In a recent attempt at consolidating literature surrounding the concept of financial health, Brüggen et al. (2017) identied that existing literatures define financial health either objectively or subjectively. Objective methods include aspects such as debt level, income level and other methods of financial ratio analysis. Subjective methods are those that measure an individual’s own satisfaction with their financial state and therefore include more cognitive, emotional and psychological elements. The authors recommend that future research will adopt a composite definition. This recommendation has been supported in other research that explores that the multi-dimensional nature of financial well-being for individuals (Fiksenbaum, Marjanovic & Greenglass 2017; Joo 2008).

In practice, an accepted measure of financial health is a composite measure developed by the American Consumer Financial Protection Bureau (CFPB). The CFPB uses a questionnaire to rank an individual’s financial health on a scale of 0–100, with below 40 being poor financial health and greater than 60 being financially secure (CFPB 2019). The CFPB defines and measures financial health in four domains, with two objective and two subjective measures, namely:

The researcher accepts a composite definition of financial health as recommended by Brüggen et al. (2017) and specifically the definition as per the CFPB as a comparison in the ‘Literature review’ section.

Financial behaviours are described as a set of observable financial activities by economic agents (Mudzingiri et al. 2018). These everyday financial behaviours influence an individual’s financial health (Dew & Xiao 2011).

For example, higher household credit consumption increases the risk of loan default and reduces the likelihood of establishing retirement savings accounts (Dew & Xiao 2011).

In 2011, researchers Jeffery Dew and Jing Jian Xiao developed and validated the Financial Management Behaviour Scale (FMBS). The FMBS is a comprehensive measure of financial behaviours across multiple domains of financial behaviour, and is psychometrically validated (Dew & Xiao 2011). The FMBS categorises various financial activities into the following four financial behaviour factors:

This definition of financial behaviours is used in practice for the National Financial Capability Study (NFCS) for collecting large sets of financial data on individuals in America as they capture both short-term and long-term financial behaviours (Hilgert, Hogarth & Beverly 2003, as cited in Wagner & Walstad 2019). The researcher accepts the FMBS as the definition of financial behaviour as a comparison for this case study.

The purpose of this section is to review existing literature that provides empirical evidence of the effect of behavioural biases on the financial health of individuals.

Behavioural biases are our irrational behaviours that are neither random nor senseless, but rather systematic and predictable (Ariely 2008). There are two types of behavioural biases, cognitive biases and emotional biases. Emotional biases are systematic deviations from rationality as a result of impulse or intuition in decision-making. Cognitive biases, unlike emotional biases, are errors in reasoning. This distinction becomes important when determining whether a certain financial behaviour should undergo behavioural modification (Pompian 2011).

The process followed for analysing each bias was to first define the bias, then explain how it is measured, then identify which financial behaviours it impacts and finally, the result on financial health. This process is illustrated in Figure 1.

The biases reviewed in this section were chosen because of their prominence in three main publications that may be regarded as seminal in the behavioural finance field: ‘Judgement under uncertainty: Heuristics and biases’ (Tversky & Kahneman 1974), ‘Predictably Irrational’ (Ariely 2008), ‘Nudge’ (Thaler, Gintis & Sunstein 2008) and ‘Behavioural Finance and Wealth Management’ (Pompian 2011).

According to the theory of planned behaviour, an individual’s perception of the ease with which an outcome may be achieved influences their willingness to perform the behaviour (Azjen 2011). Thaler and Sunstein (2008) illustrate this by proving that savings levels increase when individuals are automatically enrolled in savings programmes.

The role of behaviour modification therefore, specifically with regards to financial behaviour, is to override the influence of behavioural biases on financial decisions to improve the consumer’s decision-making process and achieve the desired outcome of good financial health. The type of behavioural bias, cognitive or emotional, is an important distinction for the purposes of behavioural modification. Either the behavioural modification technique applied is direct (such as improving financial literacy or providing incentives) or indirect (such as nudges, defaults or incentives). Direct methods of behavioural modification tend to be appropriate for rectifying cognitive biases whilst indirect methods are used for overriding emotional biases (Pompian 2011).

The first method of direct behaviour modification evaluated in this research is financial literacy. A contributing factor to bounded rationality as outlined by Simon (1955) is the lack of awareness amongst decision makers with narrow focus on the decision at hand, as opposed to a holistic view point (Kahneman 2013). It is for this reason that improving financial literacy is key to reducing the impact of cognitive biases on decision making (Pompian 2011; Wagner & Walstad 2019).

Financial literacy is predominantly measured in three domains: numeracy, knowledge of inflation and knowledge of risk diversification (Wagner & Walstad 2019). Increased levels of financial literacy reduces cognitive biases and therefore positively impacts general financial behaviour (Özen & Ersoy 2019). However, financial literacy appears to be more effective in improving financial behaviours in developing countries rather than developed ones (Mudzingiri et al. 2018). For example, Cole and Shastry (2009) demonstrated that financial education had no impact on individual savings and investment behaviours in the United States. This was because of the already high levels of financial literacy amongst the general population relative to the rest of the developing world. Whereas within South Africa, financial literacy proves to play a greater role in improving the financial behaviours of individuals because of an apparent undervaluing of its importance amongst lower income households (Mudzingiri et al. 2018; Nanziri & Olckers 2019). Therefore, the role of financial literacy is important to improving the financial behaviours of South Africans and thus their financial health (Mudzingiri et al. 2018).

Incentives are a direct method of behaviour modification, which are intended to encourage desirable behaviour (Sherman 1970). They therefore aim to alter the individual’s conscious decision making (Bruns et al. 2018).

Whilst it may seem intuitive that higher incentives will lead to better behaviour, incentives have been applied to many areas of decision making with varying results. Gneezy et al. (2011) found that excessive rewards offered to employees may not necessarily result in increased performance depending on the nature of the tasks for which they are incentivised. The need for an incentive may indicate that a task is difficult, undesirable or that the individual is not naturally suited to it (Gneezy et al. 2011).

Holt and Laury (2005) show that increasing incentives tend to encourage greater loss aversion as individuals become increasingly weary of discrepant behaviours. Shin and Ariely (2004) show the close relationship between incentives to keep options viable and loss aversion. Schwartz et al. (2014) provided further evidence when they found that consumers changed their behaviour when they were faced with the risk of losing their incentives for healthy food purchases.

Nudges, a notion developed by Thaler et al. (2008), are tools that leverage behavioural insights to influence individual behaviour (Thaler et al. 2008). Furthermore, these tools must not limit the options nor significantly change the economic incentives of the individual (Bruns et al. 2018). Unlike incentives, nudges are indirect methods of behaviour modification. Because of this, nudges are also used as a form of soft regulation or alternative to economic intervention. For example, the Save More Tomorrow™ programme, which was a default on some employment contracts, modifies employees savings behaviour by making employers automatically save a portion of any future pay rise on behalf of the employees (Thaler & Benartzi 2004). However, setting it up as a default option, rather than an opt-in term, created a lack of transparency between employers and employees. It is for these reasons that nudges are criticised as being potentially unethical. They are therefore effective, yet potentially complex to implement in practice (Bruns et al. 2018).

As dicussed in the ‘Data and methodology’ section, the Discovery behavioural bank is largely incomparable to other banks as it is a pioneer in its field. In an attempt to compare the bank’s application of behavioural finance theories, the researchers considered other international financial institutions, including asset management firms.

Prior research on behavioural funds appears to be dominated by data on funds in the United States of America, where these funds are more common compared to other countries. This finding emphasises the pioneering nature of Discovery Bank as it applies behavioural finance in the South African banking sector.

Interestingly, many asset management firms which use behavioural finance principles have founders or principals who are also key contributors to the academic research on the topic. Discovery Bank has created its products informed by a similar approach to a research by Dan Ariely, a professor of psychology (Discovery Bank 2018a).

Furthermore, Discovery Bank’s application of behavioural finance principles appears to be less focused on arbitrage opportunities as a result of irrationality amongst market participants. Its focus is on the improvement of personal financial well-being of its customers by minimising their financially undesirable behaviours.

This section used existing literature to define the key terms of financial health and financial behaviour, identifying the behavioural determinants of financial health and methods of financial behaviour modification. The results of this section will be compared to the Discovery Bank business model to identify areas where the business model aligns, contradicts or is missing from behavioural finance theory.

This section explains the business model of Discovery Bank in the context of the behavioural finance theoretical framework.

Discovery Bank aims to improve its clients’ financial health. As such, the first step was to understand how financial health is defined by Discovery Bank. According to the interviewee:

This aligns with the definition per the CFPB’s 2017 report as well as research on financial behaviour by Fiksenbaum et al. (2017) as identified in the literature review.

However, the Vitality Money programme is structured to primarily deliver on one component of the definition, which is being resilient to financial shock, according to the interviewee. This is because Discovery Bank views the remaining three components, feeling in control, being on track and having freedom to make choices as being dependent on an individual’s financial resilience. Furthermore, financial resilience is an objective domain of financial health and is therefore easier to measure and interpret, and thus commercialise (Respondent 1, Discovery employee, July 2019).

This idea of the interdependence of components of financial health is not thoroughly explored in literature.

Therefore, this approach by Discovery Bank demonstrates how the definition of financial health per the CFPB is potentially dynamic in practice, and that the components are in fact dependent on each other.

Discovery Bank claims to change five financial behaviours: spending less than you earn, saving regularly, insuring for adverse events, paying off your property and investing for the long-term (Discovery Bank 2018b).

These behaviours were selected after analysis of very large amounts of client data. This use of data science ensures that the bank focuses their efforts on specific behaviours that require modification for improved financial health (Hore 2019).

Spending less than one earns is a financial activity under Factor 3 of the FMBS. Saving regularly is Factor 1 of the FMBS. Insuring for adverse events falls under Factor 2 of the FMBS. Paying off mortgage loans falls under Factor 4 of the FMBS whilst investing for the long-term falls under Factor 1 of the FMBS. From this analysis, it appears that Discovery Bank is using a narrower definition of financial health than that used by the FMBS. Furthermore, the order of the behaviours is also important to Discovery Bank as according to the interviewee:

Therefore, according to the Vitality Money programme, spending less than you earn is the most crucial financial behaviour to be improved. The reason being that:

The approach that Discovery Bank has taken in reducing the impact of behavioural biases on their clients’ personal finance decisions is to address a financial behaviour rather than a specific behavioural bias. According to the interviewee, there are multiple behavioural biases that impact one’s financial behaviour, and so by addressing a financial behaviour successfully, they are by default overriding multiple behavioural biases. This reasoning is corroborated by the literature review on the behavioural determinants of financial health that demonstrated how behavioural biases impact multiple financial behaviours (Barberis & Thaler 2003; Thaler & Sunstein 2008).

However, self-control and loss aversion are specifically targeted by Discovery Bank. For example, the Vitality Money programme may have a monthly goal of 20 card swipes for the client. If the client utilises savings in order to attain the swipes then all prior swipes are reset to zero. This makes the client spend more reasonably at the end of the month for fear of losing their swipes and not reaching the monthly goal, according to the interviewee. This use is consistent with behavioural finance research on self-control and loss aversion (Ariely 2008; Pompian 2011; Tversky & Kahneman 1991).

The Discovery Bank business model acknowledges that financial health is a function of the level of income as well as the manner in which it is spent and/or invested. According to the bank, financial awareness and education are critical for good financial behaviour (Discovery Bank 2018b). Discovery Bank increases their client’s financial awareness through the use of comparisons. For example, once an individual completes their registration with the bank, a comprehensive overview of the client’s financial situation is generated and a comparison is made to the rest of South Africans using various information providers such as credit bureaus, according to the interviewee.

This type of personal financial reporting is consistent with literature emphasising the importance of financial literacy and awareness in South Africa (Nanziri & Olckers 2019). Therefore, Discovery Bank plays a crucial role in improving the financial literacy of its customers (Pather & Bedford 2019).

Furthermore, the provision of this type of information is suggested by Thaler and Sunstein (2008) as it tangibly illustrates how investing translates to financial wellbeing.

The Vitality Money programme uses rewards and incentives to motivate behavioural change. There are three main incentives used: dynamic interest rates on savings and borrowings, dynamic discounts on certain purchases and Vitality active rewards, such as getting a new iPhone for meeting weekly financial goals (Discovery Bank 2018b). According to Hore (2019), these incentives are used because long-term goals require ‘nudges’ in the short term. The Vitality Money programme is built using the Vitality model which is an incentives-based programme and the current data collected by Discovery is through the use of incentives. Discovery’s use of incentives in its banking product range appears to have been expected before its launch (Joffe 2017) as this approach aligns to the broader group’s business model. Whilst the nature of the Vitality Money programme may lend itself to the use of incentives rather than nudges as behaviour modification tools, it is evident that the senior management of Discovery acknowledge the usefulness of ‘nudges’ in specific contexts. The bank is cognisant of the potential for ‘nudges’ to be viewed as ‘manipulations’ (Wilkinson 2013) because of the perceived intrusions on autonomy. Furthermore, nudges may not always be practical to implement (Respondent 1, Discovery employee, July 2019).

In addition to the incentives expounded above, the Vitality Money programme offers a budgeting function to address the financial behaviour of spending less than you earn and encourages the client to use and remain within budget.

An explanation as to why these deviations exist between the Discovery Bank business model and behavioural finance theory is that

It is therefore apparent that Discovery Bank will follow research literature to the extent that it remains commercially viable. It also demonstrates how behavioural change programmes that potentially work in an academic context may present practical challenges in their implementation.

The researchers noted that the views of Discovery’s employees were aligned with regards to how the company has used behavioural finance theories in its business model. There was consensus that Discovery Bank’s products and business model are data-based. They aim to change behaviours using incentives (Gore 2019; Hore 2019; Respondent 1, Discovery employee, July 2019; Discovery Bank 2018a, 2018b). This provided the initial indication that the company’s business model and vision have been well-communicated and understood amongst its key employees.