Imagine there’s a bucket filled with red and blue balls.*
You have a choice: You can have $50, no strings attached. Or if you can blindly pick a blue ball out of the bucket you’ll get $100.
Half of the balls in the bucket are blue. No cheating – you don’t get to peek. There’s just a 50-50 chance of getting that Carolina blue ball.
Given this choice, most people choose the $50 certainty over the 50-50 chance at something better. It’s called risk aversion.
While scholars have long studied how risk aversion affects decision making, there’s another kind of risk aversion that researchers have only started to dig into in the last few years – aversion to “Knightian uncertainty” or ambiguity. In ambiguous situations, the probabilities of different outcomes are unknown – just like many real-life situations.
Now imagine there are two buckets – A and B – and both contain red and blue balls. Bucket A has 50 red balls and 50 blue balls; bucket B has 100 red and blue balls, but you don’t know the quantity of each color. You can choose either color as the “winning color” and choose a ball from either bucket. If you blindly pick a ball that’s your “winning color” out of either bucket you’ll get $100.
What would you do when faced with a 50-50 probability of getting the $100 or an unknown probability of getting $100? Most people will choose bucket A, preferring a risk they know rather than the unknown risk – the ambiguity – of bucket B. That’s ambiguity aversion.
Paolo Fulghieri, Macon G. Patton Distinguished Professor and area chair of finance at UNC Kenan-Flagler, has been studying the impact of ambiguity aversion on decision making, with his colleague David Dicks, assistant professor of finance.
“In the last 10 to 15 years, academics have started to think more carefully about the fact that when people make decisions they might face uncertainty instead of risk,” Fulghieri says. “Once you account for this, you start to get a better understanding of people's behavior.”
That distinction between risk and uncertainty – and the tendency of people to avoid ambiguity – has significant implications for corporate finance and other fields. Fulghieri and Dicks have written three papers that propose mathematical models about how ambiguity aversion affects decision making. They focus on innovation, financial system risk and corporate governance.
Innovation, entrepreneurship and uncertainty
Innovation – a new drug or the next Facebook – is one of the most obvious examples of ambiguity. Predicting the odds of success for new products is tough for both entrepreneurs and potential investors, who must decide whether to fund innovative projects based on very limited information about potential returns.
Fulghieri’s and Dicks’ model predicts that uncertainty-averse investors will be more optimistic overall when they can make other investments in other innovative ventures – similar to the notion of diversification when trying to manage risk. Thus investments in innovative businesses tend to lead to other such investments, creating “innovation waves.” As this happens, more entrepreneurs – seeing the opportunity for investment – are willing to put their own resources into an uncertain, but innovative venture.
The result? We see waves of innovation and entrepreneurship and high market values, all the result of how investors and entrepreneurs respond to uncertainty.
Implications for financial system risk
Ambiguity aversion also has interesting – and unexpected – implications for bank and financial system risk.
First, their model illustrates how negative sentiment can spread from one asset or asset class to others, even when those other assets aren’t afflicted with the same underlying weakness. Bad news in one asset class or investment causes investors to be more pessimistic about other asset classes because as the value of that one troubled investment or asset class shrinks, investors’ exposure to other assets increase. That increase in exposure, in turn, raises pessimism about those other assets.
When it comes to bank runs, Fulghieri’s and Dicks’ model reveals unexpected consequences. Their work suggests that runs on a single bank are less likely when there is uncertainty, but that same uncertainty heightens the risk of systemic bank runs, affecting all banks. This is because investors (depositors) are slower to run on a bank when they see bad news about the bank, because uncertainty-averse investors value their investment in a risky investment (the bank with the bad news) more if they hold other risky investments, a feature known as “uncertainty hedging.”
Exiting their investment in the bank with the bad news heightens their overall exposure to other risky assets in their portfolio, so running on the one bank with bad news increases their overall exposure to uncertainty, rather than decreasing it. However, if the bad news is so bad that uncertainty-averse investors do feel compelled to run on a single bank, they are likely to run on others as well, because the amount of uncertainty in their bank portfolio is now increasing, driving them to get out.
The strength of this effect depends on the overall level of uncertainty in the economy. When uncertainty is low, runs on one bank stay “local” and don’t spill over to other banks, but when there is lots of uncertainty in the economy, a run on one bank can extend to other banks, creating contagion. More uncertainty makes the financial system more fragile and prone to crisis.
For policymakers, the implications about the effects of ambiguity aversion suggest that asset sales by banks are optimal when banks need to recapitalize to meet regulatory capital requirements. In some places, such as Europe, though, central bank purchases of assets of individual banks are not permitted, which means banking crises might be more severe there.
Fulghieri’s and Dicks’ work also suggests that merely providing more liquidity to banks doesn’t, by itself, induce lending again after a financial system shock. Other interventions might be necessary when central banks want to boost lending to drive economic growth.
Impact on corporate governance
Fulghieri and Dicks also show that ambiguity aversion has major implications for corporate governance and corporate transparency. Ambiguity can create disagreements between outside shareholders and company insiders or between different levels of corporate management – headquarters executives vs. divisional leaders, for example. More disagreement typically leads to stronger corporate governance, as outsiders try to rein in insiders.
These dynamics also change based on the investment holdings of shareholders. Investors with well-diversified portfolios, who have a relatively small portion of their total assets sunk into a company’s stock, might have more reasons to disagree with corporate insiders and prefer stronger governance and more transparency. In contrast investors who have a higher portion of assets allocated to other investments that are similar to the company (as with some venture capital and private equity funds) are comfortable with weaker governance and less transparency.
Ideally, corporate governance would follow a lifecycle, based on Fulghieri’s and Dicks’ models. When a company is young and growing quickly, it benefits from strong governance. As it ages, it performs better when outsiders exert less control and insiders make more decisions – at this point the company might, ideally, be less transparent. Finally, as a firm matures and productivity declines, it benefits from stronger governance again.
Recognizing that investors’ attitude toward uncertainty is important is the take away from Fulghieri’s and Dicks’ research. The presence of uncertainty might have important effects on the economy. Specifically, uncertainty aversion could make the economy more volatile and more fragile when shocks occur – and that means the source of instability must be dealt with when making decisions about public policy and regulation.
* Credit for this example goes to John Maynard Keynes, who wrote about urns filled with red and blue balls in 1921.