Read this article and learn about:
- What behavioral finance is all about
- How active managers can improve performance
- Behavioral finance as an active management tool
- How to optimize existing investment processes
- What behavioral finance can do to boost the bottom line
About the author:
Clare Flynn Levy, founder and CEO of Essentia Analytics.
Eric Rovick is an equity markets professional based in the UK.
About the Panel
Clare Flynn Levy, founder and CEO of Essentia Analytics, has previously spent 10 years as a fund manager. Her fund management career included both active equity – running over US$1 billion of pension funds for Deutsche Asset Management – and hedge – as founder and CIO of Avocet Capital Management, a specialist tech fund manager. Clare ultimately “went native” into the software space as President of Beauchamp Financial Technology, a market-leading provider of portfolio management systems to hedge funds, which was acquired by Linedata Services S.A. Clare‘s vision for Essentia is based on years of being the client, followed by years of listening to the client.
Eric Rovick has spent over 20 years in the financial markets, initially as an analyst and then as a manager of traders, analysts, and salespeople at firms including Fidelity International, ABN Amro, and Collins Stewart Hawkpoint. He is a seasoned observer of investment decision-making across the both the buy-side and sell-side and within organizations ranging from large banks to small investment boutiques.
Faced with an increasingly uncertain outlook, active fund managers need to consider incorporating the lessons of behavioral science if they are to stand a chance of improving their investment performance and weakening the flow of assets to passive funds. While this article highlights examples of cognitive risk in active investment management, it also provides a working framework for addressing them.
Traditional portfolio analytics dissect historic performance and exposure to infer what a portfolio manager was doing during a certain interval. Nonetheless, this reverse engineering provides no real insight into the behaviors and skills actually driving investment decisions.
However, an increasing acceptance of the principles of behavioral finance, combined with advances in technology, now enables investment managers to adopt a new skills-based approach. They can continuously assess the strengths and weakness of each investment decision-maker, as well as the market, cognitive, and environmental conditions in which the best and worst investment decisions are made.
We estimate that, as a tool to optimize existing investment processes, applying a data-driven feedback loop can increase decision-making efficiency, with sustainable long-term performance gains. Indeed, portfolio analytics have shown that most fund managers cede at least 50-100 basis points (bps) of excess return annually to cognitive bias. We argue that fund managers using technology to mitigate bias and realize additional alpha will, as a logical outcome, have a better chance of reversing the flow of assets into passive vehicles.
In this article, we also provide a case study to illustrate where and how cognitive bias can be addressed to increase performance. Finally, we make the linkage between ‘skills-based investing’ and ‘Active Share’, a measure of active investment evidenced to create outperformance.
Passive funds make gains at the expense of active vehicles
Boston Consulting Group (BCG) has highlighted active fund management’s continuing loss of market share to passively managed vehicles and expects passive funds and ETFs to continue their superior growth at least through 2016.
Fuelling this trend is disappointment with performance (only 10% of US managers beat their benchmark in 2014), whilst management fees remain stubbornly high. Furthermore, investors and regulators are demanding increased transparency, in part to flush out the large number of perceived closet-indexers.
How can active management respond?
Past a certain point, the growth in passive assets will create opportunities for active managers, allowing them to trade against index funds as the latter move the markets with their fund flows and index composition changes. Meanwhile, active managers can seek to challenge the effects of market share loss by reviewing channel and product strategies and seeking cost efficiencies.
But to meaningfully counter the threat from passive funds, active managers need to review the investment process itself and may find it worthwhile optimizing it to incorporate insights from behavioral finance. Such optimization can, we believe, generate 50-100 bps of sustainable outperformance – enough to give a subset of active managers, at least, an opportunity of fighting the outflow of assets to passive.
The science of financial decision-making
Neuroscientists claim that our subconscious governs up to 95% of our behavior. Behavioral finance (see fact box) has shown that a range of unconscious cognitive biases can affect investing, diluting the role played by skill and negatively impacting investment performance.
Some of these biases are rooted in millennia-old unconscious instincts of greed and fear; hidden drivers that can manifest themselves when we’re faced with certain decisions. Active investors are accustomed to considering the ‘psychology of the market’ and the investment opportunities created by the irrationality of others. But historically only a few have applied this same filter to their own decision-making.
Aspects of traditional active management, e.g. the ‘Star Manager’ culture, can encourage this; leaving the real drivers of performance unquestioned so long as performance is good, whilst being terminally unforgiving when it is bad.
Cognitive biases found in active asset management
In our experience of working with fund managers, the overwhelming majority will display at least one persistent bias, typically costing a portfolio 50 to 150 bps of excess return per annum. This is worth addressing because it can offset or cancel out the alpha generation achieved elsewhere in the portfolio through the exercise of genuine skill on the manager's part.
Technology vs. luck
Given, then, the damage that can arise when behavioral factors are ignored, how can we optimize the investment decision framework to make sure they are included? The answer is technology.
Advances in data analytics allow us to mine a fund manager’s decision-history and obtain for him a fast, accurate picture of where cognitive biases are at work. Innovation in areas such as wearable technology allow us to further account for factors (e.g. sleep and stress levels) that may affect performance but which are not captured by trade or market data.
With insights from this data in hand, fund managers can assess and quantify where they add the most value, where their skills are strongest and weakest, and the conditions in which they make their best and worst decisions.
This technology-powered feedback loop allows them to adjust their behavior, mitigating cognitive bias through self-awareness and focusing energy on investment skill where they know it exists. Journaling, checklists, and data-driven performance coaching – all shown to work in professional sport and the military – reinforce this feedback loop, maintain discipline, and help consolidate the changes that can lead to outperformance.
Feedback Loop in Action – an example of applied behavioral analytics technoloGY
Closing the loop
As well as giving a fund manager data-driven insights about his own investment behaviors, technology tools like Essentia’s decision-support software can incorporate automatic ‘nudges’ or alerts when familiar patterns are seen arising, or when personalized stop-loss rules are breached. What the manager chooses to do at that moment is his own decision, but by prompting him to focus at the crucial time, the enabling software increases the chances of him maintaining his investment discipline.
The end-result is an investment process that is tailored to reflect the personal style and cognitive traits of the individual portfolio manager, providing for more consistent alpha generation and continuous improvement.
Making active management more active
Substantiated by academic research, the call in the investment management industry for active managers to be more ‘active’ is growing shriller by the day.
Cremers and Petajisto coined the term ‘Active Share’ as a way of measuring how close to the benchmark a portfolio is in terms of its holdings. They found that funds with a high Active Share (i.e. at least 60% different in investment composition to their benchmark) were more likely to outperform their benchmark indices both before and after expenses.
Simon Evan-Cook of Premier Asset Management has explored the outperformance available through Active Share in the context of the UK market (see Figure 1). He has found that actively managed funds with an active share of 80% or above produced 10.3% in annualized returns over 10 years, significantly outperforming passive tracker products, even after fees.
But how easy is it to run a portfolio with a high Active Share? It’s a risky business – by definition. However, a portfolio manager can take active positions with more confidence when he is keenly aware of where his skills and weaknesses lie, and of how he can reduce his own cognitive bias using technology-driven interventions.
The moving of funds to passive instruments will continue until active managers can justify their fees. Encouragingly, though, an increased focus on the investment-decision process is emerging. This is driven by developments in behavioral finance and the advances in analytical technology towards examining data, not just on the outcomes but also on the factors that lie behind the decision-making process.
As a result, we can expect to see a new breed of professional active managers emerge who seek to optimize their performance using a technology-driven feedback loop. An investment process based on the ‘science of decision-making’ can enable active management firms to realize alpha that would otherwise have been lost to cognitive bias, whilst delivering predictable, sustainable investment performance that is not dependent on a few key star managers.
The main question, then, is one of intent: are CEOs and CIOs willing to embrace these techniques as a way to improve investment performance? Are portfolio managers themselves prepared? Or will they succumb to the ‘ostrich effect’ of ignoring existing problems, leaving the door open to passive vehicles to capture the majority of fund flows?
1. ‘Global Asset Management 2014: Steering the Course to Growth’, Boston Consulting Group, 2014 https://www.bcgperspectives.com/content/articles/financial_institutions_global_asset_management_2014_steering_course_growth/
2. ‘Investment: Loser’s game’, John Authers, The Financial Times, 2014 http://www.ft.com/cms/s/0/f15a1f9c-876c-11e4-8c91-00144feabdc0.html#axzz3OiuqlGuE
3. 'Kill the Filler: The Costs of Closet Tracking’, Simon Evan-Cook, Premier Asset Management, 2014 https://www.premierfunds.co.uk/media/58892/kill-the-filler-november-2014.pdf
4. ‘The Biology of Belief: Unleashing the Power of Consciousness, Matter & Miracles’, Lipton, B., Hay House UK, 2005
5. ‘How Active Is Your Fund Manager? A New Measure That Predicts Performance’, Cremers, M., Petajisto, A., 2006 http://faculty.som.yale.edu/anttipetajisto/active50.pdf
BACK to where you left the article
Feedback loop in action – an example of applied behavioral analytics technology
Consider the example of a US$15 billion global equity active fund manager underperforming his benchmark for the last two years. Performance attribution analytics showed that the stock selection was not adding value but did not provide him with any real insight into what he should do to improve.
By using behavioral analytics technology on his trade, portfolio, and benchmark data, it became instantly clear that fund performance was being damaged by his retaining relatively small, underperforming positions for too long. A comprehensive exit analysis was conducted, including:
- Running in excess of 1,000 scenarios or simulations, deriving not only the impact of each exit decision, but also an objective score of the quality of these decisions.
- Measuring the average and distribution of the exit decisions across the portfolio for the sample time period and comparing this with the benchmark’s performance in the run-up to, and in the period just after, each exit.
- Slicing the data by a long list of contextual factors and fundamental attributes, including sector, price momentum, day of the week, and time of trade.
The analysis revealed that the manager was falling prey to the Disposition Effect: the tendency to hold loss-making investments too long, whilst selling profitable positions too soon. This behavior is common to both professional and amateur investors with meaningful performance effects; research into the bias found that winners sold outperformed losers that were retained by an average excess return of 3.4% per annum. (Source: ‘Are Investors Reluctant to Realise Their Losses?, Odean, T., The Journal of Finance, LIII(5), 1775-1798 – (1998))
Data visualizations display the extent of the bias (see Figure 1). In examining the average exit pattern for winners, the data showed that a security that had performed in-line with a (rising) index tended to continue doing so once the manager had sold it. By contrast, analysis of the average exit pattern for losers showed that the manager was tending to exit his losing positions after they had underperformed for several months, and usually when they had reached near-bottom.
In basis point terms, the Extraction Ratio (realized gains/max realized gains as shown in Figure 2) confirmed that the manager was leaving significant excess return on the table when exiting winning positions and also realizing the vast bulk of maximum possible losses when exiting his losers.
The analysis also showed that if the manager exited his losing positions even two weeks sooner he would have improved his excess returns by at least 0.50% per annum (see Figure 3).