Were The New NASDAQ Rebate Programs The Reason For Yesterday’s Excessive Quote Spam?
Yesterday, November 1st, we noticed quite a bit of quote spamming in names like BAC, YHOO, SIRI and KEY. We weren’t the only ones that noticed this -our good friends at Nanex tabulated the quote/trade ratio and also were alarmed. They wrote up their findings in this post “The Denial of Service Algo”.
“On November 1st, there were 369 seconds where the number of quotes in BAC exceeded 17,000; a total of 6.6 million quotes. During those seconds, only 1,879 trades executed. Between market open (9:30am) and 12:45, BAC had 7.8 million quotes and 116,000 trades. Which means 85% of all BAC quotes occurred in those 369 seconds. Which means it is likely that one algo from one firm (all of this quote spam is from Nasdaq) is responsible for 85% of all canceled orders in BAC.”
So what happened on November 1st to cause a high frequency trading algo to blast all those quotes through Nasdaq? We think we found the answer. Yesterday NASDAQ instituted two new pricing schemes: The NBBO Setter Pricing Program and the Qualified Market Maker (QMM) program. Both of these programs are designed to provide additional rebates for clients who are first to set the NBBO or just the NASDAQ best bid or offer. These additional rebates can range from $0.0002/share to $0.0005/share (in HFT world, this is a lot of money). The additional rebates represent about a 10% increase in existing rebates.
Putting two and two together, we believe that all the intense order submission and cancellation by market-making HFT firms was a race to set the NBBO quote first. New algorithms have been written and deployed by large HFT firms to do everything they can to get to the top of the book FIRST, and set that NBBO. These algorithms are doing this in names that are already hyper-liquid; remember that HFT “market-making” firms are trying to be 1st in line to get a rebate, but only in names in which ample liquidity exists at the same price behind them.
Margins have become razor thin the HFT industry. Their profits have shrunk alongside general market volumes; a 10% increase in profits for large speed-of-light pick-up-the-rebate video gaming HFT firms is a large carrot to be offered up by a stock exchange like NASDAQ for sure! Volumes and profits are also hurting at the stock exchanges as well, which is why NASDAQ, in their quest to maximize short-term profits, would introduce such programs as outlined above.
But we ask: Is the stressing the entire trading ecosystem, for the sole sake of padding NASDAQ’s short-term bottom line along with the bottom line of their largest HFT customers, an appropriate trade-off? Should the meteoric rise in data costs due to such activity be borne by the entire industry? Are the risks associated with market making algo race being properly monitored by Nasdaq? What might this algo race do to the market in a time of heavy volume and stress – say if the market was up or down 2-3%?
Are we all at further risk of the system crashing down because one HFT needs to beat another HFT to the exchange-proffered rebate?
This is something for us all to think about – especially the executives at NASDAQ and the SEC.