We recently talked about how smart traders use inverted venues to compress artificially wide spreads. What we found was very economically rational, but not at all that “simple” to optimize.
In general, the longer queue length a stock had, the more inverted it traded. We also saw that queue length was generally longer as the tick got wider (in basis points). Finally, we showed that because the market uses inverted venues more where spreads were wider, traders actually compressed economic spread of the most artificially wide ticks by the most (in basis points).
Inverted isn’t always better economics
However, something that we didn’t really investigate was how high inverted usage would get.
What the data shows is that even when ticks approach 20bps (a $5 stock) the proportion of inverted trading rarely rises above 25% (see Chart 1).
Chart 1: Inverted share seems to plateau around 25% for even the widest spread stocks
Source: Nasdaq Economic Research (Circle size shows queue length in minutes)
That seems to contradict the findings of the Battalio-Jennings paper that led to the access fee pilot. They suggest that inverted posts always get better executions. So why do the majority of traders use inverted venues sparingly?
The inverted offer isn’t always the right price – queue priority or not
What we need, to highlight what we mean, is imagine a scenario where your stock has a very accurate and known valuation.
Let’s start with the QQQ ETF, one of the most liquid ETFs in the world. QQQ has a price around $190 and a spread of one cent (or 0.5bp) almost all the time. We show that in Chart 2 as the green line.
Imagine then we introduce a mini QQQ, which is 1/10th the size and perfectly arbitragable. It will be priced around $19 with a one-cent spread (which is worth around 5bps).
In the chart we show the economics a trader could earn by bidding and offering in maker taker venues (blue prices, where they also collect a rebate) vs. inverted venues (orange prices, where paying to post reduces the economic spread). For a recap on how this works see Chart 3 in V is for Volume.
We can see that as the value of the QQQ (green line) moves higher over time, the maker taker offer (blue line) remains above the valuation of the mini-QQQs. However there are times where the inverted offer (orange line) is “too cheap.” Importantly, when that happens, a rational trader would cancel their offers out of the inverted venue (as we show in the chart).
Chart 2: Example of why inverted offers aren’t always worth posting
Source: Nasdaq Economic Research (descriptive only)
Spread compression isn’t the same as sub-penny pricing
Data shows that even stocks with wide spreads and long queues, the most tick-constrained stocks, traders still prefer rebates over queue priority more than 75% of the time. That’s important because it shows that although inverted venues can be used to compress economic spreads and buy queue priority, they are not the same as “sub-penny” pricing.
Importantly, although inverted venues let you “compress” the spread, lots of traders have decided (or learned) not to. That’s because they’re smart enough to know when the inverted offer may not, in fact, offer better economics than the primary market.
Interestingly, this is also consistent with an earlier post of ours that found that inverted venues are much less likely to have a two-sided quote at the NBBO.
Go here to sign up for Phil's newsletter to get his latest insights on the markets and the economy.