In Part III of my It’s All Good Until It Isn’t series on ETFs, I remind ETF investors of the pain and losses the August 2015 Flash Crash caused, make clear what really happened, and explain why it will happen again.
And, I offer advice on how to avoid the same mistakes and losses next time.
Because it will happen again.
Reminiscences of a stock operation gone bad.
The morning of August 24, 2015 provided the industry, exchanges, ETF sponsors, their authorized participants, regulators, and investors with their first opportunity to observe how market protections implemented in recent years would behave during times of significant stress.
With futures indicating a lower open for stocks, investors and traders fired volumes of sell orders at their brokerages that morning. Between 9:30 and 10:00 a.m. ET, the NYSE saw a six-fold increase in executed market order volume compared to July levels. NYSE Arca experienced an even greater proportional increase in market orders for ETPs, or exchange traded products.
A significant number of market orders were sent to exchanges by retail wholesalers, mostly the result of triggered “stop-loss” orders, which automatically convert into market orders when a specified price is reached.
Before we even get to the problems that market orders cause, it’s critical to understand what role the “wholesalers” play in general, and specifically in times of significant stress.
Wholesalers are trading firms who accept order flow from brokerages and either execute them, or route them to other trading centers for execution.
Brokerages can discount trade execution charges to customers because wholesalers pay them to route their customers’ orders to them so they can execute those orders themselves. The point and profitability in being a middleman wholesaler is that by paying (a few cents a share) for orders to come to you from many brokerages and sources, you can act as your own exchange and execute orders against other orders coming your way, or execute them yourself. In other words, you can take the other side.
By seeing order flow, wholesale shops gauge whether there are more buy or sell orders, at what prices, what kinds of orders they are, etc. This gives them an inside track on what investors and traders want to do.
They make money taking the other side of those trades or jumping ahead of them, whichever way they must go in order to make money. It’s a very profitable business.
But, in times of stress, most of these so-called (they call themselves) “liquidity providers” stop what they do on normal days. Instead of acting like an exchange themselves, they reroute the orders they get from brokerages to actual exchanges for execution. That puts extraordinary stress on the Designated Market Makers at exchanges who match orders there.
What investors don’t understand is that “liquidity providers” that make everything look good, fast and easy on normal days, shut down on days they can’t handle one-sided (sell) orders, which stresses exchange systems, market makers, and prices even more.
Worse, if they see lots of sell orders, before they reroute them to exchanges, they fire their own sell orders at exchanges, frontrunning customers’ orders, knocking down stocks and profiting as latent orders get to exchanges and knock prices down further.
The middlemen are then often the ones who buy some of those sell orders to cover the short positions they made frontrunning orders. Like I said, it’s a very profitable business.
Prices gapped down in stocks and ETPs that Monday morning because there were no bids waiting to greet huge sell order flows. I explained that phenomenon in Part II.
Investors used market orders to get out of positions and lots of investors and traders had standing stop-loss orders down with their brokerages, who had farmed them out to wholesalers. When prices fell through stop-loss price triggers those stop-loss orders became market orders to sell at the next available price.
The huge volume of sell-at-market orders in the face of very few or no bids caused prices in stocks and ETPs to gap down.
Several things came to light.
First, and most importantly, investors and traders lost untold amounts of money on their “stop-loss” and market sell orders.
For example, if an ETF investor had a sell stop order down at a price of $50, while the ETF might have been trading at $55, once it gapped down to say $40, the stop-loss order becomes a market order once the ETF trades at or through $50. Since the next available bid might be $40, the order could have gotten filled there, $10 below where the investor expected to be filled.
Since the trade was not a CEE trade, Clearly Erroneous Execution, meaning it wasn’t a mistake like a “fat finger” glitch, the trade wouldn’t be eligible for cancellation. Even if the ETF rebounded moments later to $50 or higher, the trade at $40 would be honored.
Millions of investors lost billions of dollars that morning because stocks and ETPs gapped down.
No-one knows how much money was lost and regulators never tried to calculate gross losses. Why would they? Doing so and revealing how much investors lost would scare investors out of the market.
By way of a speech on September 7, 2010 at the Economic Club of New York, about potential losses in the first Flash Crash of May 6, 2010, Mary Schapiro, the former SEC chairman, told the audience the following: “A staggering total of more than $2 billion in individual investor stop loss orders is estimated to have been triggered during the half hour between 2:30 and 3 p.m. on May 6. As a hypothetical illustration, if each of those orders were executed at a very conservative estimate of 10 per cent less than the closing price, then those individual investors suffered losses of more than $200 million compared to the closing price on that day.”
Even that was a guess, and wildly off because average drops on securities that gapped down were higher, with some stocks going to ZERO, as in down 100%!
How much was lost in the August 24, 2015 Flash Crash will never be known.
While it should matter, what matters more is that it will happen again.
Regulators and the industry are still grappling with what didn’t work, including measures like LULD (Limit Up Limit Down) rules, trading band halt rules, MWCB (market-wide circuit breaker) rules and other measures instituted after the May 2010 Flash Crash.
They didn’t work. Even though regulators say they ultimately helped calm markets and reverse losses, millions of investors lost billions of dollars, most of them on ETFs.
There haven’t been any significant new rules that will arrest the next flash crash. That’s because the fundamental cause of crashing prices, the fact that there are no-bids, doesn’t change, no matter what halts, bands, or time-outs regulators paint over the real problem.
The only course of action that makes sense for ETF investors is to not use stop-loss orders and use stop-limit orders.
While stop loss orders turn into market orders which get killed in flash crashing markets, stop-limit orders can only be filled at the limit price designated.
It’s very rare that all the underlying stocks in an ETF would ultimately fall 10% or more, but the ETF could fall 40%-50%, as some did in the Flash Crash, so using a stop-limit price that’s maybe 10% below where an investor might want to get out of a position and reassess what’s causing any selloff, is better than selling a position considerably lower and watching it bounce back up.
Of course, a stop-limit order doesn’t get filled if a falling ETF doesn’t get repriced higher at or above the designated limit price. That could be a problem if markets continue down.
In an imperfect world, especially with pricing problems affecting ETFs in flash crashes and when underlying securities stop trading, at least having some plan is better than having none.