“A formulated problem is already half-solved” (Charles Kettering). Markus Hill* spoke for FONDSBOUTIQUEN.de with Harald Schnorrenberg, GET Capital, about the interrelationships of asset management, artificial intelligence, and the importance of risk and return estimators in the investment process. Additional topics of conversation included the use of mutual funds by institutional investors as well as Robo-advisory and field hockey.
Hill: Since when have you been more intensively involved with the topic of artificial intelligence and asset management?
Schnorrenberg: We have been involved with AI since 2009/10. At that time, many investors had to realize that classic portfolio management and risk approaches do not always work well in extreme situations and the subsequent recovery phases. We were looking for better solutions. Working with universities, we became aware of pattern recognition as a subfield of artificial intelligence. It was a long process. We have completely converted the investment process to AI and automated it since 2012.
Hill: Do you have the impression that investors have developed a strong interest in this area and have sufficient know-how here?
Schnorrenberg: We see growing interest as the topic becomes mainstream and various applications have arrived in real life. You see interest especially among the generation that is not yet in decision-making roles. AI is effective, but it is still very complex. However, if you understand the principles, you are already further along as an investor than most.
Hill: You have particular expertise in serving traditional institutional investors. You launched your mutual fund years ago. How did this idea come about?
Schnorrenberg: As with most quant approaches, AI is complex to understand and investors find it easier if they can test an investment live with small amounts. Mutual funds are a good format to do that. They offer more confidence than, say, model calculations.
Hill: What is the investment approach of your fund?
Schnorrenberg: Basically, our investment process is not very different from the usual approaches. The main difference is the way it is implemented. Intelligent algorithms, not portfolio managers, provide the trading signals, and they do so throughout. This process is completely automated. This means there is no manual intervention, only checks at points critical to success. This is Industry 4.0 in asset management. Asset management, in particular, is predestined for AI, because large volumes of data have to be analyzed, forecasts made and optimization carried out. Today, technology can do this much better than humans in many areas. It is very hard to be better than the benchmark inefficient markets anyway, especially by cost. As practice shows, only a few managers manage to do this consistently over several years. We do not rely on classic analyses, but therefore use synergies from Big Data, the Internet, and artificial intelligence algorithms to generate added value from the daily flood of data.
The investment strategy is determined by three main input factors:
1. risk estimators of the relevant markets
2. return estimators of the relevant markets and
3. optimization constraints (such as a risk budget of x%).
The risk and return estimators are influenced by market changes and adjust themselves. With this adjustment, the allocation of the portfolio can also be changed depending on the market change. Likewise, the risk constraint can trigger a change in the allocation, depending on the utilization.
The investment decision is derived in three steps:
1. market analysis: via regressors and classifiers (AI methods of machine learning), about 8,000 securities of the relevant markets are checked daily to see if they are unlocked for investment and if so, which are the most interesting markets in terms of an absolute return estimator.
2. portfolio construction: In this process step, individual restrictions (for example, the risk budget) and the defined overall universe are introduced into the process.
3. portfolio optimization: Here, the relevant markets from the defined overall universe flow together with the return/risk estimators and the portfolio restrictions. An optimal portfolio allocation with all constraints and transaction costs is determined. Our return estimator logic forms the basis for market estimates. The regime-based return estimators are calculated using a pattern recognition process and their intelligent evaluation based on historical data.
Hill: How do you estimate the importance of mutual funds for institutional investors, which are launched by fund boutiques?
Schnorrenberg: You have to ask the investors that. As described, these offer a good opportunity for a “test drive,” that is, to test strategies under real conditions. Investors reflect on us that the ability to invest in mutual funds quickly, transparently, and cost-effectively is helpful.
Hill: Where can your approach be found in addition to the traditional institutional business?
Schnorrenberg: Institutional investors are our core clientele. However, due to the flexibility and scalability of the platform, it also acts as an engine for Sparkasse Bremen’s Robo-advisor: Smavesto. Incidentally, Smavesto won the Handelsblatt comparison of 25 Robo-advisors in November 2020 with a 10.2% return over 1 year by a clear margin.
Hill: What trends are you currently observing in your customer segment?
Schnorrenberg: Zero interest rates are a major problem. Here, people are desperately looking for alternatives in all conceivable formats and forms. This requires a considerable build-up of know-how because simple solutions no longer exist. Thus, capital management is becoming more strenuous. We are seeing strong demand for long/short products underpinned by our technology.
Hill: What topics do you deal with when you want to get a “fresh head”?
Schnorrenberg: As you know, the world is not just about asset management. Sports are a good balance for me. I enjoy playing field hockey and tennis.
Hill: Thank you very much for the interview.
GET Capital AG: www.get-capital.de
*) Markus Hill is an independent asset management consultant in Frankfurt am Main.
- Real Assets, Asset Management Consultants, Mutual Funds, Borussia Dortmund & “A Brief History of Humanity” (Interview – Martin Dürr, FAROS Fiduciary Management AG)
- Quantity & ESG – singularity theory, fund selection, seed money & “mirror of light” (Dr. Oliver Klehn, interview & literature reference)
- Value Investing, Aristoteles, Fund Selection & Ownership Approach (Interview & Webinar, 23.4.2020 – Dominikus Wagner)
- Value Investing, “Value versus Price” and the Importance of Family Offices as Investors (Interview & Webinar, 7.5.2020 – Stefan Rehder)