We often think of financial markets as volatile, subject to constant flux and the whims of the times. Surprisingly, some market features are stalwarts of our financial system, remaining more or less constant over time.

Alternately, there is the murky area of technological change. As smart machines take over human functions and day-to-day operations evolve, markets are moving into uncharted territories.

Let’s look at what doesn’t change. Some of it may be surprising.

Examples include: accumulation and distribution processes, market cycles, institutional players’ need to test price areas, and the wide use of mainstream media to convey a general mood about markets. Other features — particularly those related to actual market trading — change and evolve. So, the manner in which to profit from the market changes more frequently than the nature of the market and its players.

This holds true upon examination of the earliest financial markets. Let’s look at the late 19th and early 20th centuries, in 1903’s "The ABC of Stock Speculation" by Samuel A. Nelson.

The book covers the most relevant articles of the era written by Charles Dow in The Wall Street Journal. To be sure, technology has changed the mechanisms of trading. Tickers have been replaced by internet data feeds; telegraph and telephone orders have been replaced by electronic trading; and bucket shops are now called “contract by differences.” What hasn’t changed is the who and the why: brokers need to lead and drive their campaigns. These are still the factors that ultimately determine market regimes.

The pit, which has long defined the trading world in popular culture, is now history, largely replaced by High Frequency Trading (HFT), which fulfils the same mission: provide liquidity and decrease latency in price feeds. Another classic book, "Reminiscences of a Stock Operator," written by Edwin Lefèvre in 1923, is a fictional rendering of the life of Jesse Livermore, a pioneer of day trading in the early 1900s. The book provides another window into strategies that still undergird the industry. Market latency, for example, already was a key element in determining success or failure. Lefèvre also sheds light on this long-standing truth: marketing regimes matter as much as key strategies and market models. The narrative style of the book makes extracting the industry-relevant pieces difficult, but a keen eye can appreciate its seminal nature.

The examples above should not convey that there are no fundamental differences between today’s market and that of a century ago. While it must be recognised that many intrinsics remain constant, the manner of trade is quite different in a digitally-evolving world with increasingly complex modes of execution. Whatever works today may not work tomorrow. But market evolution is not traced solely to the digital revolution. Even before the advent of technology, examples of change can be found.

“Steidlmayer on Markets: Trading with Market Profile”, written by markets guru J. Peter Steildmeyer in 2002, imparts wisdom gained from his decades of experience. A director at the Chicago Board of Trade and pioneer in the implementation of market data feeds, Steildmeyer describes the evolving nature of markets with two simple examples: How the theory that associated the rail car loading in Chicago with a selling phase was outdated in the ’60s, and how the commodities market experienced a huge change in the ’70s when commodity mutual funds began trading. Both examples belong to the markets’ pre-computer era. They both show how changes are a concomitant feature of financial markets.

It is, however, the introduction of digital technology that has truly transformed the business of trading, upending practices that defined day-to-day business for well over a century. Technology has led to both quantitative and qualitative differences in market operations. The one absolute that market observers can hang their hat on is that more change is coming — and it’s coming fast.

Artificial intelligence is poised to replace at least some human function. In short order, professionals steeped in experience and market know-how may see some of their duties taken over by machines that think, reason, and communicate instantaneously with other smart machines.

This is the new frontier. No one knows exactly where it will lead.

This much is known: hedge funds are changing in 2021. Quantitative trading has evolved from basic statistical models into the realm of AI, and machine learning. These changes are ongoing, and the adaptability of reinforcement learning is a brand-new dimension of quantitative research.

Data analysts also are affected. The information flow is increasing in volume and complexity, as is the workload. The technology-driven marketplace demands more data throughput, which professionals must consume and correlate at the pace of business. In order to manage this challenge, Kaiju data analysts have had to create new systems to make the increase manageable.

Then there is the increase in data volume and workload. Solutions are available, but require new tools, mainly technology stack and scaling infrastructure that address the challenges at hand.

Less well-known are the impacts of changes to regulatory and market access, such as the increased trading volume taking place in broker-dealers and internalisers. These changes affect how and where trades take place, and alter how market data gets to other players. Despite ongoing debate around market access, fairness and transparency, practical implications can be found when analysing and using data to describe models. The buy side must adapt models and strategies as data provided by the order book contains less information than before.

To explain: in a traditional, regulated, quote-driven market, all data can be used as complementary information for trading models. However, increased use of broker-dealers and internalisers effectively hides bids and data from the rest of the market. This is because both broker-dealers and internalisers cross their customer orders against their own inventory, so trades are not taking place in the regulated exchange venues they have in the past.

A graphical representation of this change can be found in the embedded chart, which depicts the percentage of traded volume without order book information for a mid-size US equity asset. With small variations, the chart will be similar for any other asset. The graphic shows how clearly this percentage has increased during the last few years.

While this situation is not new (dark pools have existed for a long time), the increased market share these players generate changes the data, and these changes need to be taken into consideration. This is one of many examples that demonstrates how both the data and the market evolve. The changes can, in fact, lead both to competitive advantages and enhanced information, but strategies must adapt to the evolving environment. This is one of the key factors leading to profit in the modernized market.



Mario Emmanuel is an external consultant advising Kaiju on market data. He has a M.Eng. in Electrical Engineering who has specialised in market data processing and quantitative analysis. His interests include actionable quantitative models using market volume data and the technologies required to process and store order flow market data. He has almost two decades of professional experience including projects processing data in real time in Tier-1 telecom networks.