The business

Volpara Health Technologies (ASX: VHT) is a New Zealand-based, ASX-listed, health technology business which has a pure but important goal: to save lives via the early detection of breast cancer. 

As of June 2020, Volpara is a ~$350 million company — making it one of the largest Rockets on our list. However, its software is only now being found in more and more places throughout the United States, its largest addressable market, ahead of a potentially important catalyst for the business in October.

On the surface, Volpara’s technology and product set can appear very complicated, and to some extent it is. Indeed, since the business was founded by CEO Ralph Highnam in 2009, Volpara has been granted 101 national patents, lodged 9 international applications and there are a further four in progress. 

Fortunately, as the video with Ralph explains, the basic but important explanation of what Volpara does is “…help save families from breast cancer.”

It’s a simple, powerful mission.

Thanks to Ralph’s expertise, experience and calculated research & development (R&D), as well as some timely acquisitions, Volpara now offers radiologists, doctors and patients a more efficient, accurate and scalable suite of software designed to make life easier — and save lives. 

(As an aside, we said Volpara was ‘founded’ in 2009 but the origin story began well before that. Volpara CEO Ralph Highnam studied under founding Volpara director and professor Mike Brady at Oxford University in the late 90s. After that they started a company together which was subsequently taken over).

A full-stack medical technology company with AI on top

Taking a top-down approach, Volpara’s software is installed at clinics and hospitals alongside hardware from other vendors. A machine takes a patient’s breast images, known as a mammogram, then sends them into Volpara where the images are:

• Analysed on-the-spot to ensure they are of good quality
• Stored for later analysis in the cloud and on premises
• Sent across the hallways to radiologists and/or doctors for inspection, while…
• Also being analysed by Volpara’s Density, Risk and Transpara software

One of the most exciting parts of Volpara’s pipeline is its ability to use machine learning (ML) and artificial intelligence (AI) to perform, in simple terms, effectively pattern recognition to identify breast cancer before it forms.

Around 40% of women have dense breasts. Those with very dense breasts (e.g. Category D) are 4x more likely to develop breast cancer than women without dense breasts.

For years, Volpara has focused especially on women with ‘dense’ breasts. These are typically the hardest ‘types’ of breasts for clinicians to deal with because, in simplistic terms, a lot of what a radiologist does when a mammogram is received is ‘look for white spots’ (for those members who are radiologists, please note in advance that we are sorry for this over-simplification!). Take a look at the following mammogram:

The breast on the left is a typical fatty breast (low density), so you can see the star (e.g. breast cancer) quite easily. Contrast that with the dense breast on the right. It’s very easy to miss the star — and these are the women who are most at risk. Frankly, I find it alarming to think how easy it can be to miss something so important.

With Volpara’s flagship density product, a mathematical (and objective) equation is run before a radiologist inspects the images. Using raw data, years of learning and experience, Volpara’s product will:

• Rate the exam using machine learning and artificial intelligence — this tells the technician (i.e. the person who physically conducts the exam with the woman) whether the exam was up to standard;
• Rate the quality of the images — it costs around $400 for a woman to receive a mammogram in the USA, so coming back to re-do the imaging can be cost-prohibitive;
• Provide a ‘score’ for density; and
• Offer a very important risk rating, by using the density scores and integrating with industry-standard modelling techniques under the Tyrer-Cuzick 8 Lifetime Risk Model. You can hear Professor Cuzick explain why it’s so important to have density ratings in this presentation on YouTube.

In short, Volpara’s software is all about making sure women receive the best information from their time spent undertaking uncomfortable screening and imaging. That means, saving lives.

After a machine takes an image, Volpara’s software can then be used by doctors and radiologists to perform additional tests for the women.

The image above shows that a woman with and density rating of ‘A’ is rather low risk and should be ok with a mammogram, probably once per year. A woman with a dense breast might need more powerful or specialised scans for her treaters to know for sure where she stands.

Before you buy Volpara shares, I highly recommend watching our interview with Ralph (CEO & co-founder), paying particular attention to the section where he explains why and how the various pieces of the Volpara platform work together to provide an end-to-end service. The important point from a business perspective, touched on below, is that these adjunct pieces of software are only now being sold en masse to clinics and hospitals around the world. In our opinion, this presents a huge cross-sell and up-sell opportunity for the company, as measured by the average revenue per user (APRU). 

Broadly, the major products offered by Volpara include: 

1. Volpara Density – measures breast density from digital mammography and tomosynthesis data
2. Volpara Live – assesses images for quality in real-time; saving time, client inconvenience and stress, and money
3. Volpara Enterprise – the suite of breast imaging tools designed to provide the insights for breast imaging workflows
4. Transpara – sold under licence from ScreenPoint Medical BV

MRS acquisition

In addition to the above, thanks to the savvy $20 million acquisition of MRS Systems in 2019, which expanded Volpara’s reach and product set, the combined business now offers the MRS range of Aspen Breast and Aspen Lung products, which are advanced workflow software for managing patients (reimbursement, reporting, results and letters for doctors, compliance, etc.). 

This acquisition opened the door for Volpara in more ways than one. It means Volpara has one of its products in more than 27% of US imaging sites. The Aspen products also align well from a technology and growth perspective. Ralph has been touting the potential for Volpara to move into the lung cancer screening market, for example, given Volpara’s extensive experience in advanced pattern recognition and its software’s ability to work with images taken from equipment made by the likes of GE, Siemens, FujiFilm, etc. Volpara works with them all.

What we like

Chances are, you know someone who has been affected by breast cancer. I do. 

Tragically:

• Breast cancer is the number-one cause of cancer death in women between 20 and 59 worldwide, and the second-most common cancer in women overall; 
• 25% – 40% of cancers are missed by the radiologist;
• 2% – 3% of women are forced to come back for another (uncomfortable) screening because of a bad image; and
• It costs ~$400 to get a mammogram in the USA, with around $150 refunded by an insurer (if they cover it).

This is Volpara’s opportunity. 

Aside from simply feeling good in supporting an innovative business that’s doing a great thing, Volpara’s scale and economics are just as impressive — and I think they’re about to get even better. 

Right now, Volpara is loss-making, reporting a net loss of $14.9 million (all figures in NZD) during its FY20 year. But let’s keep that aside for a moment. 

In FY20 the company reported revenue of $12.6 million, up 153% from a year earlier. However, subscription sales, which represent the important ongoing majority (around three-quarters) of revenue is growing too. Year over year subscription revenue grew 106% to $9.1 million. The ‘subscription’ revenue includes ongoing maintenance agreements. The other revenue Volpara recognises is capital revenue earned from implementations ($3.5 million in FY20).

Note: Ralph explains the sales cycle around the 15:00 minute mark of the video. 

A strong foothold in 27% of US sites

One of the company’s key reporting metrics is average revenue per user (APRU), which is the average revenue of each woman screened, per year. In FY20, the APRU was US$1.04, up from US$0.94 last year. The most important bow you need to draw from the historical APRU metric is this: the $1.04 includes very little ongoing revenue from Volpara’s native products, given the majority of sites reported in the annual numbers represents the lower margin MRS products. As Volpara puts it: 

“The Group expects continued growth in subscription revenue in FY21 as the business pivots MRS away from a capital sales model to an almost solely subscription-based model, and the full end-to-end platform is available for sales.”

Though it is a rough approximation, if we take the $12.6 million of reported revenue (which itself isn’t ideal) and divide that by the APRU of $1.61 (in New Zealand dollars), the approximate Volpara user or ‘women’ count is between 7.8 million and 10 million. The US market opportunity is 39 million women per year. See our note about catalysts further down the page.

Many long-running Rask members will know that we believe revenue growth for the sake of it means little without the ability to retain customers and start pouring out valuable free cash flow. Given its deep integration, Volpara’s ‘churn’ or lost customers is negligible. This is/was evident during COVID-19. While many imaging sites paused some screening, due to safety concerns, they promptly resumed because… hey, breast cancer wasn’t waiting around to see what COVID was doing.

Dropping down a line item on the income statement, you can begin to see the potential of Volpara’s model. Volpara ran at a gross margin of 86% last year, up from 83% a year earlier. Many investors will know we get excited by companies with wide gross margins because, in simple terms, once the business covers fixed expenses, more and more of the marginal revenue will drip straight to the bottom line. 

Widening the moat

With a small but agile sales team in the US, a developed end-to-end suite of products and integration of MRS, Volpara is ideally positioned to grow its user numbers but also expand its average revenue per user as:

1. More MRS clinics adopt a recurring revenue (SaaS) model
2. More MRS clinics are sold the high margin Volpara imaging technology
3. More sites are added across the enterprise

A deep moat or competitive advantage is only truly valuable if you can apply pricing pressure to customers. Since November 2019, Volpara has provided only annual SaaS contracts to all customers and “most new deals are significantly above US$1.04 ARPU comprising multiple products – in Q4, ARPU on new deals was US$1.45 – US$3.10.”

When I spoke with Ralph he said the newest deals are priced between $1.50 and $3.50.

If we reverse our basic math from above, here’s what the midpoint ($2.50) of the latest APRU would translate to (hypothetically):

• 9 million women
• US$2.50 
• USD – NZD exchange rate ($1.56)
• 9m x 2.50 x $1.56 = $35 million revenue

This figure is hypothetical but plausible if we consider that most of the MRS sites are underpriced, relative to their potential, and the business is growing sites (I used a rough estimate). The wrinkle in this analysis, of course, is that Volpara typically inks 5-year deals and prices contracts in advance, so many of the older sites will take a few years to roll off their old pricing. 

On the cost side of ledger, in FY21 we’ll likely begin to see the benefits of the MRS acquisition costs fall off the financials. 

So, with more and more products being added to the line-up, and ongoing R&D, high levels of stickiness, extremely wide margins, relatively fixed expenses, and ARPU latency, there’s a lot of upside potential. Layer on top of that any prudent acquisitions and international growth via direct channels and distributors and you really only need one feather in your cap… a catalyst.

Catalyst ahoy

According to DenseBreast-info, co-founded by Professor Wendie Berg, in March 2019, the USA’s FDA announced it would see to modernise and enhance what radiologists and clinics tell women when they get a mammogram. As you already know, breast density is one of the most important factors that go into calculating the risk of developing breast cancer (and it’s heightened by the challenges with screening). But here’s the thing(s):

1. Many US insurers don’t cover the cost of mammograms unless a doctor requests it;
2. Many US states don’t incorporate breast density reporting into their patient’s result; and
3. Amongst the states which require breast density reporting, many will report things such as “you may have dense breasts”.

That’s an unhealthy dose of bureaucracy if we ever saw it. Fortunately, there are many doctors, organisations, the FDA — and women — who want that changed. The following map shows you what you need to know about current states, insurance and reporting.

Source: DenseBreast-info.org. Click here for the interactive chart.

At Rask, we’re not short-term catalyst cowboys running a LIC or trading strategy. That said, when you’re offered the cherry on top you may as well take it…

According to reporting and updates by Volpara in June 2020, the USA’s national standards for mammograms, including the requirement for breast density reporting, are expected to become known across the US around September/October of this year. Together with any scope for increased insurance coverage, this paves the way for more reporting, better imaging and information for women and… more images and customers for Volpara.

For us, it’s a case of:

• Heads, we win (from long-term organic growth);
• Tails, we win more (from mandatory imaging and reporting).

Financials & valuation 

A share of Volpara doesn’t come cheap at 25x sales. However, consider that revenue has risen from $1.8 million in 2017 to an annualised recurring revenue figure of $18 million in early FY20 — a 10x increase. We’re banking on meaningful top-line growth over the next three years for all of the reasons outlined above. 

Given the company’s track record, ability to scale, cross-selling, negligible churn and repricing of many MRS contracts, we don’t think it’s outrageous to expect Volpara to report annualised revenue close to $50 million in the next three years.

At a gross margin of 85%, we’re talking about $42.5 million making it past costs. If we assume a 50% increase in costs, or $10 million on top of the $23 million from last year, we’re left with a pre-tax profit of around $7 million to $10 million. That might not seem cheap but please keep in mind that valuing such a high growth business is fraught with error, so these numbers are quite rough and I’ve been conservative. Indeed, three years is a long time for a business like Volpara, so I wouldn’t be surprised if I’m being very conservative in my assumptions. 

Looking out to five years or more, management’s target is an APRU of US$10 per woman, which is realistic if Volpara’s ML/AI product delivers on its efficacy targets. Although management has a good track record, let’s be conservative and use US$5 for our APRU ‘target’ at the five-year mark.

Given our formula above, being Number of Women Screened x ARPU x Exchange Rate, we can use an (admittedly rough) guide for revenue of around $70 million in five years (9 million x 5 x 1.56). That assumes a lower APRU than expected by management, no growth in the number of women screened (unlikely) and the NZD exchange rate is mostly what it is today. And the best bit? Lots of that recurring revenue will fall straight down the income statement. 

Management

Co-founder and CEO Ralph Highnam has been with the company since… well… founding it. He owns around 16.2 million shares and 600,000 options in the company, or about 7%. It’s fair to say Volpara has a technical founder who has enough skin in the game to make him aligned. We would be surprised but concerned if he left the company and we would likely revisit our holding. 

Other notable holdings from the board include Roger Allen, who is a venture capitalist here in Australia and joined the board in 2010. He owns around 8% of the company, given his 18.4 million shares. 

All in all, the founders, management team and board owns a healthy 23% of the company. 

While it can be challenging overseeing 20 sales staff from the other side of the world, Volpara’s management team has a demonstrable track record, talent and alignment, which will help us sleep well at night.

Risks

Expect volatility with all of our rockets. These companies can rise or fall 10%, 20% or even 30% for little or no information, given the early-stage nature of their businesses. 

For Volpara specifically, the current valuation means we are placing a lot of stock in the expectations of the business. COVID-19 has shown us that even important things like medical breast imaging can take a back seat from time-to-time. 

Other risks to the demand side include a fractured or difficult private health insurance market — typically, a mammogram can cost $400 in the US, with ~$150 per year reimbursed by health insurance providers, depending on the cover, doctors and state laws.

Another risk is that the sales pipeline dries up quicker than expected during COVID-19 as the business fails to execute on its organic growth expectations. As noted in the interview with Ralph, the company has lost its number-one leads channel due to the pandemic, which is events and conferences. All in all, we see this as a speed bump or temporary roadblock rather than a permanent one. 

Finally, if CEO Ralph Highnam were to sell out completely or leave we would be concerned and consider pulling up stumps, especially if it was abrupt and a good succession plan wasn’t articulated very well.

Summary 

As is good practice in Australia, we’re happy to claim this Kiwi tech success story as one of our own. With a very large and important addressable market, a significant foothold via MRS, rising revenue, an ML/AI moonshot and better products in its stack, we think now is the time for us to act on the opportunity present with Volpara. 

The short-term is likely to be volatile, so we’ll be looking for any moments of weakness to back up the truck on this future wide-moat business.