The Future of Finance

Professors in the Finance Department in the Trulaske College of Business are publishing research that’s shaping the future of investing and trading. Here’s a look at four researchers and their findings.

By Kelsey Allen

Stocks Are Riskier Than You Think

Conventional wisdom is that stocks are a safe investment for the long run. But there’s a problem with this investing advice, says Michael O’Doherty, Charles Jones Russell Distinguished Professor in the Finance Department: The research behind it includes less than 100 years of U.S. return history. “There are known problems with studies that just rely on U.S. data,” O’Doherty says. “That data is widely available, and it’s really easy to get because the U.S. economy has done so well.”

But several developed countries have realized loss over a long investment period. For example, Japan’s stock market, once the world’s largest in terms of aggregate market capitalization, produced negative returns from 1990 to 2019. “Investors would have lost money in real terms or nominal terms,” O’Doherty says. “We have this suspicion that the U.S.-based evidence is overly optimistic, and there are some anecdotes that very strong, developed stock markets — Japan and others — had really poor performance over a really long time.”

So O’Doherty and his colleagues collected nearly 2,700 years of information about stock returns beginning in 1841 and extending to 2019 from 39 developed countries. They found that there’s a 12% chance that an investor is going to lose money over a 30-year horizon relative to inflation based on terms of the data from all developed markets. When only U.S. data is considered, there’s a 1.2% probability of a loss in buying power over the same period.

“The obvious implication is that stocks are riskier than you think,” O’Doherty says. “If you're thinking about a portfolio choice application, then investors likely should be less heavily weighted to stocks or should be positioning their portfolios in some safer asset classes. Some of the academic evidence that’s motivating how financial products are designed are based on U.S. history, so our findings would likely have some implications for the optimal allocation in those kinds of financial products, too.”

The paper, “Stocks for the Long Run? Evidence from a Broad Sample of Developed Markets,” is forthcoming in the Journal of Financial Economics.

Predicting Stock Returns from News Photos

A picture is worth a thousand words. But how much is it worth in the stock market?

“There are a lot of papers that show that textual sentiments from the news or even earnings conference calls or an annual report can impact the stock return at the market level and the firm level,” says Robert J. Trulaske, Jr. Professor of Finance Kuntara Pukthuanthong. “But no one has studied photos in the news and their impact on financial activity.”

Previously when researchers wanted to study sentiment information extracted from photos, they had to manually sift through the photos or rely on surveys and crowdsourcing websites like Amazon Mechanical Turk (MTurk). Instead, Pukthuanthong and Khaled Obaid, PhD ’19, an assistant professor in the Department of Accounting and Finance at California State University, East Bay applied machine learning to classify news photos based on negative sentiment to study how visual content in the news impacts financial markets.

“Our novelty is that we use machine learning, which allows us to go through millions of photos really easily and really efficiently,” Pukthuanthong says. The researchers applied a popular machine learning technique known as convolutional neural networks to accurately and cost-effectively classify millions of news photos from the WSJ Online Archive between September 2008 and September 2020. “Using machines to extract sentiment is less subject to bias,” she says. “If you use people from MTurk, for example, to classify the sentiment in the photo, that approach is more subjective because people have bias.”

Then they constructed an investor sentiment proxy, called Photo Pessimism, which is calculated as the proportion of photos predicted to be negative on a given day. They found when pessimism embedded in news photos is high, irrational investors will increase their demand for assets, driving up prices away from fundamentals. But then rational investors will take advantage of mispricing, leading prices to return to their fundamental levels. “Negative photos can induce negative returns,” Pukthuanthong says. “But the U.S. market is so efficient. We show the reversal within a week. The market corrects itself very quickly.”

Based on their results, the researchers propose a trading strategy that takes advantage of the reversal pattern. “Because we show that the market reverses — it corrects itself one or two days after the photo that shows pessimism is published online — we suggest investing in the S&P 500 three days after the Photo Pessimism index is above its historical mean and selling two days after that,” Pukthuanthong says.

The paper, “A Picture is Worth a Thousand Words: Measuring Investor Sentiment by Combining Machine Learning and Photos from News,” which is forthcoming in the Journal of Financial Economics, earned the researchers the 2019 Hillcrest Behavioral Finance Award. “This is a great behavioral paper showing an example of the framing bias,” commented Brian Bruce, CEO of Hillcrest Asset Management and editor of the Journal of Behavioral Finance. “It shows how the way information is displayed gives investors a biased perception of performance.”

How Fear Influences Investors

When someone feels fear, the most primitive parts of their brain take over and trigger the fight or flight response. Assistant Professor of Finance Michael Young found that when investors experience fear, they flee.

By studying terrorist activity in the U.S. over the past three decades, Young and his co-author show that an increase in the number of attacks each month leads to investors moving money directly out of equity funds and into government bond funds within the same fund families. Going from one attack to three in a month led to a $75.09 million drop in aggregate flows to equity funds and a $56.81 million increase to government bond funds. “They go from a riskier  investment into a safer investment,” Young says.

The researchers not only show that investors experience a change in risk aversion after a terrorist attack, but they also prove that negative emotions such as fear, anxiety, or depression — not changes in wealth or the outside environment — are what is driving their investment decisions.

“The vast majority of attacks in our sample are actually these localized events that aren’t going to have major impacts on the stock market or even on individual household wealth,” says Young, offering Earth Liberation Front actions and abortion clinic bombings as examples. “We find that this result even occurs when people see news about attacks overseas. After increases in attacks in G8 countries, excluding the U.S., we actually see outflows from domestic U.S. equity funds. Seeing that they reacted to news overseas was a really nice test for us to say that the shift in aggregate risk aversion is driven mainly by an emotional shock.”

The paper, “Terrorist Attacks and Investor Risk Preference: Evidence from Mutual Fund Flows,” was published in the Journal of Financial Economics. 

The Importance of Cushing, Oklahoma

A lot of finance researchers study crude oil prices and the relationship between spot and futures prices. Kate Holland, however, is interested in the physical component of crude oil storage.

Prior to becoming an assistant professor of finance at MU, Holland was working on her PhD at the University of Oklahoma, less than 100 miles away from Cushing, the world’s largest onshore oil storage and energy market hub. “I was driving through Cushing, and it’s full of these crude oil tanks,” Holland says. “Seeing it got me really interested: Where else do we have crude oil stored? Is Cushing special at all? It turns out it is.”

Crude oil is a commodity that is traded in both spot and futures markets. In October, a barrel might sell for $53 (spot), but in November, the contract might be trading for $54 a barrel (futures), and to store oil might only cost 40 cents a barrel. By buying the oil today at the spot price, paying to store it and selling it at the futures price, a trader can earn a 60-cent riskless profit per barrel — a strategy known as cash-and-carry arbitrage.

“We found in our research that if there’s a really wide spread between the prices, people will put oil into storage to take advantage of cash-and-carry arbitrage, and there is nothing unusual about that,” Holland says. “What was interesting is that we showed arbitrage-related crude oil inventory movements only happen in Cushing.”

While there are many other crude oil storage facilities in the U.S., Cushing is only the delivery point for New York Mercantile Exchange futures crude oil contracts. So for cash-and-carry arbitrage in the U.S. crude oil market to work, there needs to be enough storage capacity in Cushing.

When there is enough storage, these arbitrage-induced inventory movements are price stabilizing: Oil goes into storage in Cushing when prices are relatively low and comes out when prices are relatively high, moderating price swings.

But when there isn’t enough storage, the relationship between spot and futures is not kept in check. For example, in April 2020, there was a decline in crude oil prices, and the spot price dropped into the negatives. “So instead of you paying for crude oil, you were paid to take crude oil,” Holland says. “But there was nowhere to store the oil, so the spread between spot and futures was huge.”

And it takes 18 months to build a new tank. “Storage is a mechanism that allows the transmission of shocks from the futures prices to the spot prices,” Holland says.

The paper, “Dynamics of Arbitrage,” was published in the Journal of Financial and Quantitative Analysis.