Taking Greed Out of ‘Herd Mentality’: Part 1 of 2

Much has been said about the potential that parts of the asset markets within the US are experiencing a ‘bubble’ driven by ‘herd mentality’. Part of the driver of this herd mentality is often associated with the impact of greed on individual behaviour. 

Yet greed as a driver of individual behaviour is potentially superfluous to the dynamics within asset markets. To better understand why requires a closer look at ‘prospect theory’.

What is Prospect Theory?

Prospect theory, initially put forward by Daniel Kahneman and Amos Tversky in 1979, explores decision making under risk and uncertainty. The theory holds that when faced with alternatives that involve risk, probability and uncertainty decisions will be made based on perceived losses or gains. 

Two of the key outcomes of prospect theory are that individuals tend to (a) give more weight to the potential losses rather than gains made by taking a certain option (exhibit loss aversion) and (b) discount very small probabilities even if there is a possibility of a very high loss. The behavioural reaction function is stylised in Figure 1 where the utility from losses is materially greater from the level of utility generated by the equivalent capital gain.

Figure 1 :

Accordingly one of the key drivers of the level of risk that an individual will take is the expected loss associated with a decision. The greater the expected loss the less likely an individual is to buy into an asset.

Bringing Greed into The Equation

Where individuals are inherently risk averse to generate the dynamics associated with changing investor behaviour the concept of ‘greed’ will often be introduced to provide a balancing force. Greed refers to the ‘intense and selfish desire for something’. Once greed is introduced as a counter balance it now becomes its interaction with expected losses/gains which determines the actions of individuals and the dynamics of asset prices. 

Specifically as the level of greed alters over time the weight given to the utility of expected losses by individuals will vary thereby impacting upon the absolute level of risk aversion. Indeed at times the influence of greed may be so great that individuals completely ignore the expected losses associated with an asset as they focus purely on the expected gains. Once this occurs there is the foundation for what may be termed as ‘herd behaviour’ and a resulting asset price bubble.

Yet there is a potential issue with the introduction of greed as a balancing variable within individual decision making. Specifically the introduction of greed as an exogenous factor implies that there is no change in the individuals’s assessment of potential losses. Rather they are ‘down-weighted’ as the focus or weight of risk assessment shifts increasingly to the expected gain with the increase in greed. 

Given that the starting point of Prospect Theory is that the consideration of expected losses and gains in assessing investment decisions is logical then it follows that the failure to appropriately consider expected losses is illogical. Yet such behaviour is not viewed as the normal driver of asset prices which implies that individuals are only illogical at certain times. This requires that greed is accordingly not a normal state for individuals and is itself a variable which only impacts investment decisions periodically.

What arises is the potential contradiction that greed is on the one hand assumed to be an inherent part of individual psychology but on the other hand only impacts behaviour periodically as it is kept in check under normal conditions. So while greed within individual behaviour may exist as an explanatory variable it leaves a lot to be desired. This is not to say that greed does not exist as part of individual behaviour, rather that it is more appropriately thought of as a constant rather than a variable which will drive material changes in behaviour. 

Taking this train of thought one step further, if it can be shown that changing individual behaviour can be explained without the need for the dynamics resulting from the introduction of the concept of greed as an exogenous variable then, following Occam’s razor, greed can be excluded.

Formulation of Expected Losses

An important consideration in explaining individual dynamics when greed is a constant, is that the assessment of expect losses and gains are not constant but will change and be reassessed over time depending on the nature of information flows. The question then becomes how the incorporation of information flows in the assessment of expected losses and gains is undertaken by individuals? One answer is via the creation and impact on the psychological anchors utilised by individuals.

As outlined by Schiller information flows can be viewed as being used to form one of two types of psychological anchors.

The first are quantitative anchors which reflect some form of absolute or relative observable reference point. Such observable reference points can vary from the relatively simple such as past prices to the very complex such as detailed valuation models. The defining feature of quantitative anchors is the establishment of an ‘intrinsically right’ level of asset valuation against which expected losses and gains can be assessed. The ability of individuals to establish quantitative anchors will vary between asset classes. Indeed some asset classes may not possess the intrinsic characteristics which allow quantitative anchors to be established in the first place.

This brings investors to the second type of anchor formed by information flows namely qualitative anchors (note that Schiller refers to these information flows as moral anchors). Qualitative anchors are more nebulous and take the form of an intuitive force of stories and qualitative reasons to hold an asset. 

Put another way qualitative anchors comprise storytelling and justification. With qualitative anchors there is no quantitative dimension or ‘intrinsically right’ asset value rather individuals are effectively weighing a story/justification to assess expected losses. Though only qualitative in form qualitative anchors can be just as persuasive when driving individual behaviour particularly if the story/justification is simple and easily disseminated verbally.

Changing Assessment of Expected Losses

For individuals it is the interaction between these two psychological anchors which will determine the level of losses and gains expected when assessing an investment opportunity. Taking a closer look at the relative volatility between the two anchors it is reasonable to assume that quantitative anchors, by their nature, are relatively slow to change over time. This makes the qualitative anchor the more significant and volatile swing variable over the shorter term. The result is that at times these qualitative anchors, though only comprising stories, may seem more substantive and real than any quantitative concept. 

Indeed qualitative anchors may be the only anchor actually utilised by a material proportion of individuals. As the story driving the qualitative anchor becomes more broadly disseminated within the market the assessment of the risk of loss may decline materially. This is particularly the case where the price behaviour of the asset class provides a positive feedback loop thereby supporting the validity of the qualitative anchor. 

Importantly under this dynamic rather than individuals becoming more greedy the added weight given to the qualitative anchor is leading investors to become increasingly convinced that ‘they cannot lose’; i.e. materially underestimate the ‘expected loss’.

Diagrammatically the expected loss curve rises and in the extreme situation may become flat with the axis implying that individuals have reached the point where they do not believe that a loss is possible (see Figure 2).

Figure 2 :

With the underestimation of the expected loss individuals will be inclined to take greater risks. As this spreads across individuals the result is what is often referred to as herd behaviour (a more broadly based market wide underestimation of the risk of loss) and excessive risk taking. This shift towards excessive risk taking is also typically associated with a greater willingness to utilise leverage to amplify returns.

Greed is often used as a explanatory variable when trying to explain herd mentality and excessive risk taking. Yet as can be seen greed isn’t necessary to explain the phenomena of herd mentality and excessive risk taking. Rather the increased focus by individuals upon the qualitative anchor over the quantitative anchor can result in the systematic underestimation of the expected loss arising from an investment decision. 

As the story underlying the qualitative anchor becomes more broadly disseminated and incorporated into investment decisions the result is a rise in herd mentality. Though the distinction being made within prospect theory may appear minor it can have an important impact on expected asset market behaviour as the source of information flows evolves in the post information revolution world.