Sponsored by qplum
A 10-year bull run for stock markets makes a very strong case for passive, buy and hold investing. Active investing has recently become synonymous with high costs and extracting alpha from such strategies has been very challenging for investors and advisors.
Gaurav Chakravorty, CIO of qplum, shares his views on how to extract alpha using tactical asset allocation. He feels new, A.I. driven, holistic methods of tactical asset allocation are better and more cost efficient than relying on sought-after star managers.
Is there alpha to be made from active investing?
Yes. But it takes time, especially now with looming prospects of an equity market correction. Being smart and proactive about asset allocation is likely to yield returns. However, focusing on security selection would not work today. Our research shows that since 2004, a systematic, data-driven approach to global tactical asset allocation (GTAA) could have potentially yielded about ten times more alpha than a typical security selection model.
Are you saying tactical asset allocation generates more returns, as opposed to choosing which security to invest in?
Yes, asset allocation could potentially be a much bigger driver of returns.
In GTAA, you are not taking concentrated risk in one asset class or in one security. The alpha component comes from limited downside and the ability to switch between different asset classes based on new data.
What is Global Tactical Asset Allocation (GTAA)?
GTAA is a top-down global macro strategy. It's a dynamic strategy that decides how to allocate money across various asset classes, geographies, styles and sectors.
Traditional asset allocation models, compared to GTAA, are strategic or static. They assume that the world is the same today as it was on March 2009. The basic hypothesis is flawed.
With Global Tactical Asset Allocation, the portfolio manager is empowered to change the asset allocation based on changing market conditions, central bank policy, macroeconomic data and investment flows.
The objective of these tactical shifts in the portfolio can be client-specific. Some might seek to maximize investment returns while some might want to keep risk contained as per their mandate. The important thing is that the portfolio is ready to handle changing market regimes without constant monitoring and guesswork from the advisor or client.
Using a systematic approach, the advisor can make asset allocation decisions every day rather than once every year, and potentially generate higher risk-adjusted returns.
Where does the use of A.I. come into all of this?
There are many factors that can affect the direction of markets in the future. A.I. is much better than old-school models at taking a lot of these factors into account. The use of A.I. based methods is still very new in asset allocation and there aren’t many firms that have the experience to do it. Traditionally, tactical asset allocation has been achieved through a more discretionary approach where a team of portfolio managers and analysts pour over tons of data to guess which tactical tilts to make in the portfolio.
Using machine learning methods like deep learning, this process can now be more systematic and transparent. A machine learning tool can research relevant data points, identify which ones would affect asset allocation choices and which ones wouldn’t.
It can also potentially uncover bigger opportunities of alpha if built as a framework that can process lots of different data sources efficiently and flawlessly. After all, alpha is often about knowing what beta to bet on before others have realized it too.
What is different about the GTAA portfolios that qplum offers?
I’ve been using pure ML based systematic trading since 2005. My team and I are among the best in the world at using machine learning in trading. We bring those capabilities to qplum’s GTAA portfolios.
We’ve developed multiple Global Tactical Asset Allocation based portfolios with defined risk mandates. We use a deep learning based framework to parse through vast amounts of market and economic data, determine the appropriate asset allocation, and execute it through our trading engine. The entire process is systematic, which ultimately results in higher efficiency and lower costs for our clients.
Being systematic and data-driven, our process eliminates human biases that typically degrade traditional discretionary tactical asset allocation strategies.
Is GTAA applicable only for institutional clients? Should retail investors also focus on asset allocation?
GTAA applies to both retail and institutional clients. However, being tactical poses different challenges for both investor segments.
Institutional investors have always understood the importance of being tactical. However, they’ve found it difficult to build the research and execution infrastructure in-house. So they can execute in a cost efficient manner.
Retail investors have been offered a lot of security selection ideas, but relatively few ideas on research-driven tactical asset allocation. Let’s look at one example where not being tactical fails a retail investor, the traditional 110 minus age rule of stocks and bonds followed by most advisors and target date funds. Here we have a glide path that focuses on growth during the early phase and capital preservation during the later phase.
What happens if you have a high equity allocation and equities underperform for an extended period? Or what if the market crashes?
Since you have a static portfolio that does not consider market conditions, and your portfolio loses a lot of money, it becomes hard for the investor to stay invested for the long term.
Another problem with this 110 minus age rule glide path is that bond allocations will go up with age. However, with continually lower bond yields, and longer expected life spans, the chances of retirees running out of money is very high. This makes the asset allocation puzzle much harder to solve than you’d first think.
Why hasn’t GTAA been more widely adopted by investors?
Costs.
GTAA is a lot of work, and requires a lot of data analysis. In a chasing-after-the-star-manager model, the cost of being tactical would often be prohibitively high. With greater investor education, things are changing. Institutional investors like Norway GPFG are allocating a much larger portion of their active risk to tactical asset allocation than security selection. With transparent systematic approaches of GTAA available at a tenth of the cost, this is no longer a hurdle.
How can someone apply global tactical asset allocation to their investments?
One way is for wealth managers and institutions to work with asset managers like qplum to invest in low-cost and customized portfolios with built-in risk management.
Alternatively, they can employ a team of experts to constantly monitor the markets, derive insights from the vast amounts of data, test which ideas have worked in the past and which are inapplicable today, and allocate across various strategies like momentum, carry, value, etc. as deemed appropriate.
Gaurav Chakravorty is the Co-Founder and CIO of qplum, which he co-founded with a firm belief that investing can be approached as a science by utilizing the power of Artificial Intelligence and Deep Learning.
Gaurav has over 12 years of experience in trading and portfolio management. Prior to qplum, Gaurav applied machine learning to trading strategies in a quantitative proprietary trading firm, DV Capital LLC which he co-founded in 2010. Prior to that, Gaurav was the youngest partner at a proprietary trading firm, Tower Research Capital, where he was an early pioneer in machine learning based high-frequency trading, building one of the most profitable quantitative trading groups from 2005 to 2010.
Gaurav has a Masters in Computer Science from the University of Pennsylvania, and a Bachelors of Engineering in Computer Science from the prestigious, Indian Institute of Technology (IIT-Kanpur).
The views and opinions expressed are those of the author and are current as of the date written. This Q&A is general in nature and should not be construed as investment advice. Opinions are subject to change with market conditions. All performance information is hypothetical and there is no guarantee of any future results. Information presented is for educational purposes only and does not intend to make an offer or solicitation for the sale or purchase of any specific securities, investments or investment strategies.
