Finance in the Age of AI, Cryptocurrency, and Big Data

The transformation of finance over the past few years has been profound, as big data, artificial intelligence (AI), and cryptocurrency have merged to create a highly complex industry that continues to rapidly evolve. For Antoinette Schoar (Stewart C. Myers-Horn Family Professor of Finance; Professor, Finance), finding meaningful patterns within this structure is both a passion and a calling.

Schoar’s work reveals the extraordinary opportunities that are hidden inside this shifting mosaic: new finance models enable more entrepreneurs to launch companies, venture capitalists can lower their risk and increase their chance of success, and variations in cryptocurrency valuations by country reflect the value of trust in traditional institutions. When used wisely, data and AI can also help remove bias from hiring and other critical processes.

Along with these opportunities, however, come new challenges that will require society to engage in difficult conversations—and the outcomes will shape the future of finance.

Access to Credit in a Post-Virus World

Schoar believes that the sheer volume and detail of financial data now available to lending institutions has broad implications. In the past, for example, financial institutions found it difficult to lend to “subprime” populations because they were all considered relatively risky. As a result, institutions balanced their risk by providing an average interest rate to the entire population. Today, by pairing ubiquitous access to deep data with advanced algorithms, lenders can better differentiate “good” peers from “bad” peers. While this reduces the cost of capital for people considered a better credit risk, it puts those who are considered a worse risk at a disadvantage.

To make the challenge starker, while poor credit can be a reflection of unwise decisions by an individual or company, another major factor can play a significant role: bad luck.

Service businesses hit hard by the pandemic, for example, may suffer a poor credit report through no fault of their own. “When banks write AI targeting algorithms,” says Schoar, “they’ll have to be very explicit. If somebody has an unlucky shock—like getting sick—will we now reduce credit for them? This will be a really big issue.”

Opening New Doors for Investors and Startups Alike

In addition to facilitating the invention of new products, data-driven technologies are reducing the cost of starting a business. In the past, a new tech venture might have required extensive physical servers and other infrastructure, whereas today many of these fixed costs are now variable costs due to the proliferation of rentable cloud-based services.

For venture capitalists (VCs), the lower cost of starting a business is paired with changes that simplify how they can manage risk. Five or ten years ago, a typical VC would have spent months performing due diligence before investing in a startup. Today, VCs are more likely to make smaller investments in numerous startups, including some with similar product offerings, without performing laborious due diligence at the outset. To further simplify, VCs may make their up-front investments as “convertible notes” designed to automatically convert in the event of a future funding round. VCs are essentially letting market data prove out which startup becomes a success, spreading their risk to multiple potential entrepreneurs within a similar sector.

The impact of these shifts may be significant: Many new entrepreneurs will be women and minorities who in past years did not have access to capital on an equal footing with others.

Global Cryptocurrency Markets Seen Through a New Lens

As of 2019, more than 50 million investors were trading Bitcoin and other cryptocurrencies around the world. Working in collaboration with Igor Makarov, PhD ’06, Schoar conducted a systematic analysis of the trading and efficiency of cryptocurrency markets. Their analysis revealed wide differences in cryptocurrency valuation between markets, rooted in the foundational institutional dynamics and public sentiments in each country. For example, between December 2017 and February 2018, the daily average differential between the United States and Korea was greater than 15 percent and even reached 40 percent. What is the reason behind this “kimchi premium”? Much of the wide price disparities are based on limits to the mobility of capital between countries, which in turn limits arbitrage opportunities. Simply put, it’s not easy to buy cheaply in New York and sell high in Seoul. Deeper analysis showed that Bitcoin and other cryptocurrencies seem to have more value for citizens in countries where financial markets are underdeveloped. “The more trust there is in the financial system,” says Schoar, “the less valuable the cryptocurrency seems to be.

”It remains to be seen what the financial shocks of 2020 will mean for cryptocurrency markets, AI, and big data. One thing, however, is clear: the pace of change will likely continue to accelerate as entrepreneurs keep innovating.