Earnings season is traditionally the busiest period for us portfolio managers. Ordinary days are quite relaxed. You wake up, read a few earnings call transcripts, prepare for a call with the CEO of a company that’s down 70% YTD but, for some reason, keeps talking about leveraging up to do M&A, and finally post a meme about something that probably shouldn’t be funny.
Earnings season is different. For some reason, every company in your portfolio always decides to report earnings on the same two days. You try to update your models, listen to the earnings calls, read the transcripts, figure out what actually changed, and somehow form an opinion before the next company reports 20 minutes later.
A few years ago, we decided to start keeping score. We keep a short list of companies that we believe surprised positively, negatively, or were sort of in line. Not based on management guidance or consensus expectations, but based on our own long-term views.
If we look at earnings and believe the company is a winner, we write it down. That way, we’ve formed our opinion before the market opens and starts playing tricks on our emotions.
So with Q2 now well behind us, what stood out? And what were the losers this quarter?
Input cost Inflation and margin pressure
We have seen many companies post big earnings misses because of cost-price inflation. Especially low-value-add, low-margin companies were hit hard, while rising freight costs ate away at what little margin they had to begin with.
Is your business growing bananas in Colombia and shipping them to the US (DOLE)? Tough luck.
Do you turn milk into fermented yogurt drinks which you sell to everyone who screwed up their gut with GLP-1 injections (LWAY)? You should have kept an eye on wholesale milk prices… your company’s earnings are all gone!
Do you make cement for data centers (AMRZ)? Demand is skyrocketing, but nobody will compensate you for higher energy costs: your earnings are down!
In times of inflation, you want to own the companies with pricing power. Also, distributors tend to have an easier time passing on cost inflation than manufacturers. We’re revisiting distributors in sectors such as chemicals for a reason.
This all seems obvious in hindsight. But with inflation in Q3 undoubtedly higher than in Q2, it still feels relevant.
But there was another new theme in the earnings releases, something we expect to hear more from in the near future.
SEO and LLMs
The other new trend is being found by prospective new customers on Google in a world that is rapidly shifting to AI search.
The way customers find businesses through Google Search is changing drastically, and many businesses either face higher marketing expenses or wonder where all their customers went.
This isn’t entirely new. Companies that depend on online sourcing of customers have long blamed changing Google algorithms for quarterly losses, roughly the digital equivalent of a retailer blaming bad weather.
We still remember when our favorite nicotine e-commerce play (HAYPP SS) saw its top websites, nicokick.com and northerner.com, lose their top spots in Google Search after Google changed its algorithm in late 2024. This turned out to be a short-term hiccup.
Google ultimately optimizes for customer satisfaction. If it sends you to a website, it can measure whether you immediately hit the back button or spend twenty minutes browsing. Do you come back later? Do you make a purchase? Google has spent decades getting pretty damn good at figuring out whether it sent you somewhere useful. The best websites therefore tend to find their way back to the top.
SEO optimization needed occasional tweaks. You could hire some consultants, change a bunch of words nobody actually wanted to read, sacrifice a goat to the Google algorithm, and move on with your life. But if you had a good product, your customers would eventually find you. Haypp’s websites reclaimed their top positions within a few weeks.
But today’s discussion has shifted. It’s not about SEO; it’s about being found through LLMs.
How do you get ChatGPT to tell its users to use your product?
The consensus seems to be that you need 3rd party content to talk positively about your product. There is a reason we are long Reddit (RDDT). Having some anonymous accounts on RDDT conclude that you make the best running shoes, offer the best mortgage rates, or sell the tastiest nicotine pouches is increasingly valuable when an LLM is trying to figure out what product to recommend.
For businesses that historically relied on Google to funnel customers toward them, figuring out how to become the answer that ChatGPT gives may become one of the more important marketing problems of the next few years.
Trade Schools
Which brings us to trade schools, a recently introduced theme on which we’re bullish. Today’s youth is increasingly willing to forego the traditional four-year degree, along with the six-figure student loan and questionable employment prospects that sometimes come with it, in favor of learning a trade where employers are actually desperate to hire them.
Clearly, the performance of our favorite plays: Universal Technical Institute (UTI) and Lincoln Education (LINC), has been a real disappointment.
So what happened?
LINC mentioned headwinds in digital search in Q2. Have you tried asking ChatGPT where you should get your car mechanic degree? They’ll send you to a Community College as opposed to LINC or UTI’s more expensive results. Never mind that graduation rates at trade schools are twice as high, starting salaries are higher, and the entire curriculum takes half the time because they don’t force general education classes down your throat.
So LINC responded by leaning more into the high school channel and hired a bunch of UTI’s sales reps, which means UTI will miss in the current quarter.
In addition, Google search now shows you AI Overview, followed by a long list of paid search results. Organic search has become non-existent. The most valuable traffic for a trade school is a student who saw a LINC advertisement and then continued their research online. But if LINC has disappeared to the second search page in favor of paid search results, LINC loses out on its high conversion traffic. What if part of the problem is Google’s continued monetization? What if search is now completely pay-to-play?
While we’re annoyed with the share price reaction, we’re not concerned. The issue isn’t the product; it’s making sure the product can be found. It is like the hiccups in SEO optimization that tended to take a few weeks to a few quarters to work themselves out.
Companies will have to adapt to a world where LLMs get to decide where customers go. Fortunately, those same LLMs will tell you in detail how to do that. UTI and LINC will have to make sure that their graduation rates and starting salaries are accessible to LLMs. They need third parties to write about how good they are. Don’t talk about a Porsche partnership – let Porsche do the talking! These are minor changes. They just need to figure it out.
Meanwhile, new campuses are ramping up faster than expected. And most importantly, the trade schools keep signing B2B deals. In the past, B2B deals meant UTI didn’t have to pay for a Porsche engine to teach their students. Today’s B2B deals mean Heartland pays to build out a UTI dental hygiene campus just to put its name on the door and be first in line when making job offers to UTI graduates.
Our main concerns when initially writing about UTI and LINC were that 2026 is an investment year, and the mental arithmetic required to make UTI and LINC seem cheap.
At single-digit EV/EBITDA multiples, the valuation doesn’t require much imagination to be in value territory.