What Do We Know About Robo-Advising?

Financial automation, or robo-advising, is a quickly growing industry. Trends such as WallStreetBets, the subreddit that led the GameStop surge, have sparked interest in investing for young audiences. Apps such as Robinhood are growing rapidly and utilize robo-advising as their core services. Steven Kou, Questrom Professor in Management and Professor of Finance, breaks down robo-advising so investors can understand its advantages and downplays.  

By Steven Kou

Robo-advising is a financial automation process that is revolutionizing household finance, not just in asset allocation, but also in consumer borrowing and lending, mortgage choices, consumption, and saving. Financial automation is inevitable; we should not exaggerate it, but we must embrace it.

Robo-advising first appeared in 2008 and aims to provide automatic financial advice to the general public, generally at low cost. There is a wide range of robo-advising, from domain-specific (e.g. asset allocation only) to holistic (e.g. including tax and insurance planning), from fully automatic to hybrid, from high-frequency trading to long-term retirement planning. Assets under management of robo-advising almost hit one trillion dollars in 2020, with major players including Vanguard, Fidelity, and Charles-Schwab. Financial apps such as Robinhood have also targeted the robo-advising field and appeal to younger audiences.

What are the differences between robo-advising and traditional financial advising? Although it is difficult to give a comprehensive answer, it is useful to focus on the asset allocation part of robo-advising to glimpse the differences. First, a robo-advisor must effectively elicit clients’ risk profiles based on simple inputs. A robo-advisor does not have the luxury of a traditional financial advisor, who can talk to a client personally for an extended period (e.g., one hour) and can ask many questions to engage the client.

Second, to convince the general public to use its services, a robo-advising system needs to provide recommendations consistent with conventional investment wisdom, as clients can easily change their inputs on the internet to test the system’s effectiveness.

Third, the advice of robo-advisors typically also contains outputs that may be helpful to educate investors in a more formal way. For example, the outputs may give guidelines on the dynamics of a portfolio strategy across time, the cross-sectional portfolio strategies for investors with different investment horizons, and the changes in the portfolio strategy for an investor, should the estimated asset returns vary.

Compared to traditional financial advising, robo-advising can reduce costs, mitigate human biases such as those due to psychology or moral hazard, implement trading strategies that are difficult for humans to do (e.g. high-frequency trading), adapt the latest algorithms, and, perhaps most importantly, lead to more financial inclusion. For example, there are already robo-advising algorithms designed to help low-income people to manage debts, payments, consumption, and investments on a unified platform.

The challenges ahead for robo-advising are manifold, most of which beseech us to consider interdisciplinary approaches, such as those from psychology (eliciting risk profiles, human-machine interaction), sociology (herd behavior of similar algorithms, network effects), and computer science (algorithm biases, privacy issues of individual financial data).

Finally, caveat emptor still applies when an investor uses robo-advising. First, robo-advisors hold the same legal status as human advisors, and in the United States must register with the Securities and Exchange Commission. Second, investors should be aware of how firms use robo-advising to make money from their customers. In addition to management fees and the interest earned on cash balances, the revenue may be generated from payments for order flow (rebates from brokerage firms that execute the trades) and from marketing targeted financial products to the customers. The latter two may lead to biased investment advice via clandestine financial incentives if the fiduciary duty of robo-advising is breached.

About Our Expert

STEVEN KOU
Questrom Professor of Management
Professor of Finance

Professor Steven Kou is an expert on FinTech and mathematical finance. His research on FinTech spans the following topics: robo-advising, econometrics with privacy preservation, the wisdom of the crowd, P2P financing, blockchains, and cryptocurrencies. In mathematical finance, he is well known for the double exponential jump diffusion model, the pricing of path-dependent options, risk measures, and credit risk modeling. Some of his research results have been implemented widely in commercial software packages and terminals.