News for hackers. Stuff that matters.

A comment on Einstein and the Great Fed Robbery. Parent comment

Re:Einstein and the Great Fed Robbery by mcphilip on Friday September 20, 2013 @02:53PM (#6419758)
Market microstructure is such a massively complicated topic that few insiders, much less the SEC, thoroughly understand it.

For instance, after NASDAQ went down for 3 hours on August 22, the SEC asked the NYSE and NASDAQ for a detailed timeline of events. Each exchange blamed the other. The SEC's response? It told the exchanges to cooperate and please fix the problem, declining to make any further public comment [1].

Trading at high speeds is like the wild west right now with little to no threat of law enforcement.

[1]http://uk.reuters.com/article/2013/08/27/uk-nasdaq-halt-nyse...

[ Parent | on HN ]
Re:Einstein and the Great Fed Robbery by minimax on Friday September 20, 2013 @03:28PM (#6419945)
Market microstructure is such a massively complicated topic that few insiders, much less the SEC, thoroughly understand it.

It's actually not super complicated. A limit order book is a straightforward thing. If you can do FizzBuzz, you can probably whip up a toy implementation of a matching engine in an afternoon. People think it's complicated because of all the FUD (c.f. the comments on this story), but the basic mechanics of how a modern exchange operates are remarkably simple.

[ Parent | on HN ]
Re:Einstein and the Great Fed Robbery by mcphilip on Friday September 20, 2013 @04:19PM (#6420244)
>People think it's complicated because of all the FUD (c.f. the comments on this story), but the basic mechanics of how a modern exchange operates are remarkably simple.

I'm referring to market microstructure in terms of the complex interactions amongst exchanges, the impact of Reg NMS, etc. Obviously the underlying algorithms can be boiled down to simple components.

For those interested in learning something about market microstructure, I highly recommend "Trading and Exchanges: Market Microstructure for Practitioners" as an introduction to the basics:

http://www.amazon.com/Trading-Exchanges-Market-Microstructur...

[ Parent | on HN ]
Re:Einstein and the Great Fed Robbery by saraid216 on Saturday September 21, 2013 @03:47PM (#6424191)
> I'm referring to market microstructure in terms of the complex interactions amongst exchanges, the impact of Reg NMS, etc.

I've heard that the financial interconnections are remarkably similar to ecologies. Would be interesting to get some more of that tooling for analyzing ecologies like fisheries available to financial regulators.

[ Parent | on HN ]
Re:Einstein and the Great Fed Robbery by davidmr on Friday September 20, 2013 @04:22PM (#6420260)
What you defined is not market microstructure, but the platonic ideal of a matching engine. The microstructure of a market is defined (by a Wikipedia source) as: "the study of the process and outcomes of exchanging assets under a specific set of rules. While much of economics abstracts from the mechanics of trading, microstructure theory focuses on how specific trading mechanisms affect the price formation process.”[0]

That is fantastically complicated. As you say, anyone can write a matching engine.

[0] http://en.wikipedia.org/wiki/Market_microstructure

[ Parent | on HN ]
Re:Einstein and the Great Fed Robbery by colanderman on Saturday September 21, 2013 @12:21AM (#6421980)
And anyone can code up the Game of Life in a couple hours. But to understand the "physics" behind emergent phenomenæ such as gliders? Much much more difficult.
[ Parent | on HN ]
Re:Einstein and the Great Fed Robbery by chris_mahan on Friday September 20, 2013 @04:19PM (#6420239)
Even some people who profess to be computer programmers can't do fizzbuzz. How do you expect the general population, even those with skill in MS Excel, to be able to?
[ Parent | on HN ]
Re:Einstein and the Great Fed Robbery by JumpCrisscross on Saturday September 21, 2013 @06:08AM (#6422525)
It's rigorously documented in public repositories. I used to build trading systems that were highly reliant on (a) how microstructures were supposed to behave and (b) the expected variability in that behaviour. Figuring (a) involved poring over thousands of pages of technical documentation - tedious but doable. There is also help from the OCC, Finra, BIS, Fed, and the exchanges' helplines. Judging (b) is more complicated, and grows more uncertain the newer a particular structure is. But again, doable if tedious.

The tediousness could be relieved by centralising market microstructure documentation. But given the frequency with which we change rules, it may make more sense to have an expanded class of lawyers or risk managers who are dedicated to understanding and promulgating the documentation.

I wouldn't go so far as to say the SEC, Finra, or the Fed Market Surveillance teams don't have a strong handle on microstructure - they have smart people. The SEC's job isn't to manually fix problems. It would be troubling if every time an exchange went down the SEC felt compelled to issue a new rule.

[ Parent | on HN ]