Algorithmic Trade Execution in different Asset Classes

10 min read

By Punit Nandi

Since the advent of Investment theory, there have been different types of asset classes that have been investment avenues for Investors and traders. These asset classes have not only increased in number but also have proven to be great hedging platforms worldwide. In this blog, we will discuss a few of those asset classes and how different asset classes have different parameters of algorithmic trade execution.

Below are the asset classes that we will cover in this blog:

To begin with, let us first understand what is algorithmic trade execution and how is it different from manual trade execution for different asset classes?

Algorithmic trading involves executing orders utilizing automated trading instructions using variables such as time, price, and volume in slices or in full over time. It has many advantages over manual trade execution such as speed, zero human emotion, scanning multiple indicators in microseconds, executing without continuous supervision and many more. Let us now discuss algorithmic trade execution for different asset classes.

Cash equity

As of 2016, the equity markets comprised 87% of the global market capitalization. Since the time algorithmic trading was introduced to the markets, more and more traders have pitched into play. For equity, the steps are straight forward, you set the correct parameters and create a robust risk management system, the algorithms can do fairly well for you in the markets. Below are some of the features of algorithmic trade execution out of many for the equity markets.

First and foremost, we need to have a proper backtested strategy that has performed pretty well with both in-sample and out-sample data. This is common for all the asset classes and must be a golden principle while you read this blog.

Next, we focus on the parameters of the trade, remember, setting unusually hyped profit margins will only destroy your chances of a high performing strategy. So, tweaking the parameters of the strategy is great, but keeping in mind that the end goal of the strategy is to generate alpha, rather than costing a trade is wonderful!

Limiting losses in a strategy with the best possible stop-loss margin is another skill that is developed over time. Algorithms which buy or sell stocks frequently might lead to losses, if the parameters of the trade are not able to predict the correct market moves. Given that the volatility in equity markets is high, a perfect exit strategy can make or break the trade

One of the many peculiarities of the equity markets that sets it apart from the other asset classes is the high volatility. Equity markets are run by sentiments very quickly and that should be factored in while creating the strategy. One of the many reasons why traders are not able to generate alpha is because of weak market sentiment understanding. Understanding the psychological mindset during a bull or bear run is extremely important for extending winning periods

Another parameter for a good strategy is to factor in corporate events. Creating situation based algorithms that analyze and track corporate events help in predicting sudden market fluctuations. Quarterly profit announcements, top-management shuffles or clashes, buybacks or stock-splits always signal a probability of price change and either tremendously benefit or generate drawdown positions for traders or investors

Next, we will discuss about Fixed income as an asset class, and how is it different from the other asset classes. We will keep our parameters specific to Bonds, simply for the reason that the blog’s intention is to touch upon the peculiarities of the asset class in general rather than product by product peculiarities.

Bonds

The largest out of all the asset classes is Fixed Income. One of its major components is the bond market which represents loans made by an investor to a borrower (typically corporate or governmental) in the form of a bond. Creating a strategy that factors in the below-mentioned peculiarities will help in generating alpha.

Unlike other asset classes, each bond has a different specification altogether like different interest rates, maturity periods, terms and conditions. Therefore, the algorithm which scans different bonds and identifies the best bond investment must account for all these factors

Bonds come with a yield factor. Yields on the bond may be categorized into three parts :

1) Yield-to-Maturity(YTM) - If we hold on to a bond till its maturity and if all the coupons are reinvested, then the YTM is the measure of the return on a bond. For example - If the YTM is less than the bond's coupon rate, then the market value of the bond is greater than par value (premium bond). If a bond's coupon rate is less than its YTM, then the bond is selling at a discount.

2) Yield to call (YTC) - It is the rate at which a bond will be called if an investor foresees a slightly higher yield at a particular date before maturity. 

3) Realized yield - It is the yield for the time period an investor is willing to hold the bond. Creating a bond trading strategy that executes a long or short position based on the yields might just give an edge over other traders. For example, a $1000 bond maturing in 3 years, with coupon and ytm rate at 3% each. If the bond is sold exactly after one year at $960, the loss on principal is 4%. The coupon payment of 3% brings the realized yield to a negative 1%. If the same bond is sold one year later at $1,020 for a 2% gain in principal, the realized yield is increased to 5% due to the 3% coupon payment.

Any investor will look into the creditworthiness of the bond issuer after ascertaining that it has been positively graded by independent rating agencies (mostly AAA+ or AA+). An increase or decrease in such a rating might affect the bond prices and generate profits or losses. Thus factoring in ratings of a bond and creating parameters to resist such changes is crucial to consistent profits. Credit-rating is a very exclusive attribute of the fixed income asset class and is a key differentiator from other asset classes.

Global economic outlook and Interest rate risk also plays a key role while trading bonds. Opportunity cost is huge in the bond trading world as the other asset classes, an increase in the interest rate forces the existing bond to trade at a discounted price and vice-versa. As interest rates of a country have everything to do with its central bank’s perception over the domestic economy and macroeconomic factors affecting it, strategies involving future forecasts can be delved into with proper backtesting.

The last factor about fixed income bonds that highlights it among other asset classes is the huge variety it offers. Expecting a higher return from a government bond would not be as much logical as from a junk bond. However, given that the risks involved in trading junk bonds are way higher than a government bond, backtesting the beta and Sharpe ratio of the strategy is extremely important while creating a capital-efficient strategy.

Derivatives

Derivatives as an asset class have excellent hedging opportunities and are used by traders for maximizing profit potential with minimal risk exposure. The premium based structure of the derivatives market is another attractive feature that differentiates it from all the asset classes.

We will discuss the aspects of listed derivatives such as futures and options which are more appealing to the retail trader

First of all, as derivatives acquire value from the underlying asset, longing or shorting the same underlying asset over the period can generate extra gains in the form of dividend or bank interest and provide an extra margin to the traders. However, the requirement of extra capital and high risk involved in the strategy must be factored in.

Time value is another important factor that differentiates derivatives from other asset classes. The time value of an option, when traded with a very small maturity period can eat up all the profits contrary to what was expected. Therefore, a better option would be to analyze, backtest and create a strategy that predicts positive gamma and optimal time horizon while executing a long call option and gain fair returns. 

While creating a strategy, one important factor to notice is that the relationship between derivatives and equity is partly equivalent to gamma and delta. The equity markets might have an effect on the derivatives market and vice versa as both the markets have a causal relationship. Pair trading strategies designed to extract benefit from this causal relationship with efficient algorithms have plenty of opportunities and can be an excellent hedging opportunity. In fact, it could well be that all the major asset classes might have some causal relationships like the equity-derivative markets, it is how quickly it is identified and statistically proven for advancing to the step of trade execution.

Open Interest is another important factor that might be crucial to a derivative trading strategy. Open interest is when a buyer and seller come together and initiate a new position of one contract, this results in the open interest increasing by one contract. Should a buyer and seller exit one contract position on a trade together, open interest decreases by one contract. Open Interest is a relative indicator of the strength in the market and can be subsequently used in trend-based strategies.

Factoring in the volatility on the option expiration date is another important feature that must be explored. Given that the time value is already in its maximum value, the premium of an option is usually very minimal, creating algorithms that can take advantage of this opportunity can result in profits. However, given that most players in the market are already aware of this factor, there is a lot of fluctuation on expiry day and only a properly backtested quantitative strategy with robust risk management parameters survive such market conditions. 

Commodities

One of the fastest-growing asset classes in terms of technological penetration is the commodity sector. The Commodity Futures Trading Commission estimates that algorithmic trading accounted for 74% of orders in 2015, and 68% in 2016 for US markets. The number itself proves how the sector has been grasping algorithmic trading. Factors which are exclusive to trading in commodities are :

Local and Global economy plays a big role in the commodity markets. For example, overnight changes in oil prices or changes in government policies specific to a commodity might just push the price of the contract higher/lower the next day. So, a lot depends on how a trader prepares for such moves and how algorithms are designed to auto-hedge such positions.

Transportation and storage costs play a crucial role in the commodity market. For taking a short position in an agro-commodity trade, transportation & storage costs should be quantitatively modeled and backtested over a definite horizon before executing a trade. Strategies revolving around commodities like rice, wheat, maize, etc which require a fair amount of storage cost must be factored in. Storage cost is again one of the rare peculiarities of the commodities market, which is a key differentiator when it comes to comparing it to other asset classes.

Local and International commodities must be segregated while creating a strategy. Global commodities such as gold, silver, steel, etc are traded massively and any global news relating to this asset class has a higher probability of stochasticity than a local commodity which is mostly traded domestically.

Seasonal based agro-commodity strategies are very prevalent and can work well to generate alpha. Monsoon season brings with it either good or bad news for the commodity sector. The demand for commodities is directly proportional to the rainfall. Seasonal based strategies that are properly backtested and can predict such seasonal moves must be executed with proper precision. Algorithmic trading might just be the perfect solution to such volatile markets where the commodities can be executed at the best possible price with the required quantity. Commodities being seasonal also has effects on the companies producing them, which creates a correlation in price movements with the producers. Pair trading strategies based on such relationships can be tested for potential gain. This has again been one of the heavily profitable trading ideas for traders when compared to the other major asset classes. 

Demand and supply is probably the main reason why commodities are traded. Proper research both quantitatively and qualitatively must be undergone before taking a position. It might well be that an aggressive algorithm is executed, however, due to the CPI or quarterly production report being extremely low or high numbers, heavy drawdowns may be experienced. So, factoring in the economics of the market or even creating a strategy that considers all the relevant factors relating to the production of a commodity will help to limit losses and reap in profits.

Forex

One of the most highly traded and active market is the forex market. Let’s discuss some of the factors that differentiate it from the other asset classes.

Economic policies in a country govern a huge part of forex market fluctuations. If inflation is high, the value of the currency depreciates resulting in a weak currency. However, high-interest rates attract foreign investors and thus the value of such currencies might just increase. Algorithms that can take advantage of such rate changes statistically and are backed by tested strategies can help in generating alpha. This is a key factor that also affects other asset classes in general but to a much lesser extent when compared to forex markets.

Political tensions are another major factor that affects the forex markets. The recent Brexit tension had a very serious impact on the USD/GBP currency pair and resulted in high volatility. Political statements and judgments define how a country will be run on an economic front and thus strategies which factor in such political insights have a better chance of predicting the markets and thereby make some quick cash

Economic calendar releases among other factors are very crucial to a trading strategy’s success. Trade Balance reports, manufacturing indexes, CPI Rate Change, Unemployment rate Index, crude oil inventories are all some of the calendar releases that affect the country-specific currency. For example, when the Energy Information Administration (US) reports oil inventory numbers, market volatility is usually high right before or after the announcements. Asset classes like forex are heavily influenced by such reports and must be always on the radar of a trader.

A link among Oil, Gold and the US Dollar might also be experienced in the forex markets. Oil price is significantly and positively affected by gold and USD. The US dollar is negatively affected by the stock market and significantly by oil and gold price. Indirect impacts always exist which confirm the presence of global interdependencies. A beautiful explanation of this relationship has been presented in the paper, “Oil, gold, US dollars, and stock market interdependencies: a global analytical insight“ by Mongi Arfaoui & Aymen Ben Rejeb. I recommend reading that paper for deeper insight on the same. 

Conclusion

We can see that all the asset classes have distinct and individual factors to be considered and thus while creating a strategy and backtesting data, there needs to be a proper understanding of market microstructure and how different aspects of the financial markets are connected. We also understood certain golden principles that can work in most of the asset classes like having a realistic profit or stop-loss targets, understanding the specific economics of the asset classes, giving weight to factors that might impact externally and parameter evaluation which gives confidence in building the strategy. What we discussed in all of the asset classes can be compared to a few drops of water to an ocean that is available out there. I would recommend books like “Expected Returns on Major Asset Classes“ by Antti IImanen for exploring this topic. Thus, ideas with proper backtesting can generate alpha if properly executed.

If you want to learn various aspects of Algorithmic trading then check out the Executive Programme in Algorithmic Trading (EPAT®). The course covers training modules like Statistics & Econometrics, Financial Computing & Technology, and Algorithmic & Quantitative Trading. EPAT equips you with the required skill sets to build a promising career in algorithmic trading. Enroll now!

Disclaimer: All data and information provided in this article are for informational purposes only. QuantInsti® makes no representations as to accuracy, completeness, currentness, suitability, or validity of any information in this article and will not be liable for any errors, omissions, or delays in this information or any losses, injuries, or damages arising from its display or use. All information is provided on an as-is basis.

Machine Learning in Momentum Trading 2024