Introduction to Portfolio Manager Behavioural Analysis

What is investment behavioural analysis?

It usually starts from a richer set of data on what the portfolio manager has effectively traded.  These trades are used to build a complete database of the implemented investment strategies and the decisions behind them.  It then uses on one side the advances in the science of decision-making and behavioural analysis and on the other side the progress in data science and processing power of large data sets. The output is a suite of objective metrics providing insights on the skill used to achieve the performance, which also provide enough confidence in the ability to repeat a success.

Who uses Investment Behavioural Analysis?

1. Portfolio managers to obtain detailed and objective feedback on how well they are performing their duties and how to improve their own skills
2. Investment support functions to have the appropriate information to effectively help and support portfolio managers in making decisions and monitoring strategies
3. Senior management in asset management firms to build their confidence in the ability of the portfolio manager to deliver the required outcome and to empower them in the risk-taking activities
4. Fund distributors and investors to gather the information needed to build a trusted partnership with the portfolio manager throughout the investment journey.

What are portfolio managers expected to do?

• Conduct extensive research and data collection to find potential investment ideas with a good positive outcome
• Evaluate investment options and select those which they think have a strong probability of achieving good return
• To balance risk,  return and their level of conviction to decide how much to put into each investment.  In other words, to decide the size of each strategy
• To  build a coherent portfolio by putting together strategies that complement each other. In particular, to be mindful of how all strategies correlate to each other
• Decide when to buy, increase, reduce or when to sell a strategy

What is the objective of a behavioural analysis?

• To understand how these decisions are made
• Assess the quality of their outcomes
• Identify if behind these decisions there is any behavioural bias that may hinder the ability to achieve the full performance potential
• To improve the way these investment decisions are made
• To design a skills improvement plan through new investment discipline or a series of changes to portfolio manager behaviour

What are the two main approaches used to explore the behaviour of a portfolio manager?

1. Psychology approach: mainly uses psychometric tests
2. Data science approach: Infers behaviour from decisions already implemented in the past.

What is involved in the psychology approach?

In using the first approach we will ask the portfolio manager to fill in a psychometric questionnaire. The objective of the questions asked is to reveal personality traits which will suggest how the portfolio manager may behave when they make investment decisions. They are also designed to try to reveal any cognitive and behavioural biases that may hinder the quality of the investment decisions being made.

What are the potential drawbacks for the first approach?

1. They can be very invasive, digging into the personality of the portfolio manager and may reveal or simply discuss personal information
2. There are various ways to invest, and some behaviours simply work for certain strategies more than others. The questionnaire needs a minimum level of customisation to cope with the diversity of investment methods

What is involved in the data science approach?

The second approach for exploring these behavioural patterns starts from the premise that a portfolio manager has already faced different investment situations and contexts in the past. They have already provided answers to these situations through what strategies have been implemented and what has been effectively traded.

What we need to do is to get a history of these trades, infer the relevant investment decisions, link them to the context in which they have been made and record their performance outcomes. The result is a ledger of investment decisions.

We can then use tools driven from data science and all the advances in processing power to explore this ledger to find out how successful the decisions were, to identify if there are any behavioural biases hindering the portfolio manager and finally, to link those insights to the desired portfolio performance outcome.