The Spotlight is on HiddenLevers

HiddenLevers is usually the one shining the spotlight on our users, but this week the spotlight is shining on HiddenLevers’ co-founder Praveen Ghanta. He was recently interviewed by Manish Khatta, of  Potomac Fund Management, after HiddenLevers was chosen to power their client proposal system. After two years of failed attempts with various providers, Potomac finally found a software who fit the bill!


Can you tell us a little about yourself and your background as our readers try to understand the process behind HiddenLevers?

HiddenLevers isn’t my first tech venture – I started my career in the dot-com boom and sold an internet startup right into the bubble pop in mid-2000. I then landed on Wall St (at Deutsche Bank) and spent many years building credit risk trading and management systems. That fit well with my educational background, as I studied both computer science and economics at MIT – but I always wanted to get back to entrepreneurship.

Fast-forward to 2008, and I started to think about trying my hand at another venture – just as the financial crisis was looming…

What was the main reason you started HiddenLevers?

In 2008, prior to the crisis, I became interested in the idea of identifying the relationships between different factors in the economy and investments. If we all look at the beta to the S&P 500, why don’t we also consider an investment’s potential beta to oil, or home prices, or interest rates? The initial idea revolved around discovering the “hidden” economic levers that move your investments – hence the company name.

But I realized almost immediately that if you have analyzed these economic relationships, you can also try to answer what-if questions about the future – what happens to my portfolio if the tech sector corrects? Or if home prices pass their 2007 highs? By the time HiddenLevers hit the ground running as an actual company (January 2010), the demand for better ways of modeling risk was evident, and this kind of forward-looking scenario analysis looked like a great opportunity.

The Proposal System: 

The ability to create proposals was a recent release from HL. Why did you get into the business of creating a proposal tool?

In two words: advisor demand. We got many requests from firms that were dissatisfied with the proposal tools that they saw on the market, and at the same time really liked our approach to presenting risk. Many advisors were already using our reports and other output with clients, and they were hoping to see something more holistic. We’ve focused on making our product development process driven by client feedback, so once we started to hear the requests, we listened. Another motivating factor were the changes in DOL rules – advisors need to be able to present more clearly on both matching a client’s risk tolerance and on the value that their allocations provide.

Being a tactical, unconstrained manager made it difficult for us to find a proposal system? Why do you think Hidden Levers fit the bill?

Potomac cares deeply about risk management, and I think you needed a system that would help you illustrate that. Since HiddenLevers has built a risk-forward proposal which focuses on highlighting that capability, I think it was a natural fit.

Nerd Alert: Maximum Drawdown, Stress Testing and Expected Return:

Can you explain why maximum drawdown is a key part of your system including the client risk scoring?

Measuring a client’s risk tolerance is about trying to understand how they might react when their investments are losing money. Most risk assessment processes try to capture that, but it’s very important to look at tail risks – what might an investor do when their investment is down more than a few standard deviations? Maximum drawdown can provide a good historical perspective on this, particularly when an investment has a real historical track record, and not just back-tested or index data.

Why forward-looking stress tests vs examining the actual historical return characteristics?

HiddenLevers’ forward-looking stress tests are really about taking the concept of max drawdown and attempting to model the risks that might cause a similar drawdown in the future. The stress testing model that we use relies heavily on historical return characteristics, as it uses a strategy’s historical returns to model its relationship to various factors in the economy. But since history never exactly repeats, it’s quite helpful to be able to take the economic relationships that drive an investment’s returns, and then model an economic situation more in keeping with the present.

The next crisis won’t play out just like the last financial crisis – for one we are starting in a much lower interest rate environment. Forward-looking stress testing enables us to look at a large range of different ways that the next crisis might unfold, rather than being tied to an exact repeat of history.

Do you track sample portfolios to see how accurate the forward stress testing has proven to be?

We do track sample portfolios and we use these to evaluate the accuracy of our model across a range of asset classes, including equities, fixed income, balanced strategies, and even tactical funds.

We periodically publish white papers summarizing these results – for instance we published a white paper on model performance in recreating historical scenarios like the 1987 crash, the 2008 financial crisis, and the 2014 crash in oil prices. These studies enable us to continue to tune and refine our model to improve its accuracy. Using these sorts of historical tests, we found that the model projected results within 5% of the actual result in 85% of the cases reviewed.

The proposal shows expected return/drawdown. For the layman, how are those numbers computed?

This concept is based on some long term historical data: since 1900, major downside market events (where the S&P 500 fell more than 20%) have occurred roughly twice a decade, or once every five years. Recent crashes have been slightly less frequent but more powerful, with two 50% declines in the past 17 years.

HiddenLevers measures the risk side of Risk/Reward using its stress testing model, enabling advisors and asset managers to look at a range of plausible downside risk events over the next five years. The potential five-year expected return is calculated by first using the model to project returns in the upside case, where stock markets continue to rise in line with their long-term averages.

This upside case is averaged against a variety of different risky scenarios, with the upside given an 80% chance of occurrence (remember that the downside is roughly 1 in 5 years). Once the average annual return is computed using this approach, it is compounded over 5 years to show a total return projection for the 5-year period.

The Future: What’s on the Roadmap?

What’s next in store? Any new and exciting improvements coming up?

We have had numerous requests to include additional components in our analysis and as proposal modules – analytics like sector, style, region, and credit quality breakdowns are planned for later this summer.

HiddenLevers is also augmenting its Risk Monitor capability to monitor clients’ status throughout their whole life cycle – has a prospect completed their questionnaire? Now that they’ve closed, does the risk of their new holdings match their risk assessment?

How about integrations with our software partner Orion Advisor?

HiddenLevers is planning to start sending more data back into Orion, so that advisors using both platforms can see HiddenLevers risk/reward and other metrics inside Orion and in Orion-generated reports.

Will you add the ability to add contributions or distributions to past or future scenarios?

This is an area of great interest for us – we do have it on the roadmap, but when we tackle this feature, we’d like to really address the idea of scenarios occurring at any point during a client’s financial plan. What if a particular downside scenario occurs tomorrow? Five years from now? Six months before retirement? How does the timing impact their financial plan? We are working on that concept – stay tuned.

Click here to read the original interview

Published by

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s