"With great power comes great responsibility." No, that line didn’t come from the Spider-Man movies-successful wartime president Franklin D. Roosevelt said it first (and Voltaire did too, before him, but FDR fits better for our purposes). Please note the "wartime" part, because we are in the trenches daily as today’s trading environment continues to change. Our arsenal is technology, and broker-supplied algorithms are often the specific weapon of choice.
But what is the end result of using these algos? Trading made easy? Or further marginalization of a trader’s skill set? Your call.
Although powerful when used correctly, the available specialized algorithms are frequently rendered barely adequate, because traders may not select the correct parameters necessary to target the desired equity. In an ever more competitive marketplace, a trader is unwittingly putting his order at a major disadvantage if he’s not using his algo properly.
Based on research and discussions with many key algo providers, we believe the following scenario is all too common: Traders are increasingly relying on "go-to" desk algorithms supplied by their broker without having been properly educated about how to use them, and therefore risking exceedingly poor execution. The problem usually arises due to the misguided comfort a trader draws from using just one or two desk algorithms. These algos were thoughtfully researched, customized on launch day and are ably toggled "passive" or "aggressive" on the fly.
But after the initial setup, they’re put on autopilot because that’s simplest (the preset algo hot button is right there), and anyway, the algo has been advertised as providing a certain level of execution quality. Unfortunately, complacence here is definitely a no-no-forensic investigation will likely reveal the execution quality to be disappointing. Many algos are perfectly capable of adequate execution; but it’s how they are used-and even more important, when-that can make them ineffective.
Here’s an easy example: Returns and fills will always be poor if you’re using an unmodified VWAP on volatile stocks-as will other algo flavors in trading situations not adjusted for, no matter how good they’re advertised to be.
At several conferences I’ve attended, in the U.S. and elsewhere, I’ve heard complaints from providers, brokers and traders from both sides of the desk, stating that many algo users have not been fully educated about everything the products do and how they access the markets, not to mention that they lack the time to learn how to modify them. And even if they do understand their algos, they say, the interfaces can be less than user-friendly, especially if you need to use one quickly-as is the case on hectic market days, when adjustments are most necessary.
Focused education of traders on both sides of the desk on when to use and how to modify their algos would go a long way toward improving the quality of trade execution. On the other hand, ignoring the problem and letting improper usage continue certainly affects net fund returns over the entire investment cycle.
To understand the ramifications of the problem, consider the following, admittedly "simplistic," example: Using only the passive/aggressive toggle on a desk algo (VWAP, dark, etc.), trading 100K ABC per side round-trip once per day, at $1.50/share profit, the gross return is $37.8 million annualized. But what did it cost?
You have generated explicit costs (commissions of 25 mils) of $1.3 million annualized. If your slippage costs (away from arrival, for example) are only 12.5 basis points on each side of this trade, the implicit (slippage) costs are $3.15 million annually. What is immediately apparent is that a significant amount of your expected gross has not been realized. With all other factors constant, poor execution (due to implicit costs) will erode more than 8.6 percent of the returns on this hypothetical portfolio, easily outweighing the explicit costs.
The example above is for a relatively small portfolio executing only a single round-trip trade per day with a nice profit. With a larger asset base and more trading, the costs increase exponentially. Failing to use an optimal selection of variables for accessing the market (a loss of even 12.5 basis points for each trade, not at all unrealistic in today’s markets with poorly used tools) dramatically and painfully affects returns. Managing commissions paid is not enough; the implicit cost outcome of even slightly subpar execution can overwhelm even the most aggressive low-cost commission structures. Market access points, be they through your broker or algorithm, always need to be reviewed to manage costs, which will in turn impact net return.
In chapter 11 of "The Art of War," Sun Tzu focuses on understanding the terrain of a battlefield and how to best situate yourself to manage a preferable outcome. Likewise, even if you understand where your algorithmic weapons send your order, how they interact with those venues and how they represent the order in each venue, but not the unique character of each individual venue, that also prohibits improved execution. In certain algorithms, you may limit (but normally you cannot add to) the venues the algo points to-which you can exploit if you know that one or two particular venues contain especially strong sources of liquidity.
Market structure is yet another aspect of trading many traders overlook. That’s not always deliberate-traders have a lot on their plate-but it is certainly to their advantage to study and understand.
A few questions traders need to ask themselves and their providers are:
* How does this algorithm work and what venues does it touch?
* How do dark pools and other venues process the order?
* What internal or external flow can or does the order interact with?
* Do the venues (including dark pools) route out when there is absence of liquidity within, and where to?
* Some equities trade better in certain venues-which tool is best to find that liquidity?
These questions need to be answered to minimize market impact and achieve improved execution. Those traders and firms who consistently keep themselves informed will tend to minimize their slippage, resulting in improved pricing and higher order-completion rates, which will help them rise to the top as their portfolio manager’s trader or client’s broker of choice.
Explicit costs (commission fees, ticket charges, etc.) can be controlled, as they are much more tangibly obvious to a trader; implicit costs (opportunity cost, slippage, etc.) are the wild cards, and can dwarf the impact of explicit costs-and in our view are the difference between the winners and losers in the marketplace.
Those desks that proactively endeavor to understand current market structure, and the described implicit cost and slippage control, are destined to succeed when measured via trade-cost analysis or any other metric desired. Those still figuring out how to aim their weapons may not even notice the error until the battle is already lost.
Garrett Nenner is a managing director and head of global markets and market structure specialist at Momentum Trading Partners LLC in New York. The opinions expressed in this column are his own, and do not represent the opinions of Traders Magazine.
