The Fee for Liquidity
Vol #185: July 11th, 2024
Prop shop lore says that when you upgrade your office, get ready for a downtick in revenue. I’ve seen this enough times to know that even the most Bayesian of thinkers are not exempt from the hedonistic treadmill. Or bad luck.
Anticipating a good ogle today followed by schadenfreude tomorrow, I walked across Van Buren Street to the Fed building where our partners in a nascent business venture had just moved into new digs.
Wood paneled entryway, a floating staircase down the center; it was pretty nice. There’s only so much you can do with desks that need to support twelve monitors however, so the bullpen where we were meeting was like any other.
Up for discussion was a joint venture to back traders who brought in outside capital. The economics were fairly vanilla; fee overrides, charging haircut on haircut, and a slice of any profits. Our team would provide a stripped down version of the software, and the partner supplied office space and clearing.
As it usually does in meetings where the same thing gets said four times, my attention wandered beyond the architectural details. A few desks over, our software support lead was going through a training session for a potential trader. The gulf of experience was vast - one knew how to trade options, the other knew how to use a computer mouse.
After the pointing device was righted, the duo proceeded with their demo. That should have been an omen. You can probably guess how many distributions eventually got paid.
Caching that story for later, I focused my attention on what the in house traders were doing. Most market makers’ setups look similar. It’s a massive blink of spreadsheets and lists with prices, greeks, and positions.
Each of these has their own degree of precision. Positions only come in round numbers, you can’t have half a contract of exposure. Greeks are calculated with finely calibrated models, but their accuracy matters less than you might think. The risk manager isn’t going to change his tune if it's discovered that your short vega has gone from -10,142 to -9,987.
Directional risk is something that market makers seek to eliminate through delta hedging. They want to capture a fee for managing inventory, not make bets on the ultimate stock direction. While this suggests a need for precision, the cloud of randomness and transactional cost of constantly rebalancing swamps rounding errors.
No matter how a dealer calculates their risk, they can only make options trades in fixed price increments. Exchanges display quotes in nickels and pennies, though they might do some rounding on the backend for certain order types.
That’s why it was a bit confusing to see the NBBO priced with an extra decimal place. The ARCA offer was not at $1.45, but $1.4617. Hmm….
An options contract represents a risk profile. To some it might be an offset on an underlying stock position. Others might see that risk as cheap implied volatility, or a hedge against another contract. That risk comes for a price. Add those up and you get PnL. Now come the fees.
While the price that hits the public record will show $1.45, what these traders knew is that if they were going to lift that offer now, the price to buy costs them a little over an extra penny per share or $1.17 per contact. Hence the $1.4617.
The fees to trade vary widely by exchange. If this same trader improved the bid to $1.10 on the BZX exchange, it would show up on their screen as $1.0965. The trade in the marketplace will take place at $1.10, but they know if executed on BZX, they’ll get a credit back and net pay slightly less than the OCC trade price.
These differences are insignificant to most of the buy side. For a $1 option, the most expensive fees layer on a 1% price increase. That’s only a penny in pricing terms, which hardly moves the needle on a PnL chart. Your covered call is functionally identical at $1.45 or $1.44. Few buy side strategies require that layer of precision.
In fact for most retail traders, all of this complexity is hidden. Trades happen at round prices, and a combination of brokers and wholesalers manage the routing and pay the bills (or collect rebates) for execution because they want the random orderflow. Big wheels keep on turning.
Fees aren’t important on the trade, but rather because they foster liquidity. Customers don’t need to worry about the changes or absolute levels of trading fees, other than understanding this part of microstructure is a core contributor to the bid-ask spreads they see when firing up the terminals.
Orderflow moves like water - down the path of least resistance. The reason there is an additional decimal place on the trader’s screen (and more importantly in the algorithms), is that providing liquidity is a game of edge collection, and edge is denominated in pennies collected. Any business would love to see something that helped them manage an expense approaching 50% of their theoretical profit before any hedging, financing, or management costs.
Jockeying for pennies is what tightens the spread. Both ARCA and BZX described above follow the “maker-taker” pricing model. The idea is that traders who “make” get paid by the people that “take” liquidity. Of course this comes after the exchange has taken their slice - only about half of the fees collected get paid to the maker.
A market maker might have an option worth $1.15, and for this given product/expiration typically likes to trade for $0.03 of edge. In a simple world, the best bid would be $1.12. But exchange fees tweak the incentives, and if providing liquidity paid $1, they could stream a $1.13 bid for that option, knowing they’ll still get the $0.03 of edge.
The market place is a constant swirl of adjustments as underlying prices move and volatility assumptions shift. The transaction costs provide a subtle nudge for price improvements and routing decisions.
Different participant types pay different fees. In the maker-taker model described above, the credit/debit system applies to almost everyone. The size of the credit shifts on the margins (Specialists or LMMs on ARCA for example get a significant extra dollop), but broadly no matter who you are, you pay to take and get paid to make.
Fees on “PFOF” exchanges are structured slightly differently. With the inclusion of payment for orderflow, customers trade for free (or even small credit), while everyone else pays a flat fee whether taking or providing, incentivizing size on the bid/ask. And anytime a market maker trades with a customer, they pay an extra marketing charge that goes into a pool to allocate back to brokers.
What makes this even more confusing, is that not all of the fees are known ex-ante. While you can predict (usually) which exchange you’ll trade on and what their published fee schedule is, you don’t always know who’s on the other side. If it’s a customer on the pro-rata exchange there can be a surprise marketing charge, or any other variety of ripples in timing and sequencing that fuzz the determinism.
Further complexity comes for active participants as they navigate volume discounts. Exchanges like to reward their best customers with better rates, but the calculations for these are dynamic. Typically based on some percentage of volume (overall, customer, etc.), the exchange bill at the end of the month depends not just on what you’re doing, but what everyone else is doing.
Hitting a threshold on one exchange comes at the cost of another exchange. Is it better to trade more volume on ARCA to get an extra nickel on all your flow this month, or route an individual order to the best rebate exchange right now?
There are tiers for different order types, distinguishing between crossing/QCC volume, spread volume, and single legs. For a wholesaler with two types of orderflow (customer and market maker), there will be competing dynamics and net costs depend on multiple factors.
When prop shops get a discount based on the percent of OCV (OCC Customer Volume) that they trade, it allows them to provide tighter markets. On EDGX, trading less than .10% costs $0.20 a contract. If they trade over 1.45% of OCV, they only pay $0.02 to trade.
What that means, is they now have an extra $0.0018 of edge to play with as they engage with the markets. That’s the equivalent of getting an extra penny better in stock on a 18 delta option. That doesn’t buy a house in the Hamptons, but as stocks tick around all day, your friendly neighborhood market maker with the best rates will be the first one to lift a posted order.
All of these fees demand active management for professional traders, but require almost none from the buy side. If there’s only marginal benefit from a retail trader stressing about a pennies of price improvement, there’s even less to be gained by worrying about fees.
The layers of tables and asterixis exist but for your benefit.
Elevate your overlays with TheTape.Report.
A complete strategy tool kit for covered calls, hedged equity, and much more.
Follow along weekly with Portfolio Design.
