Research has uncovered the importance of behavioural tendencies in investing, and we may be moving toward an investment world in which sentiment plays an increasingly important role, argue Ian Heslop and Justin Wells.
The sharp correction in equity markets in early February 2018 has reignited the debate over what moves markets: rational economics or investor sentiment. We would suggest that the need to understand investor behaviour, and how sentiment can influence markets, plays an important part in active investment management. Markets are often moved by subjective forces, such as knee-jerk reactions, emotional responses, and other psychological and sociological phenomena.
Just over thirty years on from Black Monday and the stock market crash of 19 October 1987, academics, policy makers and market participants alike continue their quest to explain the forces underpinning the dynamics of financial markets. Questions abound around the likelihood, timing and prospective causes of market pull backs or corrections. Last year’s Nobel prize winner in economics, Professor Richard Thaler, is a long standing proponent of behavioural finance. Thaler has sought to research psychological explanations, in preference to the optimised mathematical modelling espoused by conventional economic theory, in explaining the interaction of investors with financial markets.
This paper seeks to highlight some ways in which market participants may become wrong-footed or demonstrate irrational judgement, and how this may provide investment opportunities for active managers. We begin with a brief overview of some long-held assumptions about the rational investor and market efficiency, before illustrating a variety of market frameworks which over time have come to challenge or even undermine some of these norms. Many of these frameworks have come to be termed collectively ‘behavioural finance’. We offer some suggested real time examples of such behaviours in the context of the smart beta phenomenon which have captured the attention of many market participants in recent times. Finally, we conclude by demonstrating some of the aspects of our process which have been designed to avail of the inefficiencies and distortions created by such frameworks.
THE RATIONAL INVESTOR
A commonly used model for the conventional, rational investor has often been referred to as homo economics. Adam Smith postulated that self-interest was an effective human trait for making markets work. Over the succeeding two centuries, economic theory adopted this as a fundamental assumption, such that competitive self-interest became the conventional investor’s natural state and also their optimal strategy for economic success. Through history, a variety of evolving but pronounced behavioural assumptions have been allotted to this mythical character, homo economics. Amongst the more resilient of these is the expectation that investors’ preferences are stable and do not change over time. In fact this assumption extended to the belief that our past decisions, and the past decisions of others, bear no influence over our current decisions. This assumption maintained that the conventional investor’s preferences are not influenced by relatives and friends, social norms, or the media: in other words, that investors make decisions as if they were making them in a bubble.
Moreover, this convention held that investors, as economic actors, demonstrated the same preferences, specifically that they should be wealth maximizers, pre-programmed to want to maximize their well-being and level of happiness. Those rejecting options offering an increase in wealth or income were adjudged to be irrational. By extension, the representative agent model assumed that all members of a given household were identical and that one person’s preferences represented the preferences of everybody else in the household. In much the same way, the efficient market hypothesis was underpinned by the belief that investors have at their command and disposal all of the salient information related to financial assets, and are endowed with the capacity to process it in the execution of optimal investment decisions. The efficient market hypothesis purported to ensure that asset prices always reflect available information, mirroring the fundamentals of the real economy upon which such financial assets are based. These underlying assumptions were themselves predicated on the belief that people are able to forecast the future consequences of their actions. These assumptions have underpinned much of the conventional explanations around major market events, while also serving to distinguish between correlation and causation across financial markets.
DEBUNKING THE MYTH
Successive financial market crises have spawned fields of research and analysis that have revealed flaws in conventional or classic underlying assumptions. The associated empirical research has not only highlighted aspects of instability within the broader system of financial markets, but has also identified patterns of repeated investor behaviours and susceptibilities to error. This new research has drawn on a variety of diverse academic disciplines, including economics, demography, sociology, history and psychology. This research has typically not featured in mainstream market analysis, but has nonetheless proved insightful in explaining market events over the years.In their quest to explain previous market crashes, proponents of the new discipline, sometimes described as ‘behavioural finance’, have sought to expose the flaws of over-reliance and dependency on mathematical modelling to forecast and rationalise investor behaviour.
An example of one such recent challenge is related to research undertaken into the causes of the great financial crisis (GFC) of 2007-08. Until this point few policymakers or commentators sought to challenge the sanctity of the equilibrium theory that dominated macro economic analysis for the better part of a half century. However, the underlying assumption that the economic world was linear, mechanical and predictable proved flawed in analysing the market environment in the lead up to the GFC. In particular, formerly mainstream economic theory was accused of having diverted attention from the banking sector, more particularly its structure and its behaviour. The post-crisis analysis shone a light on the earlier work of Hyman Minsky, an economist once shunned by his peers, whose financial instability hypothesis accounted for the role of financial institutions’ willingness to loan on easy terms, betting on ever increasing asset prices; but then call in and withdraw their loans, even from financially robust businesses, once asset prices spiral downward, thereby permitting a liquidity crisis to expand into an economic crisis.
Increasingly the worlds of business, the media and academia have begun to lend further credence to aspects of behavioural finance. It may be argued that the increasing profile afforded to behavioural finance is due, in part, to the sustained damage that the GFC inflicted on the real economy. Or, it may be that our quest for explanations, in the hope of avoiding a further destructive near term shock, has further legitimised this body of work. Either way, many of the central tenets of behavioural finance have today moved a step closer to mainstream thinking, compared to thirty years ago. From the perspective of active investment management, a great many aspects of this discipline serve to highlight how irrational and at times dysfunctional investors ‘decisions can be, particularly in respect of short-term time horizons.
Throughout the ages, investors have demonstrated how their decisions can be influenced by the perceptions of popular investing culture. In the context of equity markets, it is perhaps instructive to review some of the diverse drivers attributed to sustaining recent stock market booms. History is littered with optimistic structural explanations and anecdotes validating sustained positive market performance, while often these explanations merely map market moves themselves. Between September 1953 and December 1955, the S&P 500 increased more than 90%, and media publications of the time featured a variety of persuasive explanations. One was that the bull market coincided with the advent of television, the new technological innovation of the day. Television set ownership by US families sky rocketed from 3% in 1948 to 76% by 1955. A previous technological innovation, the advent of radio in the 1920’s, had coincided with an earlier stock market boom. As in the 1920’s, commentators emphasised the longer term planning of US corporations as a reason to sustain performance, and fend off concerns about short term market fluctuations. The Baby Boom was highlighted as a particular structural driver supporting future anticipated consumption by new parents. As in the 1920’s, the increased use of consumer credit was highlighted as a further driver of future prosperity. In fact many of the same themes used by media and market commentators to explain stock market booms in the 1920’s, 50’s and 60’s,were even recycled in the 1990’s. Stock price changes feedback into corresponding changes in long term optimism, all of which perpetuate investor confidence and expectations of future market performance.
There are a myriad of persuasive explanations of the drivers or triggers behind stock market rallies. However, the effects of these causes can often be amplified or distorted by feedback loops. Depending on specific contexts, the past behaviour of prices can have a powerful range of emotional effects on investors. The psychological rationale for such behaviours is that our brains have evolved to cope with near term responses, while remaining conditioned to expect incremental, linear change. This is evidenced by our desire, as humans, to seek out equilibrium outcomes: we even promise them in our fairy-tales with their happily-ever-after endings. These characteristics leave investors ill equipped when the world turns out to be dynamic, unstable and unpredictable.
SHORT CUTS AND SHORT COMINGS
Investors’ decisions are ultimately impacted by their limited abilities to acquire and process vast quantities of information. Investors do not always have the time or resources to deliberate carefully the costs or benefits of investment decisions. To compensate, investors’ brains have sought ways to process and appraise the information which they see and hear. Unfortunately the brain does not boast boundless capacity. The brain is not a calculating machine, and as it has evolved it has compensated for its shortcomings. Investors have developed decision-making shortcuts, also known as heuristics, which are used to make more expeditious decisions in order to circumvent such limitations. The brain provides people with a means to develop and implement these heuristics, which are often mediated by emotion, fear, and intuition, and are fundamental to how investors engage indecision making. In essence the brain paints a picture from incomplete information, filling in the gaps, which can inevitably result in distortions or misjudgements. For many behavioural economists, the development and use of heuristics is not a good thing, because it overrides people’s deliberative capabilities and results in cognitive illusions or errors in perceptions. What people see is not always what is —and this can result in errors in decision making. Yet without the appropriate heuristics in place, people cannot be effective decision makers. While we all make mistakes, we can correct them — and very often do — courtesy of the deliberative part of the brain. Correcting mistakes takes time, though, and often requires education.
Prospect theory was developed as a descriptive theory for certain aspects of average decision making behaviour. Accordingly many people use heuristics when seeking to evaluate uncertain events, such as gains and losses. Because many investors are loss averse, these same individuals could easily turn out to be risk-seeking in respect of losses. Risk-seeking behaviour in this context refers to a situation where a certain, but smaller, loss is rejected by an investor in favour of accepting the high probability of suffering a greater loss, in the hope that the small chance of avoiding any loss will occur. The theory can explain why investors keep taking risks to avoid loss, holding on to losing investments for too long in the hope of recovering their losses. They are effectively pursuing a gamble that may generate more losses than a strictly rational approach would recommend.
Confirmation bias takes place as investors tend to apportion more importance to material or data which supports or validates their views, while placing less importance on information that contradicts or undermines their own views or opinions. They are biased towards confirming evidence. This raises very probing questions around, how, when and why investors are prepared to consider information which conflicts or undermines their previous beliefs or understandings.
Generalisations occur since investors often draw broad conclusions from a restricted or incomplete sample of the entire information set. By extension this can lead to investors apportioning greater importance or significance to facts that they have derived from the incomplete or a less reliable sample. They infer or extrapolate from the facts, assuming they are representative of the larger population. Investors can therefore draw erroneous conclusions from the data presented to them.
Herding takes place when investors begin to mimic the behaviour of others, especially when faced with highly uncertain investment outcomes. In such an environment investors typically follow the crowd, hoping the crowd knows more than they do. Investors faced with such uncertainty tend to follow people they believe to be experts, or figures held out by the media as being particularly astute or savvy investors. Herding is obviously not deliberative and calculating behaviour. When investors herd, they are breaking conventional decision-making rules. People often engage in herding when they lack adequate information to make an optimal decision on their own. There is so much uncertainty that following the crowd makes boundedly rational sense to many people.
FRAMING INVESTMENT DECISIONS
Within the context of behavioural finance, Daniel Kahneman and Amos Tversky advanced a theory concerning framing, according to which people are influenced in a choice by how it is presented. The human brain processes information and fills in certain voids based on past experiences and perceptions, before applying this to current circumstances. Frames can distort the choices investors make, thereby producing market failure or inefficiencies. Framing can therefore have a disproportionate role in influencing the investment choices of investors. Framing may be seen for example in the context of the specific institutions that support a particular investment universe. It may play an integral role in moderating, influencing and facilitating investor behaviour.
Information disseminated by trusted institutions is far more likely to be accepted and acted upon by investors. Democracy, freedom of the press, and an independent judiciary, are examples of institutions that may play a formative role in informing investors’ investment decisions, particularly in markets in which information is both imperfect and asymmetric, leading some people to be better informed than others. Such institutions may serve to provide investors with easy access to credible, reliable and timely information, lessening the likelihood of such investors being duped or misled, or merely ironing out structural inefficiencies at the market level.
Conversely, in the absence of such institutional pillars, market participants may be forced to invest in less efficient environments. Less scrupulous governments may protect inefficient firms from market forces by the imposition of subsidies and tariffs. Even if productivity is relatively low, inefficient firms and economies can survive by keeping wages low. This keeps the average costs of inefficient firms from rising, propping them up indefinitely and propagating further inefficiencies within the market. Alternatively those in power in such inefficient economies may very well prefer the status quo, since these economies serve as the source of their wealth and power, despite the harm it may cause the majority of the population. Many investors in these economies may not even comprehend the actual causes of their economic deprivation. Certain investors may not even be aware that they are economically disadvantaged.
THE ETF REVOLUTION
It is arguable that certain aspects of investor irrationality, set out above, are demonstrable by the impact of the evolution of exchange traded funds (ETFs) as an investment phenomenon.
The proliferation of new and diverse ETF products purporting to offer exposure to a myriad of diverse themes across the real economy is arguably because a significant proportion of investors have been disproportionately influenced by the allure of popular investment themes. Today more than US$4 trillion dollars is invested in ETF structures. Compelling structural explanations and anecdotes have underpinned effective marketing efforts to achieve impressive market share gains for these products over a relatively short period of time. In particular this has been around the benefits of achieving cost-effective benchmark beating or tracking performance for a broader investor universe. An example of such a popular investment theme is the recently launched ‘Make America Great Again’ ETF. This US-listed tracker will invest in companies with employees or activities highly linked to Republican candidates for election to senior positions.
Limitations of time and resources mean investors have finite capacity to comprehensively weigh up the risk/return benefits of certain investment decisions. Consequently some such decisions may be reached by means of a short cut, evidenced by an apparent over dependence on some of these structures that purport to offer instantaneous exposure to specific the matics in the market. For some time now investors have shown a preference to products offering dividend yield. However, an over dependence on ‘heuristics’ may lead some to overlook certain critical risk considerations. In this particular context, it should be highlighted how income can be volatile or less volatile, growing or stable, and it may be generated by companies across the market cap spectrum. Perhaps it is not just as simple as buying exposure to companies offering income. Such a short cut could easily lead to exposure to alternative factors, other than purely income.
Further such short cuts are likely to fail to differentiate between “backward-looking” and “forward-looking” income ETFs. The relevant underlying index may, for example, incorporate an historic dividend, which could be cut, or a forward-looking one which may not be attained. Different income ETFs could also expose investors to different biases. If you are basing the selection of stocks on a historic dividend, you may find your exposure also takes on a value bias. Alternatively if you base it on future dividends, you may find your exposure assumes more of a growth bias.
The growing presence and influence of the indexers, together with the use of benchmarks to measure active managers’ performance, arguably has an impact on all investors’ behaviour. As such it represents framing which, it may be argued, is serving to distort the wider investment universe. Today many active managers compare themselves to index funds, which clients regard as their prime competition. If the weight of the stock in the index goes up, it becomes riskier for the active manager to stay out of it; often these managers buy it and thereby add further support to the stock. As index buyers are price-insensitive they will tend to buy a stock according to its weight in their index, no matter how expensive it seems to them. It may be argued that this phenomenon undermines the concept of price discovery, the process by which a market should set a fair price for a stock, taking into account all salient information. Vanguard, the passive fund behemoth, owns at least a 5% stake in 491 stocks in the S&P 500, up from 116 in 2010. It now holds almost 7% of the entire index. It is estimated that soon about half of all US equity funds under management will be passive. As money flows into indexed mutual funds, or ETFs, an increasing share of those flows will inevitably go to those stocks that have risen in value. Misallocated capital can have negative results for everyone in the economy.
In the course of recent quarters, we have highlighted how the persistently low volatility in markets has not been reflected in the realised or implied volatility at a stock level; indeed how volatility had morphed into an asset class in its own right. Consequently the rotational structure of equity markets represented a more realistic gauge of the uncertainty than conventional volatility metrics. Throughout the course of 2017, for example, markets were roiled by a series of sector rotations in March, September and November. Early in 2018, a more significant sell off, this time based on fears of inflation and rising rates, gripped equity markets. Market rotations happen all the time, but it is rare they happen with such suddenness and ferocity, without any fundamental news to shift the investment narrative. While a myriad of explanations have been given for these movements, it would nonetheless appear that the influence of conventional equity investors, who buy or sell stocks based on stock specific fundamentals, is waning. These kinds of seemingly odd technical sell-offs and rallies appear to have coincided with the proliferation of algorithmic investment strategies, such as smart beta. The most recent sell off was attributed in part to those investment strategies and products married to market swings scrambling to offset positions in managed volatility offerings. Such episodes arguably resemble herding characteristics and have become more pronounced across markets in recent years, as market participants struggle ‘real time’ to understand the drivers behind such gyrations. By way of illustration, at the time of the Q2 2017 ‘tech tantrum’, the correlation of the ‘FAAMG’ stocks – Facebook, Apple, Amazon, Microsoft and Google – to the growth, volatility and momentum factors were in the 92nd, 90th and 96th percentile respectively. While a variety of minimum volatility ETFs sold off more than the market during the week of 5 February on concerns around the rising interest rate environment.
Finally, it is arguable that aspects of generalisation bias are evident through stakeholders’ treatment and assumptions regarding ETF products. As an example, the concentration risk that is prohibited, by regulation, to an active manager is, by contrast, considered reasonable and permissible if it happens to be an index. Ucits funds cannot have more than 40% exposure to position sizes in individual issuers of 5% or greater. In the course of 3Q17, the top five holdings in the Nasdaq 100 — Apple, Google, Microsoft, Amazon and Facebook — comprised over 41% of the value of the entire index. Yet, the Nasdaq 100 is available via the iShares Nasdaq 100 Ucits ETF, a product that has attracted more than US$1.1bn in assets under management in the UK alone. It is widely argued that a prolonged market correction could well reveal the extent to which investors in index funds have been giving insufficient attention to diversification, liquidity and risk control. Yet there is a widespread assumption that indexed funds look after these things for investors and ensure that they are immune from company-specific risk.
ADAPTING TO INVESTOR BEHAVIOUR
The OMGI Global Equities team employs an active investment process designed to exploit market inefficiencies by capitalising on the divergence of stock prices from their fundamental value due in part to investors’ behavioural biases. The investment process seeks to exploit these biases in a dynamic and unconstrained way, deploying multiple style exposures, aimed at removing emotion and subjectivity from the discipline of stock selection. There are a variety of ways in which the investment process takes account of such investor behaviours. For example, our Dynamic Valuation component has been calibrated to reflect divergence in investor risk appetite, across different regions, by dynamically weighting the exposure to those companies exhibiting the characteristics favoured by investors at times of divergent risk appetite. In the context of our Sustainable Growth and Company Management characteristics, the investment process will redeploy the risk budget in accordance with market sentiment and risk environment. The characteristics within Sustainable Growth take account of the fact that in a low creating shared value (CSV) environment, investors typically attach more significance to forecast information, thereby increasing the scope for us to exploit the mispricing inherent in growth companies. By contrast the Company Management component is designed to avail of mispricing in a higher CSV environment, or when market sentiment deteriorates. In this context, the qualitative characteristics within this component are designed to identify companies whose management teams have demonstrated a conservative approach to running the business, as well as an alignment with the minority shareholder. These bring attributes which investors tend to reward in such market environments. The Analyst Sentiment characteristic, for its part, is capable of adapting to market regimes whereby sentiment may inform the market, as well as those rarer market states whereby the instability and uncertainty inherent in markets typically means that it is the market informing sentiment. Finally, the Market Dynamics component is flexible enough to avail of investment opportunities in market regimes which typically reward winners and punish losers, while offsetting this approach with mean reversion techniques which tend to be more impactful in identifying mispricing opportunities in contrasting market states.