Algorithm trading (also called as Black Box Trading) is very advanced & sophisticated trading mechanism which uses complex mathematical formulas & models to make quick decisions & transactions in financial markets.
According to Technavio, market research Company, algorithm trading is expected to grow at a CAGR of 10.3% globally by 2020. Traditional broking firms and investment institutions are building tech savvy algorithm platforms to compete in the global market. Apart from the profits, algorithm trading is providing market intelligence which is helpful to take long term trading decisions.
There are no specific rules for developing algorithms, they evolve like the nature does.

How Institutions are doing Algorithm Trading:
There are various algorithm trading strategies implemented by big trading firms and institutional brokers to take the advantage of technology in executing speedy and secure transactions. Below are some of those strategies.

1) Using Google Trends to quantify the trading behavior: Most of the advanced trading algorithms predicts the future stock prices using the past trading analysis. But the recent studies conducted by Nature Scientific Reports indicated that Google & Wikipedia searches can be used in trading algorithms during information gathering phase. Using this, weekly/monthly average and moving average for a particular term (e.g. Loans by Private Banks) can be measured to buy/hold/sell/re-buy stocks for that period.

2) Mean Reversion: Mean reversion theory states that the prices of stocks (low & high) converge to their mean value over time. So Algo traders identify the price ranges and execute the orders automatically if the price moves out of the defined range. The simple strategy is to sell high performing stocks & buy low performing stocks if they are outside the mean value.

3) Arbitrage:  Buying a stock in one market at lower price and selling it in other market at a higher price where the same stock is listed in both the markets gives you arbitrage opportunity. Algorithm can be developed to identify price differential in several markets and perform operations considering the transactional costs to generate profits.

4) Volume Weighted Average Price (VWAP): VWAP is the ratio calculated by
(∑ Number of Shares Bought x Price of the Share) ÷ Total Shares Bought
This strategy is used by mutual funds & institutional investors to break the large stock orders and execute the small ones so that market prices are not affected by large orders. With the help of advanced mathematical algorithms, VWAP will benefit the institutions by determining average price for buying and selling stocks.

5) Time Weighted Average Price (TWAP): The Strategy is to break large orders of stocks into small ones and execute orders evenly over a specified time period . The algorithm using TWAP can be useful to reduce the market impact by executing the orders close to the average price between the start time and end time.

Benefits of Algorithmic Trading:

These computer directed trading models turn information into intelligent trading decisions by analyzing every quote and trade in the stock market identifying the liquidity opportunities.

Since algorithms are written in advance, they are executed automatically as and when required. This leads to 3 benefits: Speed, Accuracy and Low Cost

• Speed: The trades are made in a fraction of seconds by executing on multiple indicators simultaneously. The execution being done at a faster pace, greater opportunities are available at better price.
• Accuracy: The algorithms are double-checked to ensure orders are made correctly. It avoids the human error of poor decision making. It also allows to run algorithms on past data (called as backtesting) to check the feasibility of the model. This allows users to remove flaws in the model, if any.
• Reduced Cost and time: As the trades are monitored by the algorithms itself, traders need not follow up on stocks frequently. This saves time and money giving more time to explore and focus on other opportunities. Moreover the models are up 24x7.

Market Based Benefits:

• Off Exchange marketplace: Algorithmic trading prevents information leakage for traders who do not wish to share quotes, connecting dark pools and creating greater liquidity, improved pricing on shares and higher brokerage.
• Cross Asset Trading (Creating arbitrage opportunities): Traders can access multi liquidity pools different various asset classes capitalizing on high frequency arbitrage opportunities. (A trader may, for example, buy equity, hedge with a derivative of the equity, and take out an FX position—all within the same strategy.)
• Risk Monitoring: Algorithms, on real time basis, calculate risks continuously and accordingly hedge a position in the market, leading to minimum losses.
• Forex Market: Algorithms continuously monitor changes in exchange rates and thus lead to faster execution of currency exchange.

Banks use them to quote market prices on real time basis and maintain a pre specified level of risk exposure in holding minimum currency level.

Risks in Algorithm trading:

Where speed is considered as one of the benefits of algorithmic trading, it is bundled with the risk of losing a great deal of money.

• High volatility: Algorithms when reacting to change in market conditions may widen the bid-ask spread, or may stop trading creating excessive volatility and hampering liquidity.
• Chain reaction: Due to high integration in the global markets, slowdown in one market moves over to other markets and asset classes creating a chain reaction. (As happened in subprime crises)
• Incorrect Algorithms: a faulty algorithm may pose risks of error trades and market manipulation causing millions of losses in a very short period of time.

The algorithms can only be accurate once the optimal selection is done.

• Lack of transparency: the trader cannot monitor the algorithms when they are executed, leading to differences in expectations and results.
• Lack of Knowledge: With brokers offerings multiple algorithms, Buy side lacks the tools to understand which algorithms would suit their portfolio. This also downgrades the quality of algorithms assessment.
• Many algorithms use similar basic functions to execute upon current market conditions which might lead to unfavorable outcomes.
• Imbalances in the market: Where few traders have the means to acquire sophisticated technology that executes orders, the others are still trading manually. This causes fragmentation in the market leading to liquidity in the short run.

Being based on computer systems, any technical default or error may cause system outage leading to backlashes in the market.

The traders lose confidence in the markets due to uncontrollable algorithmic results and repeated market issues.

Conclusion:

So the algorithm trading can pile up the profits of big trading firms and institutions and also provide the speed, accuracy, cost reduction and reliable market insights. But it is also important to build such algorithms with utmost accuracy and provide high level security to prevent it from cyber-attacks.