Could we work in an industry right now that does more to make itself look technically inept than capital markets? Facebook, BATS, Knight Capital, anything Bart Chilton says (kidding, sort of). We’ve set up a market structure with over 40 equity liquidity points in various points in New Jersey and in order to be competitive in them, you have to pay a 10x or so premium for the almighty cross connect. And, if you want to play in the futures market, you’d better be looking at microwave because that Spread Networks line you thought was the be-all-end-all is so yesterday. Cisco switches? Try Arista. Arista? Try Zeptonics. Can you see the dollar signs spinning or are they moving too fast for the human eye? Except, the dollars aren’t spinning anywhere near appropriate risk management solutions. Bare minimum seems to be the mandate in this arena.
Regulators are looking at ways to minimize the impact of the almighty HFT even though they perfected HFT with Reg NMS. As Ron White says, “You can’t fix stupid.” Don’t mandate throttles, transaction taxes, audits of algos by people who were trained to chase ambulances. Here are some ideas for improving market structure stability.
- Mandate something crazy like a market-wide testing environment. Consolidated order audit trails are useless for prevention. The technology genie is out of the bottle and you’re never going to put it back. Individual exchanges have certification environments, but those aren’t good enough. You have to simulate the entire market on Saturday to make sure the new code doesn’t blow it up on Monday. Or, open a window post market hours for testing. No reason IT should work Saturdays after all.
- Fix 15c3-5 to be what it was meant to be — a market risk equalizer. On our team, we have a buddy review process for reports to ensure quality beyond the dotted I’s and crossed T’s. Want to get pre-trade risk right? Try buddy reviews. You think Barclays, Morgan Stanley, or Citadel isn’t going to be harder on Knight’s pre-trade risk capabilities than the most astute regulator the SEC or CFTC has? Pre-trade risk controls should not be a competitive advantage for a firm.
- Revisit the top-of-book versus depth-of-book discussion from Reg NMS. The Canadians went depth-of-book. They were hit by the flash crash, but not nearly as bad. If you make systems start scanning full depth, it’ll slow the market down.
- This one is really out there: Why are the CFTC and SEC two agencies? I would speculate maybe 10% of the market only trades the domain of one or the other.
Back when I could actually code, I knew mixing business logic with presentation was bad. With the desire to be lean and mean, firms are mixing their business logic with market structure controls and that’s bad. The former should be closed and proprietary. The latter should be open, auditable, and transparent. Then you won’t have to fix stupid. You just have to make sure you tested for it and minimized its impact.