A different analogy

[PaulWBryant]

I’m certainly in the camp that the execution is what matters, but I’m glad Saul ended the Upstart thread. The below is not further UPST analysis, but I use UPST as an example to make a point, that point being: analyzing Upstart is soooooo challenging right now. I’ve created an analogy to explain why I think that is.

JonWayne, StockNovice, Saul, myself, and many others have discussed our analysis of UPST’s quarter, and here’s what we focused on:

1. What was expected
   and
2. What was delivered

We found the quarter wanting. We did not consider who was selling or what was said on CNBC. In aggregate, the market is doing the same as we are – when it comes to picking between stocks. When it comes to sector rotations and overall market behavior, like the recent pull back for all hypergrowth stocks, the “dogs running faster than a man” analogy explains the market’s behavior. But it doesn’t explain why Upstart has been hit much, much harder – why the market picks and chooses. If that’s what you want to explain, here’s a different analogy.

Not a bunch of dogs, but a bunch of scientists, are predicting when the man will get to the park, and when he will get home, and how many times he is likely to go to the park this year, and for the next 10 years. They create extremely complex models to predict all this. So complex that some scientists argue they’re all WAY off because we don’t have perfect knowledge of every atom in the universe. But all the models are largely based on how fast the man is walking.

The man slows from walking 5 mph to 2 mph. The models of course go haywire because they don’t have enough information to interpret WHY the man slowed. Maybe the man saw a friend and started walking and talking and so maybe the slow down is a random event that has no bearing on the future. Maybe the man sprained his ankle and won’t be able to go to the park as much this year. Maybe the man broke his ankle badly and is limping as fast as he can to get to the hospital – maybe he won’t ever get back to walking as fast or as much as the scientists had previously predicted.

The model is now broken and is almost surely either overestimating or underestimating the future for the man. But until the scientists get more information, it’s impossible to know which.

Bear

[XMFMonkey]

From a Monkey to a Bear:

Yes!

And notice that the part about “getting more information” requires something human animals call “time.” Which is why, in part, Saul’s knowledge base mentions we buy companies with the intention to hold forever; time simply has to be given a chance to do whatever it is it does.

Otherwise, we’re more like frogs, jumping from lily pad to lily pad, rather than the dignified furry creatures that we are.

But how much time, exactly? There’s the beef. For some who practice the investing art of ruthlessness, they are not willing to allow much time into their equations.

Those who consider themselves more in the species of “long-term-timers” are often willing to give it at least 9-12 months but often closer to 2 to 3 to 5 years to do its thing.

As investing is an art that needs to take individual consciousness and context and motivation and circumstances into consideration, there is no correct “time frame” objectively speaking.

So in this particular instance, Monkey sees Upstart as having a tremendously successful future as both a company and a stock—-given enough Time. One quarter registers as barely half an hour on Monkey’s watch. The current growth rates as cold hard facts, the profitability metrics and potential future markets and optionalities say: Patience, you must practice, young Monkey.

Likewise with Zoom: if, say, we wake up 5 years from now and Zoom is at all-time highs again, then the “temporary” dip that happened in the 2nd year sandwhich was just a blip that time needed to repair, and zooming out far enough, in terms of years, might allow for the CAGR to make it a successful investment after all, with much less jumping around in between. Of course, there’s all kinds of time cost issues, etc., and this is not a post about Zoom being a good investment; rather, just a reminder that if you identify truly great businesses and business models, you also need to identify how much time you’re willing to give them to do their thing. Forgetting that is like forgetting the art part of investing and you’re going to be either too reactive, or not reactive enough when the facts change.

Put more succinctly: complexity tends to resolve itself over time. And patience is a virtue that allows time to work on your behalf.

Hugs,

Monkey (long ZM (again) and UPST)

[Vishrant]

Interesting post. Would like to know more about why you think ZM is a buy here again ? I have great regards for CEO & his team but not able to get into Contact Center space (with the failed FIVN acquisition) have dented my expectations.

[Smorgasbord1]

When ZM dropped below $200 I looked to see if I should reinvest. But, just take a look at the YoY sales/revenue growth numbers (https://csimarket.com/stocks/single_growth_rates.php?code=ZM… ):

'20 Apr: 169%
'20 Jul: 355%
'20 Oct: 367%
'21 Jan: 300%
'21 Apr: 191%
'21 Jul: 54%
'21 Oct: 35%

What’s going to turn that around? New products? Stealing customers away from MS and Cisco?

[mekong22]

'20 Apr: 169%
'20 Jul: 355%
'20 Oct: 367%
'21 Jan: 300%
'21 Apr: 191%
'21 Jul: 54%
'21 Oct: 35%

What’s going to turn that around? New products? Stealing customers away from MS and Cisco?

I look at it a different way. I’ve actually been following ZM a lot closer lately than I had a few months ago because I consider the valuation pretty compelling right now.

Based on the above trend in revenue growth, if you are only interested in ZM if it accelerates and gets back to hyper growth, then you’re not going to be a future investor in Zoom.

But I believe the valuation right now only factors in future growth of about 15% per year for the next couple of years. So, to me, if ZM can grow just 25-30% per year for the next two years, I think today’s price is exceedingly cheap and investors are likely to do very well as shareholders. I can see the stock price doubling over two years with those results.

It’s a much larger company than many others we follow here. In the past 12 months, ZM did $3.9 billion of revenue. Compare that to $0.9 billion for DDOG or $0.3 billion for MNDY and it’s clear that it will be a lot harder for a bigger company like Zoom to grow at very high rates going forward. That’s certainly one of the negatives for them, as a future investment, as it’s a lot harder to double or triple from a very high starting point.

That being said, I’m not likely to be buying ZM any time soon for two reasons.

1. I don’t have a particularly high confidence that they will grow even the 25-30% per year going forward that I mentioned above. I think it’s very possible, but there are just a lot of variables regarding how ZM performs over the next couple of years that are far outside my areas of competence, such that I’m not going to bet on it right now.

and

2. there’s just so many great companies selling at incredible valuations right now that I can’t see reallocating funds from a much higher confidence position to buy into even a starter position from ZM today.

Most likely, no matter what ZM does from here, they aren’t going to be at the level of high growth that makes them on-topic for our discussions, but as an investment, I won’t be surprised if they exceed the market’s current expectations and provide fabulous shareholder returns over the next 2-3 years. At the end of the day, as many of us do, I emphasize revenue growth, and it’s potential for accelerating trajectory going forward as, as important, or more so than most other metrics in evaluating investments, but it’s still only one piece of a large, complex story for any company.

-mekong

[XMFMonkey]

What mekong said .

But for fear of getting permanently banned and tarred and feathered for talking about a mere 35% (and declining) grower on this board, Monkey won’t pontificate much more about Zoom as an investment for slowing-growth reasons.

However, before he shuts his bananahole, he remembers owning Apple sometime in 2014, and thinking–well, they already sold all the phones, and they’re expensive, and it’s hardware, and competition and commodification, and… and… and… That was when it was selling at a stock split-adjusted $20… and it’s $168 today. Somehow, Apple found ways to grow, and this despite already being humongous!

Is Zoom Apple? No.

But Zoom has SaaS which we adore for obvious reasons.

And it has a brilliant leadership team with extraordinary new developer talent.

And it has the context of a revolutionary shift toward hybrid work from home that’s not going away and is still in its infancy.

And it has a brand that the entire world recognizes, or will soon recognize in all the jungles of the world.

And it has 14 consecutive quarters of 130% Net Dollar Expansion Rate, which we love about all fast growers here. What are all those companies expanding into, exactly? Must be something they value…

And it’s planting banana trees like crazy: $375Million Free Cash Flow!

And it has innovation in its genes.

Therefore it has Apple-like qualities.

And anytime Monkey can more or less guarantee himself 25%+ stock appreciation for years–through growth or capital allocation stock-buy-back shenanigans or some combination thereof, he screeches.

With a healthy plop of patience, Monkey’s recent shares bought for $186 per Zoomie have a fundamental risk/reward ratio that’s highly attractive, here, so says his bananas placed on the Zoom market-roulette-wheel slice.

But only time will tell if his pea-brain will once again outsmart his big-brained homo sapien cousins. His confidence level about this is medium-high, all the while acknowledging having eaten a rotten banana or three in his investing life.

Hugs,

Monkey

[AThinkingFool]

Only on this board would we get a Bear and a Monkey talking about dogs and scientists

Context and expectations

I really like the analogy but I’m not totally convinced it applies to a company like Upstart, as used. Why? Because to get from 5mph to 2mph you need to first pass 4mph and 3mph - a linear path. But we know that Upstart does not move in a linear path. But from 2mph to 7mph to 0mph to 10mph.

"Our revenues and operating results could vary significantly from quarter to quarter because of a variety of factors, many of which are outside of our control. As a result, comparing our operating results on a period-to-period basis may not be meaningful."

And so to Monkey’s observation, the differentiator is time. Or ‘timing’.

The analogy would be more apt in my opinion for a SaaS company, which I used a similar analogy to a speeding car last year for Datadog’s usage based model:
https://discussion.fool.com/evidently-11-sequential-growth-repre…

The reason this seems important to me is that because this informs what the ‘expectation’ and therefore the ‘execution’ is. Not every company can be painted with a linear sequential growth brush like SaaS. Or ask yourself, if Upstart was a public company in Q2 20 when its revenue declined 81% sequentially, would you have sold out? Wouldn’t that have been the time to buy.

It is very true that Upstart is much more difficult to predict than other companies featured here and it is paying for this uncertainty, although I won’t claim anything about this market with 99% certainty. Though we know the stock also suffered after it blew out Q1 & Q2 beyond all expectations, irrespective of ‘execution’.

Are we therefore basing our expectations fairly? I noticed in Stocknovice’s (a great investor and a very unfitting name) expectations for example that Upstart failed all 4 of his tests; which is fair enough. But if we think about it, it seems that maybe the company only failed one of those tests and the rest by default (the Q3 revenue missed, but the Q4 +16% sequential guide was in line, the FY guide therefore missed because of the Q3, and the loans is a KPI/driver of the Q3 revenue. One is subset of the other). Or that’s just how I look at it.

Is it right that we draw arbitrary lines in the sand for companies, where we would sell out no matter what? Because it seems to me, to do this, would be only to deny yourself context. And in my opinion, context is the most important thing we can have as investor. It is what separates us from an algorithm.

We only need to look back to Q2-Q3 2020 to see this exemplified with Datadog, which ‘missed expectations’ and sold off. Would we have sold out? If it weren’t for context. Instead we listened to management comments that the slow down was due to one-off optimisation exercises and it is now most people’s top conviction position. The writing was on the wall, with 68% growth vs 38% FY21 guide thanks to this context (https://twitter.com/thinking_stocks/status/13614494379906826…). What is the context for Upstart? Resources redirected to confront fraud issues among other factors and quarterly revenue that is lumpy and not meaningful on a period-period basis.

Clearly due to its unpredictability Upstart is a riskier name than others and it does require a leap of faith, to some level, in execution into new product areas. It’s totally understandable not to invest in the company because of these reasons, but judgements of its demise because of a sequential slowdown is perhaps premature.

Zoom

To Monkey’s questions. I have been following Zoom closely, although I minimised my exposure first at the vaccine news, then again early this year when it became apparent the market did not like uncertainty over its post Covid growth. But actually, the company has performed pretty much to my expectations (although it hasn’t blown them away, which has stopped me pulling the trigger yet), and I’m looking to Q4 to rebuild a picture going into FY23.

At the beginning of the year, I asked the question ‘how much is enough’ for Zoom’s growth in FY22 (this year): https://discussion.fool.com/zoom-how-much-is-enough-34713078.asp…. I posited that analyst expectations were of 35% growth and how Zoom could get ~60% growth or $4.2bn (https://twitter.com/thinking_stocks/status/13574254509062266…) and why there could be upside to the analyst view. Well it is on course for $4.1bn or ~55% growth, although clearly the market has answered this was not enough.

Zoom has positioning, but just yet it hasn’t done enough to capitalise on that in my opinion. However, I’ll revisit that question going into FY23 because I think the answer to ‘how much is enough’ may well be different, due to the rebased expectations and positive underlying trends.

I’ll try to take a break from posting for a while so I’m not saying the same thing in different ways. At the end of the day, everyone has different opinions and ways of looking at things and this can only be good for critical thinking and for us to continuously improve as investors

[stocknovice]

Are we therefore basing our expectations fairly? I noticed in Stocknovice’s (a great investor and a very unfitting name) expectations for example that Upstart failed all 4 of his tests; which is fair enough. But if we think about it, it seems that maybe the company only failed one of those tests and the rest by default (the Q3 revenue missed, but the Q4 +16% sequential guide was in line, the FY guide therefore missed because of the Q3, and the loans is a KPI/driver of the Q3 revenue. One is subset of the other). Or that’s just how I look at it.

Is it right that we draw arbitrary lines in the sand for companies, where we would sell out no matter what? Because it seems to me, to do this, would be only to deny yourself context. And in my opinion, context is the most important thing we can have as investor. It is what separates us from an algorithm.

First off, I appreciate the compliment. To your broader point, I wouldn’t say your description about how I approach it is incorrect as much as incomplete. I don’t consider any expectation as an arbitrary line as much as an individual standard each of us gets to set for any company we own. And the only person who determines the fairness of those standards is the one setting them. It’s their money at risk after all.

In my case, I use these expectations as an initial check telling me where to dig deeper if needed for the context you mention. In UPST’s case, it did indeed fail all 4 “tests” with loan transactions being my main KPI. Looking further, I was also uncomfortable with the “context” of decreased loan size, only 12% QoQ Fee growth, lower conversions, fewer automated loans, the decrease in contribution profit margin, and an admission by management of potential fraud risk that wasn’t there before. Though I only listed my headline expectations, I considered every one of those variables other than the fraud admission (which wasn’t available until later) before drastically cutting the position.

This is the same process I described with LSPD (https://discussion.fool.com/good-question-hindsight39s-always-20…). After it failed the revenue/guide test, I double checked the context of customer growth and GMV before making the sale.

As you detailed, the difference with DDOG is the context of the underlying metrics and management’s comments kept me in the name even though it failed the Covid quarter revenue test. I did trim it though based on the new information.

In none of those examples do I feel like I denied myself context. In fact, it was the additional context I gathered that caused me to exit UPST/LSPD while staying aggressive with DDOG. However, it was the initial expectation which steered me toward the appropriate context in order to act quickly if needed. So far, I have no reason to regret any of those decisions or even more importantly the process that led me there.

[Tamhas]

And the only person who determines the fairness of those standards is the one setting them.

Which is not to say that some standards may be reasonable and some not; some may be good criteria and some not.

[stocknovice]

Which is not to say that some standards may be reasonable and some not; some may be good criteria and some not.

Absolutely, and I don’t believe I in any way implied my standards are any more or less reasonable than anyone else’s. Why? Because they are mine and mine alone.

While I 100% believe in the power of crowdsourcing, I also understand I am solely responsible for my own decisions. Honestly, the same applies to anyone running their own portfolio. What standards to set and how to use them is an entirely personal call.

Unless you were making some other point I missed.

[Texmex]

So in this particular instance, Monkey sees Upstart as having a tremendously successful future as both a company and a stock—-given enough Time. One quarter registers as barely half an hour on Monkey’s watch.

This is precisely my view as well and that is why I have not sold a single share. Let us look at the numbers. In Q3 UPST surprised their revenue projections by 8.8% (228.5/210) while DDOG surprised their revenue projections by 9.4%. Yet, many seem to think Ddog had a great Q while UPST had a poor one.

Yes, the 12% QoQ loan rev increase is lower than I would have liked but UPST business is not something you should be measuring QoQ or even YoY like a SAAS company IMHO - It is going to be lumpy and impacted by macro. There are several valid reasons for the low QoQ which have been discussed to death and I don’t wish to go into that.

Broadly speaking when I look at UPST I see a company with a humongous TAM - way bigger than our cloud companies. Personal loans TAM is $82B, Autos is $720B and mortgage is $4.5T. By my calculations, they already have 11% of the personal lending space. If (of course, it is a BIG IF) their AI model advantage can translate into autos and mortgage, UPST is going to become a HUGEE company. That is not going to happen in a linear smooth fashion. It may take a while to attain penetration in other lending markets. I can see 2022 as mostly about gaining more market share in the personal lending space. The risk of course for UPST is whether other companies can easily duplicate their AI models. Or that other lending markets are not easy to disrupt like the personal lending space. Current evidence of bank additions seems to suggest that they are executing rather well.

One final point. Even though UPST has fallen 67% from its 52 week high as of today it is still up 360% from its 52 week low! What if the stock price had progressed smoothly upwards instead? Would we feel differently about UPST? Also, despite the precipitous fall the ytd returns are way better than Ddog.

[SaulR80683]

Broadly speaking when I look at UPST I see a company with a humongous TAM - way bigger than our cloud companies. Personal loans TAM is $82B, Autos is $720B and mortgage is $4.5T…If their AI model advantage can translate into autos and mortgage, UPST is going to become a HUGEE company.

TexMex,

I’m not saying anything against Upstart here, or their business, but you are mistaking their TAM.

Their Revenue TAM in Autos, for instance, isn’t the $720 billion total amount of loans. It is the 4% to 7% of the loan that Upstart receives in fees. The loan isn’t their Revenue. Their Revenue is the fees that they get.

Granted, if they pull off autos and mortgage, they will become a MUCH bigger company than they are now. MUCH!!! But you can multiply those numbers by 0.05 to get their real TAM, and can forget about those trillions.

Best,

Saul

[qazulight]

Granted, if they pull off autos and mortgage, they will become a MUCH bigger company than they are now. MUCH!!! But you can multiply those numbers by 0.05 to get their real TAM, and can forget about those trillions.

Best,

Saul,

Upstart has the potential to truly disrupt the world. While a Crowdstrike, Cloudfare allaince may remake the entire internet and how the world communicates, Upstart can change the way the world trades. While our super SAAS companies are changing the way the connected communicate, Upstart is in the position to change the way all 8 billion people in the planet trade.

The problem with calculating TAM in that situation is it is not static. Take the graph algorithm that drives (or used to drive) the Google search. That one algorithm not only captured all the TAM at the time it came out, it created massive new markets. My guess is that the market cap of Data Dog is larger than the entire projected TAM of Google when it first showed up on the web.

So, while Upstart could be in a position to capture a percent of the entire trade of the world each year, that estimate may well be low for the TAM as the trade itself creates wealth.

Not saying Upstart will be the one. The first with the best technology do not always win.

Betamax
Amiga
etc

Cheers
Qazulight

[Conehead]

Saul:
Their Revenue TAM in Autos, for instance, isn’t the $720 billion total amount of loans. It is the 4% to 7% of the loan that Upstart receives in fees. The loan isn’t their Revenue. Their Revenue is the fees that they get.

And I’ll also note, since I was the auto loan cheerleader yesterday, that I expect that the fees to be much lower for auto loans. Rates are much lower for collateralized loans so the margin is much lower for banks, so there is less profit to go around. Also, the benefit to the lender of Upstart’s services is lower. With an unsecured loan a default is often a loss of the majority of the loan’s value. So if Upstart can improve outcomes by 10% (just as an example) that 10% almost goes directly to the bottom line. With a car loan, a default is still a negative event for the lender, but they (by definition) have more recourse. Improving outcomes by 10% is still an improvement by it won’t have as dramatic an impact on profitability.

Since I try to avoid short posts, here is another interesting read. https://files.consumerfinance.gov/f/documents/cfpb_subprime-…

Much of this document covers markets that I don’t believe Upstart plans to address, but most of the concepts still apply and much of the problem this paper tries to outline are still valid.

Section 2 segments lenders into five types (banks, credit unions, finance companies, captives, BHPH) and types (direct, indirect). This is a longer explanation of some of the things I talked about yesterday: about how the often indirect nature of the lending (because of the dealer) complicates this market.

However, most auto loans are “indirect,” i.e. intermediated by car dealerships, who arrange loans for consumers to purchase their vehicles and are compensated by lenders for doing so. In this way, the auto loan market is unique among large credit markets in the U.S.

Section 2.2 dives even deeper: talking about a fundamental problem

Auto loan prices and terms are not standardized and often depend on the lender, the borrower, the dealer, and the specific vehicle purchase to be financed. As a result, potential borrowers often do not know the kinds of loans they can qualify for or where they can obtain the best terms or rates…Economists refer to markets that operate in this way as characterized by “search frictions,” because consumers must search for prices and may not find the best prices they could in theory obtain.

While this search friction is sometimes exploited by the dealer, the dealer fundamentally wants to sell a car. In the easy cases, I’m sure borrowers will go with a captive lender. But in the cases where credit is trickier, if Upstart can reduce this friction and help a dealer present a more affordable solution there is a chance for a win-win-win scenario.

The rest of the paper, especially Section 5 is a dive into the statistics of why the interest rates aren’t well correlated with risk and default. I don’t want to try reproduce the paper here, but the key takeaway for me is that we are in a world where there are some opportunities for Upstart and their abilities to better predict and price risk.

–CH

[Texmex]

I’m not saying anything against Upstart here, or their business, but you are mistaking their TAM.

I realize that Saul. If you take their TTM revenue and divide that by the TTM loan transaction $ volume you get 7%. So, you can take 7% as their cut. Adding up the mortgage, auto, personal lending loan origination and multiplying by 0.07 gets you to $368B which is the TAM for UPST. we are still leaving out many other types of loans. Still, $368B is much higher than that cited by our cloud companies which are all in the $50B range. Everyone realizes all these TAM #s are very rosy. But the point is UPST if it executes and can maintain its AI leadership has the potential to be a very large company. We cannot judge it by weak QoQ loan originations based on 1 quarter. UPST is a rule-breaker and the industry it is trying to disrupt (lending) is very conservative and ancient. We need to give it some time.

Long UPST (11.5%), and Ddog (17.6%).

[simpleisbetter]

So, you can take 7% as their cut.

I expect the take rate of auto loans and mortgage to be much lower than that of unsecured personal loans. Assuming 6% personal loans, 3% auto, and 1% mortgage, that gets us a $70B TAM (810.06+6720.03+4500*0.01). Still plenty of room to grow given they’ve only captured 1% of that.