14. Fama-MacBeth with Shanken Correction
he Fama and MacBeth (1973) procedure is a two-step process. It involves estimation of N cross-sectional regressions in the first step. And then in the second step, it requires calculation of T time-series averages of the coefficients of the N-cross-sectional regressions. The standard errors are adjusted for cross-sectional … Read more …
13. Misvaluing Innovations
Lauren Cohen, Karl Diether, and Christopher Malloy (2013), published their study in The Review of Financial Studies and showed that a firm’s ability to innovate is predictable, persistent, and relatively simple to compute. However, usually, investors misvalue stocks when they take past success as benchmarks for future success.. Our Stata code … Read more …
12. Investor Attention and Stock Returns
Investors attention is a topic of behavioral finance. Merton (1987) presented the investors’ recognition hypothesis, which later was tested in a variety of empirical designs. For example, Fang and Peress (2009) found no-media premium. Similarly, Chen (2017) documents a significant decrease in index returns following an increase in investor attention. Our Stata code … Read more …
11. Volatility Managed Portfolios
Alan Moreira and Tyler Muir (2017) show that volatility managed portfolios can produce large alphas, higher Sharpe ratios, and significant gains for investors who take investment decisions on the mean-variance frontier of modern portfolio theory. They use a number of well-known factors such as the market, value, momentum, proﬁtability, return on equity, and investment factors in equities, as well as the currency, carry trade. The underlying rationale behind volatility timing is that volatility timing increases Sharpe ratios because changes in factors’ volatilities are not fully offset by proportional changes in average returns. Read more …
10. Stata Code for Zero-leverage Firms
Ilya A. Strebulaev and Baozhong Yang (2013) report that almost 22% of public nonfinancial firms in the US have almost zero leverage. They report that these firms have unique characteristics such as dividend-paying zero-leverage firms pay substantially higher dividends, are more profitable, pay higher taxes, issue less equity, and have higher cash balances than control firms chosen by industry and size. Read more …
9. Testing One, Three, and Five-Factor Models in Recession and Boom
We completed this project in two weeks after obtaining data from our customers. We received two separate data sets, one containing financial and other containing share prices data. We merged the two files and tested CAPM, Fama and French three-factor model, and Fama and French five-factor model under different economic conditions. For model testing, we used the methodology of Gibbson, Ross and Shanken (GRS) (1989) and other method suggested in Fama and French (2014). All the tests and data management activities were performed using Stata Read more …
8. The Persistence of Risk-Adjusted Mutual Fund Performance
We completed this project in three weeks time. This project involved the application of the methods as used in Elton, Grubber, and Black (2004). Specifically, we estimated risk-adjusted returns for each fund using factor models over a three and one year window, and then tested whether past returns predict future return; past expenses predict future returns, past betas and r-squared predict future returns. Next, we estimated how current risk-adjusted returns are related to one year and three years lagged differential returns, expenses ratio, alphas, dividend yield. This analysis further explored how betas, R-squared, and alphas are related to lagged measures of fund performance. As a next step, the analysis involved regressing mutual funds’ cash flows on lagged measures of funds performance such as alpha, beta, R-squared, expenses ratio, dividend yield ratio and other factors such as funds family sizes. And finally, we checked whether investors portfolio, past winners portfolio, and past losers portfolio perform better than others in the subsequent periods.
7. Mutual Fund Size and Performance
We completed this project in two weeks time. This project had a focus on testing a mutual fund performance predictability in relation to the size of the mutual fund. We started by writing codes for finding the percentage of funds that were in a defined size range in each year. Next, we applied Jensen model, Carhart four-factor model, Fama and French three-factor model to identify which of these models better fits the data. Selecting the best model, next we used risk-adjusted returns for each firm and estimated ranking- and evaluation-alphas from the best model and checked how evaluation alphas, fund size, expense ratios, and turnover are related to ranking-alphas’ deciles. To check association between size and alphas, we used rank correlation test. We also ranked funds by size and checked how evaluation alphas are related to fund size across size deciles. Next, we checked how funds evaluation alphas are related to fund size across ranking alphas’ deciles in the bottom-half of funds’ TNA and top-half of funds’ TNA. We also checked how aggregate expense ratio and management fee moves across different sizes of funds. As a further step, we regressed changes in expense ratios on changes in fund size. Finally, evaluation alphas were regressed on ranking alphas, cash flows, fund size, expense ratio, turnover ratio and fund family size.
6. Implied Cost of Equity Models
This project was directed at finding implied cost of equity (ICC) using different models such as the Easton (2004) model, the Gordon and Gordon (1997) model, and the Ohlson and Juettner-Nauroth (2005) model. We developed Stata codes for these ICC models. Since ICC models require forecasting earning per share, we used the methodology of Hou, Dijk and Zhang (2012). The ICC estimates from different models were then checked in association with other explanatory variables. Read more …
5. Momentum and Contrarian Strategies
This project had an objective to investigate whether momentum or contrarian profits exist? Are such profits economically and statistically significant? Can these profits be explained by proxies of risks such as market beta? Are such profits different using different samples such as small stocks, stocks with little trading volume, and stocks with zero returns? Are such profits different using different weighting schemes in the holding period? We developed Stata codes for all of the above strategies based on well-regarded methodologies such Jegadesh and Titman (1993).
4. IPOs and Abnormal Returns
We completed writing Stata codes for this project in two weeks time. The project had the objective of finding whether post-IPO significant underpricing occurs. For this project, Stata codes were developed using the famous event study methodology.
3. Fama and French (1993) model in an Emerging Economy
We completed writing Stata codes for this project in one weeks time. The project had the objective of testing Fama and French (1993) in an emerging economy. We developed the three factors in style of Fama and French. The SMB and HML factors were calculated from the returns of six factor portfolios that are created from the intersection of two size and three value groups of firms. The LHS factors were based on 25 portfolios created from the intersection of 5 size and 5 value groups of firms. The model was finally tested using the Gibbson, Ross and Shanken (GRS) (1989) method.
2. Tax Changes and Asset Pricing
This project involved writing codes to replicate the methodology of Sialm (2009), “Tax Changes and Asset Pricing”. The paper first calculates abnormal returns with well-known asset pricing models such as CAPM, Fama and French, and Carhart models, and then relates those abnormal returns with portfolios that are formed on lagged dividend yields.
1. Trading Frequency and Asset Pricing
This project involved writing codes to replicate the methodology of Florackis et al. (2011), “Trading frequency and asset pricing on the London Stock Exchange: Evidence from a new price impact ratio”. The paper forms portfolios on price impact ratios using different approaches, and then finds risk-adjusted residual returns of these portfolios using a variety of asset pricing models.