How Does AbleMarkets Pick Up HFT and Institutional Participation?

Follow us

LinkedIn Twitter Medium

Also by Irene Aldridge

The secrets unveiled

Please note the blog posts on may contain sponsored or promotional content. To sponsor or promote content on, please click here.

I am often asked about how AbleMarkets actually knows what High-Frequency Traders (HFT) and Institutions do. The answer is not a straightforward “hey, we do a feature selection” or “order-book imbalance.” I have written a few books and papers on this subject, with perhaps my most popular being "High-Frequency Trading", “Real-Time Risk” and “Market Microstructure and the Risks of High-Frequency Trading.”

An important idea that helps us understand the nature of market participants is the footprints that they leave in the markets. Consider a market to be a closed system, a cooking pot, of sorts. Different ingredients, when added, leave different traces in the pot. Some release puddles of fat, while others add particular texture or flavor. The markets are quite similar, actually, in that every agent leaves their distinct footprints in the markets, and those footprints can be seen in the data.

The data, of course, is huge. For every financial instrument, there are reams of buy and sell orders, both right-now (“market”) and wait-till-later (“limit”). There is an infinite spectrum of various combinations of orders, too. Different agents place different orders in different ways. Market makers use HFT strategies to gradually build up their liquidity supply in anticipation of their clients’ orders. Institutions come to the dark markets to meet their market makers. Other HFTs arbitrageurs (often from the same firms as the market makers) come to arbitrage the dark markets activity with the lit markets. Banks’ execution desks also filter into the mix providing execution services to their clients.

Consider a specific example: today’s morning trading in AAPL. Figures 1 and 2 show Institutional and HFT participation in the stock of Apple, Inc. (AAPL).

Institutional investors have been selling AAPL this AM. Typically, their execution involves slicing orders according to 1) time (TWAP), or 2) value (VWAP). These patterns are clearly seen in the AAPL data this morning and early afternoon. At the same time, high-frequency traders (HFT) have been 1) preparing liquidity for the institutions, and 2) trading around institutions to incorporate information into the markets. HFT patterns exhibit strong short-term mean-reversion.

But how do we really estimate these values? Well, first and foremost, AbleMarkets stays on top of the latest developments in the space by understanding all the latest research and algorithms and incorporating those in our models. Our collective experience spans decades in the academia and financial sector.

For the mathematically inclined, consider the trading universe as a Poisson process. It is the same mathematical construct as the one used for measuring and predicting the subway train traffic: some trains arrive, others leave, much like the orders in the markets. There is a certain hum to the train patterns, and when you don’t see one for sometime, you know that something is off the rails somewhere, so to speak. The same idea applies to the markets: knowing how the strategies fit into the Poisson process allows us to document them and successfully extract them for our clients.

Is this all? Not quite. With Artificial Intelligence, we are able to process vast volumes of data in seconds to arrive at meaningful conclusions about market activity. We are able to train the AI to essentially do what the microstructure researchers do all day, except AbleMarkets does in microseconds what traditional researchers take days or months to complete. Our AI algos go way beyond transformers and other large models into the realm of identification.

Would you like to learn more? Connect with us today!

AbleMarkets App:


Subscribe Now
More Real-Time Research