Winner – Best Paper: Alejandro Lopez-Lira
Read the transcript
[MUSIC PLAYING] One of the central questions in finance is why different assets earn different returns. I was trying to see if we could use the information produced by the top executives to better characterise the risks, and how does this translate into returns for investment.
The information in the annual report is very detailed, and if you have just 1,000 of companies, it’s just impossible to read all of them. We’re going to be looking at the words that people use.
Companies disclose in their annual reports a lot of very detailed information. And you can extract that using machine learning methods and get like very nice risks, in the sense that they’re interpretable.
Basically, it can be explained with just four big risk categories. So these would be innovation, or technology risk, global risk, production risk, and demand risk.
If we think about the time it takes to read a report, it’s probably on the order of hours, whereas with this algorithm it will just take seconds. And it will just tell you this firm is subject to a 20% production risk, 30% credit risk, and 50% China risk. If you give me the text of the company, I can just apply the algorithm in a second, and I will just tell you which proportion of risk they’re facing.
When I finish the paper, it was close to the line. I just decide to apply for this.
My name is Alejandro. I’m a PhD candidate at Wharton. There’s like top people from everywhere. I feel very lucky and very grateful to have the opportunity to present at this wonderful school.
If we apply math in a nice and sensible way to the information that’s out there, we’ll just get a much fuller picture of what’s happening in the economy. And we will be able to leverage that information to invest better. Thank you.
I’m feeling extremely happy. I’m very glad I had the opportunity to present, and I’m very happy I won.
Winner – Best Presenter: Kate Suslava
Read the transcript
Providing information to investors can be roughly separated into two aspects. One is informational. And another, which is very difficult to catch, is the promotional aspect. My study focuses on the promotional aspect, basically how managers obfuscate the delivery of company performance to ensure that it’s favourably impacting the value of stock.
Using natural language processing powered by machine learning, I teach computers how to read. I’d like to tell you about my dissertation project in which I taught computers to read the language of earnings conference calls.
I compiled the first dictionary of euphemisms in corporate disclosures.
Every time a company would be going through troubles, managers would talk about facing some headwinds and hitting some speed bumps.
And I applied it to more than 78,000 earnings call transcripts to show that, first, euphemisms do show up quite frequently in the earnings calls, and second, that they do have impact on the company’s stock price.
I find that investors systematically underestimate the magnitude of bad news. This results in the shares declining for several months after the call, as investors are having a delayed market reaction to the information content of the earnings call.
The information from the textual disclosures is not fully reflected in stock prices and is also not explained completely by the existing finance models. Investors can simply count how many euphemisms managers use in their answers to the analysts. Another thing they can do is compare the level of euphemisms in the current quarter to the historical level of euphemisms.
So next time you are listening to an earnings call and you hear a lot of “headwinds,” “hiccups,” “speed bumps,” “transition period,” sell your shares before it’s too late. Thank you.
I’m really honoured to win this award. It means a lot to me. And I’d like to thank the organisers for giving me this opportunity to present.
Finalist: Andrea Rossi
Read the transcript
[MUSIC PLAYING] What I ask in this paper is whether private equity firms and their manager are able to scale up effectively when they raise a larger fund. Not surprisingly, investors want to invest with private equity firms that have done really well in the past.
There are reasons to expect that private equity funds would face decreasing returns to scale. Why? Because larger funds, presumably, have to make more investments, which means that the impact of the value-adding skill of the key partners of these private equity firms might be diluted across too many investments. On the other hand, private equity firms might be able to keep increasing returns at bay for two reasons– fees and reputation.
But what happens usually is that these private equity firms raise larger follow-on funds and then tend to disappoint relative to their early funds. And this is usually attributed to a negative impact of fund growth. What I show is that this is actually, primarily, statistical artefact due to reversion to the mean, which is driven by the fact that a large portion of the variation in concessional returns across private equity firms is driven by what we could call idiosyncratic shocks, random shocks, or just luck.
In what sense luck? Well, for instance, they were in the right sector or in the right position at the right time. Now because of this, it’s only normal that the follow-on funds or the funds that were lucky in the past do not perform as well as the previous ones.
So the key finding is that private equity funds, especially leveraged buyout funds, historically have been able to scale up quite effectively. They’ve been able to keep decreasing returns at bay. On the other hand, I find that the VC funds that have grown the most did have a negative impact from this excessive fund growth on their returns.
This is a fantastic opportunity. It was really exciting. Definitely talking in front of such a diverse group of industry leaders is a really great opportunity for someone starting their academic career.
Finalist: Argyris Tsiaras
Read the transcript
[MUSIC PLAYING] In this paper, we will study the determinants of cross-country variation in equity return correlations with the US stock market, what makes one country’s equity market returns more correlated with other markets, and especially the US market.
This bar graph plots the monthly correlations between the US equity market return and a given market’s equity return. There is large heterogeneity with countries like the United Kingdom, Netherlands, France, and Germany on the high end, with high correlations with the US, and countries like Pakistan, Colombia, Peru, and Indonesia on the low end.
The main finding is a strong, positive cross-sectional relationship between the market share of foreign investors in a given market and the correlation of that equity market with the US market.
Higher positions predict lower average returns. And in particular, they predict lower alphas with respect to several different CAPM models, like the US CAPM, the global CAPM. So it seems that countries with low cross-border positions earn abnormally high returns relative to their riskiness.
One important finding in our work is that countries without much foreign presence, with a small share of their equity market held by foreign investors, appear to have abnormally high returns not justified by the riskiness. It is heterogeneity in the degree of friction, or perceived frictions, in cross-border investing that explains the degree to which a given stock market is correlated with the US market. It may be that investors are foregoing profitable opportunities just because they are not familiar with the market, even though it seems that these markets do offer a pretty good risk-return trade-off.
So in summary, we argue that the portfolio demand channel– and not the cash flow channel– explains the cross-section of equity return co-movement.
In future work we plan to study the explanatory power of cross-border holdings for foreign exchange returns. It has been an incredible opportunity. Being part of this forum has been an amazing experience for me and a first. Me and my co-author, we’re very interested in interacting with finance professionals. And the European Investment Forum is a great, great place for that.
It’s an experience that I have enjoyed very much.
Finalist: Edward Watts
Read the transcript
[MUSIC PLAYING] Really, what we’re trying to get at is if investors are presented with assets that are otherwise identical but one is used for green projects and the other one is used for non-green projects, if they’re willing to pay more for the one that actually backs green projects. So this is the term the greenium. Really, it tells you something about whether investors are essentially willing to give up returns to invest in environmentally sustainable assets.
We can actually find almost exact matches of green and non green securities that are issued at the exact same time just because of structural differences with the muni bond market. We just compare these two assets, and if we see a premium in which the green asset actually trades at, it tells us something about how whether investors are willing to give up returns to essentially invest in these. These assets are the same for risk and return standpoint. So perhaps unsurprisingly to a lot of people, there is no pricing differential between the two. And actually, what we find is that– at least in the US muni space– is not only is there no pricing differential, but investment banks appear to charge slightly more to issue the green bonds.
Do you want to issue green if it costs more? And I think that’s the important part. And we actually got a bunch of questions about this from news reporters of like, oh, is this a really bad thing? And like what would you tell people actually issuing these bonds? And I think the important part is if you’re going to issue what is effectively the same thing and you’re going to get the same pricing, you just want to be sure you’re not paying more for this.
And I think that’s important sort of policy to take away from all of this. Honestly, I was not expecting to be selected, and now I’ve got good news it’s been a little bit of a confidence boost because you know you write these papers that I couldn’t tell if I was any good or not so I think it’s great that people are sort of interested in this topic.