Mutual Funds - Science topic

I think nobody is really sure which one among the strategies is a clear winner

Dear Harmeet Singh,

check out this analysis: https://www.spglobal.com/spdji/en/research/article/what-happened-to-the-index-effect-a-look-at-three-decades-of-sp-500-adds-and-drops/

Sincerely

Tomasz Miziolek

Unfortunately the questionnaire needs re-design. For instance, the questions about mutual funds are obligatory but in my case I do not have mutual funds so cannot answer those questions hence cannot proceed beyond step one. Other aspects of the questionnaire also need re-design.

Not that hard, put them in a Excel file and do a standard deviation test and a regression analysis. You can easily see the outlier. Since you comparing mutual funds, you would have to view which one acted different on unusual trading days.

I managed via bloomberg terminal and the "corporate action" function. Name changes are published there as such.

Anders Børsting For sure, you can!

Hello Luigi,

I have my research project of Investment Analysis And portfolio management course when I was student. What I did was choosing 10 funds; 4 Mutual funds, 4 ETFs, 1 index and Bitcoin. The benchmark was MSCI Europe Index.

After testing Efficient market hypothesis and measuring the performance of the selected funds, I did my evaluation using frontier curve, Skewnes, R squar, ratios (CV, sharp ratio, Treynor ratio) and CAPM.

This is the link, I hope you find it useful:

All the best.

Mutual funds is a investment vechile to help retail investors to enter the equity market and widen the participation of retailers. But mutual funds charge higher fees and might not fulfil their role of widening participation and diversification if not properly managed. Mutual funds need a portfolio of stocks in the equity market to diversity into to succeed.

I'd like to add to this question. The data I pull from the “Monthly Returns and Fama-French Factors” page is the "Total Returns per Share". Is this the same as the mutual fund returns, Ri, that I subtract the risk free rate from to get the dependent variable?

Dera Rippo,

We have something close to what you are interested in, but it is written in Bulgarian. We have a theoretical framework and we are providing micro-crediting and consultancy support to small farmers for more then 20 years. Hopefully, this year we will have it in English. Meanwhile you can visit our web landsourceofincome.org

Under the publications and materials you can find videos (with English subtitles) describing our approach.

Keep in touch

Best

Ivan Penov

Prakash -

You want "...to know if I can compare one set of data in numbers with another set of data in percentages?" Well, I would think it best if both were raw data or both were percents, but it likely depends on what kind of comparisons you are making. Perhaps you might want to substitute rank data for each and do a nonparametric comparison, but recall that you are then 'throwing away' some information. Also, if your comparisons involve hypothesis tests, note that lone p-values, and unsupported threshold decisions are rather meaningless. Please consider a type II error probability analysis or some other sensitivity analysis involving effect size.

Regarding multiple regression, that sounds promising. A graphical residual analysis, cross-validation, and estimated variances of prediction errors can help. Also, heteroscedasticity is a natural phenomenon which you should consider. See https://www.researchgate.net/publication/333642828_Estimating_the_Coefficient_of_Heteroscedasticity

Best wishes - Jim

Nico Lippe

You can use risk-adjusted return comparison tools such as Sharpe Ratio, Treynor Ratio or Sortino Ratio to explain your work. These ratio measures risk differently, so please consider that in advance before selecting the tool.

Dear Prof. Venugopal,

You can select 5 to 10 AMCs

Each AMC will have their own funds, among different funds you have to select particular scheme which matches in all the selected AMC.

Ex: Tax advantage fund: all the AMC will be having Tax benefit fund or particular sector fund or Gold fund, Debt fund.

You can get the scheme details in the AMC scheme document.

Collect NAV data for the particular period and start applying any statistical tools to analyse. Sharpe, Treynors, Jenson model are models used in analysis of MF.

Thanks and regards,

ARMA,GARCH could be the best models for forecasting!

For research on mutual funds, prepare excel file with the following data

a) Historic NAV data of the fund

b) Historic indice values of the benchmark

c) Compute standard deviation of returns

d) Compute Sharpe Ratio, Treynors Ratio, R-Square, Jensen measure, Fama measure

e) Compare periodic returns of fund with benchmark returns

f) Compare expense ratio and other ratios as computed in d above with other funds in similar category.

Dear Valentin

For mutual funds that are actively managed (i.e. not passive stock index funds) you will have a challenge constructing an appropriate benchmark to compare the return distribution. Further, mutual funds typically describe an investment strategy that they purport to follow and then do not follow the strategy. That is, the fund managers portray the investment strategy to appeal to investors to invest with attractive words like high growth, and value stocks.

Follow this question.

FAMA is basically used in assets pricing commonly used in portfolio management of srips and more of that used in FOREX to forecast hedging against risks involved in exchange rates of currencies.

Passive investing tends to mean 'tracker' or 'index funds. These funds are meant to replicate common equity or bond indices. Eg, FTSE100.

Active funds require the fund managers to provide 'insight' into making investment decisions.

You may have to find a passive fund and an active fund from an investment house or provider and track its performance. These results are usually published quarterly.

Try to find funds which are similar. For example, you may compare an FTSE100 index fund with an actively managed UK Income based equity fund. These funds are not exactly the same but this will get you started. When you develop more experience into this area you may be able to find a close match between your passive and active funds. Good luck.

Thanks Emeka. A financial asset is a contract between creditors and a given person, firm or state. The original clauses were broken, but the claims persist. At its original market asset value was $ 100 and now trades at $ 20. The new buyers see a yield and they know how to make it happen. The enforcement of contracts is a key factor behind these markets.  

Hi, this is a complex question, because it depends on what do you want to measure. Usually, if you are trying to check if active management has offered a benefit for the end investor you should use measures that combine both return and risk relative to the benchmark. 

If this is the case, and you have a pretty homogeneus sample of asset managers (I mean, funds) you could use both the information ratio or the jensen alpha. If you want more info on these side I recommend you the paper from Bill Sharpe, "The Sharpe Ratio" 1994. The good thing about this paper is that Sharpe shows the generalized Sharpe ratio which in fact is the Information Ratio. Relative performance per unit of tracking error (relative return versus relative risk).

Regarding the jensen alpha, it is also very useful and a slightly different approach. What Mr. Jensen suggests is including in the CAPM basic formulation a new component called "the jensen alpha" wich mainly summarizes the extra return offered by an active investor (due to skill or luck) and not explained by its "beta" versus an index.

There are a lot of other performance measures, but there is enough evidence that usually if you have to rank the performance of different mutual funds with a common benchmark, different performance measures tend to offer common rankings.

It becomes more difficult if you do not have a common benchmark, because other issues arise. For instance, ranking hedge funds is not that easy, but this is another tale...

Best

Carlos

from a practitioner's point of view shareholders activities can be assessed based on the type of the companies. For publicly held companies we can see shareholders' activities in their presence in AGM as well as extra-ordinary general meeting, the questions they raise during the meeting and those which are presented in writing before the meetings based on countries' legal requirements. There are incidents in some AGMs where shareholders stop the approval of auditor's report and financial statements waiting further investigations or restatements of audited statements by the company's management. Also, how the share of the company I being dealt with in the stock markets can be seen as representing financial activity of shareholders. In other companies, some of the main shareholders can request separate meetings with the board of directors or the audit committee to discuss important affairs of the company. Also, the number of shareholders who represent other shareholders in the BOD is also another important feature of shareholders' activities. For private and family owned business, the level and extent of activities of shareholders is rather limited and restricted and based only in attending AGM and other official meetings. So you can select proxies for number of meetings both official and non-official, changes in value and quantities of shares in stock market, type and number of questions in meetings, number of shareholders represented in BOD and if you can get such details their activities within the BOD.

Hi,

You could also read this paper:

R Marquez, V Nanda, D Yavuz, "Private equity returns and performance persistence", Review of Finance, 2015, 19, p 17893 -1823.

Best regards

you can go through

Performance evaluation of Islamic mutual funds relative to conventional funds: Empirical evidence from Saudi Arabia

Performance Persistence of Islamic

Equity Mutual Funds

Islamic Versus Conventional Mutual

Funds Performance in Saudi Arabia: A Case Study

It would really help to have more details about your data. For example, Does it include prices or volume? Also, what do you mean by 10 types of schemes?

If you share more details about your data, I may have a better answer and I suspect others will share more ideas.

I hope this helps.

Rob

Dichotomy in Value Chain Upgrading Approaches: A Comparative Study in Processed Millet and Banana Wine Sub-Sectors

Hi! I think you could access the above mentioned data in http://ec.europa.eu/eurostat/web/products-datasets/-/nasa_10_f_bs. But you have to select all the correct option in order to obtain such data.

You may want to check youtube. There are a number of good videos. Try to select the most viewed ones. They are very helpful. I always use them for such calculations.

Here is the study of Norwegian Mutual funds:

Dear Konstantin

In India

Tata Ethical Fund

Taurus Ethical Fund

SBI Shariah Equity Fund

Goldman Sachs CNX Nifty Shariah Index Exchange Traded Scheme

Regards

Bikram

Good evening,

The portfolio performance depends on the opportunity costs (of the transaction, of the no-transaction, of the liquidity) linked to the portfolio yield.  According to banks and markets, these different costs are fixed, variable, direct or linked to the size of the assets. The only thing I know is : the active investors have to find their profit in the transactions, their transaction giving a free usefull information (the assets price). The number o active investor depends on their profit and the form of the opportunity costs.

Jean-François Gueugnon

Hello Jan. I think that the second approach would be more appropriate to measure active management. In this approach a constant difference in returns after adjusting for risk, would be captured by the alpha of the model. However, sometimes the tracking error is used as a measure of performance, for instance using a rolling window and estimating performance out of the back sample.

I am sure I didn't answer your question but why don't you use the same t-bill rate for the week given the information publicly available.

Fibonacci ratios are limit numbers when n goes to infinity or is large. Fibonacci methods take into account nonlinearities, (that's the common thing with the chaos theory) and silently swallow an infinite source of behavior diversity not aspiring to better understand natural phenomena and phenomena considered as complex which were adverse to any modeling before. Certainly Fibonacci methods cannot be labeled under the name "non-linear signal processing" , they are not based on the Takens theorem. It is not possible to reproduce the essential features of of a system without a reconstructed phase space of a very low dimension. Fibonacci methods are some non-parametric analysis by means of the extraction of the time-series properties allowing the reconstruction of the dynamic relation which links the time-series terms. The non-parametric modeling of nonlinear processes does unfortunately not make it possible for a great number of processes to produce robust results. In front of this lack of specification about the complex or chaotic processes new tools have been developed. Your claim how or when Fibonacci works is not a statistical test in a strict sense. And it leaves the major issues of the Fibo methods unanswered. Still it is close to nature, whatever it means, perhaps a closest to nature tool available from brokers' sites. One needs to understand its serious limitations, though, before blindly applying it for making money. I am shure that a red light alert will appear very soon for a new practitioner of financial markets.   

Hi,

Return-based analysis is very powerful to analyse mutual fund returns and portfolio management style, either through simple linear regression or more complex cluster analysis.

We may wish to have a look at my research on analyzing hedge fund style and performance and all reference therein.

Best regards,

Sébastien

If I understand your question (which is a little too brief), then Yes: in order to assess the asset allocation policy of a mutual fund you would need to consider all the asset types (quoted securities, unquoted securities, property, bonds, cash etc) whether quoted on a stock market or not.  Calculating the movements over time should provide an indication as to the mutual fund's asset allocation practices. 

Hi! I am also working on mutual fund performance but using secondary data. I have seen one thesis on this topic with primary data in Delhi University. You can search that thesis.

A good starting point is Fama-French 3-factor model. Time-series are available on their website.

BR