Customized Financial Planning

97% Annual Returns : Mutual Funds Selection - Statistical Analysis beyond Historical Returns

Below article is the 3rd in my series of articles on how I built up a portfolio giving 97% compounded annual returns (in stocks portfolio) and 37% annual returns in MFs. You can read first article by clicking on 97% Annual Returns : My Learning and Practices and second article can be accessed by clicking on 97% Annual Returns : Implementation.  

One of the most sure-shot method of convincing clients about investing in Mutual Funds (MF) employed by Wealth Managers in India is showing them Historical Returns. Typically clients are aware of 6-8% returns of Fixed Deposits, while Mutual Funds historically have returns of more than 12% and hence looks tempting. Being a Wealth Manager myself I can swear on the fact that its the most effective tool of closing the deal positively. 

I myself have been using the same methodology for selecting which MFs to invest into. It is the easiest way to select MF, just download historical data , have it in ascending order and select the top ones in the list.  

But the real change came when I started revising my concepts of Statistical Analysis learnt during my Portfolio Analysis course at IIT Delhi. Going through standard deviation , mean , Sharpe Ratio etc., I realized that same can be used to select MFs whose historical returns are not only a function of market returns and fluke but because of good investment strategy of the Fund Manager. It also makes future performance of the MFs more predictable. 

These ratios can not only be used for MF selection, but for overall portfolio as well. So before going further into the Ratios, I will dwell into some basics of the statistics beforehand.

• Standard Deviation - It is the parameter by which you can measure the propensity of any data to 'deviate' from the 'mean' i.e. average value. Hence , you can say it is a measure of 'Risk' of the MF or portfolio. 
• Correlation - It is a measure of how much two parameters are 'linked' to each other. Simply put how would movement in one parameter effect values of other parameter.  

Now lets explore the ratios one by one - 

1. Alpha - It is the difference between the returns of the portfolio or MF and the respective benchmark. E.g. If 3 years returns of Reliance Small Cap Fund is 23% and 3 years returns of its respective benchmark (i.e. S&P BSE Small Cap Index) is 15%, Alpha will be 23%-15% = 8%.
2. Beta - It is a measure of 'correlation' between a MF and its respective benchmark. In layman terms,
• A Beta value of '1' signifies that MF will move perfectly in tandem with the benchmark. So if benchmark moves up by 15%, MF will also move up by 15%.
• Similarly a Beta value of less than '1' means that MF is less volatile than the benchmark. So MF will move less 'up' and 'down' than the benchmark. 
• A Beta value of more than '1' means that MF will move more 'up' and 'down' than the benchmark. 
3. Sharpe Ratio - Sharpe Ratio is a measure of 'Returns' and 'Risk'. Numerator is the difference between expected returns of the MF and Risk Free returns like returns of US treasury bonds, while denominator is a measure of Risk i.e. 'overall' standard deviation of the MF. As anybody will tell, we will like 'Risk' to be as less as possible and Returns to be maximum, hence a higher value of Sharpe Ratio is always desirable. 
   
4. Sortino Ratio - It is a better version of Sharpe Ratio, while Sharpe Ratio takes into account  'overall' part of standard deviation, Sortino takes into account only 'negative' part of standard deviation. A parameter (in this case Returns) can 'deviate' in both 'positive' (i.e. higher returns than the 'mean' value) and 'negative' (i.e. lower returns than the 'mean' value) direction. While negative deviation is harmful to the investor, positive deviation only increases investor returns. So Sortino takes into account only standard deviation of 'negative' returns in the denominator. 
5. Expense Ratio -  It is the %age of the fund value charged by the Fund Manager as his expenses. In India it is capped at 2.5% per year by the market regulator SEBI. Needless to say lower is the expense ratio, better the MF is. 

Stay tuned to learn about stock selection in my next article. Till then you can read about my multibagger which gave me more than 300% annual returns Avanti Feeds : A Comprehensive Stock Analysis. 

Mutual Funds Selection : Beyond Historical Returns

As I mentioned in my last article , around six months back I started giving shape and direction to my own portfolio , which till then was very haphazard and heterogeneous in its allocation. So to transform this haphazard asset allocation into a scientific allocation , I undertook below steps.

1. Deciding Asset Allocation - A balanced asset allocation across Mutual Funds (MFs) (45%) and Stocks (55%) was decided.Further allocation was done among Large, Small, Mid and Micro Cap in these two categories. This allocation was completely independent of existing portfolio, market conditions, economy data etc. 
   
2. Analyzing Current Portfolio  
• Segregating MFs into liquid/non-liquid based on their Close/Open ended period, Exit Load and Taxation 
• Identifying MFs from existing portfolio which are in sync with the Asset Allocation decided above 
• Selling/Switching MFs from existing portfolio which are not in sync with MF asset Allocation and are liquid 

Above two steps gave me a portfolio which was more synchronized with the decided Asset Allocation.What I had now was a bucket with some stones already in it and some more to be filled in. 
Now big question arises how did I decide which MFs and stocks to invest into or retain from existing portfolio. So lets dwell upon my selection methodology of MFs first. 

1. Selecting the Fund Houses - My filtering starts with filtering out those fund houses which are not backed by institutes of good repute and considerable legacy. So fund houses like Sahara makes an instant exit from the universe under consideration.  
2. Identifying Schemes - Jotting those schemes which are in sync with asset allocation from the selected fund houses only. So if I have not decided to have sector specific schemes, I will straight away ignore them.Similarly, if asset allocation is based on market cap only, schemes with market cap objective (like Small Cap Fund, Bluechip Fund etc.) only are selected
3. Collating Data - Compiling of data in a readable form which can be analysed easily ( in lay-man's language Excel) was completed. Data collected consisted of -

• Assets Under Management (AUM) - I generally prefer smaller AUM schemes , as they have more opportunities and lesser constraint to invest.
• Expense Ratio  - Theoretically lesser the Expense Ratio, better would be returns 
• Turnover Ratio - One should avoid higher turnover ratio schemes , as they imply lesser cohesion in investment approach. 
• Market Cap Allocation - Among Big, Small, Mid and Tiny cap stocks, so as to get better idea whether MF is Small/Mid or Large Cap oriented. I was surprised to know that even a supposedly Small and Mid Cap scheme had 25% Large Cap allocation.
• Historical Returns - 1 Year,2 Years, 3 Years and 5 Years Returns are collated
• Inception Date - To notice how old is the legacy of the fund 
• Exit Load - Funds with more liquidity and lesser Exit Load should be given preference 
• P/E Ratio - As with stocks, MF schemes with lower P/E ratio should be looked with favour
• Number of Stocks - MFs with higher number of stocks may suffer from over diversification 
• Ratios - (covered in detail in this Article)
• Alpha 
• Beta 
• Sortino Ratio 
• Standard Deviation 
• Sharpe Ratio 

Finally based on above data filtering of the mutual funds can be done to pare down the list to the exact number of MFs planned to be invested into. This framework reduced number of MFs held in my portfolio from 22 to just 6. 

Unfortunately in a country like India where financial literacy is more of a luxury, MFs are sold based on a very unscientific and single parameter of Historical Returns. I hope that above framework moves selection of MFs beyond Historical Returns and makes the process more holistic and hence more scientific. 

While MFs selection can be completed with the help of above framework only, Stocks selection involves a very comprehensive analysis of economy, industry, shareholders, and a lot many more parameters. Will be covering the same in our next article. 

97% Annual Returns : Implementation

As I mentioned in my last article , around six months back I started giving shape and direction to my own portfolio , which till then was very haphazard and heterogeneous in its allocation. So to transform this haphazard asset allocation into a scientific allocation , I undertook below steps.

1. Deciding Asset Allocation - A balanced asset allocation across Mutual Funds (MFs) (45%) and Stocks (55%) was decided.Further allocation was done among Large, Small, Mid and Micro Cap in these two categories. This allocation was completely independent of existing portfolio, market conditions, economy data etc. 
   
2. Analyzing Current Portfolio  
• Segregating MFs into liquid/non-liquid based on their Close/Open ended period, Exit Load and Taxation 
• Identifying MFs from existing portfolio which are in sync with the Asset Allocation decided above 
• Selling/Switching MFs from existing portfolio which are not in sync with MF asset Allocation and are liquid 

Above two steps gave me a portfolio which was more synchronized with the decided Asset Allocation.What I had now was a bucket with some stones already in it and some more to be filled in. 
Now big question arises how did I decide which MFs and stocks to invest into or retain from existing portfolio. So lets dwell upon my selection methodology of MFs first. 

1. Selecting the Fund Houses - My filtering starts with filtering out those fund houses which are not backed by institutes of good repute and considerable legacy. So fund houses like Sahara makes an instant exit from the universe under consideration.  
2. Identifying Schemes - Jotting those schemes which are in sync with asset allocation from the selected fund houses only. So if I have not decided to have sector specific schemes, I will straight away ignore them.Similarly, if asset allocation is based on market cap only, schemes with market cap objective (like Small Cap Fund, Bluechip Fund etc.) only are selected
3. Collating Data - Compiling of data in a readable form which can be analysed easily ( in lay-man's language Excel) was completed. Data collected consisted of -

• Assets Under Management (AUM) - I generally prefer smaller AUM schemes , as they have more opportunities and lesser constraint to invest.
• Expense Ratio  - Theoretically lesser the Expense Ratio, better would be returns 
• Turnover Ratio - One should avoid higher turnover ratio schemes , as they imply lesser cohesion in investment approach. 
• Market Cap Allocation - Among Big, Small, Mid and Tiny cap stocks, so as to get better idea whether MF is Small/Mid or Large Cap oriented. I was surprised to know that even a supposedly Small and Mid Cap scheme had 25% Large Cap allocation.
• Historical Returns - 1 Year,2 Years, 3 Years and 5 Years Returns are collated
• Inception Date - To notice how old is the legacy of the fund 
• Exit Load - Funds with more liquidity and lesser Exit Load should be given preference 
• P/E Ratio - As with stocks, MF schemes with lower P/E ratio should be looked with favour
• Number of Stocks - MFs with higher number of stocks may suffer from over diversification 
• Ratios - (covered in detail in this Article)
• Alpha 
• Beta 
• Sortino Ratio 
• Standard Deviation 
• Sharpe Ratio 

Finally based on above data filtering of the mutual funds can be done to pare down the list to the exact number of MFs planned to be invested into. This framework reduced number of MFs held in my portfolio from 22 to just 6. 

Unfortunately in a country like India where financial literacy is more of a luxury, MFs are sold based on a very unscientific and single parameter of Historical Returns. I hope that above framework moves selection of MFs beyond Historical Returns and makes the process more holistic and hence more scientific. 

While MFs selection can be completed with the help of above framework only, Stocks selection involves a very comprehensive analysis of economy, industry, shareholders, and a lot many more parameters. Will be covering the same in our next article. 

97% Annual Returns : My Learning and Practices

I started investing into equity markets around a decade back with the guidance of my friend who was already invested in Mutual Funds for a year now. I started with below framework 

1. Tax Saving Schemes so as to fulfill my Section 80(C) quota 
2. Six different Fund houses of Indian, Foreign, Private and Public Shareholding 
3. Systematic Investment Plans of equal amount across 6 different plans

That was my approach of implementing different investment strategies and tools like diversification, cost averaging and tax planning which I could grasp through regular reading of investment and portfolio related articles. 

Lot many experiences have got collated in my life since then, in both professional and personal life. Professionally I learnt Wealth Management practices of an American Banking giant while working with Infosys, had my Masters in Finance from IIT Delhi, worked as a Senior Wealth Manager with one of the biggest wealth management player of India and finally shifted to Dubai as a Wealth Manager. On personal front, I got married , faced a huge family medical exigency, bought my first flat and became a father. All of these experiences taught me a lot about personal finances and portfolio management. 

Both of these professional and personal events made my own portfolio very unpredictable and short term. Just when I will feel that my investment approach is settling in I will either face a job change or a pressing personal issue. But one thing that always remained constant was my passion for investments. Thanks to the above discussed situations, my portfolio had become very haphazard and heterogeneous, lacking a common approach or set of principals.

So sometime in last phase of last year, I decided to implement in my own portfolio what I always advise to my clients as a Wealth Manager - Proper Asset Allocation based on Relevant Risk Profiling. 

Below is a snapshot of my portfolio sometime in Sep'16.

Scheme Name	Latest Value
Birla SL Dynamic Bond -RP (G)	90016
Birla SL Top 100 - Direct (G)	33557
BNP Paribas Long Term Equity (D)	6018
Franklin (I) Bluechip - Direct (G)	27799
Franklin (I) Prima - Direct (G)	18865
Franklin (I) Smaller Co -Direct (G)	7280
Franklin (I) Tax Shield -Direct (G)	8229
Franklin High Growth Co -Direct	32949
Franklin India Tax Shield (D)	6329
ICICI Pru Flexi Income (G)	106071
ICICI Pru Focused. Blue -Direct (G)	27769
ICICI Pru Infrastructure-Direct (G)	9630
ICICI Pru Long Term Equity	5827
ICICI Pru Long Term Equity	26134
ICICI Pru Top 100 Fund - Direct (G)	16484
ICICI Pru Value Discovery - DP (G)	11482
Kotak Tax Saver - Regular (D)	23632
L&T Tax Advantage (D)	17159
L&T Tax Advantage -Direct (G)	6149
Reliance Small Cap - Direct (G)	14928
Reliance Tax Saver(ELSS)-Direct	13687
SBI Magnum Midcap Fund - Direct	6347
SBI Magnum Tax Gain - Direct (G)	27171
SBI Magnum Tax Gain (D)	4293
SBI Pharma Fund - Direct (G)	37948

A total of 22 schemes haphazardly across asset classes, sectors and market cap. From there on my portfolio now looks like below as on date. 

A much leaner and planned portfolio spread across MFs and Stocks. But as they say, more so in the world of Wealth Management, devil lies in details. While MFs gave an annual return of 37% (compounded) , corresponding figure for stocks is humongous 97%. 

So what was the difference that brought this huge difference in the returns. Click on  97% Annual Returns : Implementation  to  learn "How I did it". Click on 97% Annual Returns : Mutual Funds Selection - Statistical Analysis beyond Historical Returns to learn which ratios you need to check to select MF.