The Securities and Exchange Commission has approved two proposals submitted by the national securities exchanges and the Financial Industry Regulatory Authority (FINRA) that are designed to address extraordinary volatility in individual securities and the broader U.S. stock market.
No, kidding, not a cure for bad jobs reports, but this is a weird phrasing isn’t it? The approved proposals are just better-thought-out circuit-breakers for single stocks and the broader market. Since getting pummeled for being an idiot about market structure, I am attempting to be less of an idiot about market structure (until now!), and one thing that seems to be settled wisdom is that trading halts have a function in reducing volatility associated with order imbalance caused by ignorance, panic, or fat-fingering, but don’t do much to cure volatility associated with, y’know, economic volatility. Not that the SEC is claiming that they do, exactly, but “easing volatility” doesn’t seem like exactly the result here.
The SEC approved two rules today. One, the single-stock circuit breaker, prohibits trading in a stock at a price more than 5% away from the five-minute moving average (10% for small caps); if the stock isn’t quoted within 5% of the average trading is halted for five minutes so people, er, computers can go find more orders. This replaces the post-flash-crash single-stock circuitbreakers, which halted trading if there were actual trades, as opposed to quotes, outside of a 10% band, which meant in practice that halts were often triggered by erroneous trades whereas the new rules in theory should prevent those erroneous trades.
The other, the market-wide circuit breaker, amends existing rules that halt trading for 30-60 minutes for a 10% market decline (depending on time of day, details pages 4-5 here), 1-2 hours for a 20% decline, and the rest of the day for a 30% decline. The new rules make it 15 minutes for 7%, 15 more minutes for 13%, and the rest of the day for 20%.
The change in numbers is less interesting than the change in calculation method, which was kind of amazing. I said the old rules called a halt for “a 10% market decline” but what that actually meant was a decline in the Dow Jones Industrial Average of more than a dollar figure that is calculated once per quarter to equal 10% of the average close of the Dow for the month prior to the start of the quarter. The new rules will update the dollar figure daily rather than quarterly, and will be based on the S&P 500 rather than the Dow.
The Dow! Once per quarter!
These things are mostly symbolic – generally speaking, if the Dow drops by more than 20% or whatever of its average price last quarter, it’s safe to say that shit is fucked up even if the S&P is only down 19.8% of yesterday’s close. But no one’s seriously used the Dow as an index of the broad market in, what, a generation?, and updating your things once per quarter is kind of symbolic of “we update our things by hand because we’re not that into computers.”
The new rules assume market speeds that are literally superhuman: they assume that a market-wide trading halt can draw in new orders and replenish liquidity in just fifteen minutes, and that a single stock’s erratic moves can be cured with just a five-minute pause to draw in more stabilizing orders. That is not a world of individual investors seeing a trading halt, studying the market, deciding it is undervalued and putting in buy orders. In five minutes. It is a world of instant communication and more than a little algorithmic involvement. Which, of course, is the world we live in.
The other world we live in is a world in which the SEC just today, in 2012, ditched its reliance on the Dow as an indicator of broad market moves. And in which its response to the May 2010 flash crash was finalized in June 2012. I feel like there may be a metaphor in the clash of those two worlds – one where market participants can respond to crashes in five or fifteen minutes, while the regulators take two years.
Oh, also, Davis Polk has its monthly Dodd-Frank Progress Report out.