Research & Career

Circular trades refer to a fraudulent practice where multiple traders collaborate to create artificial trading volumes or manipulate stock prices.

To tackle this issue, I designed and implemented advanced algorithms that analyze trading patterns, transaction volumes, and relationships between different accounts. By identifying suspicious circular trading activities, the system can raise alerts for further investigation.

Insider trading refers to buying or selling stocks or other securities based on material non-public information about the company. It involves individuals who have access to confidential or privileged information, such as corporate executives, employees, directors, or brokers(front running), using that information to make trades that give them an unfair advantage over other investors.

To detect such activities, I developed algorithms that monitor each individual trader’s profits. Unusual profit margins or suspicious patterns are flagged for review, enabling prompt actions to be taken against any potential front-running activities.

Concurrent trades or pump and dump involve artificially inflating stock prices through coordinated buying activities, followed by the sudden sale of the stocks at a higher price, leaving other investors at a loss.

I implemented monitoring tools that analyze trading volumes, price fluctuations, and trades similarity between traders. By identifying similar high volume behaviors indicative of pump and dump, the system can promptly alert relevant authorities and prevent further manipulative activities.

One of the primary objectives of the market makers is to actively manage the stock prices throughout the trading day. By continuously monitoring market conditions and analyzing real-time data to avoid misleading prices or manipulations. This helps to ensure efficient market functioning and reduces extreme price volatility.

Price management by market makers in ETFs, especially smaller ones with lower trade volume, is important to ensure that the ETF’s market price closely aligns with its Net Asset Value (NAV).

Market makers’ asset and stock balance management is crucial for several reasons, primarily to mitigate risks associated with holding excess stocks or running out of funds or stock inventory. 

Holding excess stocks can increase the risk of inventory buildup, which can lead to potential losses if the market price of those stocks declines. By buying and selling the exact amount needed in a day, market makers minimize these risks and maintain a balanced inventory.

So one of our main goal is to manage our trades in a way that at the end of each day we have nearly the same buy volume as sell volume in trades for each stock.

Enhancing market liquidity is crucial for a healthy and vibrant stock market. This is my main goal to make a significant increase in market makers daily volume trades. This involves providing liquidity by constantly placing buy and sell orders near best bid and best ask price.

By participating in the market as a liquidity provider, the system enhances trading efficiency, reduces bid-ask spreads, and promotes smoother trading transactions while avoids loss.

Another essential aspect of the system is its ability to generate signals and provide investment recommendations. By analyzing market data, including historical price patterns, supply and demand of current symbol and other factors, the system identifies potential investment opportunities. It generates signals and recommendations based on predefined criteria and use them in it’s trading strategies.

Portfolio selection is an important concept in finance and investment. The process of selecting a combination of assets, such as stocks, bonds, and other securities, with the goal of achieving a specific investment objective, such as maximizing returns, minimizing risks, or achieving a balance between the two. In choosing a stock portfolio, the most important challenge is how we define risk. In this research, we have examined four different methods for choosing a stock portfolio based on different risks. Each of these methods have been tested on various portfolios including 20 assets from different industries. Also, for more accuracy and to include more asset classes, treasury bonds (risk-free investment) and gold (as an asset outside the stock market) have also been added in the selected portfolio so that most asset classes have been examined.

The results show that each of these methods has its own capabilities and the best method is different depending on the market situation and investment situation. For example, the mean-variance method is suitable for investing in stable markets, while the maximum price reduction method works best for investing in more volatile and especially bearish markets. On the other hand, the asset method at conditional risk will be suitable for extremely risk-averse investors.

As a result of all the risk measurement methods, the combination of the Markowitz method and the maximum drawdown method seemed to be the best option. Also, VaR can always be used as a measure that can be reported to the user.

On the other hand, fixed income bonds and asset classes outside the stock market such as gold can be very effective in hedging risk. The increase of funds based on gold bonds or fixed income funds or index funds is one of the commendable developments of the Tehran Stock Exchange in recent years, which allows the audience to arrange stock portfolios with much lower risk.

Another important result of this research was the significant impact of short sales on investors’ risk coverage. It seems very necessary to add this possibility to the Tehran stock market, and it is better to welcome these tools or similar tools that create the possibility of two-way market. Futures or options contracts can also play the role of short sales to some extent.

Capital market is an essential part of modern economies. The prosperity of financial markets and high-frequency transactions have caused the increase and complexity of trading violations. Manipulation of financial markets can lead to price deviations or increase volatility and affect investors and the entire market. Despite the efforts of regulatory bodies to control financial market manipulation, there is still fraudulent activity in the market in various forms such as circular trading, pump and dump, insider trading and other methods. On the other hand, machine learning has become a popular tool in financial markets to identify patterns and anomalies.

In this research, we first introduce 3 risk-based algorithms to detect suspected fraudulent transactions, which are implemented in six months and on more than six hundred OTC symbols. Then we measure the effects of frauds on the market and share.

In this section, I just explain my algorithms that I provided for Tehran stock market for 3 different kind of manipulation.(See Work Experience / stock Market Fraud Detection)

My goal was to understand and predict occurrences of fraud in the stock market and also its effect on market. I utilized cross-sectional data and panel data analysis to achieve this.

1. Predicting Fraud Occurrences
I discovered a meaningful relationship between the standard deviation and return rates of a stock and the number of fraud cases associated with that stock. This revelation suggests that a stock’s volatility and historical performance can be indicative of potential fraudulent activities.

2. Circular Trading Insights
I found that circular traders tend to operate most actively when stock prices exhibit low volatility. This observation is underscored by a significant negative correlation between circular trading and stock price volatility.

3. Insider Trading Connection
Furthermore, my analysis unveiled a significant correlation between insider trading and the standard deviation of a stock’s return. This correlation suggests that instances of insider trading are more likely to occur when a stock’s price experiences heightened volatility.

Future Exploration
This project has opened up avenues for further exploration in the realm of financial data analysis. I am excited to uncover more insights and contribute to the development of robust fraud detection and prevention methods.