What Michael Lewis Missed: 4 Factors More Important Than Latency for Automated Traders
The praise delivered to Michael Lewis for his prose is much deserved – he is a truly amazing storyteller – but praise for his reporting in “The Flash Boys,” particularly as it relates to his understanding of electronic markets, is perhaps not so deserved. The fact is, for proprietary trading firms, market makers, hedge funds and bank trading desks that are building new automated trading strategies, ultra-low latency is simply no longer the most impactful factor influencing success and failure.
1. Trading Smarts
The best traders always know why the market is paying them – they understand where and how their strategies are capturing edge, and invest in trading capabilities accordingly. At one end of the spectrum, HFT firms are paid for bringing liquidity and narrowing markets, which require incredible investments in speed. These investments reduce trading costs for all market participants; thus, they earn a profit.
At the other end of the extreme are market participants like Warren Buffett (whose typical holding period is, famously, “forever”). Long-term investors like Buffet are paid for their intense financial acumen, along with their deep and opportunistic pockets.
But the truth is that most trading firms reside in the middle. They are not forced into the expensive race to zero since the opportunities they seek are based on longer horizon inter-relationships between products and markets. To them, execution is important, but only as a matter of logistics.
2. Model Iteration Speed
One can spend days, weeks and months taking individual trading systems through the iterative process of modeling, back-testing, trading, evaluation, revising and so on. This ongoing cycle is important, not just because the first several iterations often fail to bring returns commensurate with the risk taken, but also because each round brings the trader closer to a profitable strategy. Often, after many iterations, the trading strategy that emerges bears little resemblance to the idea originally hypothesized.
Because there can be many iterations before a trading model goes into production, the speed of each step is critical. The faster a trader can move from “first guess” to revenue-producing trading system, the faster he or she can capture profits.
3. Open Trading Ecosystems
More than a few traders pine for the days where a quick mind, a few trading sheets and some thick cards formed a complete trading environment. Today, traders often juggle a collection of specialized software packages – some for modeling, others for trading, still others for risk management. As such, knitting these systems together seamlessly is a competitive differentiator, as less time moving data between systems means more time for building strategies.
The knitting itself is important. If traders can easily share data between different applications, they can choose the tools that best fit the problem they are trying to solve. Software packages are rarely one-size-fits-all, so the ability to break up a design workflow into multiple stages gives traders a competitive advantage.
4. Market Access
Gone are the days where traders could trade one product for their entire career (“I’m a corn trader,” “I trade T-bills”). The best traders today are nimble and can apply techniques learned in one product across the entire financial landscape. This is important because a strategy that works well in one market may also work in other markets around the world.
Market access is also critical for scaling strategies, since many don’t always scale vertically (e.g., traders can’t always just trade more to make more). Indeed, some profitable strategies work with only a few contracts at a time. When this is the case, the only way to grow a trading program’s revenue is to find similar trades in other markets.
Trading in the financial markets is one of the purest forms of capitalism. The best ideas win and the firms that discover these ideas are the ones that reap the benefits. Firms that do not employ the above factors run the risk of being left behind.
This is perhaps best seen in “The Flash Boys” when several big asset managers basically tell Lewis: “Nobody told us this is what HFT was doing;” “We didn’t know how this thing worked;” and “No one said I had to do this differently.”
The truth is, nobody was stealing from anyone – those traders had simply had fallen behind in a market becoming inexorably electronic.
—Andrew Lisy is the Algo Product Manager at OptionsCity Software, a provider of electronic trading solutions for professional futures and options traders.