Marpai Inc.

Apr 23, 2024

One-liner

Marpai is a nanocap in a fragmented healthcare administration industry with an overqualified CEO who owns almost 40% and just brought in his team to make things right.

Third-party administrator (TPA)

To get your taxes done you hire an accounting firm, to get your legal work done you hire a law firm and to get your healthcare paperwork done you hire a TPA. 

Setup

Continental Benefits was a TPA firm founded in 2013. Marpai was a development stage AI startup from Israel founded in 2019. Marpai was trying to bring technology and AI into the inefficient low-tech TPA market. Continental Benefits started working with Marpai as a partner because Marpai had real-world data for AI training and testing. The two decided to merge and the duo IPOed as one in 2021 raising $25 million. After the IPO Damien Lamendola, founder of Continental Benefits, owned 34% of the shares and AI startup guys owned 18%. Management of the new company was in the hands of AI guys, Damien Lamendola was a silent partner. 

Fast forward two years and the management burned a lot of money by giving generous compensation to themselves and having made an expensive acquisition. I suspect they focused too much on the AI part. Newly formed company was supposed to be a TPA business with AI improvements and not vice versa. Management raised money at the beginning of 2023 at $1/share (IPOed at $4), did a 4-to-1 reverse split in June and wanted to do another big raise at the end of 2023. I think at this point Damien Lamendola had enough of this. He fired all the management in November, stepped in as a CEO himself, brought in a CFO, a president for operations and two new board members. 

You see, Damien is not quite a regular microcap CEO. Damien Lamendola founded WellDyneRx - pharma benefits company - in 1999 and sold it to Carlyle Group for an estimated $1 billion in 2017. He owns two private Gulfstream 450s, each of which probably cost more than the entire market cap of Marpai at the time he stepped in as a CEO. I think Damien did believe (and still does) in the concept of bringing efficiency to the TPA market with the use of tech and AI but wanted someone else to execute. This is pure speculation on my part but I think that after observing incompetent behavior within Marpai for two straight years he went like “ah, crap, I’ll have to do this myself”.  

I will talk about the market, business, competition and valuation but in reality this is simply a bet on management. 

Health insurance

The sophistication and depth of this particular topic cannot be understated so the reader should remember that the following discussion is an oversimplified view of a complex subject through the lens of a generalist who simply tries to identify a potential investment. Keeping that in mind, let’s lay out a few facts which in my opinion are important and relevant to the investment thesis.

As an employer you have two options to provide health insurance to employees: full insurance or self-insurance. Fully insured employers pay a certain premium per month per employee for a predetermined health plan to insurance carriers and don’t worry about this part of their business any further. Full insurance is often associated with paying to BUCA - the big four of medical insurance: Blue Cross Blue Shield, United Healthcare, Cigna and Aetna. Self-insured employers construct the health plans themselves to cater to unique needs of their employees and save money in doing so. They “only pay for what they need'' (liberty, liberty, liiiberty). Marpai operates in the self-insured market.  

Primary reason to go with the self-insurance route is to save money. However, it does involve some additional headaches. First of all, an employer assumes financial risk with self-funded plans. So, step one would be to get stop-loss insurance. Second, you need to take care of all the processes which include things like billing, plan design, and claim processing. Of course, no business would want to do this. Employers hire a knowledgeable entity to handle the matter - third-party administrator or TPA. The TPAs are called government-regulated organizations that act as intermediaries between providers and policyholders. They typically charge per employee per month. The fact that the self-insurance market exists tells us that even after paying TPA fees and stop-loss insurance premiums employers spend less with self-insured plans than with fully insured ones.

U.S. health care spending reached $4.5 trillion in 2022 or $13,493 per person. As a share of the nation's GDP, health spending accounted for 17.3% in 2022 compared to 5% in 1960. Spending by self-insured health plans is one-third of total healthcare in the US or $1.5 trillion. Marpai claims that the total addressable market for TPA fees from self-insured employers is $22 billion.

There are thousands of TPAs in the country and they mostly do the same thing - design health plans, handle claim benefits, answer customer’s calls, things like that. This is not a great business with low-ish margins and very little differentiation. But there are two positives: the self-insured market is growing fast and the industry as a whole is not so tech-savvy. As a result, getting more clients while optimizing the internal processes provides an opportunity to improve margins drastically. The formula is simple, it is the execution that matters. 

What’s with AI? As we can see from the numbers above, healthcare costs have grown significantly in the last decades. It is predicted that the price will increase ahead owing to chronic diseases getting more common. AI could supposedly predict and prevent costly claims related to chronic diseases and mitigate developing conditions. 

Competition

There are giant TPA companies like Sedgwick, UMR, Crawford & Co, Gallagher Bassett Services and the like. Typically these large companies offer administration services beyond just the medical field. They do property and casualty insurances, auto, marine, etc. Crawford & Co is public and trades at 6.5x EBITDA. It has four business lines and one of them is comparable to Marpai. It is called Broadspire, does $355.7M in revenue at 26% gross margin and 13% operating margin. The remaining three business lines of Crawford have less attractive metrics, so I would expect the Broadspire business to have a higher EBITDA multiple on its own, say 8-10. UMR claims to be the nation's largest TPA, serving more than 3,800 benefits plans and 6 million members. A lot of top 10 TPAs in the US grew by acquisitions. There has been a fair share of consolidation within this market in the last 15-20 years. 

It is fun to talk about the big guys. However, most of the TPAs in the country are small, regional and non-tech. There are a few tech-first national companies in the market like Centivo, Collective Health and Bind. These companies are valued at high revenue multiples as they are considered tech-ai firms. There are also smaller startups like Health Cost IQ. 

I think it is very difficult to break into the industry. If a big company already has a history of doing business with a certain TPA for multiple years the newcomer would have to promise significant savings and improvements in order to replace the incumbent. This is why it is vital to have connections in the industry and some sort of track record before any big client even starts talking to you.

Because the industry is so huge and fragmented I don’t think you compete with other tech-first companies. You compete with the incumbents and status quo.  

Actions

The most significant event which happened under previous management in a short post-IPO history of Marpai is Maestro Health acquisition. Maestro Health was a money losing TPA doing roughly $20 million in revenue with under 60 clients who employed 20,000 employees. Marpai bought Maestro Health in November 2022 for $22 million. The next action was to optimize Maestro Health legacy operations by implementing tech and AI so the business turns profitable. Then, after having developed a system for turning a legacy TPA into a better modern version of itself, the idea was to keep acquiring more and more TPAs and technologize (hope this is a word) them. The idea was correct but the execution suffered. The number of full-time employees increased from 131 in 2021 to 303 by the end of 2022 and the expenses exploded. 

Let’s have a look at the actions taken by the new management since November 2023:

1. Damien put $3 million of his money into the company through open-market purchases and two private placements.
2. Marpai secured $1.7 million in capital from Libertas Funding in February. 
3. Marpai got $12m from JGB Management through convertible notes (deal was just announced this week). Notes are convertible at $3 which is too low in my opinion because 40% dilution is implied. Cost of this loan is high - prime rate plus 5.75% which comes to almost 15% today. I am surprised by this issue. Marpai will use proceeds to pay off Libertas loan, which they just got 2 months ago, fuel growth and maintain working capital. 
4. 20% workforce reduction which is expected to save $3 million in annual costs
5. Sale of a non-core asset for $1.7 million.
6. Renegotiation of the debt with AXA (Maestro Health acquisition) which resulted in a $3 million discount. Discount relies on Damien putting $3 million in equity and Marpai maintaining NASDAQ listing.
7. A south-eastern client signed which is expected to add 20,000 households by the end of the year. Marpai paid $22 million for 20,000 employees in 2022, and they got this client “for free”.
8. Compensation setup. CEO has a $1/year salary and 600,000 RSU options, CFO has a $36k/year salary and 350,000 RSU options. Options vest over 3 years. Both get 100,000 options immediately when the company crosses $5 million in EBITDA.

“The industry has just gotten a small taste of what our relationship team is capable of”, CEO said on the latest conference call. Let’s see what they’ll do next. I expect the management to fix the issues this year, turn profitable and start making acquisitions in 2025. 

Numbers

In Q4 2023 Marpai did $8.7 million in revenue at 35% gross margin. Operating loss was $2.2 million if we exclude $3 million goodwill impairment. Cash operating expenses were $5.5 million plus $425k in interest. At the end of 2023 the company had $1.1 million in cash and we know they raised $2.7 million in equity and sold $11 million in notes since then. In TPA business the most telling quarter is Q1 because all the new contracts and re-negotiated old contracts are reflected in financials starting January 1st. Marpai also had an off-cycle client that signed up recently. I expect with additional new client revenue and expenses under control the company will be cash flow positive within a few quarters.

SP - $2.15
DSO - 10.3m
MC - $22.2m

Cash - $14m
Debt - $28.5m
EV - $36.7m

Marpai is still losing money so we can only go off revenue multiple. Today, Marpai’s EV is valued at roughly 1x revenue.

I expect the company to exit 2024 with $55 million revenue and 40% gross margin which pretty much covers all the cash expenses. How much is this business worth? Current enterprise value is around $37 million which I think could be considered a fair or even a high number if you just look at the current financials and consider poor performance since IPO. But as mentioned earlier, the thesis relies on management quality which cannot be seen in numbers yet. In my view, we will see the results of management’s execution soon enough.

Conversation with management

I’ve talked to Steve Johnson, the CFO. He is a very to the point guy, doesn’t beat around the bush, not very talkative. I can’t say there was much additional insight from the conversation. Below are a few points which were useful:

1. Seasonality. Major change on January 1st and smaller change from July to September.
2. All of the employment in Israel was terminated. Company relocated the AI efforts to the US. 
3. CFO believes TPA can be a good margin business if properly scaled.
4. Brokers are gatekeepers in the industry. To land a big client you have to go through brokers. Instead, Marpai has been talking to CFOs of larger companies directly. Sales cycle can be reduced from 6 months to 6 weeks with this approach. 

Overall, I walked away with no surprises. I think these guys want the actions to do the talking.

Risks

The most obvious risk is failure to execute. I’ll resort to the CEO's words here - “we've done it before with WellDyneRx, and we're looking forward to doing it again with Marpai. But this time, better and faster”.

There is a customer concentration risk. One customer accounted for 11% of revenue in 2023 and two customers accounted for 16.6% and 14.0% of accounts receivable respectively. Any hiccup in revenue will negatively affect the company given the debt it has accumulated. 

Marpai has been receiving warnings from NASDAQ about non-compliance. Management requested a hearing which was held on February 22, 2024, where the company presented its compliance plan. Subject to the Company meeting certain requirements by March 31, 2024, the hearings panel granted the Company an extension until May 28, 2024, to regain compliance. Here are the conditions to be compliant: (1) Stockholders’ equity of at least $2.5 million; (2) Market Value of Listed Securities of at least $35 million; or (3) Net income from continuing operations of $500,000 in the most recently completed fiscal year or in two of the three most recently completed fiscal years. Frankly, I am puzzled by this recent extension that was granted because the conditions for meeting the requirements are seemingly far away. If the company gets delisted there will definitely be a certain amount of indiscriminate selling which will push the share price down a lot.  

Conclusion

This is a bet on management. The price is not in a “no-brainer” category but it’s not too crazy either.