If you fancy James Bond movies, you probably know that The World Is Not Enough was the first Bond movie filmed by EON (Everything or Nothing) Productions instead of its original producer United Artist. For the first time, the Bond girl, a nuclear physicist played by Denise Richards, was projected to be “both brainy and athletic” making it an interesting breakthrough. I find a parallel between how the investment community balances between the Active vs Passive when talking about investment approaches. Depending on investors’ perceptions, the passive approach could be wrapped as “athletic” while the active approach as “brainy”. Yet can we make the investment process to be both athletic and brainy?
The understanding of what are Active and Passive approaches is all important for investors, allocators, and investment managers to define best methods for strategy execution and portfolio construction as well as appealing ways to marketing and investor relationships.
During the recent QuantMinds 2018 conference, I moderated the panel discussion on the topic of Active vs Passive investing. The outstanding participants of the panel covered different angles of the investment community:
- Philip Stoltzfus, CEO at Thayer Brook Partners LLP
- Michael Steliaros, Global Head of Quantitative Execution Services at Goldman Sachs
- Simon Weinberger, Managing Director, Scientific Active Equities at BlackRock
We had an enthusiastic discussion, and everyone shared his insight into topic and raised interesting thoughts. I realized there is a lot to learn from our discussion and decided to write a small summary.
Philip Stoltzfus, Artur Sepp, Michael Steliaros, Simon Weinberger
What sets apart Active vs Passive?
The very definition of Active vs Passive for practitioners depends on the field of expertise and practice. Depending on the area of application within the investment process, we can distinguish between the three key factors that clearly set apart Active vs Passive as illustrated by the featured image on the top.
- Micro-level investment making process. The goal of the active approach is to generate absolute returns. The goal of the passive approach is to minimize the tracking error between the portfolio and the benchmark. Steliaros stated the tracking error is the main thing by which passive strategies are assessed, and that is risk. Further, a passive fixed-income ETF tracking a benchmark with a few thousands of bond issues may only allocate to a couple of hundred of most liquid bonds and actively manage the tracking errors based on liquidity conditions.
- Macro-level asset allocation process.
Weinberger emphasized that in the equity space almost everything requires an active decision. When it comes to smart beta, to me you are firmly in active territory. You decide to deviate from a benchmark quite significantly.
In fact, allocating to a tracker of S&P 500 index is an active decision at the macro level even though the implementation of the tracker at the micro level is passive. The reason is because the S&P 500 index only represents around 10% of all investable assets globally so that allocating all funds into the S&P500 passive tracker is an active decision relative to the global investable benchmark. Weinberger asserted that a lot of asset allocation decisions are in fact active decisions. For example, European investors who allocate to USD assets must take an active decision whether to hedge or not to hedge the currency risk and how.
- Risk management approach. The risk management approach indicates how the risk can be allocated at both micro- and macro-levels. As a simple example, (unconstrained) risk-parity can be seen a passive strategy while volatility-targeting can be an active strategy. Both strategies do not need to have a benchmark as the risk profile is different to mainstream indices, only proxy-based comparison can be applied.
Stoltzfus noticed that in the managed futures space a passive method is one that focuses on a static rule set, which doesn’t mean that it totally lacks a dynamic element, but it’s essentially trying to tap into a pure underlying performance driver. In active there are methods employed to try and compensate for elements of trend-following that don’t work well, such as a higher focus on risk management. Stoltzfus confirmed that there still is enormous scope to use research to develop new methods that will provide benefits, such as better risk management and better drawdown control. Those active managers will be judged based on their ability to deliver something that is genuinely differentiated from the pure trend-following component.
To recap, I would like to refer to the research by Andrew Lo introducing the idea of dynamic indices which can generalize both passive and active investing. For an example: volatility targeting applied to the S&P 500 index would qualify as the macro-active/micro-passive approach. As Andrew Lo showed that with proper level of volatility target and a model for volatility estimation, the volatility target index generated much higher total returns than the S&P 500 index. At the same time, Lo stressed that the choice of the volatility model for target calculations has a significant impact on the back-tested performance of volatility-targeted strategies and thus leading to potential over-fit, which is a common weakness of quantitative modelling.
During my talk at the conference, I presented my work on machine learning for selecting an optimal model for selecting of volatility models which minimizes the back-test over-fit.
What is not about Passive vs Active?
- Quantitative vs discretionary. This is clearly not a way to distinguish between the active and passive as both approaches. Steliaros stated that systematic quantitative processes can be applied across the spectrum. He has seen a significant increase in the last few years in the fundamental, active stockpicker-type client who are using a lot of systematic risk management tools.
- Stoltzfus stressed that we are seeing assets flow both to passive methods and to cheaper trend-following strategies.
Trend towards data-centric investment approach
All participants recognized the trend towards utilization of data in the decision-making process irrespective of investment style.
Weinberger referred to the research from IBM that 90% of data have been created in past two years. Even though the traditional time series price and fundamental data have been used for decades by the investment community, we may be only at the initial stages of data-centric approaches utilizing the alternative (big) data sets.
Regarding the complexity of data analytics, Steliaros estimated that about 80% of the resources are spent on data sourcing and structuring while 20% is spent ultimately on the data analysis and exploration.
To sum up, I just came across a deeply inspiring presentation by Leda Braga, CEO of Systematica Investments, on the increasing role of data science in the investment process. She states that “Investment management is an activity whereby the pools of capital of the world get directed. That is so powerful. And if that is going to become data-driven over time then you can’t miss that opportunity. You’ve got to join in and have your say.”
I have attended the QuantMinds conference for 8 years in a row. This year the conference felt special… I find that people are very energetic and enthusiastic about increasing applications of quantitative analytics for investment processes, without regard to either Active or Passive, as well as for deeper utilization of data analytics. The quantitative data driven analytics will continue to change the way the investment community operates.