Boom and Bust, Market Brutality and Opportunity in Q3
What a Q3 earnings season. Company after company reported and sell-off after sell-off ensued. On the very same day, Thursday November 4th, both a ‘reopening play’ (LSPD) and a ‘Covid play’ (Peloton) were down 30% after earnings. A week later and another board favourite, Upstart, was down similarly.
This was then compounded by a sector rotation with many high growth tech names selling off, and market brutality again ensued. But in brutality there is opportunity.
What has been going on? Well as per my name, I have some thoughts.
Investing Background
I haven’t posted on the board since last Q3 earnings a year ago. In truth I didn’t want my words to be construed as financial advice. When I started investing semi-passively a few years ago I thought first abstractly and long term about what industries were the future, and companies were secondary to that. With no investing education at all, and without even reading earnings reports at the time, I then bought what I thought and read might be potential disruptors in those industries. With that in mind, I fortuitously picked up shares of Etsy at $35, Enph at $11, Tesla at $45 (adjusted) & IIPR at $20 among others. Great huh? Well yes, if only I had held them all until now. In April 20 with the pandemic I suddenly had more time and cash and decided to begin investing actively. Unfortunately an outcome of this was being exposed to a lot more noise, re-strategising and selling out of said names at a fraction of the profit they would’ve been now. So what?
Well soon I realised that while I could accurately model a company’s growth rates even if contrary to analysts or company guidance itself (in part thanks to my financial analysis background), that didn’t mean the market would see things the same way as I did and there was often a disconnect between my and market expectations which I struggled to bridge. And so I chose to sit on the side-lines from this board for a while, while trying to understand better the confluence of growth rates and market expectations, and how to beat them.
I hear a lot that the market is forward looking. Perhaps. But it is not forward looking commensurately. More often than not it is forward looking one or two quarters at a time at best, and one reason for the sell-offs or sudden spikes is mixing up one-off and underlying growth rates to create heightened or tempered expectations. This has at times not been favourable to investing with a long term/buy & hold or abstract mindset in recent months and has required to be flexible with positioning.
And one thing the market definitely doesn’t like is uncertainty. It’s human psychology after all. But here it’s my belief that with either finding certainty in a long term mindset, or even better, with timing – there is opportunity.
Underlying growth
What is underlying growth? I feel it is a very important principle to understand in investing, but I see it being overlooked all the time. As simple as it sounds, it is important because underlying growth will best help you interpret what the forward growth rate is. At a basic level underlying growth means growth excluding one-off impacts. Now what is truly ‘one-off’ is a grey area and can be interpreted differently according to the company. Do we have a sense of the underlying growth rates for all of our companies? Some are easier to determine than others.
For companies with short term macro impacts, underlying growth can be extremely difficult to model. Can macro tailwinds count as underlying growth? I think that comes back to how sustainable they are.
Upstart
Let’s use Upstart as an example. I’ll start by saying I like Upstart a lot as an abstract and long-term investment, I first bought at $60 and have a cost basis of $110. But I’ve seen its growth misrepresented on various platforms time & again and believe market expectations were too high going into Q3 which led to the sell-off.
First, I’ve seen some investors using a YoY growth rate of 250% in FY21 to justify forward growth. In Q2 20 Upstart was severely impacted by Covid and lost virtually all its revenue. Normalising the comparatives Q2-Q3 20 would at least adjust your YoY growth to 160% as a starting point.
To help understand underlying growth rates, we first need to break out its growth/revenue drivers. For Upstart:
Loan size x loan originations x take rate = revenue
What are some of the assumptions/drivers are that go into each of those? For Upstart there is a lot to consider.
Loan sizes: there is a general macro trend of downsizing combined with Upstart loan values declining due to lowering of FICO requirements. But assuming macro downsizing moderates, will Upstart’s impact on loan sizes moderate too? Do we have the mix between the two drivers? This seems quite a crucial assumption modelling Q4. I’d suggest we don’t.
Loan originations : I won’t profess to be an industry expert, but at a high level can we understand how much the macro environment has driven loan originations in the last few quarters versus underlying performance? The only pertinent macro comparison I saw ahead of Q3 was deceleration of Lending Club originations from 63% to 14% QoQ and Customers Bank which was flat QoQ. This didn’t strike me as macro acceleration going into Q3. Management suggested their macro assumptions were static forward looking. Other than macro, there are origination drivers in model improvements/funnel efficiencies/conversion rate and new wins.
In predicting a huge Q3 beat, did we have a clear sense of whether model improvements that drove Q2 outperformance were repeatable (management had already said the impact of these would be lumpy). For modelling contribution of new wins for Q4, is it fair to use existing bank partners to predict contribution without considering relative members/AUM or to use historical macro driven origination data. How about the impact of marketing trends or simply if the loosening of FICO requirements is at least in part a one-off impact (ie how quickly until that sub-prime segment becomes saturated)? Would there also be any impact from dormant subprime lenders during Covid coming back as competition?
The point is, it is very difficult to get a sense of Upstart’s underlying growth when trying to predict it quarter from quarter, because it is impossibly hard to strip out all these elements accurately. We simply do not have enough information, and I’m not sure relying on Google Trends data is the most reliable indicator as we saw. Therefore, for a company like Upstart, you just have to trust their guidance is accurate because they do have the information.
This unpredictability causes uncertainty and volatility. We’ve seen this phenomenon numerously with Zoom, Fastly and others before, with macro growth becoming difficult to strip out from underlying growth, the market building up expectations where it assumes the growth is all underlying and selling off when the impact dissipates.
This is where heightened expectations create a ‘risk on’ environment heading into an earning season and if you’re invested, to have cognisance of this risk.
Heading into Q3 earnings I was becoming sceptical about the fervour surrounding Upstart, the anticipation of 30%+ QoQ growth and assumption that management was sandbagging to the degree of a 20-25% beat. I saw some investors predict $300m+ vs $228m actual. The high expectations made the sell-off feel inevitable when it came. For next year some are predicting FY22 revenue of $2bn. Possibly, but what is underpinning this? What mix of personal and auto loan and what are your assumptions of each?
For me it pays to step back and think long term. Upstart claim their Total Addressable Market (TAM) is $81bn personal loan originations (shrunk from $92bn estimate in Q4 20), so extrapolating their Q3 originations Upstart today is at 15% market share. What is a realistically attainable market share in the end game? TAM is often a ‘made up’ number or misquoted ambition.
How much of that TAM can Upstart actually reach – ie what is their Serviceable Addressable Market? For example, 50% of the TAM? Taking this example this would put Upstart at 30% penetrated. In this case there is a ceiling in the next 1-3 years of ‘hyper growth’ where its personal loans opportunity could become fully penetrated. In Q4 20 Upstart’s market share was just 5%. So Upstart has gained 10% share in just 3 quarters (using the two relative TAM’s quoted). Is it realistic to expect this torrid rate of growth over FY22, putting them at ~30-37% market share in a year’s time (or 70% penetrated for eg). Probably not. Is it early enough along its S-Curve? Enough.
And so the outlook turns to the much bigger markets of auto loans and Upstart’s other new loans offerings to sustain hyper growth beyond FY22. While we ought to trust management to make inroads in these new markets, we know mortgages, small personal loans and SMB loans won’t be incremental to FY22 revenue – leaving us with auto. While the indicators are very good that auto will be a ‘meaningful growth driver’ next year, we don’t have enough information yet to model accurately what ‘meaningful’ actually means. And so there is a leap of faith that management will execute in the same style as they did for personal loans.
Sanjay Datta, Upstart CFO, Q3 call:
“We plan to continue investing heavily in our technical workforce as we ramp investments in machine learning, auto retail, fraud and as we scramble new teams to begin tackling opportunities in segments, such as the small-dollar lending, small business lending and mortgage initiatives”
“Scramble”. There is something frantic and desperate about the word scramble. A response to an emergency, as you’d scramble fighter jets to a hijacking. It might seem facetious, but I think a lot can be learned from semantics and tone of management comments in earnings season. It is, at least, something done with urgency, at pace. Are they scrambling due to the ‘emergency’ of slowing personal loan growth? Or because the opportunity is here and now and they don’t want to miss it. Let’s hope the latter.
However, if you believe notionally that an AI powered cloud platform will reform lending and that you have faith in Upstart management to execute on its opportunity at a fast enough pace, then why would you not be invested – other than uncertainty. This is a fair enough reason. But often as the market is too quick to bid up expectations, it’s too quick to bring them down. ‘Good enough’ is understated for progress to a long-term goal, it might pay to focus on the long-term with Upstart; that it is directionally on path. Everything is a question of risk reward. Boom or bust or somewhere in between.
Q3 thoughts
Mostly everything has already been said for stocks covered on this board, so rather than regurgitate what’s been said before, I’ll offer a few thoughts linked to the above notions on underlying growth.
Datadog
For the value of understanding underlying growth rates versus market expectations of growth, Datadog has been a case study. A year ago after Q3 20 Datadog sold off as it notionally ‘decelerated’ even as its usage reaccelerated. And so, in the uncertainty came opportunity, with Q4 and Q1 presenting timing:
https://discussion.fool.com/datadog-redefining-expectations-3466…
Oliver Pomel (RBC TMT conference on Nov 16th 2021) here explains the main driver of their accelerating growth rate as simply: “What we hear and what we see in the market is that the rate at which enterprises are moving workloads into the cloud and growing those workloads in the cloud has resumed to pretty much what it was before the pandemic”.
Management had told us as early as Q2 20 earnings call that growth rates were returning to pre-pandemic levels, and the market ignored it. Datadog has macro tailwinds with the cloud transition but this is adjudged to be long-term and sustainable. And importantly, its growth is predictable.
Looking forwards:
“What do you think we’re going to be most surprised with a year from now versus kind of what we know today about Datadog?”
Olivier Pomel:
”I think the surprise might be that a lot of it will look the same, but the year is a little bit further along. The thing with the heavy land and expand motion we have and the expansion for the new products too is that it is very gradual. Like we don’t have something that all of a sudden jumps to the top. The new products come in. They’re small initially and grow very fast and they gain importance and gain scale, and we layer them on top of the other products. And that’s what’s been happening in the past, and I think that’s what we want to happen in the future.”
What we might infer from this simple statement, is that ”Datadog’s growth is not going to slow down any time soon. ” I recently saw some comments that growth would slow “due to the law of large numbers”. I don’t subscribe to this thinking; there is no ‘law of large numbers’. To me that’s a cop out. As an organisation scales it will face coordination headwinds and it may begin to saturate its market, but that’s not due to numbers. My post from a year ago is as apposite today as it was then as to why Datadog can maintain its high growth rate for a long period:
https://discussion.fool.com/i-think-that-the-low-set-is-unrealis…
Another interesting snippet I took out of the conference was:
"The long-term trend, I would say, is one of data explosion. And it sometimes can get completely out of hand. Any application can generate an arbitrary large amount of logs. I think how we can deal with that in a way that doesn’t scale costs for customers in an unreasonable fashion is to give them feedback loops so that they can understand what they consume and why and create a lot of value, and then give them levers so they can decide what to consume from us to deal with that data.
But whether they send that to us or not, it is generated, it is costing them money. So we need to help them optimize that. If there is no feedback loop, this is a runaway cost on them no matter what. We don’t do ELAs. And the reason for that is we’re SaaS. And we’re SaaS in a domain where data can explode in volume. And we feel good about that.
And we help them do that within Datadog for everything we sell at Datadog. We also started helping them doing that for the rest of their cloud consumption. We announced a new product at our conference, which is Cloud Cost Management. And the goal is just that. It’s help you close the loop on where the money goes, what’s consuming dollars, what aligns with your applications, your customers, your changes, your processes and how you can actually install the right feedback loops directly between the costs, like finance teams and the engineering teams that are generating those costs. So you can drive continuous optimization there."
An aside: we know that Datadog’s growth stalled in Q2 20 with the onset of the pandemic from concerted customer usage optimisation, whereas optimisation is a continuous process usually. But how will this feedback loop turn out with their new product ‘Cloud Cost Management’? For e.g. I manage a IoT connectivity usage based platform at work and we migrated platforms to allow for more effective customer optimisation, completely underestimating the level of optimisation that then entailed which ended up being detrimental to growth. Ours is a very different platform though; Datadog can provide enough ROI with its platform, this helps with any pricing pressure from customers’ ‘data explosion’ costs – and it is yet another product to sell. But from my perspective, the less ‘optimisation’ is mentioned on earnings calls the better.
LSPD
Another company featured a lot on this board in recent months, I’ve never really been convinced by how big LSPD’s TAM is or by its positioning. However, I think it’s an interesting example of underlying vs macro growth, and the unpredictability of trying to model underlying growth rates for these hyper-macro impacted companies.
In Q1 LSPD reported organic GTV growth of +91%, including hospitality +380% YoY and organic total revenue growth of +78%. This lapped Apr-Jun 20 – the height of lockdown. The stock appreciated significantly.
In Q2 lapping non-lockdown summer 20, GTV organic growth ‘slowed’ to +39% and total revenue growth to +58%, the stock sold off even more dramatically.
Perhaps on both the way up and the way down, organic growth equated to underlying growth in the market’s ‘mind’, particularly as acquisitions have confused the overall revenue picture. But how sustainable was GTV growth of 91% given the quarters they lapped really and how justified the sell-off despite ARPU +17% QoQ (albeit slowing customer locations).
The market seemed to think the GTV growth had been sustainable, and now it thinks the revenue deceleration is indicative of a more permanent slowdown; I’d challenge both conclusions, but that’s just my take. And so we saw the mania on the way up and the brutality on the way down. Boom and bust.
Opportunity
Other than the stocks heavily featured on this board, which I also own variously, I’ll try to offer something different.
I first stumbled on this board in about April 2020 when researching what information that I could find on a little known company called ‘Red Violet’. At the bottom of the post, I noticed the poster had a 28% position in ‘APPS’, which was about $4 at the time: https://discussion.fool.com/rdvt-red-violet-34417298.aspx
Curious, I found another thread about APPS (Digital Turbine) on the board, but it was quickly dismissed as ‘a turnaround’, ‘a history of fraud’, ‘only available on Android’, ‘bloatware’, etc etc. But I was still curious, so I researched some more and ended up buying a small position.
Fast forward 20 months later to today and I have recently doubled my allocation, with APPS now at ~$55. Up over 10x in 2020, I believe there is still upside ahead. I’d not argue this is a ‘great business’ or in the league of your Datadog’s or Crowdstrike’s etc, but I believe that a ‘great business’ and a ‘great investment’ is not necessarily one and the same thing. Not all investments need to meet the same criteria either, at least for me.
It is very high risk, arguably at terminal risk to unseen privacy or regulatory changes that could come along, and I am not an industry insider – so be cognisant of that. But there is varying risk with all our businesses as we have seen, it’s a question of risk reward; and timing.
This post is already too long to go into my thesis now and I know it’s a largely unpopular stock here, but I wrote a short thread on Digital Turbine here:
https://twitter.com/thinking_stocks/status/14643730309105745…
Dead and Buried
There is opportunity in market dislocation, a market that seems to be either slow or reactive or too ambitious in equal measure, flitting from mania to despair one week to the next – ‘schizophrenic’ if you like.
Nothing has been on the receiving end of more market brutality this year than ‘Covid plays’, with many of them reverting to pre-pandemic valuations. There is plenty enough risk or opportunity in other stocks featured here and elsewhere, and although I minimised my exposure to them late last year and early this year, I am keeping a close eye on one or two Covid names which I think one day in the not too distant will re-emerge from the ashes. But it’s not quite yet the opportune time to post on them. The time to invest will be when you can identify the crossover point where underlying growth is reflected in a reversion of YoY growth comps and before the market catches on to it. Probably this won’t be apparent until H2 22 if it executes. I believe that timing can be learned.
This won’t be a ‘turnaround’; it’s my assertion that the market is misinterpreting growth pull forward as underlying and the consequent deceleration as indicative of a terminal growth rate, which has created uncertainty – the thing it hates; lest we hope none of our stocks grow too quickly to be fatalistic when running up against its own comparatives. Dead and buried to the world.
Why bother with uncertainty at all, since the market doesn’t like it? Well, for me anyway, I want to be invested in the companies of the future. The future is uncertain, I’m prepared to hedge my bets and ride some uncertainty along the way. And sometimes what is perceived as ‘uncertainty’ is only exposed after the fact.
Crowdstrike is the last of my holdings to report, on Wednesday. I haven’t touched a share in over a year, my thesis has not yet changed, in spite of the multiple re-ratings down recently. Here’s hoping for a little less earnings season brutality: more boom, less bust.
Thanks for all the great content as I’ve lurked this year 
