Saturday, 28 April 2012

The Primary Analytics Practitioner

The field of Business Analytics can be very complex. Top level analysts are experts; just like medical specialists, they have undergone years of additional training and know their area of specialty (perhaps price sensitivity, multivariate statistical modeling, survey analysis or mathematical optimization) backwards. Keeping with this analogy, most business managers are as well informed as to what business analytics can do for them as a patient heading in to see their primary physician; perhaps less so.


In the medical world there is a wealth of information specifically written for non-experts available on the web. We see legions of ads for pharmaceuticals that inform us about new conditions and their chemical solution. Many of us will routinely see a primary care physician (that’s ‘General Practitioner’ for those of you in the U.K.) and at least be familiar with our own complaints. When we have an issue with our body, we head to the doctor and let him direct us.

In the business analytic world we have done less well at making it accessible to managers. Most of the published material, easily accessible via the web, is not intended for general consumption. It’s very useful to those already initiated into the craft, but if you do not have a very strong mathematical/statistical background, good experience applying these skills in a business environment and some idea what to search for it may as well not exist at all. An excellent example, there is no equivalent of WebMD’s SymptomChecker. Perhaps also why I have seen no analytical version of a hypochondriac :-).

As a business manager when you encounter a tough problem you often just try and ‘work it out’. After all that’s what managers are paid for, right, to 'manage' ? You self–medicate, potentially causing harm or, at least, not addressing the real issue.

Well, the web is a relatively new phenomenon, whereas people going to see a Primary Care Physician is certainly not. What we are missing most, I think, is the role of Primary Analytics Practitioner. Someone a manager turns to who can:
  • Help you identify the real issues/issues and frame potential solutions         
  • Personally handle 90% of the issues that arise       
  • Source technically adept specialists when the need arises

The business analytics world is new – it has no certification process, no minimum educational requirements and no clear terminology to identify practitioners by what they do best. By my experience there are many good analytical providers out there, most of whom have expertise in specific areas. Sadly, you cannot open the Yellow Pages or search online and find a Primary Analytics Practitioner. (I just Googled the term and found nothing useful to me).  In finding one, you're on your own for now but take heart, these people are out there, it's not just me.

When you do stumble across someone with a breadth of analytical capability, an understanding of your business problem, the ability to directly handle many of your analytical needs and the willingness to say “I don’t know this well enough, but I can find someone who does” use them wisely and stop self-medicating.

Sunday, 22 April 2012

Bringing your analytical guns to bear on Big Data – in-database analytics

I've blogged before about the need to use the right tools to hold and manipulate data as data quantity increases (Data Handling the Right Tool for the Job).  But, I really want to get to some value-enhancing analytics and as data grows it becomes increasingly hard to apply analytical tools.

Let’s assume that we have a few Terabytes of data and that it's sat in an industrial-strength database (Oracle, SQL*Server, MySQL, DB2, …)  - one that can handle the data volume without choking.  Each of these databases has its own dialect of the querying language (SQL) and while you can do a lot of sophisticated data manipulation, even a simple analytical routine like calculating correlations is a chore.
Here's an example:
SELECT
(COUNT(*)*SUM(x.Sales*y.Sales)-SUM(x.Sales)*SUM(y.Sales))/( SQRT(COUNT(*)*SUM(SQUARE(x.Sales))-SQUARE(SUM(x.Sales)))* SQRT(COUNT(*)*SUM(SQUARE(y.Sales))-SQUARE(SUM(y.Sales))))
correlationFROM BulbSales x JOIN BulbSales y ON x.month=y.monthWHERE x.Year=1997 AND y.Year=1998

extracted from the O'Reilly Transact SQL Cookbook
This calculates just one correlation coefficient between 2 years of sales.  If you want to calculate a correlogram showing correlation coefficients across all pairs of fields in a table this could take some time to code as you are re-coding the math every time you use it with the distinct possibility of human error.  It can be done, but it’s neither pretty nor simple.  Something slightly more complex like regression analysis is seriously beyond the capability of SQL.

Currently, we would pull the data we need into an analytic package (like SAS or R) to run analysis with the help of a statistician.  As the data gets bigger the overhead/delay in moving it across into another package becomes a more significant part of your project, particularly if you do not want to do that much with it when it gets there.   It also limits what you can do on-demand with your end user reporting tools. 

So, how can you bring better analytics to bear on your data in-situ?   This is the developing area of in-database analytics:   Extending the analytical capability of the SQL language so that analytics can be executed, quickly, within the database.  I think it fair to say that it’s still early days but with some exciting opportunities:
  • SAS, the gold standard for analytical software, has developed some capability but, so far, only for databases I'm not using (Teradata, Neteeza, Greenplum, DB2)  SAS in-database processing
  • Oracle recently announced new capability to embed R (an open source tool with a broad range of statistical capability) which sounds interesting but I have yet to see it. Oracle in database announcement
  • It’s possible to build some capability into Microsoft’s SQL Server using .NET/CLR   and I have had some direct (and quite successful) experience doing this for simpler analytics.  Some companies seem to be pushing it further still and I look forward to testing out their offerings.  (Fuzzy Logix, XLeratorDB).
No doubt there are other options that I have not yet encountered, let me know in the feedback section below.  


For complex modeling tasks, I am certain we will need dedicated, offline analytic tools for a very long time.  For big data that will mean similarly large application servers for your statistical tools and fast connections to your data mart.

For simpler analysis, in-database analytics appears to be a great step forward, but I’m wondering what this means in terms of the skills you need in your analysts: when the analysis is done in a sophisticated statistics package, it tends to get done by trained statisticians who should know what they are doing and make good choices around which tools to deploy and how.

Making it easier to apply analytical tools to your data is very definitely a good thing.  Applying these tools badly because you do not have the skills or knowledge to apply them effectively could be a developing problem.


Sunday, 15 April 2012

Civet coffee gains popularity [Updated, April 2012]

By Jackie, Researcher
Type of business industry: Food & beverage
Product: Civet Coffee/ Kopi Luwak/ Kopi Musang

“Kopi” is the Indonesian word for coffee and the “Luwak” is the indigenous animal
who plays an “active” role in the harvesting of the raw coffee cherries.

Introduction

The objective of this week's research is to find out which type of coffee is considered the most expensive coffee in the world. A deep discovery and understanding was done through viewing a number of media news including CNN to find out the processes involved to make this coffee, the global market price of this coffee, the customers' response after drinking this coffee, the issues related to this coffee, and the potential of doing this type of coffee business.

"Kopi Luwak" or civet coffee is coffee made from coffee berries which have
been eaten and passed through the digestive tract of the civet cat.

Strange, but true!

Well, it is hard to believe that the most luxury coffee in the world is actually made from “animal dropping”. Yeah, that is true. In order to be more specific, I would say that it is made from Asian Palm Civet’s [and other related civets] dropping. Ideally, these civets are kept in cage and feed on beans of coffee berries as their major diet. They enjoy eating the berries fleshy pulp. Then, those coffee berries are passed through their digestive system and finally defecated. Amazingly, the defecated coffee beans are still in shape. This is because the civets eat the berries but the beans inside the berries which pass through the cat’s digestive system are still undigested. After that, their dropping is gathered and undergoes a series of processes such as washing, sun-drying, light-roasting and brewing. Some customers state that these beans yield an aromatic coffee with much less bitterness.


Coffee connoisseur Chris Rubin explains what makes "kopi luwak" worth the exorbitant price: 
"The aroma is rich and strong, and the coffee is incredibly full bodied, almost syrupy. It's thick with a hint of chocolate, and lingers on the tongue with a long, clean aftertaste. It's definitely one of the most interesting and unusual cups I've ever had."

Connie Veneracion, a consumer who had given a jar of civet coffee beans by her brother and family who had just came back from Indonesia in 2009 said:                                                           
“I tore the seal, opened the jar and the first thing I noticed was the glossy exterior of the coffee beans as though they were coated with oil. After dinner, I dumped half of the contents of the jar into the blender and processed the beans to a coarse grind. The aroma was decidedly fruity and sweet. The ground civet coffee beans went into the coffee percolator and, several minutes later, I was excitedly serving civet coffee to everyone who cared for a cup.”

Expert cupper and Sprudgie Award winner Stephen Vick had this to say on his kopi luwak experiences:
 “On the cups that I didn't present defect I found very mild sweetness and acidity with some grassy, iodine notes and a pretty rough finish. One of four cups was moldy and another single cup showed phenol. I tasted band-aids, iodine, and oyster.”

[Again, those comments were only based on their personal opinion and thus, controversial]


Price and availability
Ranging from US$120 and $600 (RM400 to RM2,000) per pound. 

“One small cafe in Queensland Australia has Kopi Luwak on the menu at A$50.00 (US$33.00) per cup. Brasserie of Peter Jones department store in London’s Sloane Square started selling a blend of Kopi Luwak peanut and Blue Mountain called Caffe Raro for £50 (=US$99.00) a cup” 
(Civetcoffeestarbucks.blogspot.com, 2011)

Sumatra is the world's largest regional producer of civet coffee, followed by Java and Sulawesi [This data might have been changed as Southeast Asia’s countries entrepreneurs such as Malaysia have started to invest heavily in production in their homeland due to exceptional high market price]. It is mainly sold in Japan, Taiwan, South Korea and United States. Most of us simply cannot afford it.

Issues
1. Halal or non-halal (vary in different country)
 Indonesia: Majelis Ulama Indonesia (MUI) declared that civet coffee (locally known as Kopi Luwak) is   
 Halal and can be consumed by Muslims as long as the beans are thoroughly cleansed before grinding. 
 MalaysiaThe exotic and highly-priced kopi luwak (Civet Cat Coffee) has not received Halal status from 
 the National Fatwa Council, Malaysia. A Harian Metro report advised Muslim consumers to refrain from 
 drinking the coffee due to its unconfirmed Halal status. 

2. Fake Civet Coffee
 Due to the lucrative market for high-priced Civet coffee it seems inevitable that some unscrupulous people   
 would try to capitalize through dishonest means, passing off as Civet coffee certain coffee beans which 
 indeed had not been consumed and expelled by a Civet. 

3. Maligned, abused, and beleaguered
    The civet cat has an unknown future on many fronts. The civet was traditionally hunted as a pest, but a   
    booming market in civet coffee has changed its fate, turning it from ‘pest’ into ‘producer’. 


Business Potential
In term of economic perspective, this business industry has a bright future due to its high demand, but limited supply. Franchises like Coffee Bean, Starbucks and etc which are available globally are selling it at extraorbitant price indirectly boosting more new entrepreneurs [mostly from Southeast Asia] to venture into this industry. This commodity has no price control as it is classified under luxury good, instead of essential items. It is usually considered as small scale manufacturing business as the current annual output produced from factories are not very large [usually in pounds instead of tonnes]. However, the difficulties are the methods to keep those civets healthily, halal or non halal issue, law constraint [protected species in some countries], and health benefits [some manufacturers didn’t undergoes the correct cleaning processes, resulting in producing a harmful output].     

Related links:
CNN News Team Tries Kopi Luwak Coffee: http://www.youtube.com/watch?v=KnuelLBQOxY
Tested.com Tests Kopi Luwak Coffee: http://www.youtube.com/watch?v=dk8_HabWkW0
Kopi Luwak Coffee - $65 a cup!: http://www.youtube.com/watch?v=sV7yvCoI0EY
Kopi Luwak 猫屎 咖啡: http://www.youtube.com/watch?v=fyHBi-jM5N4


Saturday, 14 April 2012

Cluster Analysis - 101

The current Wikipedia page on Cluster Analysis, excerpted below, is correct, detailed and makes absolute sense.  Then again, if you do not have a background in statistical modeling, I'm guessing these two paragraphs leave you no wiser.
Cluster analysis or clustering is the task of assigning a set of objects into groups (called clusters) so that the objects in the same cluster are more similar (in some sense or another) to each other than to those in other clusters. 
Clustering is a main task of explorative data mining, and a common technique for statistical data analysis used in many fields, including machine learning, pattern recognition, image analysis, information retrieval, and bioinformatics.
Wikipedia 4/2012 
In this post I hope to provide a workable introduction for people that need to be educated consumers of cluster analysis.
(If you want more technical detail, I suggest you go back to the Wikipedia link above, and follow up on the range of hyperlinks embedded in the document -  it's really very good.)

Let's put this in context with an example.  Assume we are working with a retailer that has 1000 stores and they want to decide which dairy products to put in each store to maximize sales.

One option would be to treat all stores as being the same, come up with one assortment list and put it everywhere.  This has the singular advantage of being easy to execute while leaving all non-mainstream products off the shelf.

At the other extreme, we could try to tailor product assortment individually by store.  Did I mention there are 1000 of them?  Apart from the work involved in building 1000 individual analyses, do we have the discipline to execute such analyses consistently across 1000 stores? Would we have sufficient organization to execute these unique selections in 1000 stores?

Most teams will end up working with groups of stores that they consider "similar" as a compromise.   These groupings may be based off single store features (e.g. stores with more than 30% of sales from premium products) or maybe geographical features (e.g. everything in the South East).  Bear in mind that, if you are trying to do this without statistical help, they really need to be simple groupings .

For assortment selection, we really want to group together stores where people buy similar products.  In this case we want to find groups of stores that have similar sales patterns.  For dairy, these sales patterns could be related to the % of sales associated with various product characteristics:
  • premium vs. value, 
  • single-serve vs. multi-serve, 
  • cheese vs. yogurt vs. creamer vs. drinks vs. milk
  • yogurt styles
(Note: I do eat a lot of dairy but I haven't worked that category yet so forgive me if I missed something big)

Cluster analysis (actually a class of statistical algorithms, not just one) is used to scan across multiple features that you think are important, to find groups (clusters) so that:
  • stores within a cluster are similar to each other
  • stores in different clusters are dissimilar
It sounds rather like magic doesn't it ?  You just throw it at the algorithm and (big fanfare) it finds clusters !  Well perhaps it's not quite that easy.
  • It does take some care to prepare the data, ensuring it's clean, accurate and in a form that works for this process (see Data Cleansing: boring, painful, tedious and very, very important).  
  • In reviewing the results you may decide to drop some features and split out others (e.g. "Premium" is split into "Premium" and "Super Premium").
  • You need to determine how many clusters are correct for your data.
  • You may want to bring in some additional data to help describe clusters
    • demographics of people living near the store (ethnicity, income, household size etc. )
    • geography (maps work well)
    • local competition
  • Really an extension from clustering, but you could build predictive models to explain why , for example, super-premium, greek yogurt is so very popular in Cluster 4.  If you can tie high sales of this product group to specific demographics, you may find other stores with similar demographics that have not previously sold it. (Could be a big opportunity).
I'll return to this topic in future posts, but for today, your takeaways are simple:
Cluster Analysis finds better groups (clusters) of similar things.  Clusters help you target your offering without dying under the weight of work.

Saturday, 7 April 2012

Reporting is NOT Analytics

Reporting is about what happened; Analytics is about answering "what if" and "what's best" questions.  Most of the materials that land on a VP/Director’s desk (or inbox) are examples of reporting with no analytical value added.


Reporting tells us what has happened: sales; orders; production; system downtime; labor utilization; forecast accuracy. Reports leave it up to the reader to digest this information and based off their experience and expertise about the world around them construct a ‘story’ as to why it may have happened that way.

Really good reports will look for known causes of routine issues. For example, if I know a store is low on inventory of a specific product, I could just report that. If I flag it as an exception the person receiving the report may even see it among the sea of other facts. To go the extra mile, it would be wise to see whether I can (automatically) find any disruptions back in the supply chain (either with inventory or flow of goods) and include that information to answer the routine questions that will be raised by my report. But, the report builder must anticipate these needs at the point they are writing the report and for more complex issues that’s just not realistic.

Analytics is all about finding a better story or, if you prefer, insight from data. We’ll talk about tools for finding insights in a moment, but much of this is about approach: develop a working theory about what may be happening and test it out with the data you have available. Revise your theory if needed, rinse and repeat: this is very definitely an iterative and interactive process.

At the simplest level a lot of really good analytics is enabled by being able to interact with the data: filter it to see specific features; sort it to find your exceptions, drill-down into more detail to see (for example) which stores are causing the issue in your region, chart it to see trends across time or perhaps even see relationships between variables (like temperature and sales of ice-cream by region). Generally available tools (like Excel) can get you a long way to intuitively understanding your data and finding some insight.
A further step (and one I fully understand most analysts cannot take - see  What is ‘analysis’ and why do most ‘analysts’ not do it? ) would be to run some descriptive statistics around the data.
  •  Measures of ‘average’ (mean, median, mode)
  • Measures of ‘spread’ (Standard deviation, percentile ranges, Min/Max)
  • Frequency histograms and boxplots to visually show the distribution of data
  • Scatter plots to view interaction
  • Correlation matrices to spot high level interactions
  • Outlier detection
  • Dealing with missing values
If these options seem strange perhaps even archaic and of little relevance to the business world, you may need to trust me when I say that these are exceptionally valuable capabilities that increase understanding of the data, uncover insights and prepare you to step into the world of (predictive) modeling. Thinking back over the last 2 weeks of work (store clustering, system diagnostics, algorithm development and even some ‘reporting’) I can confirm that I have used every one of these multiple times and to good effect.

In Predictive Modeling we build a mathematical model around the problem you are trying to solve. Once the model is built, and (very importantly) validated to confirm that it really does behave as you expect the world too, you can start asking questions like:
  • what happens if I add another warehouse to this network
  • what is the best combination of price and promotion to maximize profitability
  • how much inventory do I need in each warehouse
  • what is the best way to load this truck
  • what is the best assortment for each of my stores
  • why are sales dropping in the Western region

Everyone is familiar with the phrase “Garbage In, Garbage Out” relating to how computers will happily calculate (and report) garbage when you throw bad inputs at them. With modeling, the structure of the model is one of those inputs and many of you may have experienced the complete junk that comes out of a bad model even when you put good data into it. Picking the right modeling tools for the right job and applying them correctly is a very skilled job. Predictive modeling covers an extraordinarily broad field of statistics, mathematical and operations research. Just as for data handling (see The Right Tool for the Job), this is not something you are likely to do well without appropriate training: understanding the field you are trying to apply these tools to helps enormously too.

So why go to the trouble and expense of ‘Analytics’ rather than ‘Reporting’? Reporting is essential, but well-built Analytics or Predictive Models can find insights and opportunities that you will never find by any other means.

A really good analyst will do this work and while being willing, ready and able to take you to whatever depth of complexity you wish, will furnish you with a simple (perhaps even one page) report that answers your questions and makes the end result look … obvious.

Friday, 6 April 2012

Property speculation in Malaysia. [Updated, April 2012]

By Jackie, Researcher

The impact of property speculation towards the younger Malaysian generation, the causes and effects of this speculation as well as the future predictions and forecasting about potential property investment in Malaysia. [Updated, April 2012 with 2012 Real Property Gains Tax (RPGT)]


From stylish, contemporary homes to luxurious condominiums with lush greenery

Introduction

The objective of this research is to get a clearer picture on property speculation issues which have started to happen in Malaysia since 2007. Well, after seeing some real-life examples through buying and selling (as my family is dealing with this kind of business), doing ample researches through property news investment like attending Gavin Tee’s seminar, following up government plans [eg. Malaysia My Second Home (MM2H) programme and 2012 Real Property Gains Tax (RPGT)], studying the “conversion strategies” played by the broker and developers and analysing the increase rate of oversea investors, I have decided to share my experiences here.

Overview

The real estate market had gone through a “roller coaster ride” in the recent years, from a soaring market in 2008 to a depressed market in 2009 and skyrocketing again in 2011. There is a lot of controversial points related to this current issue whether Malaysia is currently facing “real estate bubble” [also known as property bubble as well as housing bubble] or not? Steps and measurement have been taken by both government (by implementing RPGT) and banking sector (by disallowing full-loan to Malaysian property buyers).



What is wrong with Malaysia My Second Home (MM2H) programme?

The objective of this programme is to encourage or attract foreigners to retire in Malaysia or spend extended periods here. The major advantages offered by the government include low cost of living, developed infrastructure, friendly Malaysians, attractive tourist destination, quality environment, good climate and variety of food. This programme however, has induced some foreign investors to invest instead of staying here. As most of them are very rich, they used to left their house here waiting for the price to shoot up and then sell back to earn an attractive capital gain.

“Conversion strategies” played by unethical brokers and greedy developers.

Honestly, most of them prefer to sell house to foreigner. Why? The reason is simple. This is because they have a higher purchasing power than Malaysian. Also, please note that some of them are come from Middle East, Japan, Korea, China and Singapore. Ok, now let's take at look into a deeper ground. Have you ever feel that developers always build high luxury condos, semi-detached houses, bungalows instead of low price apartments or moderate double storey houses? This is all about Economics! Land is limited. Why they want to build a lower priced house when they can build a higher priced house on the same land? They are not afraid of unable to sell all the units because now their strategy is to build in a smaller scale instead of large. Plus, they are not afraid of having no customers as they have found the “new higher purchasing power customers from other countries”.

What will happen in the future?

Well, a fresh degree graduate can only earn an average of RM2,000 to RM3,000 per month. The cost of living in Malaysia is rising plus most of them need to pay installment for cars normally when they need to go to work as there is still some “lacking” in public transport in Malaysia [Not all place is reachable, mismatch of time and inconvenient]. Recently, banks only lend up to 70% of the house in value. The new mortgage lending rule applies only to borrowers taking up a third housing loan to curb excessive investment and speculative activity in urban areas. In short, "the rich will become richer, while the poor have to work harder". Some experts spark up rumors by saying that Malaysian young generation will turn into a condition, so called "homeless generation" too, and indirectly speed up the government's plans to handle this worrying situation. 

Steps and measurement taken by the government to solve this issue.

Implementation of 2012 Real Property Gains Tax (effective from 1 January 2012) with chargeable gains from disposal of real properties are as follow:
1. Sold within 2 years = RPGT 10%
2. Sold above 2 years and below 5 years = RPGT 5%
3. Held more than 5 years will be exempted from the Real Property Gains Tax (RPGT)


Predictions and forecasting

Potential property investment in the future kindly refer to this talk given by President of SwhengTee International Real Estate Investors Club, Gavin Tee’s on Ipoh Property Talk at The Haven Lakeside Residences:
Part 3: http://www.youtube.com/watch?v=YUfSg-cgtjM




Upcoming issue***

"In order to tackle the rising of house prices in Malaysia, the government is now considering whether is it appropriate to double the minimum price of houses which the foreigners can purchase"





Wednesday, 4 April 2012

Gold Investment Forecast and Predictions [Updated, April 2012]

By Jackie, Researcher

This is a private college research on gold investment based on historical data, and analysis for the future market trend in global market. Well, after seeing a lot of theories, stories and discussion; my curiosity and doubtfulness towards gold investment grows deeper and stronger. The main idea is whether it is worthwhile to invest your money in gold investment rather than putting your money in bank's fixed deposit account to earn a fixed rate of interest. Therefore, after getting some internal news plus doing some researches on this matter, I have decided to share this knowledge and ideas with you guys.

Firstly, there are several advantages of investing in gold.
  1. The stability of gold tends to be stable from year to year. Have you ever heard of gold price fall? The answer might be yes, but most of the time is no. Why? In Economic, gold is consider as 'limited supply'. Therefore, as time passed by, the demand of gold will shoot up due to the 'increasing demand' in raw stock to make jeweleries.
  2. Not affected by inflation like other currencies do (Zero inflation effect).
  3. Can be withdrawn in the future in physical state to make jeweleries (In bank, you might have to buy minimum quantity in order to do this).
The disadvantages are as follow:
  1. You might suffer loss if the gold price drops; earn nothing if gold price stay stagnant, and eventually waste a lot of time in order to wait for the price to shoot up again.
  2. Transaction/ procedure fees.
Below are some examples of an gold investment deposit account offered by CIMB Bank Malaysia.




Perhaps, this graph will give a clear picture on the historical data of global gold price. It is positive gradient on overall.


So, what will happen next?

According to the Star newspaper which I have read on this 16 of February , it is written there "The Gold Bullion Entrepreneurs Association of Malaysia (GBEAM) expects gold price to hit US$2,000 per ounce by mid-year from the current US$1,733 per ounce as the weakening global economy would drive more funds to safe-haven investments such as gold". 

Therefore, based on my calculation there is a high opportunity of making 15.4% capital gain on investment, if the prediction is correct. Although risk is still there, but it has been minimized as ample forecasting have been made by experts. I am not saying that they will be certainly right, but their probability of making wrong is very low. 

Another popular theory I always heard is "purchase gold when the gold price is low, sell it off when the price is high". This theory is basically closely related to capital gain maximization. 

So, what is the right time to sell? Answer: during the peak price. An OCBC Bank report in January 2012, has forecast that gold price would be around US$1,800 per ounce by the end of 2012. That means the falling occurs between June till December this year.

In conclusion, it is very hard to say that gold investment is highly-profitable. This is because it depends on market situation. It is advised that investors should always keep on updating themselves with the latest market information and get some views from experts. Whenever you can foreseen the opportunity, why don't you give a try. Remember, in Economic, whenever a marginal benefit of doing something is greater than its marginal cost. Just do it! And you will never regret. Do not let the opportunity cost gone like that.

Thanks for spending sometime to read this. Any constructive opinion is highly-welcomed.    

Here are some useful links regarding this issue:
http://biz.thestar.com.my/news/story.asp?file=/2012/2/16/business/10717078&sec=business
http://www.youtube.com/watch?v=F1yjB3M8SNM&noredirect=1
http://www.cimbbank.com.my/index.php?ch=cb_per_st&pg=cb_per_st_inv&ac=12&tpt=1#