CRVL – Divestor

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.

The power of reinvestment and compounding returns

This post should not be news for anybody in finance, but it is worth refreshing fundamental principles of compounding and equity.

The most attractive feature of equities vs. debt is the effect of compounding. Stocks can rise infinitely while bonds have an effective price cap at the risk-free rate of interest.

To get on the equity gravy train and make outstanding returns, you need to have capital invested in a business with great prospects for reinvestment.

We will use an example of a debt-free company earning a perpetual $10 per year on a $100 investment, but the investment is of a style that does not scale upwards with further reinvestment. The long-term risk-free rate of interest is 2%. Our dream world also does not have income taxes or management expenses.

In this instance, the company can choose the following policies (or a blend thereof):
1) Give the money back to shareholders.
2) Bank its cash and receive 2% on that capital,
3) Speculate on other (preemptively higher-yielding) ventures.

If the policy option is (1) then in theory the valuation of this firm will be $500 (the risk-free rate of interest). The company will still generate $10/year for its shareholders whether the valuation is $100 or $500. An investor would be indifferent to sell the business for $500 and invest in the risk-free bond or just keep holding onto the business – you have magically created $400 of capital profit and you can clip dividends or bond coupons. With your $10/year you can do what you please, or put it in a risk-free 2% yourself.

This example is a constraint of reinvestment – after the re-valuation, your equity has effectively turned into a bond with no chance of compounding beyond the risk-free rate of interest.

If the policy option is (2) then you will see your returns in the appreciation of equity value. After the first year, your firm will generate $0.20 more in income and this will translate into $10 extra equity value, and this will compound at the rate of 2%.

Policy option (3) introduces the concept of risk – can management pull off the reinvestment? If there was an attractive investment at 5%, they would be able to generate $0.50 extra and this would translate into $50 of extra equity value for its holders, again, capitalized at the 2% risk-free rate.

So far we have made the assumption that the equity value follows lock-step with the risk-free rate of return. Of course in the real world, it never works that way and there are wildly divergent capitalization percentages used.

What is interesting is in this fictional example, the results you get if the initial equity investment does not trade at the risk-free return rate, but rather it trades at a higher rate, say 5%.

In this instance, the company would trade at an equity value of $200.

We will then consider a fourth policy option with the generated cash returns:
4) Buy back your own stock

This option requires a willing seller to the company (something that isn’t available to a 100% wholly owned firm!). Passing that assumption, an incremental deployment of $10 into the company’s own stock (a 5% reduction in shares) would result in continuing shareholders receiving 5.3% more returns in the future. Shareholders as an aggregate will still receive $10/share in returns, but the return per share will be 5.3% higher than before due to the reduced shares outstanding. This is a far better outcome than policy option (2).

The principle is the following: If a company is earning sustainable, long-lasting cash flows, it is to the benefit of shareholders that either the inherent business of the company has a capital outlay that offers higher returns on capital OR failing that, that the market value of the company’s equity is low to offer another conduit for reinvestment. Barring these two circumstances, returns should be given out as dividends.

This is unintuitive in that sometimes companies engage in really destructive practices with share buybacks. They are not universally good, especially if the future cash generation of the business is spotty. Likewise there are circumstances where buybacks work to massive benefit (a good historical example was Teledyne). However, in all of these cases, investors must possess a crystal ball and be able to forecast that the cash generation of the existing business (in addition to any other potential future capital expenditures) will be sufficiently positive over the required rate of return.

For example, Corvel (Nasdaq: CRVL) has a very extensive history of share buybacks:

The Company’s Board of Directors approved the commencement of a stock repurchase program in the fall of 1996. In May 2021, the Company’s Board of Directors approved a 1,000,000 share expansion to the Company’s existing stock repurchase program, increasing the total number of shares of the Company’s common stock approved for repurchase over the life of the program to 38,000,000 shares. Since the commencement of the stock repurchase program, the Company has spent $604 million on the repurchase of 36,937,900 shares of its common stock, equal to 68% of the outstanding common stock had there been no repurchases. The average price of these repurchases was $16.36 per share. These repurchases were funded primarily by the net earnings of the Company, along with proceeds from the exercise of common stock options. During the three and six months ended September 30, 2021, the Company repurchased 165,455 shares of its common stock for $25.6 million at an average price of $154.48 per share and 284,348 shares of its common stock for $39.8 million at an average price of $139.81, respectively. The Company had 17,763,576 shares of common stock outstanding as of September 30, 2021, net of the 36,937,900 shares in treasury. During the period subsequent to the quarter ended September 30, 2021, the Company repurchased 49,663 shares of its common stock for $8.7 million at an average price of $176.02 per share under the Company’s stock repurchase program.

We look at the financial history of the company over the past 15 years:

This is a textbook example that financial writers should be writing case studies about up there with Teledyne (NYSE: TDY) as this has generated immensely superior returns than if they had not engaged in such a buyback campaign. Share repurchases made over a decade ago are giving off gigantic benefits to present-day shareholders and will continue to do so each and every year as long as the business continues to make money.

The question today is whether this policy is still prudent. The business made $60 million in net income and there stands little reason to believe it will not continue, but should the company continue to buy back stock at what is functionally a present return of 2%? The business itself cannot be scaled that much higher (they primarily rely on internally developed research and development expenses and do not make acquisitions).

It only makes sense if management believes that net income will continue to grow from present levels. One has to make some business judgements at this point whether the company will continue to exhibit pricing power and maintain its competitive advantages (in this respect it looks very good).

Another example we are seeing in real-time is Berkshire (NYSE: BRK.a) using its considerable cash holdings to buy back its own stock. In the first 9 months of this year, they have repurchased just over 3% of the company. There’s more value right now in Berkshire buying its own massively cash-generating options than there would be on the external market – the last major purchase Berkshire made was a huge slab of Apple stock in 2017/2018 which was a wildly profitable trade.

In the Canadian oil and gas industry, right now we are seeing the major Canadian companies deal with the first world problem of excess cash generation. They are all in the process of de-leveraging their balance sheets and paying down (what is already low interest rate) debt, but they are also funneling massive amounts of money into share buybacks.

For example, Suncor (TSE: SU) and Canadian Natural (TSE: CNQ) are buying back stock from the open market at a rate of approximately 0.5% of their shares outstanding each month. Cenovus started their buyback program on November 9th and intends to retire 7% of shares outstanding over the next 12 months. The financial metrics of these companies are quite similar in that with oil at existing prices, an investment in their own stock yields a far greater return than what you can get through the uncertainty of opening up a major project (good luck getting through the environmental assessment!). My estimate at present is around 15% return on equity for these buybacks and needless to say, this will be great for shareholders.

It is why an investor should want low equity market values as long as these buybacks continue and the pricing power of the companies remain high. In the oil patch, this of course requires a commodity price that by all accounts should remain in a profitable range for companies that have had their cost structures streamlined and capital spending requirements that have been curtailed due to a hostile regulatory regime. The returns from these share buybacks are likely to be immense, barring a collapse in the oil price.

Corvel Corp – Annual Report

Corvel (Nasdaq: CRVL) announced their year-end results (10-K). The company has such fanfare that it has precisely zero analysts following it, so the automated news generators couldn’t even tell you whether they ‘beat’ or ‘missed’ expectations. There are few (USA) domestic companies that have billion dollar market capitalizations that have zero analyst coverage.

Their fiscal year ends on March 31st, so this was their Covid year. Revenues dropped about 7%. Margins, however, remained consistent. They made reference in their last quarterly conference call to streamlining real estate leasing costs as they were able to seamlessly ‘go online’. This can also be seen in the recent dramatic drop in IFRS 16 asset/liability for leases vs. the previous quarter. Year-to-year, lease obligations are down by half, and this is explained by the company projecting that renewals will not be exercised.

The company has nearly doubled in share price since last year. Valuation-wise, they are at about 50 times the previous twelve months of earnings. Needless to say, this appears to be rather rich, but the earnings should accelerate in the upcoming fiscal year as employment in the USA continues to expand, especially as Covid supports terminate.

Although it looks like that revenues and net income has flat-lined over the past three years, there will likely be an upward trajectory going forward. In addition, note the historical ROE numbers of 20%+. The ROE number next year will be down as the amount of cash on the balance sheet will serve to be a drag on the denominator, but this will likely be transitory.

I also believe the stock has been a positive recipient of automatic index buying. Trading of the stock is thin, with spreads typically a dollar or so. The company is currently on the S&P Smallcap 600 index, although it is well below the liquidity threshold for the index. Their market cap, over $2 billion, is creeping up to the point where they are getting into mid-cap territory.

Historically, the company has engaged in a capital allocation policy of repurchasing shares (they repurchased approximately 100,000 shares, or about half a percent of their shares outstanding in the past quarter at an average of US$104). Despite this, the company has been building up cash on the balance sheet, and the US$140 million they have at fiscal year end is an all-time high. I do not think management is of the type to suddenly declare a treasury policy of purchasing Bitcoins with spare US dollars. They also have little use for excess capital – they continue to engage in the usual R&D that software companies should be doing. In a paradoxical sense, this lack of capital reallocation is a negative in that the corporation cannot “snowball” retained earnings into more fruitful endeavours (unlike an acquisition machine like Constellation Software (TSX: CSU)). CRVL management is very happy to stay in their niche and only make the tiniest steps outwards from their strategic niche.

Operationally, they are in a dominant position, which explains the valuation. Indeed, when compared to companies like Constellation, they are trading at 85 times past earnings. On an absolute level they are expensive, but relatively speaking, it is in the ballpark. Another metric is price-to-sales, and also this makes Corvel cheaper than CSU. I would make the claim that CRVL’s niche has a much stronger competitive moat and justifies a premium valuation as a result. Their management is even more reclusive than CSU’s Mark Leonard. This is a trivial analysis, but a few of the factors swimming in my head when I look at this company in my portfolio and wonder if I should reallocate.

For many reasons, I will not be. I do not know what price the shares will rise to before I say enough is enough and seriously think about the sell button. I can easily see them staying where they are currently for a lengthy period of time, but I can also see reasons for them to head up to $200 and higher.

My only regret was not picking up a little bit more when I did, but at that time in the Covid crisis, there were many other things floating on my radar at the time. This is probably the least dramatic investment in my portfolio at present. They consume a disproportionately lower amount of attention in relation to their size in my portfolio, which is in the top 5.

Corvel Corp

Here is another quick post.

I took a substantial position in the past month in Corvel Corp (Nasdaq: CRVL). Now that it’s rocketed up, I can write about it. Corvel’s domain is primarily in the software processing of workers’ compensation, insurance and related claims processing. They are vertically integrated to the extent that they provide the software and management systems.

I have been tracking this company for nearly 20 years, and owned and sold it over a decade ago, but I’ve been keeping an eye on it ever since (which assists the due diligence effort since there is such a long institutional history in my own mind). I finally had my opportunity.

It is a very unusually managed company. The pioneering manager (Gordon Clemons) still owns 9.4% of the company and sells bits and pieces now and then. He came on board 1988. The pioneering shareholder, Jeffrey Michael, owns 37% of the company through a holding company, and this was from 1990. They have since passed the baton to the next round of management, which does provide some succession risk (what happens to those shares?). Both Clemons and Michael have kept very low profile, along with current management. Almost nobody has heard of this company despite them doing over half a billion in revenues a year. They inhabit a niche that has very high barriers to entry. Their capital allocation strategy has also been relatively unusual – the underlying business makes a lot of free cash flow (after R&D) and all of that capital (up to the end of 2019, approximately $514 million) has gone into the repurchase of shares. They do not employ debt – indeed, not only did they not use their credit facility but they got rid of it in September 2019.

Their capital allocation strategy is uncommon. They maintain a cash float, and reinvest the rest of the proceeds in their own stock. As long as you assume the company will continue making money at the rate they are doing, these buybacks are extremely value-added over time. This makes past stock charts less comparable to present day situations. As of December 31, 2019, they have bought back nearly 2/3rds of their shares outstanding, at an average of US$14.25/share.

CRVL’s sensitivity to Covid-19 was primarily related to raw employment figures – their revenues are ultimately derived from employment and insurance factors. Their fiscal year end is in March and they released the annual results, which were in-line. The April to June quarter will show a slowdown, but after that, things should normalize.

While revenues will take a minor hit during COVID-19, as employment normalizes (in whatever form), Corvel will achieve its track record in annuity-like income and will be priced at a higher multiple than it was during COVID-19, which needless to say was lower than its ambient norm.

The company maintains a significant competitive advantage. Their pricing power will not erode too much as a result of the economic turmoil.

Needless to say, this is one trade I wish I made larger. My investment style is far too cautious at times.