share interactive transcript request transcript/captions live captions
|
MyPlaylist
SPEAKER 1: Now we go to finance just to complete a portfolio of different things that happens at BEDR. Scott Yonker is an associate professor of finance in the Dyson School. He took a path that was familiar to me, but he started with a bachelor's degree in math, then moved to a master's degree in economics, and finally moved to something really applied-- namely, finance. He got his PhD in Ohio State in 2010, spent four years at another place I won't mention, came to Cornell, and has been here with us for five years. And welcome to BEDR, Scott.
SCOTT E. YONKER: Thank you. Thank you very much. Yeah, I dumbed it down the whole way. Math to econ to finance. So I finally found my niche.
AUDIENCE: OB's not that dumb
SCOTT E. YONKER: What's that?
AUDIENCE: OB's not that dumb.
SCOTT E. YONKER: No, I didn't say OB. I said finance was-- no, I was talking about my trajectory, math to econ to finance. OK. So what I'm going to do today is just talk about my general research focus. And I guess I call this-- my dissertation was called investigating human element in corporate policies, and now I'll just call it in financial policies.
And today, I'll talk about basically where I started and give you guys a little bit of context in what I do, which is empirical behavioral finance. And mostly in corporate behavioral finance. And then I'll give a couple examples of my research. These are old examples of my research, but they are all in one context, which is behavioral corporate finance.
So where did behavioral corporate finance come from? Back in the '90s, people were doing a lot of work on the biases of investors and consumers-- their departures from rationality. So investors have all kinds of biases in the way they invest.
The first step toward behavioral corporate finance was, how do rational managers then strategically respond to these behavioral biases? So Baker and Wurgler have some papers that basically said, look, if you've got investors with biased beliefs that are going to push up stock prices, well, this is when you should issue equity. When equity is abnormally high, that's when you should be issuing equity.
The next step was saying, well, maybe actually these top managers are biased. And so maybe they actually have biases. And this is where I fell into the mix. So there's this paper by now Malmendier and Tate that looks at managerial overconfidence and the effect of that on corporate investment decisions.
And the reason that this was particularly interesting to me was because, well, if you're going to argue that behavioral economics or finance matters, showing that these managers-- these are the ones who are supposed to be those rational agents. If you can show biases within this group of individuals, then it's pretty powerful.
It's also powerful because these managers make big decisions. They're making corporate employment decisions, corporate investment decisions. These are decisions that really drive the economy.
So I sort of fell into this area. And in 2008, I started working on a paper in my second-year corporate finance class, asking whether or not the backgrounds of CEOs matter in the oldest question that we ask in corporate finance, which is capital structure decisions. So does it matter what the manager's attitude toward debt is for firm capital structure decisions? And sort of motivated by this quote, but also a little bit by my father-in-law, who doesn't like debt either.
But there's this famous HBS case study where this guy, William Laporte, who is the CEO of American Home Products, says I just don't like to owe money. His firm carried absolutely no debt his entire tenure. Clearly, this was suboptimal. Later on, when he was replaced, the new CEO, first thing he did was lever up. Firm value goes up, because the tax shield benefits of debt.
OK. So we ask, how much does this matter? And can we test whether this matters? And we're going to rely on this psychology theory called behavioral consistency. And that's that, basically, individuals behave consistently across domains in similar situations. So in a personal situation as well as a professional situation, if you pose people with the same type of question, they'll make similar decisions.
And so what we did is we leaned on this theory, and we looked at a personal decision that we could get data on. We're not going to be able to ask CEOs, what do you think about debt? And so we needed to think about a way to get information on that.
And so what we did is we collected their mortgages and their purchase prices of their homes from public records. So this was also at the time when public records, you were able to start getting these data. So we were able to then calculate a loan-to-value ratio. So we figured out, how did they finance their home purchase? And maybe even 20 years ago, how did they finance their home purchase? Extract out all the rational factors that are in that, and does that then predict the capital structure that they have for their firm?
And I guess that's this slide. So we basically test, then, whether or not this personal leverage predicts corporate leverage. And we find that it does. So it's possibly related to corporate leverage.
It turns out it's very, very predictive in the cross-section of leverage decisions. It explains it as much as firm size, profitability, and more than asset tangibility, which these are some of the main factors that researchers have found matter for capital structure decisions.
There's also a whole bunch of firms in the US that are zero leverage firms. They have absolutely no leverage. It's a zero leverage puzzle. And we believe that this is potentially one of the things that's driving this. It's just the CEO preference.
What are the mechanisms driving this? One is selection. So firms are being selected to basically implement the optimal capital structure. But also, we have some evidence that suggests that they're imprinting their preferences on the firms.
I would say the identification in these types of papers is not wonderful, and so a lot of the evolution of my research has been trying to be able to identify these effects in better ways, because making a causal statement is pretty much impossible with this type of data.
OK. So then my job market paper-- this was the paper I want on the job market with-- was called "Geography and the Market for CEOs." And the question that I asked in this paper was, basically, do frictions matter in the labor market? So in the executive compensation literature, the typical assumption is that the CEO labor market is frictionless.
The CEOs will move to whatever job because they're wealthy, so they're not tied to any one place. They'll move all round. And economists have these-- what are called competitive assignment models, and they map the best CEO to the largest firm, and they get paid the most. That's how we allocate CEOs to firms.
And so I was wondering, well, how big is this? Is this actually true? This is a pretty big assumption. Are there frictions in this labor market? What type might there be? And that's why I investigated geographic frictions.
Well, you need the geography of the firm, and you also need the geography of the person. The firm, you can look at the headquarter location, because that's where the CEO is going to work. But for a person, there's a lot of choices as to how you might measure geography. You can think about where they live currently. You can think about where they went to school. But these are-- you're selected into these. The one thing you're not selected into is where you're from. So where you grew up.
So what I was able to do is collect data on CEO social security numbers. So you can get the first five digits of anyone's social security number from public records, from voter registrations, or whatever. These are a geographic locator. So the first three digits tell you the state of issuance. The next two tell you the sequence of issuance. This is the correspondence of social security numbers to state, so you can look up there, find your social security number. It probably lines up.
And then you can look up your social security number. This is me. This is when I lived in Bloomington, Indiana. And you can see, here's my social security. 270-82. And that tells me that I'm from Ohio, and I got my social security number between '84 and '85. So you get time and place of where a person is.
So I looked this all up for every CEO in the S&P 500. And what is that time and place for CEOs in the sample? Well, most of the CEOs-- CEOs are usually between 55 and 65 years old, and so most of them got their social in the 1950s and '60s. At that time, you got your social security card when you got your driver's license or your first job. You didn't get it at birth. In the 1980s, you started-- they have to get it at birth in order to claim your kid as a dependent. You can see that that's true. So about 60% of the CEOs got their social security at about age 16 in the sample.
So now I know a time and place where everybody lives. Now the question is, are there a bunch of-- are there geographic frictions? This is the distribution of-- this is the percentage of local CEOs by size decile. Now, this is size decile of S&P 1500 firms. So each bin are-- there are 150 firms in each bin. So decile 10. That's the 150 largest firms in the United States, and over 25% of the CEOs of those firms basically grew up in the state where the firm is headquartered.
If that were random, it would be 5%. So if we would randomly assign CEOs. So it's about five times random. It's a huge, huge geographic friction.
The next question then is, what's going on? Is it supply side? Is it demand side? One thing that we know-- I can test for is locals will accept lower pay. Unforced turnover's lower, meaning they're less likely to leave their firms. And the third thing is that, then, compensation of these locals tends to be load on local labor market factors. So there's local markets for these guys.
It seems like geographic preferences drive this. So the CEOs want to be close to family and friends. Very-- probably something that we can all identify with. One of the reasons I'm here, my wife is from the state of New York. Probably, if I had spousal social security numbers, the results would be even stronger.
So this was another paper. And I've gotten a lot of mileage out of this being able to identify where people are from. We've looked at familiarity bias in investment managers by using where they grew up, and seen whether they over-invest in their home state, and all kinds of other things.
So then this was-- the third essay in my dissertation was, basically, well, what's the implication of having a bunch of local managers? You might think that local managers are going to be more embedded in the community. They might be more labor-friendly. Well, that's not very-- if you're going to test that, it's difficult to test that and make a causal statement, so I ended up going to the Census Bureau and getting plant-level data.
And testing whether or not CEOs tend to exhibit a hometown bias. And why might they do that? There's this theory in environmental psychology called place attachments and place identity that basically say that people have an affinity or bonds towards special places.
What are those special places? Places they've lived a long time, where they've close friends, and especially where they're native to. So they-- and one of the things about these place attachments is people act on them. So they tend to affect behavior.
So what I do in this paper is then test whether or not managers favor hometown workers over others. So what I do is look within firm-- it's a within-firm decision-- following an industry downturn. When the firm gets hit with a negative shock and they have to make cuts to weather the storm, where do you cut your workers? Within the firm, which plants do you lay off people? And do you basically avoid these hometown workers?
And if you look at this, what you find is that they favor these workers from where they grew up. So they're less likely to lay off the workers. They don't cut the wages. They cut wages everywhere except for their hometown. And then wages over three or four years end up being abnormally high in their home town. And they're less likely to divest hometown plants. So this is sort of consistent in a much better identified behavioral bias of these corporate managers.
OK. I have a more recent paper, then, doing a similar thing about corporate investment. You could think that another way you could funnel money toward your hometown is by buying up cruddy firms. Basically throwing money toward your town.
And in fact, Warren Buffett did this. So Warren Buffett, who hates newspapers and is always saying media's a terrible investment, took Berkshire Hathaway's shareholders' money, and he bought the-- what, the Omaha newspaper. If you look on average and test whether or not managers are more likely to acquire firms from their hometowns, we see that they're 85% more likely to make these types of acquisitions. These are mostly very small firms, private firms.
And it seems that they're really bad investments. So if you look at the-- we can look at the share-price reaction to these acquisitions, and they're terrible. They lose-- they're value-destroying investments, and so we favor this agency theory of why they're doing this. This is just pet projects, favoritism, or private benefits for the CEO.
These are other papers that I've written in the same domain. I have a bunch of papers on, I would call, behavior corporate finance that have to do with investment managers, where this one, we look at the familiarity bias, this one, we look at the effect of social interactions. This one, we look at the effect of trust in the financial advisory industry.
And then some-- I have a few papers in corporate finance and other important actors, I guess I would call it. This is about firm boards, so diversity in the board of directors, and whether that's a good thing or a bad thing for organizations. And then we have one that looks at how risky is your firm when you have really concentrated human capital.
You have these-- like a very important inventor or scientist for a small firm. What's the risk look like for that firm? And then a couple other working papers where I'm working on stuff. A lot in this financial advisory industry. I used to be a financial advisor prior to working in academia. So thank you very much.
[APPLAUSE]
SPEAKER 1: Questions?
SCOTT E. YONKER: Questions?
SPEAKER 1: Suggestions? Comments? Throw quarters.
[CHUCKLES]
SCOTT E. YONKER: All right.
SPEAKER 1: That was overwhelming. Yeah.
AUDIENCE: You mentioned zero-leverage firms. What industry sectors are those most likely to be found in?
SCOTT E. YONKER: So what industry sectors? There are a lot of high R&D sectors where you probably can't get leverage, or you have more difficulty getting leverage. But they're in all industries.
But it's something like 20% of firms. It's enormous. There's a paper by some people at Stanford on this, so. Yeah.
SPEAKER 1: Have you looked at how you should recruit in terms of these geographic preferences? I remember when I was a faculty member at USC, we were trying to recruit. And we could only get people from California who moved to USC.
SCOTT E. YONKER: Yeah, I mean, I think that's-- I always argued-- I argued for that in every recruiting meeting, and people kind of laughed at me. But when I was at IU, when I was Indiana, people didn't want to come to the Midwest who weren't-- I'm a Midwestern, so I was perfectly happy to be there. But that was a difficulty there. So I think for sure. You can usually attract them and for less pay, so I don't know why you wouldn't go after them.
SPEAKER 1: But have you studied that in particular? I just thought it might be an interesting paper.
SCOTT E. YONKER: Yeah, I mean, I don't know how would I-- I mean, running experiments, or?
SPEAKER 1: Well, just data on salaries for people who are hired.
SCOTT E. YONKER: Yeah, I mean, with the CEOs, they're paid less, so they will accept lower pay. And unforced turnover is lower, so if turnover is costly-- which we all know it is very costly to recruit-- they're more likely to stay longer. So I think that is a huge implication. Yeah.
AUDIENCE: Thank you. I think it's a really interesting study, especially the last one, where it's the CEOs' favoritism towards local workers. I wonder what's the intention of those CEOs. Are they kind of charitably trying to give back to the locals, and bear some cost, or their intention was trying to make money, they just wrongfully decided to a hire--
SCOTT E. YONKER: Right. So did they believe that their hometown segments were better and that's why they did it, or was it just sort of-- so one thing that I can tell you is that in firms that have worse governance, this is where we see it. So it seems that's consistent with it being more of a perk, a private benefit, than anything else. But we don't really have performance of those segments. And so it's hard to know. Yeah. All right.
SPEAKER 1: All right, thank you very much.
SCOTT E. YONKER: All right, thanks.
[APPLAUSE]
Scott E. Yonker, associate professor of finance, presents current research findings regarding behavioral economics and human decision-making Sept. 3, 2019 as part of the BEDR Workshop Showcase. Sponsored by the Behavioral Economics and Decision Research Center at Cornell University.
