Stronger Investment Theory Thinking Ahead Institute

Thinking Ahead Institute Stronger investment theory Embracing complexity We believe there is an intellectual clash on finance theory that requires attention. We also believe that the transmission of mainstream theory to mainstream practice needs attention. If theory guides practice, why do so many investment practitioners appear to disregard modern portfolio theory? Table of contents Stronger investment Introduction...... 2 theory The ‘mainstream’...... 3 The market at an equilibrium...... 3 Embracing complexity Representative agent with rational expectations...... 4 Efficient market hypothesis (EMH)...... 4 Modern portfolio theory...... 5 From mainstream theory to practice – a critical appraisal...... 7 Behavioural finance...... 10 Adaptive market hypothesis (AMH)...... 11 Rational beliefs equilibrium...... 12 Reflexivity...... 13 Delegation...... 14 Investment industry as a complex adaptive system...... 15 Complexity introduces multiple disciplines...... 19 A review of other theories through the lens of complexity...... 20 VUCA environment...... 21 Bibliography...... 22 Limitations of reliance...... 24

1 Stronger investment theory Introduction

In 2007, global financial markets defenders such as Professor Jeremy We believe there is an intellectual experienced the first signs of the Siegel at Wharton, who blamed the US clash on finance theory that requires Global Financial Crisis (GFC) in the US Federal Reserve for not cracking down attention. We also believe that the subprime lending market. The damage on the housing bubble (Siegel, 2009). transmission of mainstream theory to to economies and markets was so mainstream practice needs attention. profound and prolonged that even The Royal Swedish Academy of If theory guides practice, why do so today the healing process is far from Sciences fuelled the debate further many investment practitioners appear complete in many parts of the world. when it announced the winners to disregard modern portfolio theory? An over-leveraged Main Street, a of the 2013 Nobel Memorial Prize greedy and irresponsible Wall Street, in Economic Sciences: Lars Peter Our goal in writing this paper is, firstly, and exploding complexity in financial Hansen, known for his statistical to gain clarity on this intellectual markets matched by under-resourced breakthrough (the so-called debate by providing a synopsis of and under-effective regulators all ‘generalised method of moments’), mainstream finance theory and its joined with devastating results. Eugene Fama and Robert Shiller. challengers, including behavioural The interesting, yet puzzling, fact finance, adaptive market hypothesis, Not only did capital markets fail, is that Fama is widely recognised reflexivity, rational beliefs equilibrium but mainstream/prevailing finance as the father of the efficient market and a model that attributes the theory was seriously challenged in hypothesis (EMH), while Shiller source of market inefficiency to the GFC’s wake. Jeremy Grantham once famously stated, “The efficient delegation. We then put forward our used the subtitle, ‘The story so far: market hypothesis is one of the assertion that equilibrium economics greed + incompetence + a belief in most egregious errors in the history is really a special case of non- market efficiency = disaster’ in the of economic thought” (Wallace, equilibrium economics, as we believe first GMO quarterly letter for 2009 2010). Reacting to the news, Paul economies and financial markets are (Grantham, 2009). Martin Wolf, the De Grauwe, professor at the London complex adaptive systems. Rather chief economics commentator at the School of Economics, tweeted, ‘Nobel than treating financial markets as Financial Times, suggested that it was Prize for Fama who led millions to mechanistic systems that are always indeed the ‘mistaken ideas’ of market believe financial markets are efficient operating close to or at equilibrium, efficiency that ‘helped to bring the and for Shiller who showed opposite. we believe they are perpetually economy down’ (Wolf, 2009). However, What a contradiction.’ changing ecologies of beliefs, actions, prevailing finance theory did have strategies and cultures competing for survival.

2 willistowerswatson.com The ‘mainstream’

We define mainstream or prevailing finance theory as the theory taught It is worth stating that a stationary market means that at major universities and business statistical properties such as mean and variance that are schools. It is part of the core Chartered Financial Analyst (CFA) used to describe the movement in prices in that market curriculum and is what appears are constant over time; it does not mean that prices in major journals. Mainstream themselves are constant over time. theory forms the foundation of the investment industry participants’ knowledge base. It is important to note that the beliefs of investment Asset pricing models are a bit tricky. gravitational force that pulls markets practitioners can vary significantly We cannot avoid them completely back to equilibrium. It is worth stating from these theories, and in that sense because market efficiency and asset that a stationary market means that the ‘mainstream’ could alternatively be pricing models are inseparable statistical properties such as mean defined as ‘where the money is’. We twins. EMH comprises two distinct and variance that are used to describe refer to this as mainstream practice. hypotheses: (1) financial assets have the movement in prices in that market We will explore both in this paper, but fair prices that are determined by are constant over time; it does not start with theory. asset pricing theories, and (2) market mean that prices themselves are prices coincide with these fair prices. constant over time. Traditionally, many Laurence Siegel (Fabozzi et al., 2014) However, while the discussion of financial economists have held the suggests that the basics of investment various risk/return trade-offs (such as view that only the state of equilibrium finance theory can be distilled down the size of the equity risk premium) is interesting and worth interrogation. to the following eight ideas: is clearly very important, we will Many believe that unstable systems treat it as outside the scope of our do not generate many interesting Time value of money current paper. What we will address observations, a viewpoint possibly Discounted cash flow within asset pricing models are the encouraged by Paul Samuelson’s widely used but seriously challenged comment, “How many times has the Bond mathematics assumptions, which include (but are reader seen an egg standing upon its The no-arbitrage condition not limited to) the state of equilibrium end?” The mainstream finance theory Option pricing and the representative agent with is built on general equilibrium models rational expectations. Portfolio (the latest variation being dynamic Portfolio efficiency and optimisation optimisation/construction will also stochastic general equilibrium – Market efficiency receive some coverage given its DSGE) which support idealised Asset pricing models (such as prominence in almost all investors’ mathematical models and analytical the capital asset pricing model – processes. solutions, the so-called ‘physics envy’. CAPM) In a stationary world, bubbles should not exist, and financial crises ought to The market at an equilibrium Our review of this theory will mainly be extremely rare (that is, only caused focus on a very fundamental area – In short, the market at an equilibrium is by unexpected external forces such dynamic financial market behaviour, defined as the goods/services/assets as war). and the capture of information and clearing at a price (or set of prices) aggregation of individual participants’ at which the quantity supplied equals actions and interactions. This allows the quantity demanded. When viewed us to ignore the less controversial in this way, markets are deemed of Siegel’s ideas – the time value of stationary by nature and can only money. Market efficiency falls right be disturbed by large, unpredictable into our scope. and exogenous events. Even after these events occur, there is a strong

3 Stronger investment theory Financial markets, however, might Empirical evidence lends its support not be stationary at all. Financial Because the expected to EMH as interpreted by some crises are more commonplace than value of that random academics. Fama (1970) conducted exceptional according to Romer a survey of empirical work and (2013), who counts six distinct shocks forecast error is zero concluded that the results are strongly in the US markets during the past (that is what random in support of the weak form of market 30 years. Possibly one of the most means), from a efficiency and the evidence that well-known criticisms of general supports the semi-strong and strong equilibrium theory comes from John modelling perspective forms is extensive, and contradictory Maynard Keynes: “The long run is a all participants in the evidence is limited. A more recent misleading guide to current affairs. financial markets can be study by Dimson and Mussavian In the long run we are all dead. (2000) claims that there is strong Economists set themselves too easy, combined into a single support for the weak and semi-strong too useless a task if in tempestuous representative agent. forms, and even the strong form if seasons they can only tell us that the focus is on the performance of when the storm is past the ocean is professional investment managers. flat again.” Needless to say, these assumptions It is important to note that EMH is face some serious challenges. For simply a statement that security prices Representative agent with starters, Kenneth Arrow (2004) fully reflect available information and rational expectations points out the representative agent does not specify how markets are assumption is incompatible with supposed to reflect all information. Rational expectations require economic exchanges – why are As a result, tests of market efficiency participants in the financial markets to there both buyers and sellers of the are always joint tests of market have a perfect probabilistic knowledge same asset at the same time if you behaviour and models of asset prices of the future. Technically, that means and I are the same? A similar point (for example, CAPM), and rejection of the market or economy has only one was made by Keynes in his 1936 such tests does not necessarily mean equilibrium, and everyone in the market magnum opus, The General Theory that the market is not efficient. In fact, or economy forms expectations of Employment, Interest and Money, the joint hypothesis problem suggests around this unique equilibrium. The when he suggested that if all agents that it is never possible to disprove individual deviations from this perfect had identical expectations, market market efficiency. foresight can only be random and prices would fluctuate between zero offset each other in aggregate. and infinity.

The representative agent assumption EMH is simply a then follows as a result: because Efficient market hypothesis statement that security the expected value of that random EMH refers to the informational prices fully reflect forecast error is zero (that is what efficiency of financial markets. There random means), from a modelling are three forms of market efficiency. available information and perspective all participants in the The weak form claims that prices does not specify how financial markets can be combined of traded assets fully reflect the markets are supposed to into a single representative agent. information implicit in the sequence Because everyone has perfect of past prices, the semi-strong form reflect all information. knowledge about themselves, asserts that prices reflect all relevant everyone else and the market information that is publicly available, While some economists might view environment, an optimal investment while the strong form claims that EMH as the foundation of finance strategy based on deductive prices instantly reflect even private theory, the increasing evidence of reasoning is possible. (insider) information. ‘anomalies’1 suggests that EMH needs to be viewed critically.

1The use of the term ‘anomalies’ might shed light on the development of finance theories. In the hard sciences (for instance, physics), the discrepancy between reported results and what theories would predict will almost always lead to a process of theory revision – for example, the development of relativistic mechanics (physics for super-fast travelling items) and quantum mechanics (physics for super small items). Results that do not fit theory in finance, however, are often ignored or simply coined ‘anomalies’, making theory even harder to change.

4 willistowerswatson.com Grossman and Stiglitz (1980) High-profile practitioners have also costs, low management fees and so argue that strong-form efficiency joined the critical chorus. Warren forth). As a result, the choice between is impossible – it has to admit Buffett (1984) famously argued indexation and active management the possibility of minor market that the preponderance of value is driven by the beliefs about the inefficiencies to provide an incentive investors among the world’s best existence and potential of manager for investors to acquire and act on money managers rebuts the claim skill, the scale of the opportunities, the information. Shiller (1981) shows that of EMH proponents that luck is time preferences and risk aversion of price fluctuations in stock markets the reason some investors appear the investor, along with the expertise are too large (5-13 times) relative to more successful than others. EMH and incentive structure of the specific the variation in dividend payments, supporters also struggle to make manager. which could imply market inefficiency. sense of financial crises. Former However, this claim is not free of the Federal Reserve Chairman Paul Modern portfolio theory joint hypothesis issue (that is, it is Volcker claims it is “clear that a joint test of market efficiency and among the causes of the recent In 1952, Harry Markowitz published his validity of Shiller’s dividend model). financial crisis was an unjustified paper ‘Portfolio Selection’ in the Mehra and Prescott’s (1985) equity faith in rational expectations, market Journal of Finance, and that paper laid risk premium (ERP) puzzle also poses efficiencies and the techniques of the foundation for one of the most some serious challenges to EMH – modern finance.”2 influential investment theories of all theory-implied volatility (and therefore time: modern portfolio theory (MPT). ERP) is much lower than observed Richard Thaler (2009) elegantly Markowitz himself, however, never liked (ERP of 0.35% versus 7%). summarises EMH as having two main the word ‘modern’ and prefers to messages: (1) no free lunches, and (2) classify it simply as ‘portfolio theory’ Paul Samuelson formulated the well- the price is right. He argues that the (Markowitz, 1991). In a nutshell, it is a known hypothesis that: “...modern GFC should raise our respect for the framework to quantify risk and return markets show considerable micro no-free-lunch supporters given that and determine the trade-off between efficiency (for the reason that the not many investment strategies have them. MPT is all about diversification. minority who spot aberrations from successfully outperformed the market Key to the theory is the concept that micro efficiency can make money on a risk-adjusted basis. At the same any investment opportunity should be from those occurrences and, in doing time, the GFC provides ample evidence assessed by its contribution to the so, tend to wipe out any persistent that the price is not always right. overall risk/return profile – or efficiency inefficiencies). In no contradiction – of the whole portfolio instead of on to the previous sentence, I had It is important to note that EMH has its own merits. For example, MPT hypothesized considerable macro been refined over the past decades to breaks the risk of individual stock inefficiency, in the sense of long reflect the realism of the marketplace, returns into systematic and waves in the time series of aggregate including costly information, unsystematic components and indexes of security prices below transactions and financing costs. Ang proposes that the benefit of and above various definitions of et al. (2009) conduct an extensive diversification can be achieved by fundamental values.” Jung and Shiller review of the theoretical and empirical increasing the number of (not perfectly (2005) provide empirical evidence to literature concerning EMH and point correlated) stocks in the portfolio. The support this dictum. out that the most recent expressions purpose is to reduce and potentially of EMH allow a role for arbitrageurs in eliminate all unsystematic risks. the market who may profit from their Mean-variance optimisation (MVO) is comparative advantages (for example, the mathematical tool of selecting and specialised knowledge, lower trading choosing the proportions of various assets to be held in a portfolio (using their mean, variance and covariance) to achieve efficient diversification.

2http://www.nybooks.com/articles/archives/2011/nov/24/financial-reform-unfinished-business

5 Stronger investment theory James Montier (2012) MPT and MVO do not escape The risk-parity approach takes it even criticism on their idealistic underlying further – the weight of each asset in argues for abandoning assumptions: homogenous and the portfolio is inversely proportional the pursuit of the optimal, rational investors, perfect information, to its own variance. Constructing a given no one has the normally distributed asset returns, portfolio that way completely avoids constant correlations and many the need to estimate expected perfect foresight about others. The consequences are some returns or correlations. An equally the future: “Ex ante practical challenges in using the weighted portfolio (1/N portfolio) is the optimality is inherently theory. For example, this approach ultimate ‘naïve’ strategy that requires can be highly sensitive to estimation estimation of nothing. When there is fragile; it is only optimal errors (the so-called ‘garbage in, no need to estimate anything, nothing for your best guess of garbage out’ effect) and small can go wrong, but in the meantime, the future.” Instead, he changes in the inputs, particularly the diversification effect can still with regard to expected returns. In be achieved by pooling together calls for building a robust other words, it fails to recognise the not-perfectly-correlated assets in the portfolio fit for a wide difficulty (or impossibility) in accurately spirit of John Maynard Keynes’ remark forecasting expected returns and that “it is better to be roughly right range of possible future covariances. James Montier (2012) than precisely wrong”. In fact, using outcomes. argues for abandoning the pursuit data from 1978 to 2011, Andrew Ang of the optimal, given no one has the (2014) shows that both risk-parity perfect foresight about the future: “Ex and equal-weight strategies perform ante optimality is inherently fragile; it is better than the full mean-variance only optimal for your best guess of the procedure during that period. future.” Instead, he calls for building a Using data from 1978 robust portfolio fit for a wide range of Factor investing is one of the possible future outcomes. latest developments in building to 2011, Andrew Ang well-diversified portfolios. The focus (2014) shows that To address the issue of inevitable on asset classes can be too narrow both risk-parity and estimation errors, alternative and miss the rationale of why assets optimisation strategies have been earn returns and why seemingly equal-weight strategies proposed, and various asset uncorrelated assets moved side by perform better than management companies have side during the financial crisis. In his the full mean-variance launched products to employ these book, Ang (2014) proposes that factor strategies. For example, variances investing is closely associated with the procedure during that are believed to be much more concept of ‘bad times’. Investors each period. predictable than expected returns. have their own definition of bad times, High-frequency sampling, readily given their liabilities, income streams, available with today’s technology, constraints and beliefs. Investors’ can significantly enhance estimation wealth will take a hit in these bad times, accuracy for variance and correlation, and as a result they demand a risk although it does not help achieve premium for compensation. Examples Ang uses the analogy a better estimation of expected of factors which can be fundamental returns. A strategy aiming to build and economy-wide are the market (the of food and nutrients to the portfolio with the lowest overall beta in CAPM), inflation, economic encourage investors to variance bypasses the need to growth or volatility. Factors can also look beyond asset class estimate the expected returns (by be linked to tradable investment styles, assuming that they are all the same). such as value, growth or momentum. labels to underlying A single asset class can consist of factor risks. multiple factors. Ang uses the analogy of food and nutrients to encourage investors to look beyond asset class labels to underlying factor risks.

6 willistowerswatson.com From mainstream theory to Financial modelling is another example “In theory there is no practice – a critical appraisal of the impact of mainstream theory on practice. While we believe the difference between It is clear that mainstream investment financial models being developed and theory and practice. theory has found its way into the used in leading financial institutions In practice there is.” mainstream of investment practice, are somewhat more sophisticated for better or worse, despite some than a completely stationary process —Yogi Berra strong criticism of its assumptions and with a representative ‘super-intelligent’ mathematics. agent, there are nevertheless signs that modelling’s roots in equilibrium For example, EMH implies that thinking are too deep, and that one cannot consistently achieve they are struggling to live up to the (net-of-fee) returns in excess of speed of change in the real world, average market returns on a risk- particularly in times of crisis. A good “Practical men…are adjusted basis. In Fama’s 1991 study, example is the comment from David although some mutual funds achieved usually the slaves Viniar, the then-CFO of Goldman minor abnormal gross returns, pension Sachs, that “we were seeing things of some defunct funds underperformed passive that were 25-standard deviation economist.” benchmarks on a risk-adjusted basis, moves, several days in a row”.4 Given a result that is in line with Grossman that the likelihood of one 25-standard —John Maynard Keynes and Stiglitz (1980). As a result, an deviation event5 is comparable to the investor should only expect to make probability of winning the UK National a return equal to the market return Lottery 21 or 22 times in a row (Dowd and should thus focus on minimising et al., 2008), this is more likely a result costs. As such, an index fund would of bad models rather than bad luck. be the most appropriate investment vehicle, allowing the investor to Mainstream investment theory achieve the market rate of return in a also influences how the industry is cost-effective manner. Willis Towers regulated. Proponents of market Watson’s research shows that in 2013, efficiency believe that regulators of the assets managed by the world’s should not try to interfere with largest 500 asset managers (US$77 markets. Robert Lucas, who received trillion in total), over US$10 trillion the Nobel Memorial Prize in Economic aimed purely to mimic the movement Sciences in 1995, notes that the 3 of the market. main lesson from the EMH “for policymaking purposes is the futility The fact that active managers are the of trying to deal with crises and larger part of the investment universe recessions by finding central bankers should imply that mainstream theory and regulators who can identify and has not affected practice, because puncture bubbles. If these people managers believe that they can add exist, we will not be able to afford value by exploiting market inefficiency. them” (Lucas, 2009). Adopters of this However, while not strictly in line with viewpoint are blamed by many for EMH, the measurement of professional indirectly causing asset bubbles and managers has its origin in CAPM, as their subsequent inevitable bursts. shown by the industry-wide use of terms such as ‘alpha’ and ‘beta’.

3http://www.towerswatson.com/en-GB/Insights/IC-Types/Survey-Research-Results/2014/11/ The-worlds-500-largest-asset-managers-year-end-2013 4http://www.ft.com/cms/s/0/d2121cb6-49cb-11dc-9ffe-0000779fd2ac.html#axzz3n9eNbskO 5One 25-standard deviation event is expected to occur every 1.309e135 years. For comparison, the entire history of the universe is around 1.2-1.4e12 years, and a mere 8-sigma event should occur less than once in that period.

7 Stronger investment theory We believe practice can be improved by breaking a theory down into two parts – ‘if A’ and ‘then B’ – and then paying more attention to the former, which is arguably more important.

At the same time, the significant limits life. A meaningful theory always testability is high, the consequences of how mainstream investment theory asserts that certain forces are can be severe. One classic example is used in practice are evident. The more important than others is the role of mispriced structured UK’s National Employment Savings in understanding a particular products such as collateralised debt Trust stated: “Investment theory and phenomenon. The key criterion to obligations (CDOs) and credit default practice have evolved considerably judge whether a particular departure swaps in propagating the GFC. over the last fifty years. Despite from realism is acceptable or not this, there is no generally agreed should always be whether any Even worse, as Blommestein (2009) objective framework for investors that empirically meaningful predictions can suggests, academic economists adequately describes how to view be produced with these unrealistic can grow attached to their favourite capital markets, or how to apply these assumptions and whether there are theories and models and dogmatically insights for investment purposes.”6 We any alternative theories delivering bend reality to fit them. Einstein’s echo that by strongly suggesting that better predictions. mocking remark “If the facts don’t fit prevailing theory fails to represent and the theory, change the facts” is taken predict the world; it has proved to be That, however, can be very too literally in some cases. While of little practical use. challenging in the world of investing, there is increased understanding where inferential power from empirical of the limitations of current models, We do not think that too much methods is very low due to its high some academics dogmatically cling emphasis should be placed on level of noise (and due, in our opinion, to past success. We believe practice criticising assumptions that seem to its non-stationary nature). In can be improved by breaking a theory unrealistic. As Friedman (1953) wisely sharp contrast, financial economists down into two parts – ‘if A’ and ‘then pointed out, the assumptions of any and investors usually act as if the B’ – and then paying more attention theory cannot be thoroughly realistic inferential power is high. In other to the former, which is arguably more in the immediate descriptive sense. words, there is not enough proper important. The distinction should be made recognition of ambiguity in the system. between descriptive accuracy and The testability of any theory is low analytical relevance. The formula for and the empirical discussion is often gravitational force assumes a vacuum, highly selective. If practitioners use the pure form of which does not exist these theories and models as if the in the real world. And yet the formula is universally accepted and has far-reaching influences on everyday

6https://www.nestpensions.org.uk/schemeweb/NestWeb/includes/public/docs/statement-of-investment-principles,PDF.pdf

8 willistowerswatson.com Most theories completely ignore They suggest that in a ‘thin’ the influence of institutions or political process (in other words, “In the absence of clear ethics. Institutions should matter, self-interested corporate political guidance from existing as group decision making can activity influencing or even capturing analytical frameworks, depart from rational economic politicians and regulators without the assumptions. Institutional and balancing force of well-represented policy-makers had to governance interactions create and diverse views), it is both the place particular reliance even more dynamics that need to ethical duty and in the self-interest of be captured, or accommodated, by managers to preserve the legitimacy on our experience. theories. On the other hand, ethical of market capitalism, even at the Judgment and standards are crucial facilitators expense of financial profits. The thesis experience inevitably for successful financial markets, can be viewed as a challenger to and yet the entire set of investment Milton Friedman’s shareholder value played a key role.” theories is based on assumptions of maximisation ethos. ethically neutral behaviour. “Although —Jean-Claude Trichet, these financial theorists have made To summarise, the relevance of former president of the important progress in analysing the mainstream academic theory to European Central Bank consequences of important classes of practitioners has declined over time. incentive problems (adverse selection, Because of the fault lines in mainstream moral hazard, and principal-agent theory, the theoretical foundations for It is evident that not everyone agrees complications), they do not address practitioners have been very limited. on all tenets of the mainstream the full set of ex-ante and ex-post We, as an industry, have inevitably investment theory we have reviewed ethical transgressions: to lie, cheat, developed norms of decision making so far. The paper will now consider a steal, mislead, obfuscate, distort that are essentially backward-looking selection of alternative world views. and confuse” (Blommestein, 2009). and convention-based. Henderson and Ramanna (2013) further highlight the importance of ethics in market capitalism.

9 Stronger investment theory Behavioural finance

Behavioural economists do not believe Prospect theory (Kahneman and EMH supporters dismiss behavioural markets are perfectly efficient, and Tversky, 1979) uses cognitive finance as an alternative theory to attribute the imperfections in financial psychology to explain various EMH – stating that ‘people aren’t markets to a combination of cognitive divergences of economic decision rational’ is not a theory, rather biases (such as, overconfidence, making from neoclassical theory. merely an empirical observation. An overreaction, representative bias Behaviour is vulnerable to the alternative theory would need to and information bias) and various arbitrariness of psychological framing, offer an explanation, including causal other predictable human errors in and small changes in context can processes, underlying mechanisms understanding, reasoning and action. lead to profound differences in human and testable propositions. More Robert Shiller, widely recognised as behaviour (for example, people value importantly, behavioural finance keeps one of the founders of behavioural gains and losses differently). Using the analytical focus on individuals finance, seemed to be staging a prospect theory, one can explain and falls short of specifying an direct attack on EMH by claiming it is the excess volatility of stock market aggregating mechanism – markets a ‘half-truth’.7 returns by understanding that in aggregate can still stay error- and investors derive direct utility from bias-free if individuals’ errors and One can argue that Keynes’ ‘beauty fluctuations in the value of their wealth biases are random and offset each contest’ theory (1936) is highly (Barberis, Huang and Santos, 2001). other. We will attempt to address this behavioural in nature: “There are when we discuss complexity theory. some, I believe, who practice the Other contributions in the behavioural fourth, fifth and higher degrees.” finance area include the ‘house The implications of behavioural Investors in this framework are trying money effect’ – the tendency for finance on investment practice to buy into their predictions of the investors to take greater risks when are obvious. It provides legitimacy conventional valuation of assets in the investing with profits – which can for investment strategies to near future, not the true value, and this help explain bubbles in asset prices exploit mispricing opportunities describes how speculative markets (Thaler and Johnson, 1980). Investors’ due to behavioural factors. These function. ‘narrow framing’ (Barberis, Huang opportunities are not short term in and Thaler, 2006) – for example, nature, as behavioural economists mental accounting – can explain would argue that investors are other evidence against efficient hardwired to make the same mistakes markets. There is also evidence of over and over again, and cannot be a general human tendency towards arbitraged away completely due to the overconfidence that causes investors limits to arbitrage. Richard Thaler, a to trade too much (Odean, 2000) well-known behavioural economist, set and a tendency for people to anchor up his own investment management their opinions in ambiguous situations firm in 1998 to exploit these mispricing on arbitrary signals (Tversky and opportunities. Kahneman, 1974). Over the past two decades, the number of fund managers that integrate behavioural concepts in their investment process has steadily increased. Wallace (2010) points out that according to one estimate, half of the 200 listed small-cap value funds in the US use some form of behavioural finance concepts in building their portfolios.

7http://www.nytimes.com/2013/10/20/business/robert-shiller-a-skeptic-and-a-nobel-winner.html?_r=0

10 willistowerswatson.com Adaptive market hypothesis

Adaptive market hypothesis (AMH), are no longer suited to the new, proposed by Andrew Lo (2004 and ‘behavioural biases’ are observed. “The flopping of a 2011), is an attempt to reconcile Rather than labelling them ‘irrational’, fish on dry land may EMH with behavioural economics these behaviours should be labelled by applying the principles of as ‘maladaptive’ – “The flopping of a seem strange and evolution to financial interactions: fish on dry land may seem strange unproductive, but under competition, adaptation and natural and unproductive, but under water, water, the same motions selection. Lo suggests that much of the same motions are capable of what behavioural finance refers to propelling the fish away from its are capable of propelling as irrationality – for example, loss predators,” Andrew Lo (2004 the fish away from its aversion, overconfidence and mental and 2011.) predators.” accounting – is, in fact, consistent with an evolutionary model of individuals That leads to the conclusion adapting to a changing environment that market efficiency is highly via simple heuristics. context-dependent and dynamic. In Lo suggests AMH has several a highly competitive market where investment implications, including: The concept of ‘satisficing’ (or multiple market players chase bounded rationality) – first introduced scarce opportunities, the market The relationship between risk and by Simon (1955) – argues that is likely to be highly efficient (for reward is unlikely to be stable over individuals engage in making choices instance, the market for 10-year US time. that are merely satisfactory rather Treasury notes). If a small number of There are opportunities for than optimal, as humans are naturally market participants are competing arbitrage. limited in computational abilities. for abundant opportunities in a Investment strategies will perform This concept is integral to AMH. given market, that market will be well in certain environments and The hypothesis suggests that when less efficient. Efficient and irrational poorly in others. the environment changes and the markets are two extremes, neither of heuristics of the old environment which fully captures the state of the The primary objective is survival; market at any point in time. profit and utility maximisation are secondary. The key to survival is being adaptive.

11 Stronger investment theory Rational beliefs equilibrium

We now introduce a less well-known theory that argues financial markets Key to this theory is the concept of a stable process that are not as efficient as EMH supporters defines market movement (Kurz, 1999), as opposed to would suggest, for reasons that are not driven by behavioural factors. a stationary process widely embedded in mainstream theories. Key to this theory is the concept of a stable process that defines market movement (Kurz, 1999), as opposed to a stationary process widely embedded average P/E ratio for that index is 15. past data on which they are based. in mainstream theories. A stationary Under a stationary process this would However, it does not necessarily rule process is defined as a system in which represent a sell signal, as the P/E ratio out a societal belief structure and the all the joint probabilities of variables will necessarily fall or revert to the mean shift of that structure at the aggregate remain constant over time. The (albeit a short-term rise is possible). level. For example, the societal belief statistical properties of such a process Under a stable process, there is no structure can be either optimistic or can be discovered completely through such sell signal: (1) this could be the pessimistic and when one particular data crunching the past and as a start of an extended regime of higher- mood dominates the society, financial result agreed by all participants in that than-normal P/E ratio readings, and/ markets can indeed be stable, even system. For a stable process, empirical or (2) there is no certainty that the exhibiting the characteristics of an properties are also well defined, or observed historical average of 15 is equilibrium. Regime shifting manifests in other words, statistical analysis the true average, as it could drift to a itself when the societal belief can be successfully carried out just higher true average value. structure switches between the two. as for a stationary process. The key difference is that the true statistical The next term to explore is ‘rational In this framework, mistakes are properties of a stable process can beliefs’. M. Kurz chose the term allowed (mistakes here are defined as never be learned, even when endowed to contrast with the ‘rational the acts or thoughts that are wrong with all possible empirical information expectations’ of mainstream theory. on the ex-post basis but completely about the process. A stable process The complete information held by rational and unavoidable on an ex-ante can be characterised as a sequence all investors, as discussed above, basis according to their beliefs) and of regimes and, in practice, no one leaves only one (rational) probabilistic it is when these mistakes become knows exactly the parameters of the view of the future. In the rational highly correlated that we can observe prevailing regime, or its start and end beliefs framework, the assumption excess market volatility or bubbles dates. Let us look at a simple example of complete information is relaxed building up/bursting. where investors are witnessing a so that different investors have price/earnings (P/E) ratio of 20 for a different information (for example, The theory also allows room for pricing major equity index, while the all-time individuals may have access to data model uncertainty (PMU), referring to of various lengths or frequencies and the fact that there is a divergence of may interpret them differently). As opinion on how to price a new event investors know their information is (for example, quantitative easing). incomplete, they must form beliefs In other words, market participants to guide their decisions, and the only cannot agree on the ‘map’ from requirement of the theory is that exogenous variables to prices. This these beliefs must be rational, in the represents another source of market sense that they do not contradict the volatility.

12 willistowerswatson.com Another contribution of the theory is In recognising that markets can be that it makes a distinction between inefficient, Brock (2014) proposes two A rough estimation exogenous market risk (risk due ways of legitimately outperforming suggested by Brock to news about fundamentals, as markets: Shiller observed in his 1981 paper) (2014) is that one-third and endogenous market risk (the Use deductive logic to arrive at a of total volatility is arithmetic difference between deeper understanding of structural exogenously determined, total observed market volatility and changes than other investors do for volatility due to the news, a result the long-term trend. leaving two-thirds to of overshooting and difference in Exploit short-term endogenous risk endogenous risk. beliefs). A rough estimation suggested (excess volatility) by understanding by Brock (2014) is that one-third correlated mistakes and PMU. of total volatility is exogenously determined, leaving two-thirds to endogenous risk.

Reflexivity

We now focus on insights from feedback is self-reinforcing (similar to investment practitioners and Reflexivity can then correlated mistakes discussed earlier). review two investment managers’ connect any two or more Equilibrium is a possible end result for philosophies. George Soros (2008), a self-correcting negative feedback influenced by his tutor Karl Popper, aspects of reality, setting process. On the other hand, positive has been an active promoter of the up two-way feedback feedback cannot go on forever relevance of reflexivity to economics loops between them, because eventually Stein’s law8 will and financial markets. kick in when the gap between views which can be either and reality becomes unrealistically To unpack the concept, it helps to negative or positive. large. Such initially self-reinforcing but introduce two different functions eventually self-defeating boom/bust of one’s thinking: (1) the cognitive processes are perfect explanations of function to understand the world, world is always partial and distorted. financial crises. and (2) the manipulative function Confronted with a reality of extreme to change the situation to one’s complexity, we are obliged to resort The concept of reflexivity clearly advantage. In the cognitive function, to various methods of simplification presents some serious challenges a market participant’s views reflect – generalisations, dichotomies, to equilibrium theory. Equilibrium the reality, while in the manipulative metaphors, decision rules and moral should be viewed as a special case of function they shape the reality. When precepts. reality rather than the general truth. both functions operate at the same It is also very important to recognise time, they can interfere with each Consequently, while there is only that financial markets can affect the other, or at the very least cannot be one external reality, there will be fundamentals they are supposed to disentangled objectively due to two many different subjective views. reflect (especially when leverage is cause-and-effect flows. Reflexivity can then connect any two involved) – for example a market fall or more aspects of reality, setting up can create a negative wealth effect, The complexity of the world can well two-way feedback loops between reducing consumption and quoted exceed our capacity to comprehend it them, which can be either negative company earnings. Soros believes this (‘fallibility’, to use Soros’ term) and as or positive. Negative feedback is the point that behavioural finance is a result, the participant’s view of the is self-correcting while positive missing.

8“If something cannot go on forever, it will stop,” Herbert Stein.

13 Stronger investment theory Delegation

The last theory we review has its Vayanos and Woolley (2008) The model suggests that the principal/ origin in the work of Paul Woolley, argue that momentum, ‘the agent problem is the biggest one, another successful fund manager. premier unexplained anomaly’, therefore the solution lies in having The basic premise is relatively can be explained by this model. the principals recognise the nature straightforward: financial markets Trend-following strategies are a and the extent of the problems, and are inefficient and exploitative, as natural consequence of short-term then change the way they contract investors do not invest directly in incentives (for example, short-term and deal with agents. Woolley (2010) securities but through agents such as performance fees) and once has a long list of policies that he urges asset managers, creating a principal/ momentum becomes embedded in leading asset owners to adopt, some agent problem. The act of delegation markets, it is in managers’ interests to of which are rather radical, including creates a series of issues: asymmetric ride, and further reinforce, the trends. capping portfolio turnover at 30% information, misalignment of interests Asset owners contribute by firing or staying away completely from and asymmetric pay-off structures for underperforming managers and alternative investing. asset managers. investing in outperforming managers. The end results are mispricing and market inefficiency, along with rent capture, that lead to misallocation of both financial and human capital.

14 willistowerswatson.com Investment industry as a complex adaptive system

Having reviewed mainstream finance instead study the components of Q: Some say that while theory and five other theories/ the system and observe how the the 20th century was economic thoughts that challenge, system-level behaviour emerges supplement and, to some extent, from the interactions between those the century of physics, replace the mainstream, we components. A complex system is a we are now entering conclude that a coherent framework system where the whole is greater the century of biology. to understand and interpret the than the sum of the parts. Complex intellectual debate is needed. We systems exhibit emergent properties What do you think of believe that not only do we need to arising from the interactions of the this? embrace a new mindset that the world agents that cannot be deduced simply is not machine-like, but investors also by aggregating the properties of the A: I think the next century need a new set of tools to handle agents. more realistic descriptions of the will be the century of world. We postulate that financial markets complexity. are fascinating examples of complex We believe complexity science – an systems. There are many types of —Stephen Hawking emerging discipline originating at agents (investors, intermediaries and the Santa Fe Institute9 – offers a regulators) using existing technologies new way of thinking about finance. (the theory under discussion, Complexity science is the study of governance and regulation) who meet complex adaptive systems. Rather in markets (exchanging securities) and than impose on the system our marketplaces (for products and talent), conception of how it works, we can as illustrated in Figure 1 on page 16).

9http://www.santafe.edu

15 Stronger investment theory Figure 1. Investment industry as a complex adaptive system

Participants Technologies

Institutions and individuals Theory and practices Asset owners and asset managers Beliefs and norms Banks and shadow banks Systems and tools Culture and incentives Governance and regulation

Markets

Investment markets and marketplaces and their relationships to fundamentals

Each major market has hundreds, in financial markets, particularly in has created new asymmetries of if not thousands, of institutional respect of changes in regulation. knowledge and understanding for investors, and multiple thousands of A Google search on ‘unintended exploitation, as vividly illustrated by individual investors. The investors consequence’ and ‘financial markets’ Michael Lewis in his latest book titled span multiple markets, resulting in yielded 1,760,000 results.11 Flash Boys. interconnectedness of apparently separate systems. Despite complex Financial markets are full of examples We believe that ‘endogenously aggregate behaviours, there is of emergent phenomena. For instance, generated non-equilibrium’ is the underlying simplicity – people buy and bubbles and busts are not controlled natural state of financial markets. This sell a security. Transactions cause by any single actor or group. challenges the mainstream assumption a security’s price to change, which Consequently, they can be considered of equilibrium, that the market can only triggers subsequent transactions a behaviour of financial markets that be disturbed by exogenous shocks. and further price changes – typical emerges from multiple individual There is no proof that can be offered to of interaction. In addition, news flow interactions. Financial systems are demonstrate conclusively that financial can cause coupling when unrelated also adaptive and evolving. If we markets do not have an underlying parties trade on the same story. compare financial markets today equilibrium. Ultimately, it comes down Stock price charts over different with any previous point in time, we to an individual’s belief about how the lengths of time cannot be will see remarkable changes, and market operates. However, the clear distinguished, meaning that stock yet seldom have these changes exhibition of the other characteristics price movement is scale-free with appeared material at the time. The suggests to us that financial markets respect to time.10 Unintended underlying simplicity remains – the do not exist in equilibrium. It may consequences are very common buying and selling of securities – but appear in many situations that financial now the trades are in decimals, occur markets as a whole correspond in fractions of a second, and may not loosely to equilibrium, but beneath Despite complex happen on an exchange. The number the surface we also see mechanisms aggregate behaviours, and variety of securities available that produce forces and as a result have also ballooned, particularly cumulative changes which at a certain there is underlying in derivatives. This complexity point lead to a non-linear jump. In simplicity – people buy evolutionary biology, this is referred 12 and sell a security. to as ‘punctuated equilibrium’.

10This is an alternative expression of Benoit Mandelbrot’s idea (first formed in the 1960s) that stock market prices follow a fractal pattern. See for example http://www.scientificamerican.com/article/multifractals-explain-wall-street. 11As of 23 July 2015 12http://en.wikipedia.org/wiki/Punctuated_equilibrium

16 willistowerswatson.com Hyman Minsky’s financial instability sudoku puzzle. Counterparty risk hypothesis is a good example of is not just unknown; it is almost these forces in play – the financial unknowable”.14 system may look as though it is stable, or in equilibrium, but those are Another potentially much more precisely the conditions that cause important force is well illustrated in debt to increase, which creates future the Keynesian beauty contest and instability. Brian Arthur (2015) offers Soros’ principle of reflexivity: there a compelling analogy with the sun. It is no optimal move, and investors appears to be a large ball in uniform have to confront a genuine lack of spherical equilibrium when viewed knowledge by applying subjective from the earth. But underneath this beliefs about subjective beliefs, ‘equilibrium’, there are extremely and then apply third, fourth and higher dynamic phenomena such as gigantic degrees. Financial market participants magnetic loops, X-ray bright spots, who, via the manipulative function (to and mass plasma ejections moving at use Soros’ words), cannot only shape up to 2,000 kilometres per second. other participants’ subjective beliefs The sun is never at equilibrium. but also cause market fundamentals The sun is never at Neither are financial markets. to change. Uncertainty becomes equilibrium. Neither are self-reinforcing. That leads to financial markets. Complexity finance works with the historically contingent and assumption of heterogeneous agents path-dependent financial markets. that have imperfect information. Because prices of the financial In most cases, to some degree, instruments can affect the participants in the financial markets fundamentals that they are supposed Knowing the ultimate do not have complete knowledge. to reflect, the historical path of prices counterparty’s risk This might be caused by insufficient will influence its future trajectory. becomes, in the words of resources to process all the information regarding an extremely Where does this leave an investor Andrew Haldane, chief complex product. If an investor has who has to make an investment economist at the Bank of to read over one billion pages to decision? In a world of ‘we simply England, akin to “solving understand the make-up of a CDO2 don’t know’, the pure deductive contract fully,13 it is safe to assume logic of mainstream theory looks a high-dimension sudoku that these pages will not get read and increasingly anomalous. This is where puzzle. Counterparty risk investment decisions must be based complexity finance reaches out to is not just unknown; it is on an incomplete understanding. Or behavioural finance for help. Brian consider counterparty risk. In a highly Arthur (2015) argues that human almost unknowable”. connected banking system, each bank deductive rationality is bounded has a large number of counterparties (limited) and when faced with ill- and each of those counterparties defined situations with irreducible itself has a large number of uncertainty, people use inductive There is no optimal move, counterparties. Knowing the ultimate reasoning. They look for and match and investors have to counterparty’s risk becomes, in the patterns (for instance, it is common words of Andrew Haldane, chief to hear people saying “this financial confront a genuine lack economist at the Bank of England, crisis reminds me of the Great of knowledge by applying akin to “solving a high-dimension Depression”). It also means simplifying the problem by building temporary subjective beliefs about subjective beliefs, and then apply third, fourth 13This calculation is sourced from Andrew Haldane’s April 2009 speech at the Financial Student Association (http://www.bis.org/review/r090505e.pdf). CDO2 is used as a shorthand for and higher degrees. CDO of CDOs (typically 125 of them), which are themselves bundles of normally 150 RMBs (residential mortgage-backed securities). There are around 200 pages in a typical RMB prospectus. That is 22,000 years of reading, assuming one can read 50,000 pages a year. 14ibid

17 Stronger investment theory hypotheses to work with. Deductive reasoning can still be applied based on current hypotheses/beliefs (for example, given that I believe equities are overvalued, I should underweight them in my portfolio), which are constantly shaped by feedback from the markets and environments. As a result, hypotheses can be discarded when they cease to perform, and they can be replaced with new ones. In this sense, financial markets act as an ‘ongoing computation’ or ‘search engine’ looking for the next profitable opportunity. Financial markets become algorithmic. Arthur (2015) views the presence of both positive and negative feedback as a defining property of complex systems.

Another key concept is positive feedback, introduced and discussed earlier in both the rational beliefs equilibrium and reflexivity sections of this paper, particularly self-reinforcing asset price changes in financial markets. Arthur (2015) views the presence of both positive and negative feedback as a defining property of complex systems. Equilibrium theory is only compatible with negative feedback. However, if a system only contains positive feedback, the system will run away and show explosive behaviour. A ‘live’ financial market results from the interaction of both forces. When negative feedback becomes the dominating force, financial markets can converge to temporary equilibrium, and in that sense, the equilibrium-heavy mainstream finance theory is just a special case of complexity finance. Even so, bubbles and crashes are entirely possible when the positive feedback force is dominant.

18 willistowerswatson.com The transition from a negative-feedback-dominated With a better understanding of how agency and state (self-repairing) to a competition issues present themselves, game theory can positive-feedback-dominated state (self-destructing) can be highly then throw light on how solutions can be improved. nonlinear and as a result financial markets can exhibit a ‘robust yet fragile’ property. Only three years concept of the ‘interconnectedness of other active strategies. This is 16 prior to the GFC, Ben Bernanke of all things’. The United States simply a mathematical fact. With a celebrated over 20 years of ‘the great National Research Council defines better understanding of how agency moderation’,15 and yet since 2007 network science as ‘the study of and competition issues present the global financial system and world network representations of physical, themselves, game theory can then economy has been embroiled in an biological and social phenomena enlighten how solutions can be acute period of instability. This is leading to predictive models of these improved. Over time, competition can 17 also called regime shifting or phase phenomena’. The world economy and will evolve, and fitness landscapes transition. The tipping point is a key consists of highly interdependent will test organisational attributes of concept when regimes shift. Haldane commercial and financial networks, the market participants. That is the (2009) argues the shock that caused and as a result, network theory can research subject under evolutionary the tipping point in the GFC – the be a good framework to improve our theory. Within an evolutionary subprime crisis – was rather modest understanding in areas such as how framework, participants, technologies by global financial standards. organisations in the financial markets and markets interact and evolve Scientists and financial economists are connected or how markets are dynamically in accordance with suggest it is important to watch and could be structured. the ‘law’ of selection. In that sense, out for some early warning signs, sound investments need to be well although it is difficult to predict when Management science, also known adapted to the changing environment the tipping point will be reached. as operational research, explores instead of just having the highest For further reading in this area, we what organisations do and how they Sharpe or information ratios. Andrew suggest Kritzman and Li’s work (2010) work. With the help of mathematical Lo’s adaptive market hypothesis in financial turbulence. analyses and sometimes computer mentioned above is a good example of simulations, insights can illuminate establishing links between finance and management issues and solve evolutionary biology. Complexity introduces managerial problems that normally multiple disciplines manifest themselves in an In this lengthening list of relevant Complexity theory represents a environment where co-operation disciplines, we also include novel scientific approach across and competition co-exist. As a anthropology – the study of humans traditional discipline boundaries consequence, game theory can be and their behaviour, past and present. and creates opportunities for particularly helpful in understanding It builds on knowledge from the social research from multiple disciplines. the landscape where strategic and biological sciences as well as It can be viewed as an overarching decision making needs to be the humanities and physical sciences framework that describes the conducted. A good, simple example to understand human societies and participants, technologies and of applying game theoretical thinking cultures and their development. A markets, and their interaction. in financial markets is the zero-sum central concern of anthropologists Network theory, a multidisciplinary game nature of active investing. is the application of knowledge to academic field itself, is closely Gross of fees, active outperformance the solution of human problems. In linked to complexity and its central relative to the benchmark must the context of investing, it means exactly equal the underperformance understanding the self and others in the group and individual contexts, and understanding cultural dynamics.

15http://www.federalreserve.gov/BOARDDOCS/SPEECHES/2004/20040220/default.htm 16A phrase the author Douglas Adams put into the mouth of his fictional detective in Dirk Gently’s Holistic Detective Agency, UK, William Heinemann Ltd (1987) 17http://www.nap.edu/catalog/11516/network-science

19 Stronger investment theory Lastly, behavioural economics and neuroscience have a special focus Markets can be persistently inefficient due to human error on the cognitive process and how and cognitive biases. In our opinion, what is missing is the that process might produce certain biases. For example, a neuroscientist aggregation mechanism that turns individual actions into can trace the circuitry our brains use market behaviours. to make the kinds of decisions we rely on as investors. It sheds light on why investors choose to make certain investment decisions, in the context ‘search engines’ looking for the next Rational beliefs equilibrium theory of emotions like greed or fear. Why profitable opportunity, and are not produces a lot of sensible concepts do investors chronically buy high and efficient all the time. The concept of – rational beliefs, endogenous market sell low? Why is there more money market efficiency is not necessarily risk, pricing model uncertainty and chasing alpha than harvesting beta? wrong, but it is incomplete. Financial correlated mistakes, among other More importantly, by understanding models need to be built both from ideas. In our view, these concepts how the brain works, we hope to top down (DSGE) and bottom up tie in practitioners a lot more than overcome certain biases to get better (agent-based modelling, examined the mainstream theory counterparts. results. shortly). It is, however, still an equilibrium theory with its implied aggregation Behavioural finance provides great mechanism. The assumption of A review of other theories insight into how participants in the time-invariant mean and variance through the lens of complexity financial markets make individual (for the stable process) can be If by now we have successfully decisions. Markets can be persistently challenged. convinced you that financial markets inefficient due to human error and are complex adaptive systems, how cognitive biases. In our opinion, Reflexivity is an insightful concept, do we make sense of the ‘mainstream’ what is missing is the aggregation but the theory is still being developed. and ‘non-mainstream’ investment mechanism that turns individual As a result, most practitioners theories that we have previously actions into market behaviours. instinctively understand parts of reviewed? In this section, we try to do the theory, but rarely in ways that just that. AMH has so many complexity can be used in practice. It is a very elements in it that it seems reasonable important concept that is closely Unfortunately, but unsurprisingly, to categorise it as one of the related to interconnectedness and the mainstream theories are least complexity models. Market players path dependency, and is possibly the compatible with the complexity learn and adapt via simple heuristics. most important force driving financial worldview, indicating a sticky but, in Competition drives adaptation and market dynamics. our view, highly rewarding evolution innovation, and natural selection of beliefs. The belief in a natural state shapes the market ecosystem. The work on delegation also of equilibrium only subject to large Evolution determines market contributes significantly to our external shocks needs to be replaced dynamics. It is very in line with the understanding of one of the less by the concept of endogenously complexity view that market efficiency understood investment issues – how generated non-equilibrium with both is merely a special case of market asymmetric information, misalignment negative and positive feedback behaviour that is a function of market of interests and asymmetric pay-off forces in play. Investors cannot be ecology – a context-dependent and structures create market distortion, equipped with complete information dynamic concept. AMH has addressed and how these issues can be and make the optimal move all the the issue of underweighting the managed. However, the work’s narrow time. They differ in requirements, influence of the institution in focus on a single factor leaves it ability and constraints. With mainstream theories. It has picked unintegrated with other aspects of imperfect information, they employ up some elements from behavioural finance. Many practitioners support a combination of inductive and finance, but their integration remains a the thesis, but this driver’s impact deductive reasoning. Markets are work in progress. Overall, it is certainly remains a question. a plausible framework that works well in the complex adaptive system space.

20 willistowerswatson.com VUCA environment

Today’s investors are operating in This worldview will inevitably call for a an extraordinarily volatile, uncertain, step up in practice across the whole We urge investors, both complex and ambiguous (VUCA) spectrum of investment activities asset owners and asset environment. In stark contrast, the for all types of investors, as detailed theoretical foundation to guide in our companion research paper, managers, to recognise investors’ actions is unhelpfully weak, Going above and beyond: stronger and embrace the as we have detailed. So we urge investment theory and practice. complexity in our system investors, both asset owners and asset managers, to recognise and and move towards a embrace the complexity in our system complex adaptive system and move towards a complex adaptive worldview where more system worldview where more realistic phenomena can emerge. realistic phenomena can emerge.

21 Stronger investment theory Bibliography

Ang, A (2014). Asset Management: Dimson, E and M Mussavian (2000). Haldane, A (2009), “Rethinking A Systematic Approach to Factor “Market Efficiency”, The Current the financial network”, Speech at Investing, Oxford University Press State of Business Disciplines, 3, the Financial Student Association, pp. 959 – 970 Amsterdam, 28 April 2009 Ang, A, W Goetzmannand, S Schaefer (2009). Evaluation of Active Dowd, K, J Cotter, C Humphrey and Henderson, R and K Ramanna (2013). Management of the Norwegian M Woods (2008). How Unlucky Is “Managers and Market Capitalism”, Government Pension Fund – Global, 25-sigma? March 24 Harvard Business School Strategy December Unit Working Paper No. 13-075 Fabozzi, F, S Focardi and C Jonas Arrow, K (2004), Edited by D (2014). “Investment Management: A Jung, J and R Shiller (2005). Colander, R Holt, and J Barkley Science to Teach or an Art to Learn?” “Samuelson’s Dictum and the Stock Rosser. The Changing Face of CFA Institute Research Foundation Market”, Economic Inquiry, 3, issue 2, Economics: Conversations With pp. 221 – 228 Cutting-Edge Economists, The Fama, E (1970). “Efficient Capital University of Michigan Press, Ann, Markets: A Review of Theory and Kahneman, D, J Knetsch, and R Thaler p. 291 Empirical Work”, Journal of Finance, (1979). “Prospect Theory: An Analysis 25, pp. 383 – 417 of Decision Under Uncertainty”, Arthur, B (2015). “Complexity and the Econometrica 47: 236−291 Economy”, Oxford University Press Fama, E (1991). “Efficient Capital Markets II”, Journal of Finance, 46, Keynes, J (1936). “The General Theory Barberis, N, M Huang and R Thaler pp. 1575 – 1617 of Employment Interest and Money”, (2006). “Individual Preferences, London: MacMillan Monetary Gambles, and Stock Market Fama, E and K French (1992). “The Participation: A Case for Narrow Cross-Section of Expected Returns”, Kritzman, M and Y Li (2010). “Skulls, Framing”, American Economic Review Journal of Finance, 47, pp. 427-465. Financial Turbulence, and Risk 96(4):1069: 1090 Management”, Financial Analysts Friedman, M (1953). “Essays in Journal, 66(5): 30-41 Blommestein, H (2009). “The Financial positive economics: Part I – The Crisis as a Symbol of the Failure of Methodology of Positive Economics”, Kurz, M (1999). “Endogenous Academic Finance? (A Methodological University of Chicago Press Uncertainty and Rational Belief Digression)”, Journal of Financial Equilibrium: A Unified Theory of Transformation Grantham, J (2009). GMO Quarterly Market Volatility”, Stanford University letter, January 2009 Brock, W (2014). “Resolving the Lo, A (2004). “The Adaptive Market Market Efficiency Paradox”, SED Grossman, S and J Stiglitz (1980). Hypothesis: Market Efficiency From an profile March, number 125 “On the Impossibility of Informationally Evolutionary Perspective”, The Journal Efficient Markets”, American Economic of Portfolio Management, Vol. 30, Buffett, W (1984). “The Superinvestors Review, 70, pp. 393 – 408 No. 5: pp. 15 – 29 of Graham-and-Doddsville”, Hermes: the Columbia Business School Lo, A (2011). Adaptive Markets and Magazine: 4–15 the New World Order, December 2011, SSRN

Lucas, R (2009). In Defence of the Dismal Science, The Economist, 6 August

22 willistowerswatson.com Markowitz, H (1991). “Foundations of Thaler, R (2009). Markets can be Portfolio Theory”, Journal of Finance, wrong and the price is not always Volume 46, Issue 2 (June 1991), right, Financial Times, 4 August pp. 469 – 477 Thaler, R and E Johnson (1990). Mehra, R and E Prescott (1985). “Gambling With the House Money “The Equity Risk Premium: A Puzzle”, and Trying to Break Even: The Effects Journal of Monetary Economics, 15, of Prior Outcomes on Risky Choice”, pp. 145 – 161 Management Science, 36, 6: 643-660

Montier, James (2012). “The Flaws of Tversky, A, and D Kahneman (1974). Finance”, 65th CFA Institute Annual “Judgment under Uncertainty: Conference in Chicago (6 May 2012) Heuristics and Biases”, Science, 185: 1124−1131 Odean, T (2000). “Do Investors Trade Too Much?” Hersh Shefrin, Editor, Vayanos, D and P. Woolley (2008). Behavioral Finance, Northhampton “An Institutional Theory of Momentum MA: Edward Elgar and Reversal”, The Paul Woolley Centre for the Study of Capital Market Romer, D (2013). “Preventing the Next Dysfunctionality, Working Paper Catastrophe: Where Do We Stand?” Series No. 1 Presentation at the IMF conference “Rethinking Macro Policy II: First Wallace, C (2010). Using Behavioral Steps and Early Lessons” (9 May) Finance to Better Understand the Psychology of Investors, Institutional Shiller, R (1981). “Do Stock Prices Investor, 12 May Move Too Much to Be Justified by Subsequent Changes in Dividends?” Wolf, M (2009). How mistaken ideas American Economic Review, 71, helped to bring the economy down, pp. 421 – 436 Financial Times, 28 October

Simon, H (1982). “Models of Bounded Woolley, P (2010). “Why are financial Rationality”, Volumes 1 and 2. markets so inefficient and exploitative Cambridge, MA: MITPress – and a suggested remedy”, The Future of Finance, LSE, pp 121 – 143 Siegel, J (2009). Efficiency Market Theory and the Crisis, The Wall Street Journal, 27 October

Soros, G (2008). “The New Paradigm for Financial Markets – the Credit Crisis of 2008 and What It Means”, Public Affairs

23 Stronger investment theory Limitations of reliance

Limitations of reliance – Limitations of reliance – completeness of this data and Willis Thinking Ahead Group 2.0 Willis Towers Watson Towers Watson and its affiliates and their respective directors, officers and This document has been written Towers Watson Limited (Willis Towers employees accept no responsibility by members of the Thinking Ahead Watson) has prepared this material and will not be liable for any errors or Group 2.0. Their role is to identify and for general information purposes misrepresentations in the data made develop new investment thinking and only and it should not be considered by any third party. opportunities not naturally covered a substitute for specific professional under mainstream research. They advice. In particular, its contents are This material may not be reproduced seek to encourage new ways of not intended by Willis Towers Watson or distributed to any other party, seeing the investment environment in to be construed as the provision whether in whole or in part, without ways that add value to our clients. of investment, legal, accounting, Willis Towers Watson’s prior written tax or other professional advice or permission, except as may be required The contents of individual documents recommendations of any kind, or to by law. In the absence of our express are therefore more likely to be the form the basis of any decision to do written agreement to the contrary, opinions of the respective authors or to refrain from doing anything. As Willis Towers Watson and its affiliates rather than representing the formal such, this material should not be relied and their respective directors, view of the firm. upon for investment or other financial officers and employees accept no decisions and no such decisions responsibility and will not be liable should be taken on the basis of its for any consequences howsoever contents without seeking specific arising from any use of or reliance on advice. this material or the opinions we have expressed. This material is based on information available to Willis Towers Watson Towers Watson Limited is authorised at the date of this material and and regulated by the Financial takes no account of subsequent Conduct Authority – reference number developments after that date. In 432886. preparing this material we have relied upon data supplied to us by For more information, contact: third parties. Whilst reasonable care has been taken to gauge the Tim Hodgson reliability of this data, we provide +44 1737 284822 no guarantee as to the accuracy or [email protected]

24 willistowerswatson.com

About Willis Towers Watson

Willis Towers Watson (NASDAQ: WLTW) is a leading global advisory, broking and solutions company that helps clients around the world turn risk into a path for growth. With roots dating to 1828, Willis Towers Watson has 39,000 employees in more than 120 countries. We design and deliver solutions that manage risk, optimise benefits, cultivate talent, and expand the power of capital to protect and strengthen institutions and individuals. Our unique perspective allows us to see the critical intersections between talent, assets and ideas — the dynamic formula that drives business performance. Together, we unlock potential. Learn more at willistowerswatson.com.

Towers Watson is represented in the UK by Towers Watson Limited. The information in this publication is of general interest and guidance. Action should not be taken on the basis of any article without seeking specific advice. To unsubscribe, email [email protected] with the publication name as the subject and include your name, title and company address.