Trading is technology driven. These technologies themselves live at the forefront of innovation, as they are are subject to ever increasing demands of capacity and speed. To meet these demands and others, various technologies have emerged which can be seen as playing or potentially playing a role in trading.

1. Cloud computing and data centers

Computing in the cloud is letting remote clusters of computing machinery process instructions, offering advantages of scale and scalability. For the trader this means that their broker can offer fast services, and more directly for them, it means that they can run robots on machines other than their own computer, offering greater stability and processing speeds, compared to that is possible on their own machine.

Cloud computing essentially farms out high-end performance to the many. Cloud computing lives in data centers, where computing devices are clustered to reduce latency and to take advantage of algorithms for routing and processing efficiency. For example, at a data center, the servers of major financial institutions and liquidity providers may be co-located with the servers of the broker, offering proximity and its attendant improvements in speed and efficiency.

2. Artificial intelligence and machine learning

Artificial intelligence (AI) and machine learning (ML) can be seen as technologies which can look at large amounts of data and see patterns and from this data make inferences. While AI and ML have been behind the scenes in many technological processes (including data centers), in recent times it has come more into the open, via generative AI (gen AI).

Gen AI uses machine learning to create a non-deterministic, predictive inference engine, which has shown dramatic results in natural language processing (NLP). The non-deterministic and probabilistic nature of gen AI has helped create outputs which seem human like, but also can produce inaccurate output, unlike a rule based system, or one which uses search to find an optimal result.

However gen AI has the potential to analyse large amounts of market data and in effect process and make inferences on this data, in the way a human would not. Gen AI is relatively static in that it is pre-trained and may need to be partly trained again, using fine-tuning or methods intermediate between fine-tuning and full training. In effect there is a foundational model (itself a neural net) which may be tweaked to fit a particular case.

However as the neural net itself is a black box of extremely complex higher dimensional connections, weights and biases (from its training phase), it cannot be deconstructed and modularised to fit into other tasks. This is a limitation for the fast and complex world of trading, where models failing is a fact of life. So gen AI can be seen as having applications in the pre-trading phase of trading, particularly areas like sentiment analysis.

3. Quantum computing

Quantum computing seeks to utilise the underlying nature of matter to create improved computing machines and algorithms, with the aim of solving intractable problems and seeing speed-ups in other problem types. Quantum technology makes use of a number of features apparent on the scale of particles, including superposition, where a particle can be in multiple states at the same time.

The impetus of quantum for computing is the possibility of performing a kind of massively parallel computation instantaneously. Massively parallel computation is a way to speed up processing, but doing it instantaneously offers the tantalising possibility of performing computations which are intractable even for massively parallel machines. However quantum computing itself has encountered a steep climb, from factors such as decoherence, which is the tendency for the quantum states to decohere from the myriad influences of its environment.

As well as the machinery there are also quantum algorithms, although given the nature of quantum computing these are tied to the underlying machinery much more closely that in conventional computing. Quantum algorithms may however offer improvement in ways to simulate market conditions and markets themselves.

4. Internet of Things

The Internet of Things (IoT) is an important but perhaps less visible technology which is making great strides and inroads in various areas. It is a type of enhanced connectivity, where many devices are connected together for a task allowing data to be collected with an increased level of granularity and density.

Trading is based on data, so it can be seen that increasing the density of this data and its cope could provide enhancement. For example, instead of relying on a wheat report, the trader could potentially access real time data about wheat. This data itself could be analysed through techniques such as AI and ML, to make sense of it and to spot emerging trends.

IoT is being used in the supply chain industry, and this is a way to access real time infomation about possible disruptions and improvements, which can directly affect companies and economies.

5. Augmented and virtual reality

When Google Glass came out, there were those of use who wondered about its utility for trading. While Google Glass was discontinued, research has continued and has been enhanced by augmented and virtual reality. At Google I/O 2024, there were hints at a new route for Google Glass, via AR glasses.

AR and VR headsets already exist as do AR glasses, but one could speculate about something more glasses like (i.e. an everyday wearable) which could allow for tailored AI mediated information (perhaps from IoT !) to be delivered to the trader via their AR trading enabled glasses.

Of course more information is not always better in trading, as the market itself digests inputs in complex unpredictable ways. But perhaps AI could smooth this and allow for an enhanced delivery of pertinent information.

6. Blockchain

The blockchain is new but it is already well developed. This technology allows for trustless decentralised processing. Currently the blockchain does have a use case in trading, as some brokers offer deposits via the blockchain, but there are potential use cases of the blockchain as an alternative backbone for processing financial data. This said, traders already deal directly with the blockchain if they trade cryptos, as these exist on their respective blockchains.

7. Theoretical physics and number theory

Theoretical physics has constructed advanced mathematical structures to explain complex natural phenomena, with great success. However the application of these to financial markets has not met with similar success. One area which has a clearer use case in finance is number theory, particularly in the field of cryptography. Cryptography is important in finance for securing data and is key to the way the blockchain works. Modular arithmetic has a use case in High Frequency Trading, is it is a way to compute large sets of numbers more efficiently.

One of the issues with financial markets and modelling them is the randomness. Traders look for correlations and they look for causality. However the market, particularly through volatility, can break up an correlations and the causality can be lost in the complexity of interactions in the market.

However it might be expected that continuing research and development in physics and number theory might yield improved methods for understanding and making inferences from markets. It might be added that underlying neural nets are mathematical models, so these already have a practical role when using AI to make inferences on financial data.

It may be that the future is what these advances in physics, engineering and mathematics can produce to enhance trading and its myriad technologies, rather than using them predictively, as the market is inherently adversarial and works to arbitrage away any advantage.

Putting it all together

We have identified technology which can be used to make trading faster and more efficient. This is mostly at the level of the processing which goes behind the scenes, however there can be potential improvements for the individual trader from technological improvements to enhance their access to information. We have noted that expecting technology to make trading easier is perhaps not the way to look at it, rather it is a way to enhance what trading can do in terms of speed and the type of information available.