Hollis Johnson/Business Insider
Clare Flynn Levy founded Essentia Analytics to create an evaluation system for investors. Not everyone's convinced.
- Essentia Analytics, founded by a former hedge fund manager turned software company president, wants to help traders identify decisions that make money and avoid those that don't.
- By analyzing under what circumstances portfolio managers have underperformed - through a combination of machine learning and human analysis - the startup seeks to nudge them toward different choices in the future.
- Essentia often appeals at first to endurance athletes and academically-minded traders interested in overcoming their own biases. But it can be hard to convince some veteran traders to look in the mirror.
- Visit Business Insider's homepage for more stories.
Clare Flynn Levy knew she was doing something right picking technology stocks in the late 90 , but she didn't know exactly why.
Just out of Barnard College at the time, the then-portfolio manager at Deutsche Asset Management was "the toast of my team," she recounted in a recent interview in New York. Her portfolio was soaring in tandem with the tech sector, but she couldn't find much information to explain why she did so well. Was she holding stocks for the right amount of time, adding appropriately, or exiting particularly well? Could she have made any of those decisions better and outperformed even more?
"As a fund manager, I was looking for companies that maximized the return on capital deployed," Levy said. "Yet I couldn't convince myself that I was maximizing my return on my own capital - my energy capital. Therefore how could I possibly produce consistent, positive performance?"
Even in her subsequent roles, first as a tech-focused hedge fund manager, then as president of a software company for hedge funds, followed by internal and external advisory positions, Levy was frustrated by the lack of detailed performance feedback for stock pickers. In 2013, she founded London- and New York-based Essentia Analytics to create the evaluation system she had wanted.
Now, Essentia assesses historical trading patterns for fund managers at more than 30 clients globally. The startup uses machine learning to dig into when traders out- or under-perform and connects them with former fund managers for quarterly analysis and coaching. Essentia's software also prods managers to reconsider potentially money-losing decisions in the moment, based on their past trading behavior, and sends them alerts about which companies to focus on to narrow their attention. Levy likened it to athletes reviewing game and practice tape.
"Some people trade around too much. Some people hold onto losers for too long. Some people get sucked into the momentum. Everyone's a little different, but there are some common biases we've seen documented in academic literature that we can see in investing," she said.
Essentia's clients include long-only and hedge funds, ranging in size from $1 billion to hundreds of billions. Since the company raised a £2.5 million Series A funding round in January, it's exploring expanding to other asset classes, working with allocators, and improving benchmarks, Levy said.
Essentia doesn't work for every trader, but as investors flee actively-managed funds, it's more important than ever for investment firms to articulate their competitive edge by showing how they're generating outperformance beyond beating a benchmark. And as the buy side faces ever-growing margin pressures, they need every one of the 94 basis points of foregone alpha that Levy said Essentia identifies, on average, for each manager. For a $500 million portfolio, that's $470,000 that Levy said managers leave on the table because of their decision making.
FitBit wearers and CFA holders
The startup's approach fits with increasing interest in examining behavioral bias within finance. Dr. Enrichetta Ravina, a finance professor at Northwestern University's business school, explained behavioral biases as "rules of thumb" that help make decisions quickly. In investing, they can manifest as overconfidence or chasing trends, among other attitudes.
"Evolutionarily, behavioral biases are efficient - you don't want to calculate a problem every time you make a decision," she said. "Sometimes in finance, those same shortcuts that help you everywhere else in life are going to be your enemy. Everyone has behavioral biases. They're very natural and they actually overall are optimal from a life point of view, but they're problematic in investing."
Ravina said that she's heard of large managers similarly keeping track of under what conditions traders do well and when they don't. If a portfolio manager makes a decision that's historically led to underperformance, the manager's firm might take an opposite position against the trade - without telling the manager that the firm is betting against him or her.
Internal risk managers also typically analyze much of what Essentia captures, but Levy said traders are more willing to work with an independent group than a company risk manager, who they view as "the police."
Sign up here for our weekly newsletter Wall Street Insider, a behind-the-scenes look at the stories dominating banking, business, and big deals.
Some fund managers are particularly keen to analyze their performance and mitigate those behavioral biases, Levy said. Early adopters at any given firm typically include "continuous improvers," often endurance athletes and people who wear monitors like FitBits - they're accustomed to quantifying their training and health, and want to translate that attitude to their jobs. Academically-minded fund managers, typically those who hold the Chartered Financial Analyst designation, are also early adopters.
Those groups "don't need any selling on the concept of using data to do a better job, and they have no qualms about looking into the mirror and seeing that they're not perfect," Levy said. "As this concept of behavioral alpha is being embraced by the industry more and more, it has to trickle down to people who aren't as excited right off the bat, who are worried this is going to humiliate them in some way, that they're going to get fired."
'A little bit scary in the beginning'
Levy said reluctant managers include veteran traders who have decades of experience and no desire to look at their history, thinking they understand themselves better than a computer or a former trader could. Sometimes, such traders come around when they see colleagues' performance improvement, or when allocators and consultants ask them to prove their investment process.
"That's a killer question right there," Levy said. "If you're a human active fund manager, you're under a ton of pressure now to prove that you're worth the fees you're charging, when the alternative is a cheap index fund that demonstrates zero skill, but it's also not pricing in any skill."
For Seth Wunder, who founded $750 million long/short fund Black and White Capital, his six months using Essentia so far have underscored that he does particularly well with position sizing - identifying big ideas and leveraging them well.
"For myself, it was a little bit scary in the beginning because the reality is you perceive why you're making money, but to potentially be told you're not making money for the same reasons you think you are, or you would have made more had you done X, Y, or Z, is interesting," he told Business Insider. "I could see why people wouldn't care to know or deal with it."
Next, he plans to overlay his firm's internal research with Essentia's data for what he expects will lead to even better trading outcomes.
Essentia isn't applicable for all strategies, like quantitative investing. Olivia Engel, the State Street Global Advisors active quantitative equity chief investment officer, said she was pulled to the quant side because it's less susceptible to behavioral biases than fundamental stock picking, where she got her start as an analyst.
"You may think you're good at trading something because all you remember all the ones that went well, but if you objectively measure all of your trading, the story may be different because what you remember is colored by your experience and emotions along the way," she said. Of Essentia, she added, "I think it's a novel and interesting thing they're doing."
While not all investment strategies are the right fit for her startup, Levy said one good target would have been a hedge fund she advised during the recession. The fund was led by a manager who fully believed in his own competitive advantage but couldn't articulate it to investors.
"Maybe it wouldn't have made a difference in the context of the financial crisis, but investors pulled their money out quite rapidly," she said. "If we could've provided them with more data that gave them piece of mind about leaving their money with us, it would've made a difference."
- Read more:
- UBS is now operating 1,000 robots, and it's betting they can help create 'super human' employees
- Artificial intelligence is transforming a $22.9 trillion investing strategy - but the cutting-edge technology comes with a new set of problems
- Credit Suisse's CTO says that AI could create huge opportunities on Wall Street and that banks haven't even scratched the surface