Based on an interview with Praveen Ghanta
Founder and CEO of HiddenLevers
Site: www.hiddenlevers.com
Founded: 2009
Clients: RIAs, broker-dealers, banks, trusts, hedge funds, asset managers
Value proposition: Portfolio stress testing, economic data, risk analytics, risk monitoring, macro analysis, and scenario modeling
The executive team: Raj Udeshi, founder; Praveen Ghanta, founder and CEO
HiddenLevers is a technology platform providing business intelligence, risk analytics and economic research for the wealth management space. It enables institutional-level stress testing that one can get from very expensive platforms such as Bloomberg or BlackRock Aladdin, and delivers it in a way that helps advisors talk about risk with their clients. Whatever is going on in the world, HiddenLevers claims to have a scenario to show the client how their portfolio will perform.
Praveen Ghanta, founder and CEO, is responsible for the HiddenLevers engine and product development. Ghanta has a deep background in financial technology, having built credit derivatives trading and risk management systems at Deutsche Bank. Previously, he founded and led SmartWorkGroups, an online collaboration startup until its acquisition by Intralinks in July 2000. He also serves on the Board of Councilors for the Carter Center. He is a graduate of the Massachusetts Institute of Technology with bachelor’s degrees in computer science and economics, and a master’s degree in computer science.
In this article, we discuss what HiddenLevers offers to clients and what challenges influence the company’s roadmap. We also take a closer look at their corporate culture and talk about trends.
What does the platform provide?
The basis of HiddenLevers is portfolio stress testing. The platform provides access to a scenario library that covers 100+ outcomes, or economic “what-ifs.” A scenario about the Yield Curve models 4 outcomes on how a yield curve inversion would affect U.S. and global markets. These outcomes are expressed as a set of over 50 economic indicators, like oil prices, interest rates, where the S&P 500 is at, etc. The variety of outcomes gives advisors and portfolio managers broad insight into the risks of their investments.
“We typically use historical precedents or macroeconomic theory," Ghanta says. "We do a lot of research, look at what economists are saying, and try to construct our factors into a scenario that is fitting with the possible outcome.”
To increase the usability of their product, HiddenLevers built integrations with many platforms including Orion, Tamarac, Addepar, Black Diamond, eMoney, MoneyGuidePro and more. The types of integrations they provide can be divided into three groups:
• Sync integrations. Typically, sync integrations are done through API. HiddenLevers has 30+ such integrations, including the leading custodians, financial planning tools, performance reporting tools, CRM systems and account aggregation tools. The sync integration runs every night, pulling in all the account-level data and building them into clients’ profiles.
• Single sign-on integrations. This kind of integration has been built with 10+ firms as a different way to get information to the system. It allows advisors to launch data from an outside tool right into HiddenLevers.
• File upload integrations. Finally, there are a handful of integrations where there’s no API in place or a way to sync. For them, HiddenLevers has built file uploads where people can download all their information from a third-party system and upload that to the platform.
Stick to feedback or think strategically?
HiddenLevers finds it crucial to catalog every piece of feedback they get from subscribers. They rank the information based on the size of the firm, their AUM, number of requests and other factors to define the most impactful potential requests. However, there are two sides of that coin—Ghanta reminds me of a quote from Henry Ford: “If I had asked people what they wanted, they would have said faster horses.” So there are also strategic things they do to differentiate themselves from other firms.
“Feedback is very valuable, but there are also things that people aren’t asking us for or haven’t even thought of. Nobody was asking for stress-testing eight or nine years ago when it came to the market,” he says.
To define what to work on next, the whole organization gets together for a Town Hall, a meeting where they review all the feedback and strategic goals. During the event, they form a sprint backlog. They strive to release every 10 business days, which means about 35 to 40 releases per year. Ghanta asserts that tech can go stale pretty quickly without constant upgrades,.
Specialization drives productivity
To build a stress-testing platform, one should know how a risk-monitoring and business-intelligence tool should work. At HiddenLevers, product managers and product developers come from the financial industry. The rest of the employees spend a couple of hours every Friday talking about different topics in finance, such as options, asset management, the wealth management ecosystem structure and so on.
The team members communicate with each other through Slack, Asana, Pipedrive, and Aircall, which makes them feel like they are right next door. If someone has an industry-related question, it’s easy to get the answer on the fly. Also, they tend to educate the advisors, too.
“We have a knowledge base for our current subscribers," Ghanta says. "Some of the things that we do on our platform are a bit foreign even to them. The knowledge base is also a good resource for our development team, product managers, and even some of our business development team.”
AI-based risk analysis, prospects and other trends
Artificial intelligence and machine learning (ML) have a lot of potential benefit for HiddenLevers. ML can help identify underestimated risks and positions that are exposed to a heavy loss. However, there are challenges arising from using ML, such as explainability.
“How do we explain why the results of machine learning make sense?" Ghanta asks. "In fact, there’s a machine-learning ETF that owns a lot of funky stocks, and nobody can figure out why it owns those. Bringing the results of ML in a way that can help advisors to better explain things to clients is what we should focus on.”
Another prospective application of ML is to help get quicker insights and conclusions.