Skip to main content

Fiduciary Standards for Semi-Autonomous Algorithms?

By April 1, 2015May 21st, 2019Leadership News

Kim and I are back in the office this week, having spent last week in Orlando. We didn’t go to Disney World, but to “INSIGHTS 2015”, the 11th annual conference of fi360. Together with the Center for Fiduciary Studies, they are the folks behind the AIF® designation that Kim and I have both proudly earned. The conference is a key part of their ongoing mission to encourage the investment industry to embrace a strong fiduciary standard, and to provide ongoing education to those of us who have already committed to take our fiduciary duties seriously. We weren’t asked to do a presentation this year, but we always learn so much from the sessions we attend. We look forward to going back next spring!

What new insights did we gain this year? The unofficial theme of this year’s conference seemed to be technology: both of the keynote talks, and a fair number of the breakout sessions, were about the opportunities and dangers of the powerful technologies that sometimes seem to be running our world.

The first keynote talk was by Kevin Slavin, a professor and founder of the MIT Media Lab, on “How Algorithms Shape Our World” … and, in part, on how the “flash crash” was almost certainly caused by electronic “high frequency trading” (HFT) systems ran themselves very far out of control, very quickly. Analysis of the trades placed on the NYSE’s computer trading systems appears to show that “algorithmic trading machines” have been responsible for as much as 70% of the daily trading volume — though Slavin said that this has probably dropped back to about 55% more recently. The key point, Slavin said, is that “we have written something ultimately unreadable” in our algorithms, and we are at significant risk of losing control of them.

The Magic Shoebox

The Magic Shoebox

The second keynote talk was by Brad Katsuyama, the founder and CEO of IEX Trading, on “Ethics and Capitalism”, and the fallout from the “flash crash”. HFT algorithms may only make a fraction of a penny in profit on any given transaction, but the average daily volume on the NYSE is approaching 1 billion shares traded, so the “rounding error” differences can add up to significant sums. Recognizing that the algorithms that HFT firms rely upon require just a few microseconds — millionths of a second — to gain the advantage they need, Katsuyama and IEX have set up a system (known playfully as “the Magic Shoebox”) which slows trading almost to human time scales. This, they hope, will make trades placed on their exchange fairer to both buyer and seller, by removing at least some of those computer traders from the mix. Katsuyama is clear that while IEX is a business, the decision to create this new, fairer marketplace was a deeply moral decision.

There was a breakout session about the latest trend in investment advice, the “rise of the robo-advisor”, and what flesh-and-blood advisors can do to compete against them. Another session was about the increasing use of software to attempt to analyze, and maybe even understand, short-term market market trends. And there was a session in which sophisticated market risk analytic software was used to demonstrate that a human touch is, ultimately, needed in the investment selection process, if we’re to avoid the worst possible results.

And that’s our big take-away from this year’s conference: We are a couple of smart people with some pretty good computer systems and a disciplined process. Our market strategies are focused on the results our clients will experience over many years — not on the twitches of the most recent micro-second. There’s no way for us to compete with HFT algorithms, but we believe that our focus on long-term results, on prudent processes, and on the real human needs of our clients is a decisive competitive advantage, which we bring to all our clients.