This is what I did about UPST

Going into the earnings I had the following plan:

Revenues < $250m & loans < 400,000 = reduce position
Revenues ˜ $250 & loans ˜ 400,000 = hold
Revenues > $300 & loans > 400,000 = increase position

By having a clear plan, it helps me take the emotion out of it. Bear in mind that I was overweight going into earnings and run a very concentrated portfolio (2-3 core positions) hence I can’t just sit back and let time take care of my portfolio.

So as soon as I saw the $228m revenues and 362,780 loans I hit the sell button. Even though this was my only concern going into earnings — that the expectations were high, and we might not hit that mark — I have to admit I was surprised by the results.

Some people might say that I focused too much on intangibles instead of following the numbers. Well, the only reason I do that is to try to add some color into the numbers we get between earnings. That is only to set my expectations and try to come up with what, I believe, would be solid earnings.

As I already said in a previous post, I won’t hesitate to pull the trigger no matter how much I like Upstart as a company if they came below my expectations. Well, they did. Of course, don’t get me wrong, I still like the company. I reduced my overweight position, but Upstart still carries a decent weight in my portfolio. Around 20% as of now. Yes, it might be accelerating down the line but for now, I just can’t justify my previous allocation.

Dave — for the first time — didn’t sound as confident and excited as other times. I always listen to the calls as it helps me get a feeling of the overall situation.

The revenues went from

86.7     121  194    228   255-265 (guidance)

QoQ increase

39.5% 60.3% 17.5%  23% (top guidance plus 6% beat as this Q)

[There is a clear deceleration here. I prefer the QoQ instead of YoY growth for fast growers as this gives me a better understanding of what is really happening. Even if they beat the top guidance again by 6% as this Q the QoQ would be 23% which annualizes to 128%. Still solid, but maybe not worth the much-added premium? However, with the AI/ML models being somewhat choppy in their performance we might see several more Qs of beat and raise as we did in the past.]

The loans went from

123,396   169,750  286,864   362,780

QoQ increase

37.5% 68.9% 26.4%

The reason I focus on these metrics is simply because the company defines these as key operating metrics:
We review a number of operating metrics, including number of loans transacted and conversion rate, to evaluate our business, measure our performance, identify trends affecting our business, formulate business plans, and make strategic decisions.

We define the number of loans transacted as the transaction volume, measured by number of loans facilitated on our platform, between a borrower and originating bank during the period presented. We believe this metric to be a good proxy for our overall scale and reach as a platform.

So far, they beat guidance by

2.5% 21.25% 6%

[Obviously, this is not an easy to predict company. Not a SaaS, not recurring revenue. But there is a bigger opportunity when there is uncertainty. The AI/ML models might improve significantly soon and the reaction would be good. However, with such high expectations it’s easy to go either way as we’ve seen.]
As I mentioned in previous posts, the market seems to get excited/frustrated not by earnings per se but by the acceleration/deceleration that happens.

The way I think about it is this: When you ride a motorbike, the exhilarating feeling (market’s reaction) comes in direct proportion to acceleration (QoQ). Not by speed alone (YoY). When you let go of the throttle (QoQ drops) it doesn’t matter if you go with 300 (YoY) or 100km/h you will still feel the deceleration (market’s reaction) instantly. Especially when the initial speed was higher. That is how I try to think about it when trying to predict the market’s reaction.

Another way I see investing is this: You are at the train station (stock market) waiting to go from A (initial value of your portfolio) to B (portfolio times x). There are several trains coming by (stocks). Some cost more some cost less (market cap). Some will get you there in 30 mins (hypergrowth) others in 30 hours (slow growers). You decide to hop on what seems to you the best for your own needs (investing style) and while going to your destination the driver (company) announces some news (earnings) that your journey will be delayed because of some issues (like logistics/supply chain).

Automatically what’s the reaction of the passengers? Well, depending on their personality (investing style) they might not care or focus on such small concerns (investors with large portfolios) or they might get furious (investors with concentrated portfolios). So, what do you do? You have the choice of getting off the next stop (sell/trim) and hop on another train (reallocate funds elsewhere) that is up and running (growing/accelerating) and is expected to reach its destination on time.

The only reason I make these parallels is because I have zero background in investing. Zero background in economics/accounting. Zero background in the stock market. Just a few months ago I had no clue what dividends were. I read the knowledgebase maybe more than 10-15 times before I could start getting a grasp of what it was about.

I would read 1 line and then spend 10 times as much time googling and trying to understand all the metrics used and whatnot. The only reason I post this is to be accountable for my own actions and to see in retrospect if those actions were right at the time being with the info available at that specific time.

I judge myself not by my YTD return (as I might just follow someone and get decent returns) but by the actions I take based on my understanding. I believe that this is the only way to become a better investor no matter if you manage 10K, 10M or 10B.

For this reason, I admire Saul and others who carry concentrated portfolios with a much bigger portfolio. In order to execute without emotion at these levels is not an easy task. I’d love to hear your thoughts. Thanks :slight_smile:

Note: I reallocated the funds into Semrush (SEMR) as it was another solid Q which drove the price up by about 20% which is not even close to where it was 2 months ago. I’ll do another post soon about it when I listen to the earnings call this afternoon (15:30 GMT+2 time).


Upstart presents us with a complex decision, which is why I haven’t said anything yet. I was going to wait until my decision was finalized before speaking up, but it seems like there is a big demand to hear from people who post their portfolios, so here is where I’m at:

On the one hand, revenue growth had been truly off the charts since covid. Starting with Q3 2020, this is what revenue growth has been QoQ:

282%, 34%, 39%, 60%, 17.5%

Basically growth went from off the charts to “on the charts”. 17.5% is great for any other company I own, except Upstart. I started to feel like we are witnessing another Zoom Video situation, where growth consistently fell after they had a couple big time blow out quarters.

The difference here between Zoom Video and Upstart, is that Upstart is providing strong guidance of 16% growth next quarter. By comparison, in Q2 Upstart guided Q3 to be 11%. So we are seeing comparatively strong guidance, and I would bet growth for Q4 will come in over 20%, which is pretty great, although not unheard of (Snowflake comes very close to this).

Unlike most of my other companies, Upstart is not a SaaS company. Its revenues are mostly not recurring on any basis. It’s going to have to make a veritable **** ton of loans next year in order to keep growing like this. The absolute growth in loans transacted slowed down from 117,114 (69% QoQ) in Q2 to 75,916 (26.5% QoQ) in Q3. And where things really slowed down was in the growth of the loan transaction total, from 62% QoQ growth in Q2 to 12% growth in Q3.

I also expect Upstart to give us guidance for 2022 on their next earnings call. I would imagine that they will have to be ultra conservative here (since it’s for a full year) and may even try to tell us growth will be flat for 2022. This is going to weigh down on any potential upside we may get from the Q4 results. So it’s possible we sort of have dead-ish money until they report Q1 2022, which would happen approximately in May 2022.

I’m not sure where this leaves me on Upstart. On the one hand it seems silly to sell a company consistently growing over 17% each quarter. And I do think there is potential for Upstart to come in much higher than 17% QoQ going forward. On the other hand, non-SaaS weaknesses are highly apparent here, and the latest quarter has presented us with a very material slow down. My other holdings may offer better risk/reward trade offs.

For now I have trimmed the position to 12% of my portfolio. Unlike Lightspeed, I’m not done with my decision yet. I may sell more or I may buy some of my position back.