Skip navigation Scroll to top
Scroll to top

Are investment algorithms crisis-proof?

October 15, 2021

reading time: 4 minutes

by Tilmann Schaal, LGT

Expert for Investment-Algorithms: Andreas Vetsch, LGT Capital Partners

What role can systematic trading strategies play in your portfolio? Analyst Andreas Vetsch has the answer – and explores possible risks and knock-on effects.

In almost no other sector is IT as important as in the financial sector. Whether as a support for fund managers, when providing financial advice or when investing directly using automated strategies: it’s likely that investors are using algorithms, no matter how or where they invest.

What are the advantages and disadvantages of automation – especially in the event of a sudden market correction? Research Analyst Andreas Vetsch, one of the experts in this area at LGT Capital Partners, answers this and other questions.

How widespread are algorithms in financial markets today?

The use of algorithms (algos) has become widespread in the industry. They take emotions out of the trading process, allow for strategy back-testing and offer advantages in terms of implementation, particularly when operated at scale.

There are two main types of algos. The first are investing algorithms, which can be extremely useful in identifying investment opportunities across many securities in a systematic manner. The second are trading algorithms, which are used to enhance trading efficiency through automation.

While actively managed portfolios are still primarily managed by discretionary managers, the share of systematic approaches in asset management is growing. An increasing number of managers use algorithms to apply a more repeatable and data-driven investment approach. When it comes to trading, algorithmic execution already accounts for over 80% in liquid futures, large cap developed market equities and credit default swap indices. In recent years, electronic trading has also grown significantly in markets such as corporate bonds.

Can algorithms react independently to large, unforeseen price fluctuations or does this require human intervention?

It’s important to make the distinction between model types that anticipate and model types that react. There are scenarios where anticipation models that rely on behavioral heuristics could comfortably navigate seemingly “irrational” market dynamics. Similarly, models that react to market behavior by utilizing timely, accurate and irrefutable information are designed to operate well in a myriad of unseen scenarios.

Investment-Algorithmen
Research Analyst Andreas Vetsch: "It’s likely that investors are using algorithms, no matter how or where they invest."

On the flip side, highly fitted, static or slow models, or models calibrated only on historical patterns may struggle with regime changes and events that have never occurred before. They can give a false sense of security and may be wrong-footed or completely malfunction during events or times that are driven by unprecedented factors or circumstances.

Historic data cannot, therefore, be extrapolated to the future if the underlying fundamentals of these data are not well understood, recognized and taken into account. Furthermore, systems or algorithms can play a useful role in trading if, and only if, it remains clear that the humans who are behind these systems and algorithms are responsible. This responsibility encompasses the recognition that market regimes can change, and that systems that are designed for one regime may not be effective in another regime.

Will investment algorithms do better on these singular market events in the future?

During the Covid-19 market crash, we saw large performance differences between some segments of the quant industry. Directional short-term strategies utilizing timely information generally did well at the expense of slower, less reactive and more backward-looking strategies. Behind every algorithm is a model and some models will handle market crises better than others, depending on their design characteristics. This is comparable to a self-driving car that is programmed to drive on highways and may, therefore, not function if it ends up on winding mountain roads.

That is why it is key to distinguish between different systematic strategies and understand what each one can and cannot provide. If investors understand the nuances of different investment algorithms and select more crisis-proven approaches, they will be able to better navigate through “irrational” market dynamics going forward.

Information and insights

Andreas Vetsch is a research analyst at LGT Capital Partners and in this role, he also observes digital technologies. He and his colleagues in Asset Management look for attractive investment opportunities and identify the best portfolio managers. They also manage a substantial portion of the assets of LGT’s owner, the Princely House of Liechtenstein.

If you are interested in the latest global market and economic developments, we recommend you read the insights provided by our research experts of LGT Private Banking.

More from the LGT Online Magazine?

Enter your E-mail address and get regular newsletter updated of the LGT Online Magazine.
Subscribe to newsletter