WSM’s year-end portfolio Dec 2022

I do these write-ups as a type of journal which helps me reflect, and hopefully improve. That requires identifying what I got wrong, so that I can do less of that. Given that this is my last write-up for the year, that’s what I’ll be focused on. First, what I did in December, then a fairly critical review of what I’ve done this year and what I could have done better, followed by a review of my current positions. I hope my reflection on what I could have done better helps others as well.

RESULTS (post changing to a focused portfolio in Jan 2020)

2021 +53%
2020 +220%


**Dec	+53.2%**
Nov	+70.6%
Oct	+101.1%
Sep	+83.9%
Aug	+79.4%
Jul	+22.1%
Jun	+24.1%
May	+6.6%
Apr 	+4.6%
Mar 	-6.0%
Feb 	+7.7%
Jan	+7.6%


My portfolio lost some more value in December. Still, I managed a return for the year, at 53%, which handily beat my twin benchmarks of the S&P and aiming to treble my portfolio in 3 years.

In December I sold Confluent, Crowdstrike and Digital Turbine (again).

I sold Confluent because of the problems inherent in the business model of companies that grow from a foundation of open-source/… software. Conehead’s excellent post on open-source businesses () and Bear’s thoughts on Confluent (…) focused my mind. Thanks both.

I sold Crowdstrike to buy into SentinelOne. SentinelOne is growing faster than what Crowdstrike was growing at when they had similar revenues, and they seem to have an edge with their GTM strategy (as I wrote here:…). Also Crowdstrike’s revenue growth just does not want to accelerate, their absolute number of customers added decreased sequentially, and they seem to be increasingly fighting on all fronts (…).

Snowflake made its way back into my portfolio due to the good revenue and RPO results they posted.

Lastly, I exited Digital Turbine because I feel owning it is basically hoping for a pop based on a thesis that can take three years to play out, and driven by only one of their products (mobile). Or maybe it’s that I just do not like the CEO…I don’t think that I will dip my toe into this one again.

So three sells: Confluent, Crowdstrike and Digital Turbine, and two buys: SentinelOne and Snowflake. I also added marginally to most of my positions, and I leveraged them all a bit via call options, except for Amplitude. I’m including the options as part of the holding and not splitting them out separately.

Of my current positions, I will be looking to add to SentinelOne and will be looking into


Here is the composition of my portfolio, split per conviction tier, at month end as well as the two months before that (Dec first):

Monday			24.2%, 21.5%, 8.8%
Datadog			20.4%, 20.0%, 19.3%

Upstart			10.6%, 10.4%, 22.3%
Zscaler			10.5%, 9.3%, -
Zoominfo		10.7%, 9.1%, 8.7%
Amplitude		7.9%, 7.3%, -
SentinelOne		7% (new)

Pubmatic		4.2%, 6.2%, 2%
Snowflake		4.5% (new)

Crowdstrike		-, 10.1%, 11.9%
Digital Turbine		-, 1.7%, 3.0%
Confluent		-, 3.8%, -


Clearly I got a couple things right, given my return for the year, but this part is about improving, not back-slapping.

Last year in December I reflected that I did 4 things poorly in 2020 and would like to improve on them in 2021, namely trading in and out of stocks, buying too late, selling too early and not following through on my convictions.

This year I managed to improve on all of those points, and I still got sucker-punched a couple of times.

I bought into Upstart relatively early, and followed through on my conviction for Q2, which richly rewarded me. However, I did the same in Q3, albeit with a smaller percentage of my portfolio, so a little bit more cautiously, and got hammered as a result. I also followed through on my conviction on Lightspeed and got duly beaten up (and I was late to the party so that hurt doubly).

One lesson for me is that for both Lightspeed and Upstart there were times when I could have/should have sold or trimmed. For Lightspeed it was after the short attack and reading some of the more bearish analysis about the company on our board, and for Upstart it was after Bear and Jonwayne posted their thoughts on why they sold out completely after results.

Both times, based on where the stock went afterwards, I was evidently wrong and I should have sold out (I did sell a lot of Upstart after Q3, but still kept a sizeable position, and I sold Lightspeed immediately after their results, but I should have sold/trimmed before). Had I done so, I would have had a return for the year around 80%. In stead I’m 30%pts lower, at just north of 50%.

So I guess this leads me to a bit of a self-evident observation, namely that when following my conviction, I need to be right as well…and that takes more diligence than what I displayed on these two positions.

Maybe the lesson is that I must be more cautious when running a concentrated portfolio. Betting so big that it can take me 30%pts down is a risk that should be avoided. And I need to accept that I will be wrong sometimes even when I’ve got absolute conviction.

Accordingly, even when swinging for the fences it is prudent to limit the size of the portfolio at risk. I need to be extremely cautious, going forward, with any position, especially the larger ones, and need to have a more sane limit to any one position than the 30%-ish positions I took in Lightspeed and Upstart.

It is also worth reflecting on my last 3 months results - going from 101% up to 53% up in three months is terrible. Even though I ended up more than many for the year, and double the return of the S&P, my results these last 3 months sucked.

So in these last 3 months, what went wrong? In November I lost 30% of my starting capital, and in December 17%.

November’s answer is easy - it was mostly due to taking outsized positions in Lightspeed and Upstart, and leveraging both positions, and getting my fingers burnt twice.

And in December it was due to the general malaise in growth stocks, and aided by having kept a 10%+ position in Upstart.

Of these four things which really dented my portfolio - Lightspeed, Upstart, tech rotation and keeping Upstart after Q3, what could I have done differently?

I can think of 4 things:

  1. Going into the anti-tech rotation I could have kept more cash. This would have limited my losses. However it also would have limited my upside earlier on in the year. Saul tells us that he’s always 100% in. I’m not sure I can get this one right…so I might as well stay all in and not try to time the market. So I will pass on doing this part differently and just remain fully invested.
  2. Lightspeed. I could have been more critical of the thesis and focused less on the short attack (which was baloney but also a red herring). Because I got into Lightspeed way after most on the board had bought in, I did not know the company as well as others. Accordingly I should never had taken such a big position in a company that I did not know that well and knew only for a relatively short period of time. This one is clearly one to improve on.
  3. Buying aggressively into Upstart just before results. I did this twice this year - before Q2 and before Q3. The first time I made a killing; the second time I got killed. Trying to trade before results (incl. using leverage) requires a lot of effort and causes a lot of angst. Giving it a rest, like Saul does, is probably a good option.
  4. Keeping Upstart after Q3 results. When the narrative changed for the worse, I could have sold out completely for a while, even though I still believed in the company longer-term. I could then have bought back at current levels in stead of riding it down. Not sure I could have successfully done this for Upstart. For me, I either believe in the thesis or I don’t, and Upstart’s prospects continue to look great to me. Still, this is one to focus on to try to improve, as clearly others have the ability to time the stock.

Putting all of that into a pot and stirring gets me to this: 1) Limit the size of positions I take in any company to around 20%, and be extremely careful with those positions, 2) seriously consider stopping fooling around with options - it’s a zero sum game after all 3) think about the fact that I can always buy back later when the narrative changes for the worse in companies that I have a lot of conviction in (Like Upstart).

I think if I were to aim for one new-years’ resolution therefore, it is to focus more on the company and less on the stuff around the company: short reports, analyst up/downgrades, macro factors, valuation. It feels to me like when I try to focus on all of these things together I invite madness. And poor returns.


I’m sticking to the format I used last month - a condensed number of KPI’s for each company, an overview of why it’s special and why the CEO is the right person for the job. In all of the summaries below, the numbers are for the last 4 quarters (except where I don’t have 4 q’s history), with the most recent last. (MNDY)

LEADERS in workflow management with their Work OS product. The co-CEO’s Roy Mann (41, ex CTO of Wix) and Eran Zinman (37) are two developers from Israel who know each other very well. They speak in ensemble without interrupting each other and have a passion for making their products simple and easy to use. Clearly customers are loving it.

Revenue growth: 88%->85%->95%->95% yoy / 18%->18%->21%->17% qoq
Gross margin: 89.0%->89.7%->90.2%
FCF margin: -24%->-3%->-2%->+3.5%
NRR: 105%->107%->111%->115% / 119%->121%->125%->130% for >10 employees
Customer growth (>$50k ARR): 247%->219%->226%->231% yoy / 43%->27%->40%->30% qoq

These are truly exceptional financial metrics - just look at the improvement in margins and NRR and the meteoric growth of customers. I believe they are the leader in their category and they have two exceptional CEOs. The combination of increasing margins - FCF turned positive! and GM is above 90%!!, NRR acceleration and insane customer growth will undoubtedly continue to lead to hyper-scale revenue growth. And bear in mind that they define NRR as a 4 quarter moving average which likely means the current quarter’s NRR for 10+employees is above the 130%.

My notes about their latest results:…

Datadog (DDOG)

Datadog is the clear next-gen observability category LEADER, with an act two in security coming soon. The CEO Olivier Pomel (43) is a computer scientist - educated at one of the elite universities of Paris who felt the pain of having siloed tooling for observability and dev in a previous company and set out to fix the problem. He builds for developers, with developers and is an exceptional leader, speaking with clarity and calm while relentlessly, almost inevitably building newer and bigger things to “solve the whole problem” for his customers.

Revenue growth: 57%->52%->67%->75% yoy / 15%->12%->17%->16% qoq
Gross margin: 78%->77%->76%->78%
FCF margin: 9%->22%->18%->21%
NRR: >130% for all historical Q’s I could find
Customer growth: 35%->32%->36%->34% yoy / 8%->7%->8%->7% qoq
Large (>100k ARR) customer growth: 43%->51%->60%->66% / 13%->14%->12%->15% qoq

Exceptional financial metrics, with revenue accelerating from the 50%’s to 70%’s in the last 4 quarters. If I were to call out one thing that really impressed it is the acceleration in growth in large customers coupled with fantastic NRR. Their guide for Q4 Revenue growth at 64% is also very bullish, and they called out an expected improvement in GM for Q4. Lots and lots of momentum here.

My notes about their latest results:…

Upstart (UPST)

AI lending disruptor and clear LEADER in the space, and miles ahead of any competition, currently conquering unsecured lending in the US, with auto and mortgage debt in the US next. CEO Dave Girouard (54) is experienced in massively scaling tech businesses, having previously done that for Google’s Gsuite of products. He is calm, focused and passionate about the problem he is solving - fairer credit based on true risk.

Revenue growth: 38%->89%->1041%->250% yoy / 33%->40%->60%->18% qoq
Contribution margin: 48%->46%->52%->46%
Op margin: 12%->13%->19%->13%
NRR: Lumpy…
Customer growth (bank partners): 210% yoy / 20%->50%->39%->24% qoq
Loans transacted: 57%->102%->2316%->348% yoy / 53%->38%->69%->26% qoq

Astounding growth, margins and cash generation and superb growth in the number of bank partners. I think this last part of their story is under-appreciated. Ultimately their aim is to sign up lots and lots of bank partners - these are their “customers”. And this grew by 210% yoy in the last quarter! The CEO is the real deal and is in this for the long term, and dead set on building a massive financial services company. I take the point that many metrics slowed in Q3, of which the most impactful was the number of loans, which grew “only” 348% yoy / 26% qoq vs 2316% yoy / 69% qoq in Q2. But hey - that’s still supersonic. Clearly the yoy numbers are COVID-impacted so the qoq is the one to focus on, and yes, 26% qoq is slower than 69% qoq but 26% qoq growth is astounding however you slice it, and this business is more lumpy, so we should perhaps not expect the cycle to fall nicely into 3-month buckets, but perhaps rather 6, or 12 month buckets. Even though, as I argue above I could have sold and rebought now, I’ve held this far and see no reason to sell at these levels.

Some more thoughts on Upstart:…

Zoominfo (ZI)

B2B GTM Sales intelligence LEADER for the online/remote working world. CEO Henry Schuck (37) seems like a dyed in the wool sales professional who built a company that solves the problems that B2B sales professionals encounter daily. But, like his company, there’s more to him than is apparent at first glance. He’s also a doctor of law (cum laude) and Washington State and Nevada registered lawyer. The company has a red-blooded sales culture and the CEO is passionate about the problem his company solves and seems relentless in driving execution.

Revenue growth: 45%->48%->56%->60% yoy / 13%->10%->14%->14% qoq
Gross margin: 88%-89%->89%->88%
FCF margin: 55%->64%->53%->37%
NRR: >115% (my estimate; not disclosed)
Customer growth (>$100k ARR): 47%->51%->69%->74% yoy / 18%->12%->16%->14% qoq

This company just prints cash, and they’ve gone from a 40%-ish grower to a 60% grower in the last 4 quarters. Their large customer acquisition machine is strongly accelerating, clocking in at 74% yoy growth in the latest quarter, significantly higher than revenue growth. Add to this that many people under-appreciate that their NRR has already significantly ticked up from the last disclosed % of 108% a year ago (just because it isn’t disclosed does not mean it’s still at a year-ago level!!), and this is a very strong conviction pick for me. They will disclose NRR in Q4, and I’m predicting that a lot of people will then go “ah!”.

My thoughts about their latest results:…

ZScaler (ZS)

Zero Trust LEADER, with huge macro tailwinds and best positioned to capitalise on those. CEO Jay Chaudhry (62) is an Indian immigrant electronic engineer who completed three different masters degrees in the US after becoming an engineer in India and immigrating in the 80’s. He had a long career at IBM before founding and selling a string of security-focused companies and then fouding ZScaler in 2007.

Revenue growth: 55%->59%->56%->61% yoy / 10%->12%->12%->17% qoq
Gross margin: 81%->81%->80%->81%
FCF margin: 11%->32%->14%->36%
NRR: >125%
Customer growth (>$100k ARR): 46%->51%->52%->53% yoy / 12%->11%->13%->9% qoq

Just look at that latest qoq revenue growth and the accelerating trend! 17% qoq for Q1 is the highest it’s been in the 3 years of history I have for them, with Q1 for the last 3 years going from 9%->13%->17% now. The company is led by a hardworking and seasoned leader with an inspiring story and a career building security companies. Their growth and excellent cash flow margin is further fuelled by accelerating NRR and RPO and expected further tailwinds due to their strong positioning in zero trust and government.

Amplitude (AMPL)

These guys are young, energetic and the Digital product improvement insight LEADER. Co-founders CEO Spenser Skates (33) and CTO Curtis Liu (31) won MIT’s AI competition battlecode twice - meaning they can both code and know AI. Well. (

Revenue growth: 50%->48%->66%->72% yoy / 14%->10%->19%->16% qoq
Gross margin: 71%->71%
FCF margin: -15%->-35%
NRR: 119%->121%
Customer growth: 51%->54% yoy / 11% qoq

This company is led by a young but highly accomplished CEO with his CTO friend from university. Revenue growth and NRR accelerated even before their first two new products which were announced in the last Q (which will bode well for future NRR). Customer growth is also seemingly on an upward trajectory and they are surfing digital transformation as a huge and long-term wave/tailwind. FCF is still negative, though, and it’s not year clear to me if there is operating leverage kicking in and they don’t have a lot of history as a recent listing. Given these points, this may be a candidate to trim, should I want to increase another position (such as S).

SentinelOne (S)

SentinelOne is a CHALLENGER in the next-gen endpoint security market. They see their main competitor as Crowdstrike as per co-founder Tomer Weingarten, an Israeli computer scientist. On this one, the numbers really do say it all.

Revenue growth: 95%->108%->121%->128% yoy / 22%->25%->23%->22% qoq
Gross margin: 54%->53%->62%->67%
FCF margin: -86%->-87%->-98%->-37%
NRR: 117%->124%->129%->130%
Customer growth: 79% yoy / 11% qoq

$100k ARR customer growth: 138%->140% yoy / 27%->26%->25%->21% qoq
$1m ARR customer growth: 183% yoy

Wow, what a blistering pace of revenue growth, coupled with strongly improving gross margins and NRR and exceptional customer growth which is starting to move to larger customers (for a deeper look, have a read here:…). A look at Sentinel is not complete without also looking at how it compares to Crowdstrike when CRWD was at a similar revenue scale for me, because I sold Crowdstrike to buy SentinelOne.

			**CRWD	S**
**Q2 2019	Q3 2022**
Revenue			55.7	56.0
Rev g qoq%		19%	22%
Rev g prior Q qoq	18%	23%
ARR			208	236
NRR%			127%	130%
GM%			71%	67%
FCF%			-64%	-37%
FCF% prior Q		-35%	-98%
Op margin		-50%	-69%
Customers		1,800	6,000
Cust adds qoq		309	600
Rev/cust		33,850	9,825

SentinelOne’s growth velocity is evidently quite a bit higher than CRWD’s was back in 2019: it’s revenue growth rate is higher, ARR higher, NRR higher and customer adds higher. Hence, I would actually venture the opinion that SentinelOne is executing much better than CRWD was at the same stage of growth. The flipside is that cash burn and gross and operating margins are slightly worse. But for me the higher growth velocity weighs more and points to them continuing on a potentially even better trajectory than CRWD.

And what’s in store in Q4? They guided for 104% revenue growth. And to get a sense of the potential beat, from a Dec 8 Barclays TMT conference, quoting the COO:

“If you look at the growth there, 140% year-over-year growth with customers over $100,000 ARR.” and

“So what we’re finding is, large deals are also consuming more of our technology platform upfront. And so, if you combine that with further acceleration in terms of new customer ads, we’re really, really in a great place as far as customer acquisition is concerned.
And then entering into Q4 we have record pipeline. Typically from a seasonality perspective, Q4 is the busiest quarter of the year in cybersecurity in terms of purchases, especially at the higher end of the market. And so, we feel like we’re in a really, really great place going into Q4.”

Snowflake (SNOW)

Snowflake is led by Frank Slootman, who is essentially a hired gun, vs the founder-leaders of all of the other companies in my portfolio. It is the LEADING next-gen data cloud platform growing at triple digits at scale.

Revenue growth: 117%->110%->105%->109% yoy / 19%->20%->19%->23% qoq
Gross margin: 70%->72%->74%->75%
FCF margin: 9%->10%->1%->6%
NRR: 168%->168%->169%->173%
Customer growth: 73%->67%->60%->52% yoy / 16%->9%->10%->9% qoq
Fortune 500 customer growth: 52%->37%->36%->30% yoy

I bought SNOW again due to the accelerating revenue growth at this massive scale and an uptick in RPO. However, I was thinking a bit more about this position as I was writing this and went back to my previous posts on them, which made the point that I thought I should be focusing more on their very large customer additions than their total customer additions (here: and that I was under-appreciating their international growth. So I had a look at international growth from Q1 to Q3:

EMEA: >300%->185%->174%
APAC: >200%->170%->219%

So it seems this part of the story is still intact, albeit declining in EMEA.

However, this line of thought highlighted an issue I have with Snowflake: declining growth in the Fortune 500 customers as well as declining growth in total number of customers. When thinking about large customers for SNOW, the number of customers with >$1m in ttm revenue is a trailing indicator and therefore less interesting (vs customers with ARR over a certain threshold, which our other companies report, which is leading). The two customer numbers above - total and Fortune500 - are leading indicators, and not going in the right direction. Add to that that this is the only company that is not founder-led in my portfolio, and that it already has a very large market cap, and I’m wondering anew about this position. So, even though I know Snowflake will be a very important company in future and it’s still growing like mad, I will be looking to trim here to fund any increases elsewhere.

Pubmatic (PUBM)

This is my only ad tech investment. It’s a supply-side advertising platform CHALLENGER, with big macro tailwinds as the industry undergoes a fundamental change in how it is organised and works. The son of Indian immigrants, CEO Rajeev Goel (42) founded the company with his brother and the CTO (all 3 computer scientists). The CTO is based in India where they maintain a big developer presence to build and dev on their own infrastructure, which is part of their competitive advantage.

Revenue growth: 63%->54%->88%->53% yoy / 49%->-22%->14%->17% qoq
Gross margin: 80%->72%->74%->72%
EBITDA margin: 48%->33%->37%->42%
NRR: 122%->130%->150%->157%
Customer growth (publishers): 25% yoy / 5% qoq
Customer growth (CTV): 35% qoq!

This is a young, energetic team who come at the ad-tech problem from the technology, and not the media side (like leader Google, not like their large but slower competitor Magnite). It is a small company with strong growth fuelled by increasing numbers of high-quality publishers, strong NRR and super-charged growth from CTV. If there is one gap in the thesis it is customer growth, which has been relatively slow, and there is not too much quarterly information on this metric that I could find; however 2018->19 customer growth was 47%, 2019->2020 was 43% and Q3 2021 vs Q3 2021 was only 25%. So on the face of it a marked slowdown. However the business is seasonal so I’ll reserve judgment until Q4.

My detailed write-up on the company and Q3 results:…


Again, as was the case a year ago, I’m writing this from a country in lockdown. I truly hope that 2022 brings us lots of investing success and an escape from this virus-infested reality that we’ve inhabited for the last two years.

All of the very best to you all, wherever you are in the world in 2022!

  • WSM

Previous reviews:

Nov 2021:…
Oct 2021:…
Sept 2021:…
Aug 2021:…
July 2021:…
June 2021:…
May 2021:…
April 2021:…
March 2021 Q1 ytd:
Dec 2020 full year: