To most traders, it’s hard to spot algorithms in the market if you don’t know what they look like. 
Most algorithmic trading strategies are so advanced that they mimic the order flow of all the other market participants, and to the untrained eye this can be very difficult to distinguish, especially in liquid markets. Liquid markets are those with the most volume; the products that everyone knows about and everyone wants to trade.

With futures products like ES, CL, and TY, there’s so much action going on every day that this noise masks the movements of the robots. So in order to spot the automation in these markets, we first need to familiarize ourselves in markets that they are more obvious. Look no further than the overnight sessions of some of the most illiquid products out there.



Most traders also only focus on day sessions and are told to proceed with caution when holding positions past the close, or attempting to trade overnight. I’m here to tell you that those warnings are real. Take it from someone who has traded algorithmically in the grains, equity index, energy, and currency markets overnight for over 2 years. I may not have the amount of experience that most algo traders possess, but nevertheless, I’ve seen my fair share. Now there still is plenty of activity in the equity index futures overnight, so let’s use something a little more illiquid as an example.  What I mean by illiquid is markets where there’s very little volume, and even less bids and offers resting in the order book.

A great example would be the grains futures (ZW wheat, ZC corn, ZS soybean, ZM soybean meal, and ZL soybean oil) and the refined energy futures (RB Gasoline and HO Heating Oil). When you’re watching these markets through the Asian or European trading hours, it’s a whole different ball game. Markets that you’re used to looking at are now very thin in volume, with very few orders in the book, and maybe even some gapped prices depending on what product you’re looking at.

The smooth motions that you’re used to seeing during the day have been replaced by jumpy /sporadic quoting. Many times you’ll see an order being placed and removed at the same price over and over again like someone is clicking the mouse as fast as they can, and you’ll even see a rapid move up for just a couple of ticks just to watch the market go right back to the price it was trading at a second ago.

Algorithms in the Overnight Sessions

Welcome to the algo dominated market. This is why professional traders tell you to be careful, because you’re not trading with other humans at this hour of the night. I vividly remember the night in 2014 when Argentina defaulted on its outstanding debt. I was trading the soybean market at the time and we had orders resting and a large position in every single month of the current crop year.

Now for those not familiar, the biggest exporter of soybean products, such as soybean oil and soybean meal, is Argentina. The country exports almost half of all soybean derivative products in the world. Argentina consequently plays a big role in derivative soybean prices, and so any news out of Argentina is worth paying attention to when you’re trading US soybeans. I’ll never forget what I saw overnight when the news broke. The markets literally disappeared. All the bids and the offers that were just in the market were now gone.

Since the algos in the market immediately forecasted massive volatility from this breaking news, the market making programs in these products instantaneously widened their bid ask spreads to prices so far off the previous last traded price that you had to scroll multiple times through the book to even find an order still out there. In the back months (those that are subject to the most volatility), the orders that were there were now completely gone. I was the only one with algos still running in these back markets because I needed to get filled in order to hedge some of my overall position.

I remember staring wide eyed at my screens, heart beating out of my ears thinking, “I AM THE MARKET”. This happened so fast that there was no conceivable way that any manual trader could have reacted. In fact, it took me a good 10 seconds to even figure out what was going on, even more to compose myself and think of a game plan.

Turns out that this was largely market overreaction, and after a while the market makers slowly materialized back into the market. Thankfully, I had some risk logic in my algos that were resting in the current month of the soybeans that caused them to pull my orders immediately when the other orders disappeared, and those algos I had on in the back months were intentionally there to pick up any sharp movements. I just never expected to see movements that sharp.

So hopefully my story resonates to you readers so you can understand how common algorithmic trading really is. Be curious, and pull up some of these illiquid markets overnight so you can familiarize yourself with them and gain the experience needed to identify algos trading in the day session. Once you understand how the algos react and trade, you’ll find that for better or worse, you need to plan your strategies around them.

You may also like

Fixed-Income Trading Desks Making the Switch to Algos
The “Real” Traders
Virtu Extends Services to Large Investment Institutions

Professor Algo