Quantitative Investing Takes Away Human Emotion

Quantitative Investing Takes Away Human Emotion

Our way of investing and processing information at Copia can’t be carried out by a human being alone

Reviewed by: Evrin Erdem
Edited by: Evrin Erdem

The word “algorithm” is an increasingly popular term used in finance. Along with robo-advice and smart beta, it has captured the interest of industry pundits and journalists. But what really is an algorithm? And how and why is it used across the financial services sector?

Definition At Hand

The Oxford dictionary defines the word algorithm as “a process or set of rules to be followed in calculations or other problem solving operations, especially by a computer”.

In simpler terms, an algorithm is just a series of steps to perform a certain task. There are three main ingredients to an algorithm: inputs, a set of rules for the process to follow to perform the task at hand, and an output.

Considering that in essence an algorithm is a well-defined process of steps for accomplishing a certain task, it is easy to see how computers and algorithms go together like bread and butter. With the advent of computers, more and more sophisticated algorithms have been devised and become available in our daily lives.

But algorithms are not limited to the financial sector – we see them being used everywhere, whether it be a Google search algorithm, auto-matching couples on online dating sites, driving instructions from a GPRS device, online loan approvals based on the information supplied about an individual’s financial background, or even cooking a dish by following a recipe.

Applying Algorithms To Finance

For a financial adviser, the risk-rating process of a client is also an algorithm. The task is to assess the risk level that would be most suitable for that particular client. The inputs are the answers given by the client about his or her financial background and goals, time horizon for investment, preferences, and risk appetite.

Algorithms can crunch heaps of data in a fraction of a second. It is unsurprising, then, that algorithms are used extensively in finance ranging from the simplest to the very sophisticated. Quants and financial engineers have developed algorithms to predict price movements, develop arbitrage strategies with a view to exploit price discrepancies, price certain financial instruments, identify and mitigate various sources of financial risk in a plethora of various scenarios, match orders and route trades, for market making.



Understanding Quantitative Investment

Quantitative investing is an area of finance where by definition, mathematical algorithms are used to make investment decisions. By using a variety of mathematical techniques, vast amounts of data are crunched to look for patterns between price behaviour and other factors to predict asset returns.

One of the best known algorithms in finance is mean-variance portfolio optimisation, which won its inventor H. Markowitz a Nobel Prize in 1990. In layman terms, this can be described as “not putting all your eggs in one basket”. The objective (output) of this optimisation algorithm is the optimal percentage allocations of assets in a portfolio for a given level of risk tolerance, such that the expected overall portfolio return is maximised whilst the overall portfolio volatility is minimised. Constraints such as minimum/maximum holdings in a certain asset group and limits on portfolio turnover can also be added to this algorithm. The inputs to the mean-variance portfolio optimisation process are the expected returns of the assets, the correlations between them, each asset’s individual volatility and the risk tolerance factor.

At Copia we use the mean variance optimisation algorithm to determine the optimal asset allocations for each of our risk-rated model portfolios. The expected returns, which are one of the inputs of this algorithm, are also determined using a mathematical prediction algorithm.

Can A Human Analyse 500 Factors?

This prediction algorithm looks at 500 economic and market factors to determine which factors are affecting the returns of a particular asset most in that particular time period. Once those factors are identified, a mathematical prediction formula is found between those factors and the expected returns of the said asset. This is done for each potential asset in that portfolio.

No matter how great a portfolio manager may be, to mathematically number crunch through 500 factors for each particular asset is beyond the capability of any human. We have made use of computational power and mathematical tools to design a systematic, rules-based investment strategy. This takes away human biases and emotion in investing, takes away key man risk, thereby reducing costs and increasing scalability. The investment decisions are not made by star human portfolio managers who are expensive; instead the investment decisions are made by the algorithms that are built and rigorously tested by clever people. This makes it possible for us to bring value to clients by supplying solid asset management at a modest cost.


Evrin Erdem is head of investments at Copia Capital Management