%
Bill 10% <1% <1%
Crowdstrike 9% 11% 10%
Cloudflare 9% 8% 8% 10%
SentinelOne 9% 7% 6% 13%
Transactions:
After earnings and going thru the reports, trimmed some DataDog to bring it from about 15.5% of my portfolio down to about 12%. Added these funds to Bill as my Bill position was tiny and is still very small.
Leading up to Bill earnings, trimmed Snowflake and added to Bill. Trimmed Snowflake because it was outsized (>25% and I’m trying to keep to under 20%). Bill then became a 9% position then grew after earnings.
Made a small move at the end of the month. Trimmed a bit of ZScaler and put that money into Sentinel One. I wanted to build up Sentinel One to closer to 10% (its at 9). It was tough to decide what to trim, I decided on ZScaler since its in the Security bucket. I almost trimmed MongoDB instead as I had wanted it to be closer to 10% from its 13% after my last earnings call but couldn’t get myself to pull that trigger. WISH I HAD!! Trimmed in the after-hours but that’s technically September I guess.
Individual position thoughts
Snowflake:
Wow, just wow. Surprise higher beat on revenue. In my opinion, Snowflake is the closest thing we have to investing in a pure hyperscaler at an earlier stage. Data is different. Every company has their law of large numbers but… AWS is still growing 40% with revenue north of 62B. I’m not sure if Snowflake will get to those numbers but I believe its slowdown will be much more gradual and bumpy (with accelerations and decelerations).
Financially, the upcoming (by end of year) GA of Python on Snowflake will bring “meaningful consumption contribution”. Python is a key in the data science world. This is bigger workloads than the traditional warehouse type of workloads. And what is “meaningful”? Like any new feature, expect a bump and for it to continue to grow over time.
New customers are important, of course. But they aren’t as important over upcoming quarters & usually years (with some exceptions as they stated on the call) as say Crowdstrike. 90% of the revenue in a year is from customers ALREADY on the platform. A couple quotes from the call,
“Most customers still ramping”
“I don’t see ANY of our customers fully saturated”
RPO
I’ve read a lot of consternation about their RPO again. We learned last year that its not really a metric to follow too closely. However, if you insist on following it, you should probably take the historical RPO numbers for Snowflake and apply them forward.
Warning, math ahead.
For fun (yeah, I’m a geek like this), I looked back at quarters where they provided xRPO (expected RPO consumed next 12 months) and we have 4 following quarters to see how they ended up doing. Unfortunately, this is a small sample size as we only have 3 quarters with these numbers, however, the multiplier came back for those three between 1.86 & 1.95. How did I calculate the multiplier? Take their xRPO (for example, 4 quarters ago, they said RPO 1.5B with xRPO 56%. This comes out to 840M. If you multiply this 840 by 1.95, you get to what the actual revenue was for those next 12 months.
For the most recent quarter, RPO = 2.7B & xRPO = 57%. This comes out to 1.54B. If we apply our multiplier (1.86-1.95), we get a range of 2.88B-3B. This is in the range of 75%-83% YoY.
I have NO idea if this is foretelling and am not putting much emphasis on this. Like I said, for those who insist on RPO being a useful forward looking metric for Snowflake, they should probably do this calculation.
Bill
I’m not the best source as to what Bill is doing for a couple reasons. First, I’m newer to following them and second is its my wheelhouse of knowledge (I’m a data and tech person). However, we don’t have to be experts in a field to invest and the beauty of Saul’s is it enables crowdsourcing. I’ve learned a TON about it from my friends here and on Twitter. Enough to build decent confidence in them.
One thing that’s interesting to me for Bill is that it seems that the platform for their FI partners (banks) is almost treated as a separate product (although its the same). I say this because they profit from the FI partnerships differently. On the call they said the FI partnerships have transaction revenue basically at cost. The subscription revenue is higher for them. For me, I think this means that the ramp-up of revenue might be faster but the top end revenue might be lower. The FI partnership grew faster last Q (6000 v 5000) and is a much smaller base and much smaller percent of total transactions (I calculated about 9% of transactions are from FI partners but 20% of “customers” are FI customers).
MongoDB
This is technically September but I was very disappointed after listening to the call. The larger slowdown in Atlas especially. I thought their workloads were much more mission critical but now I’m wondering if a large percent of applications running on them are more ancillary rather than mission-critical. To the point where their usage was effected. The only other rationalization I can come up with is that the Serverless that came GA had a bigger negative effect than they expected. This allows companies to only pay for what they used (similar to Snowflake). However, the difference being that a company can change from the old model where they’re paying for the instance to be up to the new model where they’re paying more granularly for what they use… I don’t know. The call was a lot of talk about Macro macro macro & tough comps. None of my other companies talked about macro like they did.
I’m mad at myself because I had this conversation after the last cc, “FF, lower your allocation going into the Q3 call” and as discussed above, I did not do it. Anyway, not shown here but I have closed out my MDB position and put most of that money to Sentinel One. I moved quicker than normal on this because $S hadn’t yet spiked in price as they should have. Will probably lower my $S but its now my #2 spot.