5 Questions with Joseph Signorelli of RCM-X
On May 30, we announced the availability of RCM-X execution algorithm strategies on the TT platform. We sat down with Joseph Signorelli, founder and managing partner of RCM-X, to learn more about the company and their offering.
Prior to forming RCM-X, the technology and execution division of RCM Alternatives, Joe served as a managing director and head of futures for Wedbush Futures. He previously served as managing partner of Wedbush Market Neutral Hedge Fund.
For more, follow RCM Alternatives on Twitter at @rcmAlts.
Tell us about how you came to work with TT and the offering you recently launched on the TT platform. How is this unique, and how does this benefit our mutual customers?
Joe: The team at RCM includes some of the most entrepreneurial folks in the industry, and when looking to expand our business into the algorithmic execution space, partnering with a forward-looking group like TT with their advanced trading tools was top of mind. Working with TT on deploying the algos is a natural fit—with their technology a long-time favorite of hedge funds and institutional customers who demand the best in trading technology, and the ability for TT users to implement both benchmark and custom algos into their tool set on the TT platform was an exciting development. TT users demand the best, in our experience, and the ability to access not just RCM-X’s market-specific algorithms directly from the platform, but also allow traders to adjust algos on the fly from the platform, adjust parameters and set unique defaults per user, was a big consideration.
You’re sitting down with two people, one who has no idea of what algos are and one who has been using algos for years. What do you tell each of them?
Joe: For those new to algorithmic execution, I sum it up by saying your current execution is like bringing a knife to a gunfight. A less violent analogy might be: you’re using a wooden driver or tennis racket while the competition is using graphite and carbon fiber, with billions of dollars out there working hard, not just figuring out how to generate profitable trading signals, but also how to make those signals more profitable via better execution. There’s HFT, prop firms and billion-dollar behemoths on the other side in this zero-sum game, so you better make sure you’re armed with the best equipment as you step onto the field.
For the experienced traders who’ve already been using algorithms, I would say always be testing. Always be comparing. There are small differences in speed, parameter settings and algorithm logic that can make small differences in your execution prices, which can add up to big differences in performance over time. It’s the leading edge of technology and innovation in our space, where leading firms need to be plugged in to see what’s new, who’s doing it differently and what the latest advancements are. It isn’t enough to say you have a solution. You have to be continuously innovating alongside the competition, testing your in-house algos alongside third-party algos alongside broker algos—in a continuous horse race to determine where pennies can be shaved.
There are plenty of algo execution offerings these days at the bank clearing firms. What makes the RCM-X offering different?
Joe: The main difference here is that our algos are built by traders, not just quant/programmers, and that they are built on the Strategy Studio technology stack, which we believe is materially faster than typical solutions. That speed can be the difference between getting your desired price and material slippage as the market moves away from it. Further, our ability to customize algos to unique customer parameters—say, adding a trigger to a standard iceberg, which does x, y, z—adds the alpha element back into the algo space. Many clients are asking: if everyone has the same algo execution (like standard VWAP), does anybody have an edge using it? Custom algos can find sources of alpha in and around the benchmark algo suite. Finally, we’re greatly expanding the market coverage, moving into commodity markets and not just financials.
You mentioned commodity market coverage. Why is that appealing to managers and commercial traders?
Joe: Commercial traders and managers of all sizes who are trading grown-in-the-ground commodity markets have the same problems as big shops trading tens of thousands of bonds. It can be argued that algos are even more important in the commodity markets, as the depth of book is smaller and liquidity periods more sparse and spread out. Our team has researched the specifics of every market, taking care to incorporate the unique matching engine characteristics and volume and liquidity profiles for each market. We now offer benchmark and custom algos across the traditional commodity markets list at CME, including energies, meats and grains.
What are the questions algo users should be asking to make sure they’re best set up for success in the space?
Joe: This is a great question, because those new to using algos sometimes assume that merely turning an algo on will perform some trading magic and the savings will be instant and easily observed. Unfortunately, just using an algo may not be enough. Just as important as using an algo, if not more so, is using the right algo. For example, a lot of traders come in looking at TWAP and VWAP algos in the beginning, partly because they are most widely known, and partly because they do get the job done more days than not. But it’s as much of an art as it is a science, and just any old TWAP/VWAP won’t get the job done. Without setting the parameters that line up with how you see the market and with how you view execution, results won’t match your expectations. One large CTA we worked with thought they wanted a TWAP, but upon further analysis of the trading and a deep conversation with our team on the trader’s expectations of how the order should be filled (and a lot of testing), it turned out they were actually better off with a customized implementation shortfall algo.
So whether you are brand new to algo execution or a long-time user of them, the question you should be asking is: have we analyzed the algo we’re using (or thinking of using) against a custom solution—or even against a different benchmark algo?