Dustin Shindo: The story behind what happened at Hoku and his new venture in healthcare IT
Thirteen years after he and his partner, Kaleo Taft, founded their fuel cell startup in a Waikiki apartment. Nine years after he put the company on NASDAQ in a $21 million IPO. Seven years after he and his colleagues gave up on fuel cells and broke ground on a $400-million polysilicon plant in Idaho. Four years after he stepped down as CEO and a Chinese conglomerate acquired controlling interest in Hoku. And a year after what was left of the company finally declared bankruptcy.
People still think of him as the wunderkind from Hoku, but Shindo’s now 41 and he hasn’t been sitting still in the four years since leaving the company. Today, he presides over a pair of healthcare startups from offices in a Moiliili strip mall. Don’t be fooled by the modest setting. He may be older, but he’s still the smart, ambitious entrepreneur who once sold over $2 billion worth of polysilicon before the manufacturing plant was built. And he still expects to change the world. The question is: Are the lessons Dustin Shindo learned at Hoku the right ones?
First, the present: “We have two businesses running out of here,” Shindo says. “One is Pono Corp. and the other is IDEAS Health. IDEAS stands for Information and Data Exchange and Analytic Solutions. Both of these companies are very healthcare-centric. They’re two legal entities, but we really run them as one business.”
Pono Corp. is a drug development company that Shindo founded with his old Hoku partner Kaleo Taft, though Taft has since moved to Dallas and become president of METIS Scientific, which makes DeConGel, another Hawaii-born product. “We have two drugs,” Shindo says. “The main one is a cancer drug for multiple myeloma, which we originally bought from the University of Hawaii and have continued to develop. Multiple myeloma is a blood cancer. It’s pretty rare, but it’s one of those you don’t want to get. There’s no cure. It’s a blood plasma cancer and, once you get it, it’s just a matter of time.”
Treatments for rare diseases like multiple myeloma have become profitable for drug companies. “The primary treatments for multiple myeloma are proteasome inhibitors, which is what we have. Right now, there are only two in the marketplace. The main one is called Velcade, which was developed by a company called Millennium Pharmaceuticals, in the U.S., which, in turn, was acquired in 2008 by Takeda Oncology, in Japan. Velcade now does about $1.5 billion a year. It’s an expensive drug to take, which, in part, is because there aren’t very many alternatives for this cancer.”
Despite the potentially lucrative future for Pono’s cancer treatment, most of Shindo’s efforts are focused on IDEAS. Its goal is to build a digital platform that can provide sophisticated analytics for healthcare. According to Shindo, the company has an interesting pedigree. “Several years ago,” he says, “HMSA started investing in an analytic platform. Using funding from something called a Beacon Grant, it started going through data from East Hawaii on the Big Island. Beacon was a very competitive program; this was one of only 17 communities in the country to get the Beacon Grant. I’m from Hilo, so I followed that project for the last couple of years. Last year, the Beacon funding ended, and HMSA decided it wasn’t going to be the developer of this system; it’s not really their business. Instead, HMSA wanted us to develop this analytic platform, so we ended up buying it, and HMSA signed up as our first customer. So, HMSA is our largest customer, although, since then, we’ve signed up a couple of others.”
One of the biggest trends in healthcare is the conversion from paper to electronic medical records. While crucial to a more efficient healthcare system, IDEAS is interested in something much broader. “Just to be very clear,” Shindo says, “this isn’t the same thing as the Hawaii Healthcare Information Exchange, which is a secure system to share individual patient health records. This isn’t at the individual level. This is more about asking questions like, ‘What does the community spend on diabetes? What treatments are most effective? Which treatments cost the most money?’ Maybe, at a more detailed level, it could be, ‘Which physicians are most successful at treating diabetes?’”
The analytics part of IDEAS is designed to answer those questions, but it won’t work without the data exchange part. The key is to create a system that takes in data from as many sources as possible – insurance companies, hospitals, physicians, etc. – cleans it up and packages it so those same entities can analyze it. Increasingly, healthcare experts see this kind of data analytics as crucial to identifying problems and controlling costs.
“Of course, you have to follow all the healthcare (privacy) rules,” Shindo says, “but you want to trade data from multiple places for the purposes of analytics. That’s definitely cutting edge as far as healthcare data goes. If you only have your own data to run analytics on, which is what we have today, it’s already beneficial; but running analytics on data from multiple sources would certainly take it to the next level.”
Mike Sayama, the executive director of IDEAS’ Learning Health Home division, is in charge of coordinating the healthcare community in East Hawaii. Before coming to IDEAS, Sayama was an executive VP at HMSA and was in charge of the Beacon project, so he understands the complexity of the process. He contrasts what IDEAS does with Kaiser’s EPIC system, perhaps the industry’s leader in analytics. But Kaiser is a self-contained healthcare system; all its hospitals, clinics, physicians and patients comprise a single entity, which makes it easy to share data. IDEAS, on the other hand, is building a community system. “The question is,” Sayama says, “Can a community model of delivering healthcare remain viable compared to the Kaiser model? For it to remain viable, my belief is we need the information platform that we’re trying to develop.”
This is the complex and high-stakes scenario that Shindo and his team will have to negotiate. That’s what makes the Hoku story – one of the most complicated and high-stakes business stories in the state’s history – a meaningful comparison.
The most fascinating part of the Hoku story is how many times the company changed course. When Shindo and Taft founded Hoku in 2001, it was a fuel cell company whose technology was based on a proprietary fuel cell membrane. With a big investment from Hawaiian Electric’s venture arm, the company grew quickly. Because Japan had a large fuel cell market at the time, early Hoku customers included Sanyo Electric and Nissan Motors, both of which ended up as investors in the company. A $3.5 million deal with Sanyo, in particular, set the stage for Hoku’s success. Subsequent deals with Nissan totaled more than $8 million.
Steve Iwamura, a business consultant in Japan who helped set up the meetings between Shindo and Japanese executives such as Sanyo R&D head Fusau Terada, thinks these Japanese activities highlight Shindo as a deal-maker. He lists the challenges: the language difference, the relative size and strength of these giant conglomerates compared to Hoku, and Hoku’s lack of money. Shindo not only closed these deals, he was able to secure cash up front and have most of the rest of the money parked in escrow at Bank of Hawaii.
“Even with that first deal with Sanyo,” Iwamura says, “that was like Hannibal coming over the Alps with elephants. I still find it hard to believe we did that.”
In the early 2000s, many renewable energy experts believed the “hydrogen economy” was only 10 years away. Dirty, inefficient internal combustion engines in cars, for example, would be replaced by fuel cells, which work by taking hydrogen and passing it through a sophisticated membrane – Hoku’s main innovation – producing electricity and pure water. But, even though fuel cells were clean, but they weren’t cost effective, especially compared to the new electric and hybrid car engines entering the market. Even after Hoku’s IPO in 2005, the hydrogen economy still seemed to be 10 years away.
“By 2006,” Shindo says, “a lot of the fuel cell companies were struggling and the market wasn’t coming to fruition as we had hoped, so we started looking for ways to diversify.”
One diversification was commercial solar development: designing and installing large solar farms. “We did a lot of great projects,” Shindo says. “The second largest solar farm on this island – still, today – was built by Hoku out at the Pearl City peninsula. The largest solar farm in the state was built by Hoku on Kauai. All the solar at the outer island airports was built by Hoku.” According to Shindo, solar helped Hoku turn a profit – a rarity for a tech startup. But solar was just a way to provide cash while Hoku moved to something more ambitious: producing polysilicon, a key ingredient in making solar cells.
Karl Stahlkopf, who served on the board of Hoku “from the time it was born until it died,” describes how the company got into polysilicon. “It was pretty clear that there was a cliff up ahead for the hydrogen car. … So, the team of scientists and some of us from the board got together and said, ‘What are we going to do now?’ ” The conversation turned to polysilicon, a hot commodity in the booming solar industry. At one point, the spot-market price for polysilicon was more than $300 per kilogram; but, by Hoku’s calculations, it would only cost about $25 per kilogram for a new plant to produce it. That’s an attractive margin. Maybe more important: industrial polysilicon production is essentially chemistry, which was Hoku’s scientific strength. Also, there was some customer overlap between the fuel cell industry and polysilicon. Sanyo, for example, made both fuel cells and solar panels.
“So,” Stahlkopf says, “the company made the strategic decision to shut down the work we were doing on hydrogen and put all our time, energy and money into developing a long-term plan to do two things. One was commercial solar development in Hawaii, basically selling solar energy in the commercial industrial space. The other was making polysilicon.”
In March 2007, with great fanfare (and criticism from people who wanted to keep jobs in Hawaii), Hoku began construction on a $400 million polysilicon plant in Idaho. Shindo approached funding the construction with characteristic ingenuity. The company still had millions of dollars in the bank from the IPO, plus steady revenues from its solar business in Hawaii. Shindo ultimately secured more than $2 billion in contracts – essentially commitments to buy Hoku’s future production – with solar companies in Japan, China and Germany. These deals generally included some cash up front, with the rest becoming available in increments as Hoku reached various milestones. For example, Sanyo, which, like other early investors, made a handsome profit on the Hoku IPO, signed a $370-million deal that included $2 million in cash and $108 million in escrow, again, at Bank of Hawaii. These efforts raised nearly $330 million in cash toward the construction and development costs for the polysilicon plant.
“All we needed to do was get $70 million in project funding out on the financial market,” says Steve Iwamura. “That’s not a lot of money in a $400- million project.”
Then Lehman Brothers collapsed.
In 2008, not yet realizing the severity of the financial crisis, Shindo began knocking on doors for financing. “We had a great deal to offer,” he says. “We had the plant structured. We had an engineering company to build it. We had customers. We had some money in the bank. We had contracts. This is what project finance companies lend on. And they all said no. At the time, we had Merrill Lynch as a bank that we were counting on. It was tough because we thought, ‘We’ve done all we can do.’ Then, of course, we found out why they were saying no: They didn’t have the money to lend.”
Like Merrill Lynch and many other companies, Hoku became a victim of the great recession and its deals unraveled. The big solar companies, under duress themselves, demanded that Hoku renegotiate their contracts (most of which were not guaranteed). Payments came in late. Construction delays sent the price of the plant spiraling, ultimately exceeding $700 million. Hoku fell further and further behind with its own creditors. Worst of all, the market for polysilicon tanked.
“There are a couple of reasons for this,” Shindo says. “One, it is a commodity, so it’s driven by supply and demand. And, when demand was high, supply just shot up. Existing manufacturers added capacity, and the Chinese built a ton of new capacity, so supply increased like crazy. In part, that was predicated on the belief that demand was going to continue to grow. So, the second half of the story was that demand didn’t continue to grow, because the global crisis not only cut off the debt spigot, it cut back or curtailed the government incentives for solar. As government incentives, particularly in Europe, dried up, demand for solar dried up. Thus, demand for polysilicon didn’t grow nearly as fast as everyone predicted. So, it was supply and demand that drove prices down, just as it was supply and demand that had driven them up. It was the flip-side of the same coin.”
But Shindo still had one last trick up his sleeve. In 2009, he arranged a deal to sell controlling interest in the company to the Chinese energy conglomerate Tianwei, which agreed to convert $50 million in polysilicon prepayments into stock, and to provide an additional $50 million in loans to help finish the construction of the plant. That effectively ended Shindo’s relationship with Hoku.
“In Japan, we have a saying,” Iwamura says. “ ‘When a bird lands on a lake, the lake is supposed to be calm. And when the bird leaves, the lake should be calm, too.’ When Dustin left Hoku, he left calm waters behind him. What happened after he left he had no control over.”
“I woke up the next day feeling good,” Shindo says. “The poly price was still in the $35 to $50 range, so everyone thought, ‘Well, it’s not $200 anymore, but it’s still good enough that we can start the plant and move forward.’ Here’s the interesting thing, though: After we sold it to the Chinese, the spot price for polysilicon fell to $15. This is why there’s really no one to blame for Hoku’s failure. The spot price fell below the price to make polysilicon for any new plant, which was roughly $25. In the end, a lot of those new Chinese plants just closed up. Today, the spot price sits around $20, so it’s probably going to bea long time before polysilicon is profitable again.”
Last December, the $700-million Idaho plant was auctioned off for little more than $8 million.
Karl Stahlkopf compares the end of Hoku’s saga to a Shakespearean tragedy. “One of the things that distinguishes Hoku from other startups that I’ve worked with,” he says, “is that the original path it started on was very different from where it ended up. And it was able to execute an absolutely brilliant U-turn, which very few people have successfully done. It went out with one market and technology, which the whole company was founded on, then went in a completely different direction. But, because of the intelligence, the diligence, the thoughtfulness and the hard work of Dustin and his staff, they were able to make that U-turn and make it successfully. But, again, the Shakespearean tragedy takes over and Juliet dies.”
So what does the Hoku tragedy have to tell us about the prospects for Shindo and IDEAS?
On the one hand, those who know him personally and have worked closely with him say it shows his intelligence and resourcefulness as an entrepreneur. They certainly don’t view him as a failure.
“Hoku means star in Hawaiian,” says Steve Iwamura. “Dustin had to look at his stars to navigate the company toward the horizon. I think he does that better than anybody I’ve seen in my lifetime. Some people think leadership is a function of age. I’ve never counted how much older I am than him, but I’ve got several years on him. But that guy is a CEO in every sense of the word.”
From another perspective, Hoku wasn’t a failure, but a success, because a lot of people made a lot of money on the IPO. “Hoku has two stories, as far as investors go,” Shindo says, “You have pre-IPO investors and post-IPO investors. My pre-IPO investors did very, very well. Some of them weren’t in very long, but as long as they sold their stock, they had very handsome returns for their support. And I value their support; we couldn’t have had Hoku without it. So, I had people thank me for the effort I put in to create the company. That meant a lot to me.
“The post-IPO investors – well, we were a public company. People could buy and sell whenever they wanted to. So, there were people who loved me and people who were frustrated. We got calls all the time: either somebody saying, ‘Thank you so much,’ because the stock went up, or, ‘What did you guys do?’ because the stock went down. I think that’s the case with most public companies. But the pre-IPO investors were pretty happy. Even today, when I see those investors, they remember – maybe, in part, because the other high-tech investments in Hawaii didn’t go as well.”
Rob Robinson says the Hawaii Angels investors were a good example. “Most of us were invested at about 75 cents a share. If you sold on the day the SEC lockup ended, as some of us did, you could sell for just under $15 a share. That was a pretty significant profit. That’s what you call a home run. Actually, the baseball metaphor is appropriate. In baseball, most of the time, when you go to bat, you strike out or ground out. Nothing good happens. But one or two times out of 10 at-bats, you get that double or triple or home run. And you need that to keep your average up. It’s the same with angel investing. Hawaii Angels has been around since nearly 2002. We’ve invested in about 50 companies, and we’ve never had another company produce the kind of return that Hoku has. We’ve had a couple that were pretty good, and a couple in the process that we think will be successful, but nothing as profitable as Hoku.”
Ironically, Hoku insiders didn’t do as well. “A lot of people made money,” Stahlkopf says. “Unfortunately, I wasn’t one of them. Being on the inside, with things jiggling around like they were, many of us were kept by SEC rules from selling. So, I think most of us lost a lot of the upside potential.”
No Big Payoff to Shindo
Even Shindo never really got to cash in on Hoku’s early success. Although his paper wealth approached $70 million when the stock was at its peak, insider-trading laws prevented him from selling more than a small percentage of his holdings at substantial profits.
But it wasn’t just about the financial returns, Robinson says. Like many of those familiar with Hawaii’s startup community, he views Shindo as a particularly competent entrepreneur. “What was unusual was that Hoku did go as expected. The usual experience with tech investment is it’s a pretty wild ride, with lots of twists and turns, and a lot of times it doesn’t go right. But Dustin had a pretty clear vision of where he wanted to go, and he more or less stuck to that timeline. He really executed well. That’s what’s unusual in the business world. Nobody really expects that to happen. Of course, they certainly had some dramatic changes in their business model.”
Nevertheless, the company did go under. And, given the dramatic swing in prices for polysilicon, it’s fair to ask: Was the whole Hoku saga based on a bubble? That’s certainly what some analysts have said, most of them with the benefit of hindsight. But, as early as 2007, Citron Research’s stocklemon blog pointed out the difficulty a small company like Hoku faces competing in a commodity market:
“The problem with commodity manufacturing is simple. What is now a high-margin business will soon turn into a low-margin one as undercapacity gives way to overcapacity. The company is not producing natural resources; rather they are turning raw materials into a manufacturing component. Hoku doesn’t seem to have acknowledged that access to capital and lowest-cost-of-production rules this game.”
Not surprisingly, Hoku insiders have a different view. “It wasn’t a commodity market when we started,” says Karl Stahlkopf. “It was a value-price market. And there were a lot of people who agreed with us – the early companies in China that invested in the long-term future of polysilicon. Maybe we could have seen the market swing coming, but I doubt it. I find that looking into the future is hard to do. Every sign we were getting from the smart people in the market was that there was a need for this. And we got multiple validations along the way as we signed several long-term contracts with international players.”
Even so, it’s hard not to wonder why Shindo and the Hoku team didn’t anticipate a ramp-up in supply as competitors raced to get a piece of those outrageous poly prices.
The bubble issue is relevant because there are signs that today’s healthcare IT space is also overheated. When asked to describe the potential size of the market for IDEAS, Shindo cites a recent IPO: “I’ll give you a sense of how hot or exciting this market is,” he says. “If you look up a company called CastLight Health – it just went public last month – its annual revenue was only $13 million, but its market cap at IPO was $1.5 billion, which never happens. Now, it’s climbed over $2 billion. It’s a healthcare data analytics company. We’re a smaller version of it. That gives you a sense of how big and meaningful this market could be.”
Of course, those figures look pretty alluring to a startup in the healthcare sector. But industry observers are (suddenly) skeptical. In May, Aaron Pressman of The Exchange described Castlight as the “Most overpriced IPO of the century.” He wrote: “Castlight’s insane level of valuation – 107 times revenue (not profits, as they had huge losses last year) – of the original IPO pricing hasn’t been seen for a tech deal since the year 2000, the twilight of the 20th century. Of the prior 13 deals priced at 100 times revenue or more and sales of at least $10 million, the average three-year return was negative 92 percent.”
That bodes ill for IDEAS, if the business model depends on these kinds of valuations. Probably more important, though, is what all this attention means for competition in the healthcare IT sector. Just as the big players were drawn to sky-high poly prices, giant healthcare tech companies are swarming to the same space IDEAS is trying to stake out. Inevitably, there will be a shake out.
In a Wall Street Journal interview, veteran healthcare investor Anne DeGheest explains how the bubble will burst for many small analytics companies: “A lot of these companies say they are going to pull data from a variety of sources, run algorithms on the data and then present the data to the customer. At the Series B level, we would ask, ‘Why is this data actionable? Who is meant to take action on the data? How do you measure the impact of this data?’ ”
These are questions that get to the core of what it will take for IDEAS to succeed.
There is one big difference between IDEAS and other startups, though. Because of how it was formed, Shindo has been able to grow the company without investors, and IDEAS has been break-even from the start. Having that steady, growing suite of customers is insulation from heady swings in the marketplace. It’s also a competitive advantage.
“There are certainly other companies that are doing healthcare data,” Shindo says. “The next question is: What makes us different? We have three things going for us, I guess. One is, via what Mike Sayama and HMSA have already done, we’ve been doing this much longer than those other companies. That means our understanding of the technology issues is better.
“The second thing we have going for us is people – not just within our team, but in the community. The relationships we have in East Hawaii, in large part due to Mike and his team, allow us to have that trust, that foundation that brings people together. Anytime you’re sharing data, you definitely have to have that trust.
“The third component is we’re developing custom algorithms here that help further our system. That’s a differentiating factor as well. It’s a competitive advantage because it takes data to come up with algorithms. If you’re starting from scratch, then you don’t have data, so you don’t have algorithms. And, over time, as these companies get bigger and bigger, it will become harder and harder to break into this market. It becomes the chicken and the egg thing. If you don’t have the algorithms, you can’t get data; if you don’t have data, you can’t get algorithms.”
Shindo also emphasizes the difference in culture between IDEAS and its potential competitors. “My philosophy, in every business I’ve created, has always been to grow it to be the best business it can be. As far as how do we grow it, I have a philosophy called ‘Silicon Valley meets family business.’ I grew up in a family business in Hilo. We had the Pepsi Cola franchise and distribution network for the Big Island and, as a kid, I worked six days a week. My dad used to make me stock shelves as soon as I could hold one six-pack in each hand. So I bring a lot of those principles into how I run a company. It was the same thing with Hoku. People always looked at Hoku as a company that raised a lot of money, but we actually raised very little money before the IPO. We were actually profitable before the IPO. I’m growing this business the same way, with a lot of the same principles any corner mom and pop store would have: You have to work hard. You have to make your customers happy. You have to control your costs. And you have to build a good team. All those things are consistent here.”
Broadly speaking, IDEAS enjoys the advantage of an owner who’s trying to build something that will last, not make a quick killing in the market – a concession, perhaps, to critics who said Hoku went public too quickly.
Investing in People
For better or worse, IDEAS also gets the benefit of a leader who’s run the gauntlet. Stahlkopf thinks that’s crucial. “The thing that you learn, if you’re around venture capital for very long, is they really invest in the management team as much as they do the technology. The world is littered with ideas that were better than their competition, but didn’t make it to market because their management team wasn’t as strong.” He cites the example of Sony’s Betamax technology, which, he says, lost out to the inferior VHS technology largely because of Sony’s poor management.
“That’s why you see venture capital folks, when they evaluate a company, they vet the management team as closely, or more closely than they do the technology. So many deals go before their faces, but most of these guys are financial guys, they’re not technical guys and, as an early investor, it’s really tough to tell if a technology is going to make it or not. So, you really have to look at the credentials, the work ethic and the personality of the management team so you end up with the VHS as opposed to the Betamax.
“The reason I bring this up is because Hoku was not a case of management making a mistake. Not the kind of mistake that let an inferior product take over the marketplace. It was a case of externalities – the downturn in several markets at the same time. I continue to be a big supporter of Dustin. I think he’s shown it with what he’s done since then. He’s clearly a bright, energetic, hard-working entrepreneur, who’s kind of dusted himself off from Hoku and has done some extremely interesting things since then. He’s still well thought of in the financial community. He’s still well thought of in the Hawaii technology community. So, the business failed; but, in my opinion, management didn’t fail.”
“Mostly a Success”
Shindo himself doesn’t see Hoku as a failure either. “It was mostly a success,” he says. “I say that for many reasons. First of all, we started that business in an apartment in Waikiki. How many businesses start that way and get to being a NASDAQ-listed company? We were the only profitable fuel cell company ever publicly traded, that I know of. And, when the fuel cell market didn’t happen, we reinvented ourselves as a solar company. We were constantly overcoming challenges. When we had trouble with financing – in large part, because of the decline in the polysilicon market and the global financial crisis – we found a buyer, an investor, in China. So, insofar as my involvement goes, it was a big success. It’s certainly something I’m very, very proud of. The reason I say that so adamantly is: that was my baby. I built that for 10 years. And when the company was closed, even though I was four years removed from the company, it was still sad. It’s something that, obviously, I would never have wished to happen.
“Another way it was a really big success is, I think, it changed the beliefs, if not the perceptions, of entrepreneurs in Hawaii of what could be done here. A lot of times, it’s really easy for people here to say, ‘It’s Hawaii; we can’t do certain things.’ And I don’t think they should. I think, with the right people and the right situation, we can do almost anything here. Certainly, for at least a couple of years after the IPO, there were things that people thought they could do in Hawaii that, previously, they would have moved to the mainland to do, or just not done at all. In that way, I think the ripple effect of Hoku was much bigger than people realized.”