Thanks a lot for the insights in this Forum. I’m highly interested in exploring market inefficiencies and correlations. Currently I’m backtesting certain automated stratgies under ideal conditions. Is it still worth to dive into the theories like Spread Trading or do HFT algorithms eliminate any opportunities? Where do you still see potential? I’m happy looking forward reading your thoughts.
Thanks
Chris
@sonaht, your interest in exploring opportunities in futures trading is a great starting point. When considering method X or any other trading strategy, there are essential aspects to keep in mind:
Having the Right Tools: Do you possess the tools necessary for the method you wish to pursue? Understanding the tools and how to use them is vital.
Understanding Risks: Do you comprehend the risks associated with the method? Every trading strategy has associated risks that must be well-understood.
A common misconception is that “spread” trading in futures is less risky due to potential market correlations. In reality, spread trading has the same risks as outright futures. High Frequency Trading (HFT), which refers to execution speed, is part of this complexity, but it’s not the only factor. HFT is a specialized field where speed is essential, especially when scalping for 1-2 ticks. Since most individual traders don’t have the technology to compete at this speed, focusing on more accessible strategies can be a wiser approach.
Trading isn’t merely about executing trades; it involves continuous learning, self-awareness, and real experience. While starting with hypothetical or demo trading is helpful, nothing replaces the insights gained from live trading. It helps you understand your approach to risk and adapt your strategies based on real-time experience.
Finally, it’s important to recognize that futures trading involves leverage, meaning that it can magnify both gains and losses. If this doesn’t align with your comfort level or financial situation, exploring other avenues or asset classes may be advisable.
Your journey into trading is unique, and there’s no one-size-fits-all method. Embrace the process, learn from experiences, and never hesitate to seek guidance or clarification. There are no “perfect” methods, only continuous growth, learning, and adaptation to find what resonates with you.
Feel free to reach out with further questions.
Best,
Matt Z
Optimus Futures
There is a substantial risk of loss in futures trading. Past performance is not indicative of future results.
Thank you Matt for pointing out important fundamentals. Live experience and risk awareness is essential.
As you mentioned “HFT for individuals”, let me add one question here: Some of my algo-strategies are based on short term trades including scalping, that require speed and minimum slippage. Optimus Futures got my attention as I found out that you are offering HFT soluitions in cooperation with rithmic and Aurora data center.
Please, let’s ignore my personal skill, budged or strategy for a while and focus on the technical potential of this offer. I assume one of the best setup is a dedicated linux server within Aurora datacenter. The prices seem not to bee unreachable for individuals, especially when the infrastructure (including optimized code) is fast enough to execute working strategies. Can the provided setup compete with other HFT soultions in terms of speed? Please share your best guess. Does anyone has experience?
Thanks in advance
Chris
The fastest off the shelf software are the Autospreader products from companies like Trading Technologies, CQG, Stellar, BTS, Cunningham/CTS etc to name a few. These products are fast, and I mean really fast. They can change an order over 100 times a second and are all based upon being colocated (which is included in the pricing of some of the products like TT and CQG). Should be noted that this is still significantly slower than true HFT. The problem for retail though is the price tag. Not only will the software put you back $2k+/month but the CME will also probably classify you as using a Semi-Automated Trading System and as such charge you data as category A-1 Non-Display Usage which will add about $500/month PER EXCHANGE to your data fees.
An alternative is to write directly to the exchange through an API. With the exception of TT’s ADL if your looking to do anything more advanced than autospreading then it will involve this route. Many companies offer API access including TT, CQG etc etc. Just like data where not all data is the same quality/speed, not all API access is the same either. Cost of API access varies a lot from the virtually free to >$2/month. While I don’t have evidence to support this next statement, I am sure you will find that you get what you pay for, and that cheaper APIs will be a lot slower than expensive ones. Of course you can always write directly to the exchange itself but that will involve colocation (~$1k/mth) and direct connect data feeds (>$2k/mth).
Getting back to the trading itself, as somebody who considers themselves a spread trader (and even a spread of spreads trader, ie Butterflies) I can confidently say that if your trying to go the lower risk route of trading spreads on a relative value basis, buying what you think is cheap and selling what you think is rich (as opposed to the higher risk outright speculating on spread direction) things get more and more competitive by the year and has become a very crowded space. There are a lot of people trying to do this now since it’s so easy to do. It’s still possible to find trades where you can make an easy 1 tic, but only for odd lots/day - which doesn’t add up very quickly. Finding a consistent edge, that you can consistently make money on is now very difficult. I think I discussed this in the original thread you posted in yesterday. It’s just got a lot more difficult.
Welcome back!
Thank you for taking the time to write this.
It’s great to hear from you.
Matt Z
Optimus Futures
@sonaht If you have developed strategies that rely on scalping and are highly time-sensitive, Rithmic’s R|Diamond solution could help you achieve sub-millisecond execution speeds. However, to realize such low latencies, you would likely need to co-locate your equipment in the same data center as Rithmic’s matching engines.
Ignoring costs per your request, I would point out that the frequent trading required for scalping strategies can generate enormous market data and exchange fees. Retail broker commissions would almost certainly prevent profitability given the slim margins in HFT trading. You would likely need to obtain a CME lease to get competitive institutional rates.
To implement your strategies successfully, you will need to closely analyze and optimize all expenses including market data feeds, commissions, IT infrastructure, networking, etc. The margins in high-frequency trading are razor thin so every microsecond and penny counts.
The key challenges will be reducing latency as much as possible, gaining access to exchanges at a competitive cost, and managing data and trading costs. If executed well, Rithmic’s solutions could provide the speed you need. But making HFT trading profitable demands cutting edge technology, infrastructure, and incredible attention to minute details.
Matt Z
Optimus Futures
@Mod-MattZ @Digital_Jester Thanks again for taking your time and providing this valuable knowledge.
At first glance 2-3k/mo sounds like a lot. As a human trader I would not like to step into the ring with so much psychological pressure. Luckily, algorithms don’t have emotions and don’t care about the lot size. As already mentioned, they need to be on point and optimized down to the last bit. I guess on HFT level strategies need to primitive to guarantee speed and fast response times. I need to evaluate the limits and trade-offs here and I guess I have to bite into the bullet and pay a high price to archive certainity. At least it’s fun.
Have a nice weekend!
Chris
Yes, but humans have emotions when they look at performance.
When you’re trading, whether it’s using “fancy” things like algorithms, high-frequency trading (HFT), or even simpler methods, you’re going to face ups and downs. Every trader does. Sometimes, you’ll see your money go down, and you’ve got to figure out if something’s wrong with your strategy or if it’s just a bad day in the market. As @Digital_Jester said, it is becoming harder and when it does you face decision making, and as much as we’d like to believe we are all rationale, emotions will play a role.
I really believe, and I’m saying this from experience, that every new trader should start with point-and-click trading. That means making trades manually, one at a time. It might seem old-fashioned, but it can teach you a lot about how emotions play into your decisions. Understanding this can potentially make you stronger mentally and help you come up with strategies that work for you.
Matt Z
Optimus Futures
Yes, but humans have emotions when they look at performance.
Belive me, I monitor my live algorithms all the time. Monitoring, Testing, Maintining, Optimizing are essential. On HFT level it will be a full time job that might even not performed by one person only and I need to create corp. around.
I agree, I’m happy that I started manual trading 7 years ago and I’ve made all the mistakes many beginners make: Martin Gale, Grid-Trading, Averaging, Revange Trades, No/Bad Stop Loss, Revange Trades, Over Trading, Over Leveraging, Impulse trades and any combination of it… you name it. I blew up a lot of Demo accounts and lost quite a lot of real money, but I’ve learned valuable lessons.
However, I claim, whenever you find a clearly defined ruleset, clearly defined setups as well as clearly defined entry and exit strategies, there will always be an algorithm that can execute those rules way better than a human trader can. Also, you don’t want to backtest your strategy many times with different parameter setups e.g. on a 15 Minute time frame over many years.
People could argue that backtesting is not essential but it tought me valuable lessons. For instance, many rulesets based on predicting price movements can work perfectly over years but then completly get murdered by the market at some point of time. It must be traumatizing for a human trader experiencing this. Even worse, they can come up with wrong conclusions and start questioning themself for no reason. Other traders might belive they found the holy grail but are not aware how much risk they take and after some month a few trades take all their profits away.
It’s fine if people want to classify me as a beginner, since I did not experience all failures in real time with real money. But I’m happy that I did not always spent month or years to blew up my complete account. I got already traumatized after seeing the backtesting results of my algorithms. Since then, I got more realistic expectations.
Today, everybody teaches rules and mindsets but it is hard find people that teaches the upredictability of the market. As market rithm and structure can change any time unexpected there will always be a phase where many (if not all) prediction based strategies get broke. Whatever your stratgey is, it doesn’t matter how fancy or complicated, executed manual or automated, on live or demo account, using trend or price action, the market can ( and will at some point of time ) do exact the opposite 100 times in a row. Almost no one will tell you that. Most people tend to measure the future against the past or follow their own conflicting interests.
As I’m trading from central europe there is no chance for me to test and execute some strategies manually. Especially if they require speed and low slippage I need to be co-located and automized. Frankfurt/Eurex is 1.5h away from my place but the RRT are still not sufficient. People could argue I should focus on “normal” predictive strategies like other retail traider do. But I love programming and optimization, so maybe co-located HFT fits best to my personal trading style. I belive that best results come when people do what they love. How can I find out without trying?
Have a nice sunday.
Chris
In futures trading, it’s crucial to find a strategy that works for you. Consider off-the-shelf software or direct API access to exchanges. However, keep in mind the costs and competitiveness of the market, especially in spread trading. Despite the challenges, with careful research and strategy development, success is possible.