Testing of assets pricing models requires time series returns of portfolios that are formed on size, book-to-market, leverage, beta, or any other criteria and factor returns that might include market factor, SMB, HML, momentum, profitability, liquidity, investment, etc. One of the challenging task is to sort assets on a given criteria, make portfolios, calculate the portfolio returns in each period, and then rebalance the portfolios periodically if the underlying criteria of the portfolio formation changes with time.
We overcome such difficulties with the help of dedicated programs that we write in Stata language. These programs not only reduce the chances of error that might occur if one tries to do all the analysis manually in a spreadsheet program such as MS Excel, but also expedite the process and saves considerable amount of time. Some of the well-known asset pricing models are given below:
Testing CAPM using Fama and McBeth cross-sectional regressions
Testing CAPM using time series regressions and then applying different tests based on regression intercepts
Asset pricing models using generalized methods of moments (GMM) technique
Any other model that requires factor development from portfolios of assets
While testing these models, the most common approach is to use portfolios to reduce noise in asset returns and to isolate effects of different risk factors. Portfolios are created both for the left hand side (LHS) and right hand side (RHS) factors. Such portfolios are created from the intersection of selected variable using dependent and independent sorts of the given variables.
We do not depend upon the ready-made factors as available for the US market from Fama and French website. Instead, we develop both the RHS and LHS factors using our Stata codes. The primary reason for doing so is that the ready-made factors are available only for limited markets.
Frequently Asked Questions
Q1. Will this paid help negatively interfere with my learning process?
Answer: Not so quite. In fact our experience with previous projects led us to believe that researchers learned a lot and far more quickly than what they could have learnt on their own. During the process, we not only develop relevant Stata/Excel codes for doing a specific task, but also we enable the researchers to do the same on their own. Researchers can ask questions related to the steps/techniques being taken. So their is no chance of missing the crucial learning process and things that researchers ought to do and learn during a research project. What makes us different from other websites offering illegal solutions such as those selling ready-made essays, we believe in student learning and hence we develop and share codes for statistical analysis and not just the final results tables.
Q2. Shall I have to provide the actual data?
We encourage researchers to provide actual data, though it is not a strict requirement. We respect confidentiality of the researchers identity and their data. Actual data is helpful in applying the statistical analysis and preparing and cleaning it before analysis. This way final analysis is directly linked to the already processed data. However, if researchers choose not to provide their data, still we can develop codes on dummy data which researchers can apply on their own to the actual data.
Q3. How long does it take to finish the analysis?
Duration of the project usually depends upon how quickly we get and prepare the data for final analysis. It also depends upon the nature of analysis and number of steps involved. Our previous experience indicates that it takes almost from one week to three weeks time to complete a project.
Q4: How the interaction/coordination takes place?
We share a Dropbox folder and share our codes, data, and results through that folder. Researchers can access this folder in real-time. Any question / updates are communicated through emails.
Q5. What is the preferred method of payment?
We accept payment through bank transfer and Paypal, which ever method is convenient and less costly for a customer.
Q6. Do I get suggestions related to methods and tests?
We do provide suggestions on best practices and leave the final decision to the researcher and their supervisors.
Q7. What is the refund policy?
Before the start of a project, if a customer wants to cancel his project, we shall refund the full amount after deducting any bank paypal trnasfer fee. Once a project is in the completion stages, refunds are not possible. Similarly, after we have shared our paid data with the customers, refunds cannot be issued.
Q8. Are the codes and programs used in paid projects for personal use?
Yes. Codes, techniques, programs, macros or any other material shared during the project are strictly for personal use and cannot be shared with friends, supervisors, or the general public through emails, internet or any other means.
Paid Help Pricing
We charge fairly affordable prices, when compared to the complexity and number of steps involved in data cleaning, data preparation, and estimation of the relevant empirical models in finance. Following are our price lists for a single model and bundle offers.
Data Cleaning and Management
Raw data needs proper cleaning, merging with other data sets to create desired set of variables, and usually reshaping from wide format to long. For example, to test a standard asset pricing model such as Fama and French model, one needs to merge financial and share prices data, t-bills rates, and market index data before rest of the analysis. We provide such services for $50 or Rs. 5000
Price of a Single Model
If data is already in ready to use format, then the standard rate is $100 or Rs. 10,000 for a single model. However, the rate might vary with complexities e.g., comparisons of models in two or three different markets, comparisons of models in different economic conditions, or estimating a model on two different time frequencies such weekly and monthly. For each such additional step of analysis, an additional $50 or Rs. 5000 will be charged. Examples of single model definitions are given below:
▬ Application of CAPM using cross-sectional regressions (Fama and Mcbeth, 1973) or time-series regression (Jensen, 1983)
▬ Fama and French three factor (1993) or Fama and French five factor (2015) model
▬ Carhart (1997) four factor model
▬ Event studies models using market model
▬ Momentum portfolio returns
▬ Panel Data models
▬ Models used in earning management research such as Jhones model, Kanzik model, Dechow et al, and Kothari model
▬ Mutual fund performance evaluation using factor models
▬ Time-series analysis including co-integration, GARCH/ ARACH/ VECM/ VAR etc
▬ Implied Cost of Equity models
▬ 2SLS and GMM regressions
▬ Credit risk models, Merton Model, KMV-Merton model
You can ask for a quote through email if a desired model/method is not mentioned in the above list. The email address is email@example.com
In the bundle offer list, service fee charged per table/model declines significantly. For example, if you choose to follow methodology of a given paper and want to replicate results of a table that has 10 different regressions, then the pricing formula is as follows:
▬ Fixed $100 plus $30 for each regression / variation
The bundle offer reduces the price per model from $100 to $65 if there are two models / variations in the analysis; to $55 if there are four models, and so on. This offer is best for PhD and MS level students as the rigor in empirical analysis lends significant support to the hypotheses being tested from a whole lot of angles.
For any additional information, please feel free to contact at:
Misconception about paid help
Every now and then, we receive odd requests, such as request for ready-made proposals, thesis write-up, paper write-up as a ghost writer, and so on. As before, here we want to clarify that we do not entertain any such requests. Our paid help is only related to writing Stata and Excel codes, data provision, data management, and guidance on selecting appropriate analytical tools. We do not do any of the following activities: doing students assignments, writing thesis, conducting literature review, developing hypotheses, selling ready-made results, publishing papers for others, or any other activity that might negatively affect students’ learning.