Fernando Chague

associate professor of finance at the São Paulo School of Economics

curriculum vitae: english, portuguese (lattes)

email: fernando.chague | at | fgv.br

research interests: short-selling, day-trading, behavioral biases, household finance, emerging markets, empirical finance


working papers:

Broker's equity lending desks have a privileged position: they talk every day to short-sellers in the over-the-counter lending market. Consistent with prior evidence showing information leakage by financial intermediaries, we find that clients from brokers that are connected to informed short-sellers trade in the same direction and benefit from short-seller's superior information. 

There is solid evidence showing that retail investors end up holding stocks with falling prices. This is the result of a combination of several behaviors: contrarian investing, preference for distressed stocks, preference for lottery like-stocks and the disposition effect. In this paper, we use detailed data to discipline (i) a reduced-form contrarian retail demand and (ii) an equation that limits arbitrageurs. Then, we show that (i) and (ii) can produce severe overpricing of popular high-risk stocks.

To answer this question, we examine how losing day traders decide to quit. Using detailed id-level data from the Brazilian futures and equity markets from 2012 to 2018, we find that day traders look at the proportion of profitable daytrading days to decide when to quit and not to the their actual accumulated financial loss. This surprising behavior is also problematic, as the proportion of profitable days can be inadvertently inflated by the individuals’ disposition effect and give the impression of superior performance. Indeed, we find that those day traders who persist the most exhibit a stronger disposition effect; their daily losses tend to be much larger than their daily gains, but their proportion of profitable days is well above 50%.

Individuals over-invest in familiar stocks. We document the existence of familiarity bias also in day trade, a trading activity that lasts hours at most. Living close to a brick-and-mortar firm's local store in a small city more than doubles the likelihood of an individual day-trading its stock. Information, the usual explanation for the familiarity bias, is unlikely to explain our finding: a single local store in a small city could give useful information for day-trading only in truly abnormal events.

Can one make a living day trading stocks? In this short note, we hope to inform retail investors who are thinking about leaving their jobs and become day traders. We base our evidence using high-quality data from the entire universe of Brazilian retail investors who decided  to do just that. 

published papers:

Are some investors prone to behavioral biases? We find that retail investors who forget to take advantage of a simple and widely known tax break opportunity in Brazil are also more prone to biases. Importantly, our results hold even after controlling for a large number of investor sophistication measures such as trading longevity, professional occupation, and whether the investor has traded derivatives or sold short stocks in the past. Our results support theories that relate investor's inherent inattention to behavioral biases.

Some claim that there is no equity risk premium in Brazil; they arrive at such conclusion after looking at stock returns and risk-free rates from the mid 1990's until now. We argue that if we imposed the same data limitation to the US, we could arrive at the same puzzling result. Conclusion: if you want to use realized returns to infer the size of the Brazilian equity risk premium, you better wait a couple more years...

Short-sellers are informed investors who make prices more efficient. However, they are restricted by the frictions in the equity lending market; the equity lending market is an over-the-counter markets with opaque prices (i.e. loan fees). We explore a transparency shock that took place in the Brazilian equity lending market in March, 2011, to show that a less opaque equity lending market results in more trading volume, lower loan fees, and improved stock price efficiency. We conclude by suggesting regulators to adopt "loan-fee benchmarks" as an effective way of improving transparency in equity lending markets worldwide.

Aggregate short-selling predicts future returns, but what can we say about individual short-sellers? Relying on individual trading data on both opening and closing of lending positions by all Brazilian short-sellers, we find that 30% of short-selling volume comes from investors who are able to profit consistently. The evidence of superior performance is consistent out-of-sample (both over time and across stocks), an indication of skill. Finally, we examine what the skilled short-sellers do different; among other things, they are more likely to pick already losing stocks (i.e., momentum investors) as opposed to correcting overpricings.

Short-sellers have to borrow the stock they want to sell short in the equity lending market, an over-the-counter market. The price they will pay to borrow the stock is not clear, however. We show that those short-sellers who are well connected to brokers, who in turn are well connected to lenders, are able to pay lower loan fees. Our results are consistent with search costs being an important source of price variation in opaque markets.

We estimate an augmented VAR model for a system that includes not only the Nelson–Siegel factors of the Brazilian yield curve, but also the principal components of a large number of macroeconomic and financial indicators. The main finding is that forward-looking variables, such as market expectations about inflation and GDP, are crucial to improve short-term forecasts of the yield curve.

In a single regression, we test the two main hypothesis concerning short-sellers: (H1) short-sellers are informed investors and their trading activity predicts lower stock prices; (H2) when short-sellers face restrictions and there is dispersion of opinion, stocks become overpriced. We find that both are true: short-sellers are informed but at times restricted.

published papers in Brazilian journals:

Structured products designed for Brazilian retail investors have negative expected excess returns.

We provide an update on the odds of day trading stocks for a living using reliable data.

We show that retail investors trade stocks that released "information-void" news.

We propose a VIX measure for the Brazilian stock market. You can download the "IVOL-BR" time-series here: https://nefin.com.br/data/volatility_index.html

We implement a text-reading algorithm that reads COPOM meetings minutes and produces an "optimist factor" that predicts changes in the term structure of interest rates.

Trader ou Investidor?

Aprenda a investir na bolsa sem cair nas armadilhas dos vieses comportamentais

Disponível nas melhores livrarias!!

media coverage: Folha de São Paulo, Valor Econômico, Jornal da Noite, Brazil Journal, FGV Podcast, VC S/A, Zero Hora, Folha de São Paulo


PhD, University of North Carolina at Chapel Hill, 2012

MA, São Paulo School of Economics FGV, Brazil, 2007 

BA, University of São Paulo, Brazil, 2003


Brazilian stock market risk-factors

check out the nefin webpage for data on Brazilian stock market: