REFRESHER - INVESTMENT IN THE STOCK MARKET - Quick Overview by Nik Zafri

Disclaimer : The information comprised in this section is not, nor is it held out to be, a solicitation of any person to take any form of investment decision. The content of the article does not constitute advice or a recommendation by the author and should not be relied upon in making (or refraining from making) any decision relating to investments or any other matter. You should consult your own independent financial adviser and obtain professional advice before exercising any investment decisions or choices based on information featured in this article. The author can not be held liable or responsible in any way for any opinions, suggestions, recommendations or comments made nor do opinions of contributors necessarily reflect those of the author

Please always remember, stock market has nothing to do with fortune telling

WHICH PLATFORM IS THE BEST TO INVEST?

It’s been a while since I last talked about the stock market, but I’m still in the game, just as an investor, not an expert. I usually rely on my humble decades of experience to guide my decisions. So, here we go.

Whenever I'm asked the question, which platform should a person invest? Actually, I have no preference as there are many platforms that you can invest - it all depends on your goals, risk appetite, and time horizon. Common platforms include:

Stock Market (e.g. Bursa Malaysia, NYSE, NASDAQ): Suitable for medium- to long-term investment and active trading.

Unit Trusts/Mutual Funds: Good for passive investors,

ETFs (Exchange-Traded Funds): Offers diversification with lower risk,

Robo-Advisors: Automated, algorithm-driven platforms (e.g. StashAway, Wahed Invest),

REITs (Real Estate Investment Trusts): Good for income-focused investors,

Cryptocurrencies (e.g. Bitcoin, Ethereum): Very high risk, suitable only if you understand the volatility.

Real Estate: Tangible asset, but requires large capital and longer holding period.

For this explanation, I'll focus on the stock market.

2. TECHNICAL ANALYSIS

(example) Let’s take Maybank stock listed on Bursa Malaysia. 

(Note : Please read the above disclaimer before proceeding)

You want to know whether to buy, hold, or sell.

Tool Used: Moving Average Crossover

• Short-term MA (e.g. 20-day Moving Average)
• Long-term MA (e.g. 50-day Moving Average)

Scenario :

If the 20-day MA crosses above the 50-day MA ➜ Golden Cross ➜ Buy signal

If it crosses below ➜ Death Cross ➜ Sell signal

Result: On 15 March 2025, you observe that the 20-day MA of Maybank crosses above its 50-day MA. The trading volume also spikes. This may suggest a bullish trend and a potential buy signal.

But, you need to confirm this with other tools like:

RSI (Relative Strength Index) : to check if it's overbought or oversold.

MACD (Moving Average Convergence Divergence) : to confirm trend strength.

Support/Resistance levels : for better entry/exit.

Here's a more thorough analysis :

Current Stock Information

Last Traded Price: RM10.42​

Day's Range: RM10.16 - RM10.40​

52-Week Range: RM9.52 - RM11.04​

Market Capitalization: RM122.84 billion

PE Ratio: 12.18

Dividend Yield: 5.89%​

​Data as of March 20, 2025. ​

Technical Indicators:

Moving Averages:

• MACD (Moving Average Convergence Divergence):
• RSI (Relative Strength Index):
• Support and Resistance Levels:
• Support: RM10.34​
• Resistance: RM10.50​

​These levels are based on recent trading activity and volume. ​

Recent Performance:

Maybank's share price reached a record high of RM8.93 on September 18, 2024, before experiencing some profit-taking.

Analyst Outlook:

As of February 27, 2025, Maybank was downgraded to a 'Hold' with a target price raised to RM11.42. 

Conclusion

The technical indicators suggest that Maybank's stock is in a consolidation phase, with key support around RM10.34 and resistance near RM10.50. The MACD indicates bullish momentum, while the RSI suggests the stock is approaching oversold conditions. Investors should monitor these levels and indicators closely, considering both technical signals and fundamental factors before making investment decisions.

3. WHY DATA IS SOOOO IMPORTANT?

Accurate data is the backbone of any investment decision. Here’s why:

• Historical Price and Volume Data: Needed for technical analysis,
• Financial Reports: Show the company’s performance, debt level, profit margins, cash flow, etc,
• Economic Indicators: Inflation rate, interest rates, GDP growth influence market behavior,
• Sentiment & News: Public perception and events (e.g., CEO changes, legal issues) affect price.

Without reliable data, you’re investing blindly - it’s like driving in the fog without headlights. (I had few unpleasant experience on this one)

4. TRANSPARENCY BY LISTED COMPANIES 

Transparency builds trust and allows investors to make informed decisions.

Besides publishing prospectuses and undergoing audits, companies must:

a. Disclose Material Information

• Major contracts won/lost
• Changes in top management
• Corporate exercises (e.g. bonus issues, share splits)
• Legal disputes
• Mergers and acquisitions

b. Provide Forward Guidance

• Future outlook or business direction
• Expansion plans

c. Consistent Investor Relations (IR) : 

• Quarterly briefings
• Press releases
• Investor presentations

Finally...Does this all matter? The answer is a big "YES"

• It will reduces risk of Insider Trading
• It helps maintain fair valuation
• It gives investors confidence

A company that hides bad news or delays reporting is a red flag - transparency reduces uncertainty and helps maintain stock stability.

MAKE SOME INCOME (By Nik Zafri)

The strategies employed to invest and increase income are often complex and multifaceted when it comes to "practical".

However, I found some common themes emerge all the time which I sum up as the core investment strategies.

1) Diversification

Spreading investments across various asset classes (stocks, bonds, real estate, commodities) to reduce risk is one good move.

2) Long Term Perspective

Always focus on building income over time rather than quick gains. In short, greed will make you lose, so be patient.

3) Leverage

When you borrow, increase them on potential returns - it's risky but once you strike, it's worth. You can also get ASB loan and let them invest or reinvest for you. In the long run, you might get handsome returns

4) Get a good advice

Although Warren Buffet might not agree with me, but if you can afford it especially when you have no experience, it's logical to employ financial advisors or wealth managers to handle investments.

5) Always pay tax but optimize them

Draw up a good strategies to minimize tax liabilities but never avoid from paying tax

There are quite a number of investment vehicles.

1) Stocks and Bonds

Investing in publicly traded companies and if you have heard of government debt, go for them.

2) Real Estate

Own properties for rental income or appreciation.(REIT)

3) Private Equity and Venture Capital

Investing in startups and private companies.

4) Hedge Funds

Pooled investment funds that use various strategies to generate returns.

Alternative Investments can be purchase of assets like art, collectibles, and commodities.

Additional Soft Income-Building Sources

1) Business Ownership - Creating and growing businesses.

2) Intellectual Property - Developing and protecting patents, copyrights, and trademarks.

3) Inheritance - if you come from a rich family.

4) Risk-Taking - Pursuing high-reward, high-risk ventures.

5) Networking - Building relationships with influential people.

Also, think the way successful people think.

You must have access to information for e.g. exclusive investment opportunities and market insights

You must adopt a risk tolerance attitude - allowing yourself to pursue higher-return investments.

Have a long term perspective (Time Horizon) - enabling yourself to weather market fluctuations.

Good luck, just sharing my perspective from my so-called research. No guarantee of success unless you really work hard and do not let failure stop you. Remember, investing involves risk, and past performance is not indicative of future results. Just conduct thorough research or consult experts before making investment decisions.

HOW EXACTLY DO ARTIFICIAL INTELLIGENCE ANALYZE THE STOCK MARKET - From the draft notes of Nik Zafri

Knowledge Sharing (from my draft notes) :

Question : How exactly Artificial Intelligence predicts the stock market?

Nik Zafri : If I have to make an intelligence guess, it should look something like this

#artificialintelligence #AI #stockmarket

PREDICTING THE STOCK MARKET? HERE'S SOME TIPS (NIK ZAFRI)

Predicting the stock market with absolute certainty is an elusive goal because it's influenced by numerous unpredictable factors such as economic indicators, geopolitical events, investor sentiment, and even natural disasters. However, many analysts and researchers attempt to forecast stock market movements using various methods and tools.

I believe statistical models such as linear regression, ARIMA, GARCH are some of the most popular ones in analyzing historical price data and identify patterns or trends. There are few other typical analyses such as

(a) quantitative (maths (Black-Scholes) + statistics = pricing or Var for loses),

(b) fundamentals (company's financial health, quality, trends, and macroeconomic = intrinsic value and predict future stock prices),

(c) technical - past market data, primarily price and volume = identify patterns and trends = predict future price movements. Tools : moving averages, chart patterns, and technical indicators such as RSI and MACD and,

(d) sentiment - alternative news from social media, news, anything of public sentiment to gauge investor mood and predict market movements. NLP should be useful to extract insights from textual data.

And now here comes Artificial Intelligence (my gen used to call it machine learning - we don't have sophisticated tools like AI then) - the principles are much the same except the approach is far more advanced. Algorithms such as neural network, SVM, random forests (data mining) and make predictions. AI can handle non-linear relationships and capture complex patterns better than traditional statistical methods (although I may disagree with that :-) )

INSIDER TRADING IS A CRIME - Nik Zafri

Insider trading is a criminal offence under the CMSA 2007. Upon his or her conviction under either sections 188(2) or (3), the insider shall be punished to imprisonment for a term not exceeding 10 years and to a fine not less than RM1,000,000.00 (section 188(4) CMSA 2007)

Insider trading refers to the buying or selling of stocks by someone who has access to non-public, material information about a company. This practice is illegal because it undermines the fairness and integrity of the financial markets.

It erodes trust in the fairness of financial markets. When insiders use privileged information to gain an advantage in trading, it undermines the level playing field that is crucial for the proper functioning of stock markets. This can deter investors and reduce market efficiency. They give access to non-public information an unfair advantage over other market participants. This allows them to profit at the expense of uninformed investors, distorting the true value of securities and leading to market inefficiencies. Furthermore, it can create mistrust by investors in the financial system as if it has been rigged where investors may not participate, leading to decreased liquidity and reduced investment activity.

Insider trading can distort the allocation of capital by diverting investment to companies based not on their fundamentals but on insider information. This misallocation of resources can hinder economic growth and reduce the overall efficiency of the economy.

Overall, the danger of insider trading lies in its potential to undermine the fairness, efficiency, and integrity of financial markets, as well as erode investor confidence and distort capital allocation. As such, stringent regulatory measures are necessary to deter and punish insider trading and uphold the integrity of the financial system.

BAILING OUT TROUBLED CORPORATIONS IS NOT THE RIGHT ANSWER - THOUGHT FOR THE DAY BY NIK ZAFRI

 Bailing out troubled corporations is not the right answer. Most of the money is used for shares buyback - hence not helping with the employment prospects or development. Regulations must be improved to limit shares buyback. Get them to commit using the money for employment opportunities/development as well.

QUOTE FOR THE DAY - BAILOUTS AND SHARES BUYBACK ARE NOT HELPING THE PEOPLE

 When bailouts happen or corporate loans being approved with low interest rates and restructured, I think the money is intended to boost the economy by building more infrastructures and/or ease unemployment problems.

Instead; the "money" is used to buyback own shares/stocks to stabilize the price (by the "corporate borrowers"). As a result we see : two shoot ups - (1) corporate debts, (2) stock market price.

While this can being applauded, one thing remain constant, this "glory" happened while so many people including graduates are still unemployed.

Although share buybacks is legal, yet, it does not contribute to the people.

And yes, providing stimulus may be a good thing, but to those who are able to work - to have them employed is far more better.

Basics : When money is overflowing, the goods are lesser and higher in price, inflation will happen. (or is it happening?)

I hope this will not be a trend in Malaysia but I do see signs.

TYPICAL PITFALLS IN CORPORATE GOVERNANCE - Nik Zafri

Lesson No. 1 : The CEO must be well-informed of any risks (both technical and financial) when deciding to invest in "new innovation"

Innovative ideas may sound "juicy" but potentially may also become a great threat to sound governance practices. Few gigantic motor corporations thought innovation was a good investment hence a lot of money went to Research and Development (R and D) to produce "innovations" to only knew later that they have breached the law.

a) The VW emissiongate/dieselgate

Quoting Ian McVeigh, Head of Governance at Jupiter Asset Management wrote in the Telegraph UK in 2015 : 

"The revelation that the car giant (VW) has been using so-called “defeat device” software to get its diesel cars to pass strict emission tests has been an unmitigated disaster. Since the scandal broke, the value of the company has fallen by around €30bn (£22bn)"

Source : https://commons.wikimedia.org/wiki/File%3AVW_stock_price_after_emissions_violations.png (Analysis by Dennis Bratland)

He wrote further about the significance of the CEO being well informed of investment analysis without the investors finding out about it first. Ian also mentioned about the corporate structure of German companies - having split boards - a supervisory and a management board making it potentially unclear of the "who is responsible or authorized for what?". UK however adopted the principle of putting the final responsibility with only one board.

The Volkswagen dieselgate or emission gate case started on 18 September 2015, when the US EPA issued a notice of violation of the Clean Air Act to German automaker Volkswagen Group for intentionally programmed turbocharged direct injection (TDI) diesel engines to activate certain emissions controls only during lab emission test. The programming caused the vehicles' NOx output to meet US standards during regulatory testing but emit up to 40 times more NOx in real-world driving. 

The CEO sportingly have since apologized to the customers, users and the general public. Later, he resigned.

Although Volkswagen is embarking on new policies and slowly regaining the trust of stakeholders and the car buyers, but the price they have paid is high.

Lesson No. 2 : Hiring "external party" to advise the Board can also lead to disaster.

I once heard somewhere that billionaire Warren Buffet; at some point; do not fully trust "external party's" judgement. He would rather make his own final decision.

Be careful when appointing external "subject matter experts" especially authorizing them to speak in a Board Meeting. If it is not necessary to hire, then DON'T! 

Many important decision cannot be made unanimously because the "so-called guru" says "NO". Ironically, the CEO agrees and subscribes to the "gurus" advice rather than seeking majority opinions from the experienced board members. (I've seen this happened before - it was a disaster!)

Hiring consultant assisting in corporate governance implementation is fine but these practices should be restricted to mid and support levels where CEO leads the initiative - not vice-versa. Board Members meeting must always be treated as CONFIDENTIAL.

Remember : Not everyone in the boardroom are smart guys!

Lesson No. 3 : There are always good rationales of why laws are implemented and enforced. 

Those who try to avoid the law are deemed to have "cheated" the government/ lawmakers and other stakeholders. Or put simply, betraying everyone's trust. 

One of the top cases that I've bumped into is misappropriation of funds (breach of trust) due to shareholders "manipulating" their unclear roles, responsibilities and authorities to serve their own ends. Other cases include : board ineffectiveness, unethical audit and risk practices, "unfair" pay, bad relations with external parties such as accountants and company secretaries. (aah yes, accountants and company secretaries can tell stories too - they know your secrets!)

We've witnessed how prolonged "conflict" between corporate entities and the lawmakers ending up in series of investigations and prosecutions. Long battles in the court of law not only endanger the company's reputation (even if you win the case) but the sad consequences that follow it - from people loosing jobs to reduction of market values.

Lesson No. 4 : Risk is always PROACTIVE

Never wait until problem arises. Learn the lessons from others who have failed. Take proactive measures by adopting Risk Management. (and don't just simply say....DO IT!) Proactive Risk Identification will provide you with a guide on Plan of Action (mitigation), brainstorming and develop a strategic plan. 

All these activities must happen prior to operation/project implementaton, assessed during the operation (comparing with the Risk Register for proactive risk identification stage) and mitigate during post-operation stage.

Adopt good practices of Risk Management but do not try to be too bold by taking unwanted risks!

Lesson No. 5 : Do What You Say and Say What You Do.

When you have given solemn promises or have put them in your customer charter or have addressed them in Manuals and Procedures - deliver them. Don't just file everything up and ignore them, or displaying what needed to be displayed and "that's it...end of story" - hoping that your PR will do the rest for you.

Sometimes, as a shareholder, you need to realize what you think is "impossible" can happen and do happen - whereby other shareholders (your own friends) can turn against you due to one or two angry customers who feel that you are not delivering what you have promised.

There are so many reasons of why ISO 9000 or Total Quality Management is introduced on the first place. Surely they are not to burden you but to help you.

Lesson No. 6 : Non-Conformance and Risks are opportunity for improvement

Don't scare auditors with cold bullying remarks such as :

• "I'm your paymaster",
• "Make it looks like there is more "loss" than "profit",
• "I know your boss",
• "Don't try to be too smart - remember Enron?" 

Remember, auditors are human beings too. If they feel threatened, despite the Non-Disclosure Agreement, they still can adopt the "Need to Know Basis" vs your "Transparency" - they can even get court-orders. Believe me, the end-results won't be good.

Treat auditors as friends who are trying to help you not destroy you. Treat any non-conformances as opportunities for improvement and not some "fault-finding activities". In the end, you'll be in every auditors "good book"

Don't make your company becoming one of the case study being discussed in front of other auditors in their association.

Lesson No. 7 : Be careful when setting limits on shareholder voting power.

One study (by Ken L. Bechmanna, Department of Finance, Copenhagen Business School and Johannes Raaballeb Department of Economics and Business, Aarhus University) shows that the bad corporate governance in the banks is visible in the shape of severe restrictions on shareholder rights, including voting and ownership ceilings, etc. These severe restrictions are quite unique, especially from an Anglo-Saxon perspective. In the U.S., ownership ceilings are not allowed and in a sample of 4,399 U.S. firms, only 24 had voting ceilings (see Commission of the European Communities, 2007). 

Similarly, voting and ownership ceilings are not among the 24 corporate governance provisions carefully examined by Bebchuk, Cohen and Ferrell (2009), who show that restrictions on shareholder rights lead to significant reductions in firm value.

Out of the six entrenchment provisions found to be value destroying, four of them set limits on shareholder voting power and the ability of a majority of shareholders to impose their will on the management.

And FINALLY!!

When you are having problems, try avoiding easy paths such as buyover, mergers, bailouts etc or worse - let yourself willingly to become "victim" to "hostile takeover" or ending up in government intervention. 

Even if you feel like selling your shares or the company itself, don't give up yet - try your best to first to solve your problems - who knows, you might succeed against all odds. 

Be transparent to your employees - let them know what is happening, many companies have gone down to earth and surprisingly, their employees have very good ideas on how to save the company. If the idea works, give these employees the rewards that they deserve!

KAEDAH ALGORITMA PEMBELAJARAN DIPANTAU DALAM MENGESAN MANIPULASI PASARAN SAHAM – SIRI PENGENALAN

Walaupun terdapat maklumbalas dan tindakan yang pantas oleh pihak berkuasa serta terdapatnya pindaan dalam undang-undang, manipulasi pasaran adalah antara kebimbangan utama pelabur pasaran sekuriti/saham.

Kaedah mengenalpasti unsur penipuan dalam pasaran saham yang ada pada masakini adalah bergantung kepada undang-undang, peraturan dan kepakaran. Ianya berdasarkan pendekatan “atas ke bawah” yang menggunakan set corak yang dikenali dan ambang (threshold) yang telah ditakrif terlebih dahulu.

Data pasaran seperti harga dan volume saham (jumlah saham/kontrak yang didagangkan) dipantau menggunakan peraturan yang berkaitan dan notifikasi red-flag. Kemudian sebarang transaksi dalam tempoh pengesanan akan disiasat secara lanjut bagi mengenalpasti sebarang aktiviti penipuan.

Cara ini mempunyai dua kelemahan :

i) Mengesan tempoh yang luarbiasa yang tidak berkaitan dengan simtom (contohnya skim manipulatif yang tidak dikenali)

ii) Kesesuaiannya dengan keadaan/suasana pasaran yang berubah dengan pantas, transaksi data yang tinggi kerana kenaikan jumlah pelabur serta syarikat/saham yang disenaraikan). Ini menuntut pindaan terhadap undang-undang untuk memantau data yang begitu luas dan amat mencabar sekali.

  

Pasaran sekuriti sedang berubah dengan drastik. Suasana ini berpotensi mendedahkan loopholes  yang akan digunakan oleh pedagang untuk meraih keuntungan.

Pihak berkuasa mungkin memerlukan sejenis modul dan algorithma bagi menyokong pengenalpastian aktiviti manipulasi pasaran.

Secara amnya, penipuan saham adalah merujuk kepada kaitan antara tawaran dan jualan saham contohnya :

a) Pelaburan Hasil yang Tinggi (High-Yield) – merujuk kepada pulangan yang dijamin berdasarkan “risiko rendah” atau “tiada risiko langsung” dalam pelaburan instrumen sekuriti.

Penipuan ini bermula dengan “broker” mengambil peluang ke atas kepercayaan pelabur dan mendakwa pulangan yang tinggi di atas modal mereka. Pelaburan Hasil yang Tinggi dapat dikesan menerusi Skim Piramid, Skim Ponzi, Skim Perbankan, Skim Wang Pendahuluan, Komoditi (Forex dan Logam Berharga) dan Nota Promisori.

b) Penyelewengan Broker – Skim ini melibatkan broker bertindak melanggar undang-undang bagi mendapatkan keuntungan daripada pelaburan klien. Biasanya ini melibatkan perdagangan yang tidak sah dan penipuan dokumen.

c) Manipulasi Pasaran – Skim ini melibatkan individu atau sekumpulan manusia yang mengganggu-gugat pasaran yang teratur dan adil bagi mendapat keuntungan. Cara yang dilakukan biasanya melibatkan percubaan mengawal harga pasaran. Faktor inilah yang paling ditakuti pelabur.

Menggunakan pendekatan pelombongan data (data-mining), sekurang-kurangnya 5 kaedah dapat dilakukan.

a) Mengenali Corak (Pattern Recognition) – penggunaan kaedah ini adalah sama dengan melihat pada tren (trend) yang menimbulkan syak bahawa aktiviti penipuan sedang berlaku iaitu :

i) mengesan pedagang yang tatalakunya (behaviour) menampakkan aktiviti penipuan,

ii) mengesan saham yang berkaitan dengan aktiviti penipuan berkenaan,

Sebagai contoh, aktiviti berikut mungkin menampakkan unsur penipuan :

a)    Kecairan, pulangan dan volatiliti adalah lebih tinggi untuk saham yang dimanipulasikan berbanding dengan sampel yang dikawal.

b)   Keuntungan juga dapat dicapai dengan manipulasi harga saham bagi deflate harga (jualan jangka pendek – short-selling) tetapi kebanyakan pakar manipulasi akan cuba untuk menaikkan harga saham,

c)    Waktu Tutup (Closing Hours), Suku Berakhir, dan Tahun Berakhir adalah “ruang bagi manipulasi berlaku”

d)   Kenaikan drastik dalam volume dagangan dan volatiliti pulangan selalunya akan diikuti oleh manipulasi harga (dalam banyak kes)

Apapun, kaedah mengenalpasti corak adalah antara cara terbaik di mana pihak berkuasa dapat membekukan dagangan saham berkenaan.

Namun, proses pelombongan data secara talian dan real-time diperlukan dalam kes ini. Set datanya dalam konteks ini melibatkan secara data dagangan untuk setiap akaun pedagang atau setiap saham dan set corak/tren yang dikenalpasti sebagai penipuan (dilabelkan)

b) Pengesanan Outlier  - mengesan pemerhatian yang tidak konsisten dengan baki data. Ini akan membantu dalam mengenalpasti corak penipuan yang tidak dikenali. Pengesanan yang efektif dapat dilakukan dengan melihat Pancang (spike) berdasarkan keadaan pasaran dan bukannya menggunakan Ambang (threshold) untuk menapis Pancang.

Seperti grup pertama, kaedah ini dapat dilakukan ke atas pedagang dan saham yang didagangkan. Set Data yang diperlukan ialah maklumat mengenai sejarah perdagangan oleh pedagang atau saham yang berkaitan.

Pengesanan anomali akan dapat dicapai menggunakan kaedah pengelompokan (clustering) dan tidak memerlukan data yang dilabel.

Contoh senario : Data diwakili oleh variable statistik (min dan varian) untuk pembelian dan jualan pada tempoh yang tetap (fixed).  Satu parameter akan ditetapkan yang menerangkan jumlah objek dalam peer group dan kawalan ke atas sensitiviti model. Satu objek sasaran akan menentukan ahli dari peer group samada ahli-ahli peer group adalah objek yang paling mirip kepada objek sasaran.

Pada setiap tetingkap masa (contohnya setiap 5 minggu) peer group ini akan diringkaskan bagi mengenalpasti sentroid peer group. Kemudiannya, jarak ahli peer group dengan sentroid peer group akan dikira menggunakan statistik t dan objek yang “lari” daripada peer akan dikutip sebagai outlier.

Akaun dagangan akan dikaitkan dengan objek yang yang dianggap sebagai pedagang yang dicurigai di mana tatalakunya berbeza dengan peer yang lain.

c) Induksi Peraturan (Rules Induction) – mengestrak peraturan yang dapat diperiksa dan digunakan oleh auditor dan pihak berkuasa pasaran saham. Set data yang diperlukan ialah maklumat sejarah dagangan untuk setiap akaun pedagang yang dilabelkan sebagai “disyaki aktiviti penipuan”. Juga ekstrakkan peraturan yang mengenalpasti corak dan ketidakaturan (irregularity) yang tidak dikenali menggunakan data yang tidak dilabelkan (kaedah pembelajaran yang tidak dipantau)

Contoh senario : induksi peraturan menggunakan pelombongan data secara temporal.

Pertama, data siri masa dibersihkan (pra-pemprosesan) menerusi dua langkah :

i)       Penentuan Tempoh Seterusnya (subsequent) dan

ii)     Ektrak corak temporal menggunakan algoritma pengelompokan (min EM dan K)

Juga atrribut data yang relevan dipilih secara manual atau menggunakan pemilihan atribut algoritma.

Kedua, algoritma induksi peraturan (contohnya AQ15) untuk menentukan “if-then”. Satu suasana akan dibentuk bagi menggunakan kaedah yang dicadangkan dan diuji menggunakan set data yang mempunyai data harga temporal (harga, volume, tinggi, rendah, secara keseluruhan 13 indeks trend) untuk 9 saham dari negara X selama 6 bulan.

Keputusan menjual atau membeli setiap saham ditentukan menerusi kaedah pengelompokan dan digunakan untuk menguji ke atas saham yang berbeza. Hasil kajian menunjukkan kaedah Ekstrak Corak adalah yang terbaik di mana ianya melangkaui prestasi garisan asas (baseline)

d) Analisa Jaringan Sosial (Social Network Analysis) – mengesan akaun pedagang yang dikaitkan dengan manipulasi pasaran. Set data yang diperlukan juga maklumat sejarah dagangan untuk setiap akaun pedagang. Di sini, sumber data tambahan seperti sejarah pekerjaan pedagang dan perhubungannya diperlukan untuk disatukan dalam dataset ini. (contohnya – mendapatkan akses kepada mana-mana depositori pendaftaran pusat yang mempunyai maklumat mengenai broker yang berdaftar)

Cara pelombongan data secara tradisi menggunakan klasifikasi, pengelompokan dan peraturan perkaitan selalunya mempertimbangkan sampel poin data yang berasingan (independent). Cara ini tidak dapat leverage kaitan antara sampel set data yang distrukturkan secara komprehensif dan heterogen

Analisa Jaringan Sosial dapat mewakili penstrukturan data seumpama ini di mana nodnya adalah berkaitan dengan sampel data samada objek atau individu, dan sudut yang mewakili perhubungan dan pergantungan sesama objek. Pemetaan, kefahaman, analisa dan mengukur interaksi di jaringan dipanggil Analisa Jaringan Sosial.

Ianya berkesan untuk mencari korelasi yang menandakan berlakunya penipuan dalam pasaran saham bermula dengan perubahan peristiwa (event) dalam pasaran ke bentuk graf (pra-pemprosesan)

Selain itu, analisa ini juga dapat mengesan broker yang bersubahat dalam :

a)    Inflate/Deflate harga saham dengan meletakkan pesanan yang telah diaturkan terlebih dahulu dengan broker-broker lain diikuti dengan aktiviti manipulasi volume,

b)   Memindahkan saham antara akaun bagi mengelak cukai,

c) Mendapat kredibiliti dalam pasaran dengan transaksi jumlah yang tinggi.

Ada yang menggabungkan analisa jaringan sosial dengan kaedah visualisasi interaktif bagi mengenalpasti akaun yang dicurigai.

Antara aktiviti berbentuk penipuan ialah :

a) Perdagangan Kitaran (Circular) – konsisten dalam pembelian dan penjualan volume saham dalam jumlah yang sama samada belian/jualan tinggi atau rendah,

 b) Indikator Primer/Sekunder – menanda akaun yang rendah beliannya dan tinggi jualannya. Jaringan sentraliti akan membantu mencari akaun primer (utama), fungsi ‘f’ akan dikira untuk setiap vertex yang mewakili saiz akaun dan membandingkan harga transaksi dengan harga purata transaksi terdahulu,

  c) Sudut Indikator Utama : Memindahkan stok dari satu akaun ke satuakaun yang lain yang berlaku bila sudut (transaksi) muncul beberapa kali pada 2 vertices.

Analisa Jaringan Sosial terbukti berkesan dalam mencari aktiviti subahat dalam memanipulasikan pasaran dan kaedah bagi memantau interaksi pedagang di dalam pasaran.

e) Visualisasi – ini bukan terhad kepada carta semata-mata di mana auditor perlu berinteraksi dengan data pasaran dan mencari corak yang menampakkan penipuan.

Visualisasi terhadap data pasaran adalah sama pentingnya dengan pemantauan secara real-time dan penyiasatan ‘luar-talian’ (offline)

Kaedah visualisasi akan membantu auditor untuk mengenalpasti aktiviti yang disyaki penipuan dalam transaksi pedagang dan saham yang didagangkan.

Ini dilakukan dengan melihat pada corak di dalam data atau maklumat yang belum disedia untuk dibezakan (discernable). Apa yang diperlukan ialah sejenis kerangka analitikal visual yang menggabungkan treemap 3 Dimensi untuk pemantauan pasaran dan visualisasi berasaskan Tatalaku menggunakan Analisa Jaringan Sosial bagi memantau aktiviti broker.

Dalam Visualisasi 3 Dimensi, setiap sel mewakili satu saham dan saiz sel adalah proportional kepada pemodalan pasalan dan kod warna sel menunjukkan perubahan harga (warna “hijau” untuk “harga naik” dan “merah” untuk “harga turun”)

Ini akan melahirkan satu peralatan yang memantau secara real time dengan tangguhan 15 minit untuk aliran perdagangan asas (harga dan volume)

Keterangan dagangan akan dibandingkan dengan satu set parameter dan pengawasan akan dilakukan ke atas sebarang keluarbiasaan. Analisa jaringan dagangan bertujuan untuk mendedahkan struktur sosial sesama pedagang dan mengenalpasti corak dagangan yang luarbiasa. Nod-nod pula mewakili pedagang, kawasan sekitar setiap node mewakili nilai dagangan dan sudut berarah (directional edges) menandakan aliran dan keseimbangan dagangan.

Satu pangkalan data yang merekodkan setiap corak yang berpotensi ada elemen penipuan akan digunakan sebagai rujukan dan membandingkan peristiwa dalam jaringan dagangan serta mengenalpasti aktiviti yang boleh menimbulkan syak.

Kecairan, pulangan dan volatiliti adalah lebih tinggi untuk saham yang telah dimanipulasi. Oleh itu, mencartakan parameter secara parallel dengan alignment masa yang sama akan membantu penguatkuasaan bagi mengenalpasti corak dan tren yang boleh disyaki.