Investing in AI – Corvel

I unintentionally made an investment in an AI company. You can read my original thesis on May 27, 2020. Like most companies around that time, it was trading heavily down during the Covid crisis. Indeed, my intentions were everything other than investing in AI at the time.

Fast forward nearly four years, the stock has rocketed upwards to valuations that are difficult to rationalize.

The company is operated by a reclusive management and they have very terse press releases. They quit doing their no-question conference calls (which gave slightly more colour to their business operations) after their January 2023 quarter. No analysts follow the company, there are no EPS estimates, and almost nobody knows that this company has a huge competitive advantage in its niche.

Amusingly, I have noted they have been using the two-letter “AI” phrase in the last few press releases.

On May 27, 2021, they first used the words “artificial intelligence” in their press release:

CorVel Corp. applies technology including artificial intelligence, machine learning and natural language processing to enhance the managing of episodes of care and the related health care costs. We partner with employers, third-party administrators, insurance companies and government agencies in managing workers’ compensation and health, auto and liability services. Our diverse suite of solutions combines our integrated technologies with a human touch. CorVel’s customized services, delivered locally, are backed by a national team to support clients as well as their customers and patients.

January 31, 2023:

The Company has also continued work in the area of digital transformation. Most recent efforts have focused on enhancing CorVel’s document repository system with AI-centric technologies. The advancements being implemented automate the extraction and codification of critical data, which can then be leveraged dynamically within systems. The development roadmap for the quarter and beyond includes increased automation and augmentation, which will further optimize bottom-line results and outcomes.

May 25, 2023:

In other areas, CorVel is moving forward quickly and intentionally, using generative AI in a closed-source data environment. The technology will be incorporated into CogencyIQ® service offerings and has extensive benefits. Most importantly, generative AI will elevate the work of claims professionals and allow more time to be spent interacting directly with injured workers. The reallocated time will ultimately improve the experience of injured workers and enhance partner outcomes.

August 1, 2023:

CorVel’s 1st generative AI initiative will be released in the September quarter. The release will reduce mundane, repetitive tasks and provide decision support at critical inflection points. This automation will add to the existing machine-learning tools with increasing capabilities within the system. The Company also views generative AI as an effective tool to mitigate labor challenges and provide guidelines for future generations of professionals. In the quarter, investments in the foundational systems and workflow processes continued to strengthen the results achieved with CorVel’s products and services.

October 31, 2023:

In the payables market, developments were made in both the revenue cycle management arm, Symbeo, and the treasury services department. At Symbeo, hyperautomation, a combination of AI, machine learning, and robotic process automation technologies, presents an expanded opportunity in the market. By using Symbeo’s payable solutions, partners receive the benefits of touchless digital invoices, AI enabled optical character recognition, a machine learning Document Classification model, configurable AP rules, and standard ERP integrations via Robotic Process Automation which provides faster invoice cycle times, lower total cost of ownership, and an enhanced user experience.

Finally, today, on January 30, 2024:

The implementation of generative AI initiatives has been proven to boost the efficiency and effectiveness of both the P&C and Commercial Claims teams. These updates have gradually reduced time spent on routine tasks, thus creating more time for essential activities that require critical thinking, directly impacting the user experience and results achieved.

…

With today’s earnings release (the 3rd fiscal quarter for the company), we have a past 12-months EPS of $4.32/share. The company historically has been able to slowly increase its per-share earnings both through a combination of higher net income and also through the deployment of a significant share buyback program which has been running for a couple decades – the capital allocation strategy appears to be holding about $100 million cash in the bank and dumping the rest of the free cash flow into buying back shares. While the company has been able to reduce its shares outstanding by 139,000 shares from December 2022 to 2023, much of the $55 million spent was to offset prior option issuances. Needless to say, with the stock price as high as it is, those equity options are all entirely in-the-money.

Focusing on the $4.32/share, this gives a backward-looking P/E of 55, which makes Corvel the richest (highest valuation) stock I currently own. The valuation has me concerned, but I can easily see a scenario where for some reason the hype decides to bid it up even further. It is very difficult to predict these things. Who is to say that the capitalized value of their software technology is not worth well greater than the $4 billion market cap? Maybe some insurance giant wanting to internalize their own software operations (and de-licensing competitors) would be a strategic bidder for $8 billion in stock? My guess is about as good as anybody else’s.

While this is still a top-5 in my portfolio, I did pare a little at the 200 level to justify my sanity a little bit. But instead of chasing Nvidia, I will take solace in this – at least the company itself is set to generate cash for a very, very long time.

Yellow Pages – What now?

By virtue of two mandatory share redemptions (post 1, post 2) and some other dispositions, my position in this company is a lot less than it used to be (which at one point was over 10% of my portfolio!). However, what is remaining is not an insubstantial sum as measured in dollars and cents and I also have some sentimental value with this company as I regard it as one of my best calls ever which I made public on November 2019. Indeed, it is my longest-dated portfolio holding.

Other than capital return decisions (which has reduced the company’s shares outstanding from 26.6 to 14.2 million), the other significant event appears to have happened on March 21, 2023:

On March 21, 2023, the Company was the target of a cybersecurity incident. The Company immediately activated its internal network of IT professionals and retained the services of cybersecurity experts to assist in securing its systems and to support its internal investigation. The Company also suspended its operations and IT systems to contain the situation.

As of May 10, 2023, the Company had restored all its operations and IT systems and has taken steps to further secure all systems to help prevent a similar occurrence in the future. The Company is working with its insurance providers to make claims under its policies.

In their Q2-2023 release, they stated when discussing their decline in revenues:

… (b) a cybersecurity incident which resulted in the Company’s operations and IT systems being suspended for approximately three weeks of the second quarter of 2023.

While this did not have a material impact on Q2 (I was jokingly thinking that the company is so old-school that it generates cash even when they don’t have any computers operating), it certainly had an impact on Q3’s metrics, which significantly underperformed the decline regression model I had.

Now that this incident is over with, presumably the company can get back to its regular decline.

The only difference is that instead of working with a base of 26.6 million shares outstanding, there is about 47% less shares out there to deal with.

Management explicitly wanted to get the last buyback out of the way in calendar 2023 because of the 2% buyback tax that was going to be imposed in 2024. It would suggest that future buybacks are unlikely – perhaps this is why the stock is edging down lower due to the lack of anticipation of future demand.

This is an odd case where a company has removed nearly half its shares outstanding, but the share price has actually decreased despite the underlying financial metrics being relatively stable (the graceful decline downward). The company continues to trade at less than 4x cash generation – if there is any hints that profitability will stabilize, this multiple will rise. This has always been the thesis and 2024 is probably a good a year as any to see if it occurs. If there was ever a case study to deal with a company that should just go private to alleviate itself from public company hassles and expenses, this one is it. In the meantime, the dividend yield is around 7.5%, so at least we’re being paid a few bucks to wait for the inevitable.

Corus Entertainment

Corus Entertainment (TSX: CJR.B) has been a slow melting cube business case, dealing with the woes of competing against Netflix and streaming media, and the internet in general. Despite these competitive factors, they have still been able to generate a prodigious amount of cash flows.

They released their fiscal quarter results on January 12, 2024 and the following was their quarterly cash flow statement:

I’ve circled some relevant numbers. The key parts are that they have huge capital requirements to acquire program rights (this part makes the EBITDA look very attractive!) but these costs are necessary and increasing; they also have non-controlling interests which get claims to dividends that ordinary shareholders do not receive.

When netting this out, you have about $15 million in effective cash to work with. Management has a very strange interpretation of “free cash flow” which I won’t get into here.

This effective cash is contrasted to the $1.08 billion debt they have. Management is very fortunate to have $750 million of this at much lower fixed rates (5% and 6%) due 2028 and 2030 (the rest of it is in a credit facility) but needless to say, this leverage coupled with non-controlling interests do not leave the prospect of equity returns to be very good given the melting ice cube. The fixed rate debt is traded on the open market at a 13% yield to maturity at present.

There is some residual clout in ‘traditional media’ and Corus does still reach a lot of television sets and radios across the country. Perhaps some deep-pocketed individual will want to take control for strategic/political and not necessarily financial reasons.

The entity does appear to be more viable if they got rid of about half their debt.

Some quick thoughts at the beginning of 2024

Two data points, I am not adding any value to the universe with this post:

The Nasdaq 100 had a +55% year, while the Nasdaq Composite was +45%.

I don’t think there is any degree of active portfolio management that would match this number.

The correct strategy in 2023 was to put your portfolio into 5 equal-sized chunks, in NVidia, Facebook/Meta, Tesla, Microsoft, Amazon and Apple. (You can sub-in your favourite large-cap darlings here, including Google and the like, but you get the idea).

No sane portfolio manager would do this.

It is very similar to the times in 1999 how if you weren’t in technology stocks (whether large cap, or the trashiest of the trash dot-com companies) in 1999 that you were guaranteed to under-perform.

The question is whether we will see “momentum”, or “regression to the mean” going forward.

I truly don’t know anymore. I note that, separate to the investment world, I have been receiving some email correspondences that are worded like they were generated by ChatGPT. Indeed, when I entered the text of the email and asked for it to form a response, it spit out some language. I then asked ChatGPT to simplify it, and what came out looked like a carbon copy of what said individual emailed to me.

I think from this point forward I’m just going to resort to in-person face-to-face communication.

However, others will opt for the convenience of not having to parse language in their heads, let alone spill it out on a keyboard (or god help those that can use touchscreen phones to do their typing). They will not have to deal with grammar, or even have to think about anything. AI will take care of it.

So perhaps it isn’t too late to buy those deeply out of the money calls on NVidia, it is banking on the intellectual laziness of people – sadly a safe bet.

================

In 2024, Canada will have an active stock buyback tax of 2% applied on net share repurchases. While the legislation is more technically worded, essentially “net” in this case means that share buybacks beyond offsetting share issuances (whether through SPOs or option issuances, etc.) will be taxed at 2%. I don’t wish to get into the stupidity of how meddling in capital structure is going to cause perversions (indeed, prior to this, share buybacks are tax-preferential to dividends and this in itself was a perversion), but nothing makes the government more happy to claim to the public they are sticking it to greedy corporations with their share buybacks than applying an additional tax.

While 2% is not a huge penalty to pay, it is one more unnecessary friction impeding the return on and of capital. I am looking at the companies in my existing portfolio and am wondering if the ones voraciously performing share buybacks will be more keen to compensate through the issuance of stock options. Again, 2% is not a huge penalty to pay, but it is yet another annoyance.

Changing Course for 2024

You’ve got to know when to hold ’em
Know when to fold ’em
Know when to walk away
And know when to run
You never count your money
When you’re sittin’ at the table
There’ll be time enough for countin’
When the dealin’s done
– The Gambler, Kenny Rogers

This seems to be the most appropriate song to describe the feeling I am getting at the close of the calendar year.

Probably the biggest sign of whether you have a grasp of reality or not is whether your mental model of the world can correctly predict the future.

Unfortunately, as I have written here before, one does not necessarily need a truthful model of reality in order to survive. Donald Hoffman is a cognitive psychologist that has made some very interesting lectures on the matter and this disconnection between truth, reality and survival – this has always been in the back of my mind.

In particular in our 21st century age of information, mis-information and dis-information at our fingertips, coupled with AI bots, deepfakes, etc., it is getting very difficult to distinguish truth from fiction.

Ironically during the Covid era (from March 2020 onwards) I felt like (who knows whether I actually did!) I had a better grasp of reality than most, and indeed as it translated into the financial markets, it was a very rare time where most participants were so strongly positioned for disaster that it was possible to make reasonable gains when people’s perceptions of reality normalized to some semblance of truth (i.e. we’re not all going to die – just look at a chart of Moderna (MRNA) or Alpha Pro Tech (APT)!).

The financial marketplace is actually a reasonable place to measure the perception of one’s reality against others. Note I did not say “truth” – the old cliche of “the market may remain irrational longer than your ability to remain solvent” is true in many cases, say for those that wanted to short the economically unprofitable cannabis sector in 2016, the dot-com market in 1998, or the short-lived upsurge of plant-based meat companies IPO’ing around 2018-19. Never mind Gamestop! Even if your sense of reality is closer to the truth than others, the market can dictate reality for longer than one thinks.

Going back to present, the Covid effect has slowly abated over time and sometime around 2022 the markets began to be the same old efficient market machine that we have been used to – the primary difference between 2022’s market and onwards was that we exited out of the zero interest rate environment.

In retrospect, I was inappropriately positioned for 2023. It seems increasingly likely that the reality in my brain is not corresponding with what is actually going on out there, and as a result, I need to discard these narratives out of my head and start from the foundation again.

You could already tell that I had significant amounts of self-doubt in the middle of 2022 – where I mercilessly started to cull elements of the portfolio and raise cash, and I was soothed by the fact that cash had obtained a reasonable yield once again. The baseline performance for doing nothing (or rather resting on the sidelines) was actually pretty reasonable.

My doubts on my grip on financial reality have continued to increase even further – one should never invest when you are flying blind. The other rule is a break in thesis – one of my tenants going into the fossil fuel trade back in 2020 was that North American production would not be able to eclipse their Q4-2019’s highs for various reasons (resource exhaustion, drilled and uncompleted numbers low, lack of capital spending, later on – Russia being cut out of the oil equation, costs of drilling, self-induced ESG restrictions, etc.) but this is a failed and broken thesis. US production in particular is now at all-time highs, despite all the narrative.

World demand also continues to be very high but for whatever reason, there seems to be ample production capacity.

My continuing fossil fuel trade in 2023 has been incredibly offside with reality. I consider myself lucky to still be marginally positive YTD performance in what has been a very uninspiring year financially riddled with errors of omission (i.e. not listening to my instincts earlier and getting out at higher pricing).

Instead, what we are going to get might be similar to what happened in 2014, and is typical of cyclical industries – a terrible race to the bottom. Low cost producers and those that can provide additional value (e.g. refineries, mid-stream, etc.) will survive, but returns are likely going to be muted going forward. This is the conventional financial playbook and it is one that is telling me to fold ’em. And that I have been doing.

I did venture away from my Covid playbook a tiny bit in 2023, but not to a significant degree. 2024 will also exhibit a change in focus. The markets have always been about adaptation and survival and I am fortunate to begin the new year with half a clean canvas to work with. I am not in any rush to deploy capital, however – I am not at all sure that my grip on reality or the truth is quite where it needs to be. I am worried that my thoughts are currently too close to the consensus out there. As a result, if there are going to be any movements, they will be baby steps.