A crowd can be wrong in perfect harmony. Leigh Drogen built a company around the possibility that it might do better if its members were allowed to disagree before they compared notes. Estimize invited people outside Wall Street’s usual circle of analysts to forecast corporate earnings. Then came a useful embarrassment: showing those people what everyone else thought could make the collective answer worse.
That discovery gets close to the recurring question in Drogen’s working life. Who gets to supply the number everyone treats as reality? A bank analyst? A hedge fund? A student? And what happens when the people supplying it would rather belong to the conversation than contradict it?
Today, he is general partner and chief investment officer at Starkiller Capital, working in digital assets. He founded Estimize, ran it for a decade, and moved into an advisory role when it combined with ExtractAlpha in 2021. The markets have changed. His interest in the distance between what people believe and what their incentives encourage them to say has held up rather well.
A wagon, some shells, and a ferry
Before there were earnings estimates, there were painted shells. As an eight-year-old spending summers on Fire Island, Drogen and his sister sold them from a wagon. When the inventory ran out, they helped ferry passengers carry their luggage. Their parents weren’t financing the candy-and-pizza budget. The children found customers instead.
His teenage employment history had less managerial approval. At sixteen, he lost a boating-counselor job after about two weeks. In his own telling, the newspaper and a pastry had occupied him on the dock while the children sailed. He then taught private tennis lessons for $50 an hour, enjoying the direct connection between his effort and his earnings. A performance review, administered by the customer, apparently suited him better.
He grew up in Chappaqua, attended Horace Greeley High School, and later studied at Hunter College. Hunter identifies him as a member of its class of 2009. He credits his classes there with developing critical thinking, a less photogenic entrepreneurial asset than a shell wagon but considerably easier to carry into adulthood.
Hockey supplied another education. He has described it as a team system in which success depends on working together. The connection to company-building was explicit: leadership, getting teammates to believe in you, and learning how to win. He also surfs. An earnings call is unlikely to improve the waves, which may be part of the attraction.
The number behind the number
Drogen’s first finance apprenticeship was at Geller Capital in White Plains. David Geller taught him to look for inefficiencies, including the relationship between earnings expectations and price momentum. He later ran Surfview Capital, his own investment firm, and joined the young Stocktwits in 2009. Product development and business development gave him practical experience in building a financial community.
The problem he wanted to tackle was familiar to anyone watching quarterly results: a company can report growth and still disappoint investors. Markets respond to expectations as well as outcomes. If the benchmark forecast represents only a narrow slice of participants, beating that benchmark may tell an incomplete story.
In a 2015 explanation of Estimize, Drogen argued that traders had shared information long before the internet. His proposed change was to widen the group supplying it. Estimize asked for specific earnings-per-share and revenue estimates from investment professionals, independent analysts, students, industry specialists, and individual investors. A forecast could be evaluated against an eventual result. Opinions acquired a due date.
That distinction matters. An online conversation can reward confidence, speed, or a memorable insult. A recorded estimate creates a different opportunity: compare the prediction with what the company actually reports. Someone with an impressive title still has to get the number right. Someone without one can build a record worth examining.
The idea took organizational form in 2011. By 2014, Drogen appeared on Forbes’s finance 30 Under 30 list. Recognition helped make the experiment visible, but visibility and usefulness are separate tests. The more interesting achievement was getting people to keep contributing information that other people could use.
Independence needed a design decision
The weakness in a public forecasting community is easy to overlook. Give a contributor access to earlier estimates and the contributor may improve a personal guess by borrowing from the group. Repeat the process often enough and the group loses independent judgments. A roomful of agreement can contain surprisingly little additional information.
Researchers Zhi Da and Xing Huang examined this problem with Estimize, randomizing the information users could see. Their study covered 2,516 users, 2,147 earnings releases, and 730 companies. When contributors couldn’t see other estimates before making their own, the resulting consensus beat Wall Street’s consensus 64 percent of the time. With other estimates visible, the figure was 57 percent.
Estimize switched to a blind platform in 2015. Drogen had helped facilitate the research, and the result became a product decision. That is a revealing kind of founder story: the useful move was to reduce the information available at a particular moment. A community designed to collect more voices needed to keep those voices apart long enough to hear something new.
Other academic work tested whether crowdsourced forecasts added useful information alongside existing estimates. A 2016 Journal of Accounting Research paper found incremental forecasting value and stronger results with larger contributor groups. The work thanked Drogen for supplying data and answering questions. The company was becoming both a business and something researchers could interrogate.
“I’m extremely opinionated”Leigh Drogen, describing himself in 2013
Getting institutions to listen
Collecting forecasts was one task. Getting institutions to use them was another. Estimize’s early customers included large quantitative hedge funds. By late 2014, demand also came from discretionary investors, bankers, investor-relations teams, and corporations. The business had to turn a participatory website into data that fit professional workflows.
Expansion brought experiments beyond earnings. In 2014, Estimize launched Mergerize, inviting predictions about acquisitions and their prices. The same year, it raised an additional $1.2 million from investors that included existing backers Contour Venture Partners and Longworth Venture Partners. The question was spreading from a company’s next reported quarter to what a company might do next.
Euromoney’s 2015 investment added a different connection. It paid $3.6 million for a 10 percent stake. Drogen saw value in the publisher’s media properties and the relationship between financial information and distribution. A dataset needs contributors, customers, and a route between them. Those routes are part of the product, however unromantic they sound.
He also urged finance chiefs to collect internal expectations from their own employees. People close to a business may know something a centralized forecast misses. The practical challenge is finding whose judgments deserve attention. That is the Estimize question brought inside the company itself, where an organizational chart can be an inconvenient substitute for evidence.
In May 2021, Drogen announced the combination with ExtractAlpha. Vinesh Jha, an early Estimize advisor and ExtractAlpha’s founder, took over the combined organization. Drogen reported a community exceeding 100,000 contributors and said he would remain involved as an advisor. Handing over the reins also gave him room to explore again.
Estimize begins
Blind forecasting adopted
ExtractAlpha combination; Starkiller founded
A different market, familiar human habits
Starkiller returned Drogen to managing money. He had bought a small amount of Bitcoin in 2013 as a way to understand the technology through participation. His views developed unevenly: curiosity, skepticism about utility, and later interest in decentralized finance. He approached the market with experience in momentum strategies rather than a lifelong allegiance to a particular token.
The firm now describes two discrete strategies: trend following in liquid tokens and market-neutral decentralized-finance yield. It also makes selected seed investments. His colleague Dennis Qian focuses on protocol research and on-chain opportunities. Their public writing considers how an expanding technology can still be a poor investment for holders of a particular piece of it.
In a September 2025 essay, Drogen and Qian argued that value was moving toward applications and customer relationships as blockchain infrastructure became easier to obtain. Stripe served as their case study. Their claim was about where businesses capture economics, rather than whether the underlying technology deserves applause. Useful infrastructure and lucrative ownership are questions that need separate answers.
A December 2025 Starkiller essay turned to prediction markets, where participants may pay for excitement or sell because they prefer deploying their capital elsewhere. The firm examined how those preferences could leave near-resolution contracts priced below their eventual payouts. Its interpretation placed patience alongside forecasting skill. A market can be full of informed people who want different things from the same trade.
In August 2026, Drogen’s “The Cycle Turns” described a change in Starkiller’s market view, drawing on quantitative signals and qualitative observations. He acknowledged that conviction wasn’t certainty. That distinction belongs in a career spent examining expectations: even the person studying everyone else’s beliefs has to decide how much confidence his own deserve.
Whitefish, with an argument still in progress
Drogen moved from New York to Whitefish, Montana, describing a life near the lake, ski mountain, and Glacier National Park. Hockey remained part of it, along with snowboarding, snowshoeing, sailing, and windsurfing. The scenery changed without requiring the markets to close.
His public voice remains willing to revise itself. His Bluesky biography includes a small, sensible admission: “even I don’t agree today with everything I’ve ever tweeted.” For someone whose businesses concern predictions, a permanent archive of opinions is a fairly demanding colleague.
The shell wagon, the trading models, and the forecasting platform are different endeavors. Looking across them, a pattern emerges: Drogen prefers to participate, see what happens, and adjust the arrangement. Estimize’s blind forecasts are the clearest example. He wanted to hear more people, learned something about how people influence one another, and changed the rules. The disagreement was useful. The design made room for it.