My email inbox took such a fusillade of stock volatility halts yesterday that I set two rules to sort them automatically. Emails rained in well after the close, girders triggered hours before and stuck in an overwhelmed system.
As I write, volatility halts Mon-Thu this week total 2,512. Smashing all records.
You need to understand these mechanisms, public companies and investors, because high-speed trading machines do.
On May 6, 2010 the market collapsed and then surged suddenly, and systems designed then to interdict volatility failed. They were revamped. Finalized and implemented in 2013, new brackets sat dormant until Mar 9, 2020.
Wham!
They were triggered again yesterday, the 12th. At Level 1, the market in all its forms across 15 exchanges and roughly 31 Alternative Trading Systems stops trading stocks when benchmarks fall 7% from the reference price in the previous day’s closing auction.
To see exchanges, visit the CTA plan and exclude Finra and CBOE (17 members becomes 15 exchanges). You can track ATS’s (dark pools) here.
The Level 1 pause lasts 15 minutes and trading then resumes. Say the reference price was 2,400 for the S&P500 the day before. At 2,242, it stops for 15 minutes. Down 13% to 2,123, it halts again for 15 minutes. At 20% down, the markets close till the next day (that would be SPX 2,000 in our example).
Here’s the kicker: Levels 1-2 apply only till 3:25p ET. If the market has been off 5% all day till 3:25p ET and then it swoons, it won’t stop falling till it’s down 20% – SPX 2,000. Girders apply only down, not up. Stocks could soar 30% in a day but couldn’t fall 21%.
Then there are single-stock guards called Limit Up/Limit Down (LULD) halts (the stuff inundating my inbox). When a Russell 1000 stock (95% of market cap), or an ETF or closed-end fund, moves 5% away from the preceding day’s reference price in a five-minute span, the security will be halted.
Russell 2000 stocks (add the two and it’s 99.9% of market cap) halt on a 10% move from the reference price in five minutes, applicable all the way to the close. Prices for all securities must be in the LULD range for 15 seconds to trigger halts.
For perspective, high-speed machines can trade in microseconds, millionths of a second (if machines can find securities to trade). Machines can game all these girders.
Boeing (BA) was volatility-halted three times yesterday (market cap $87 billion, over $220 billion of market cap in April last year) and still declined 18%, 80% more than the DJIA (and it’s a component).
Our friends at IEX, the Investors Exchange (the best market, structurally, for trading) tracked the data. Full-service broker-dealers handle customer orders, as do agency brokers (like our blood brothers at Themis Trading). Proprietary traders are racing their own capital around markets. Look at this.
It matches what we see with behavioral analytics, where machines outrace any indication that rational money is coming or going. It’s why real money struggles to buy or sell.
How have stocks lost 25% of value in two weeks with no material change to shareholdings (widely true)? This is how. Machines are so vastly faster than real money that it’s like shooting fish in a barrel.
A word on futures: The Chicago Mercantile Exchange triggers halts overnight if futures move 5%. But that tells machines to bet big on the direction prices were last moving.
Let’s bring in Exchange Traded Funds (ETFs). They depend on predictable value in ETF shares and the underlying stocks. If ETFs have risen above the value of underlying stocks, market-makers short ETF shares (borrow them) and return them to ETF sponsors to get stocks worth less than ETF shares. And vice versa.
With a low VIX, this trade is easy to calculate. When volatility soars and ETFs and stocks move the same direction, market-makers quit. They can’t tabulate a directional gain. The market loses roughly 67% of its prices, which come from ETF market-makers. Machines then yank markets up and down thousands of points without meaningful real buying or selling.
Which leads us to next week. Options expire. This pandemonium began with Feb options-expirations, where demand plunged. If the market puts together two solid days, there will be an epochal rush to out-of-the-money call options before Mar 20. Stocks will soar 15%.
I’m not saying that’ll happen. It’s remotely possible. But we’re on precarious ground where ETFs subtracted from stocks suggest another 35% of potential downside.
Last, here’s my philosophical thought, apolitical and in the vein of Will Rogers or Oscar Wilde on human nature. A primitive society ignorant of the Coronavirus would blithely pursue food, clothing and shelter. Life going on.
Now our global self-actualized culture in one breath proposes we change the climate, and in the next paralyzes over a tiny virus. I think Will and Oscar would suggest we learn to live (with viruses and the climate).
Whether we lose 35% or gain 15%, market structure is crushing human thought and shareholder behavior, and that fact deserves redress after this crisis.
