Floaters and fills
Vol #266: March 26th, 2026
I have no idea how many bottles to order. Two cases? Twenty cases?
With a lot of data we can triangulate exactly to whom a cell phone number belongs to. Ads can be served so well that you think your phone is listening to you. Options prices are calibrated within a penny given a thousand points of light.
Other times all it takes is a handful of datum to make a well informed decision. If you’re replacing the roof, two or three quotes give a good idea of what something should cost. A recommendation from a friend or Facebook forum might be singular, but it comes with a lot of context.
Then there are problems with fuzzy data, or no data at all. Whether it’s brand new, or just new to you, the best effort is a guesstimate. Hiding behind the false precision of a few rows in Excel, is nothing but a hunch.
I’ve been helping my brother and sister in law organize the wine list and club for their new venture in Omaha. It’s a lot of fun talking to distributors, and centering on the mission and aesthetic of the bottle shop. We’ve been intentional in selecting both the glassware and the glass list, and what will be available for retail. But come opening day in a few weeks, who knows what will actually sell?
I have a pretty good idea about what taste profiles are interesting. And I’ve seen enough menus to know what balance looks like across regions, price points, and flavors. But quantities are difficult even to rough in based on the bounds of expected traffic.
How many glasses are getting filled with what - that’s all going to depend on orderflow.
With a year’s worth of weekends, things will be a lot more predictable. But for now we need to adapt quickly. The fun little Chenin Blanc that wine geeks love - might have to be replaced with something a little more mainstream. The Bordeaux blend from Lebanon is exceptional, but if you don’t know about Château Musar, each bottle is sold before it’s bought.
Every customer and vendor relationship is a balance of supply and demand. Liquid and efficient markets mask just how much data and experience go into determining quantities or fair values.
In options market making the specialist is the first person burdened with this exercise. For brand new issues they’re tasked with setting price across potentially dozens of different strikes and expirations. Legal width binds them to setting markets at a certain spread, and there’s no history of orderflow.
Save for the blockbuster IPOs, new listings are anyone’s guess as to who is going to come and play or when and how they’ll do that. A little bit of arithmetic can describe the historical volatility of the underlying stock, but that history isn’t always long or predictive about the future.
So inputs move quickly. A ten lot will fade prices a lot faster than in a legacy name with ostensibly similar characteristics. Every point of data is adding quite a bit of information.
After a big move, be it earnings or an energy crisis, we see similar effects even in the most liquid markets. As pricing adjusts to a new normal, the inputs move quickly. Even going into the broadly telegraphed and overanalyzed FOMC announcements, spreads in SPX tick a little bit wider.
So how does a buy side trader know if they’re getting a good price?
If you’re trading for risk transfer, avoid transacting when uncertainty is higher than average. The big known unknowns are best avoided - don’t set your overlay position in GOOG to expire on the Friday after they announce earnings.
It’s harder to predict when the President is going to tweet out a new war, but the best defense is to keep your eggs in several baskets. Tranche out positions so that even if you’re caught in a whipsaw roll coaster, it’s only for a portion of the overall size.
The best time to trade is when neither you or your counterparty are forced to do so. That’s why blowing out positions in a margin call or stock loan buy in is always worse than it should be. But even less ham-fisted transactions will pay an urgency tax.
Independent of volatility pricing, for the actual mechanics of execution, it’s always worth thinking about the constraints and opportunities of your counterparty. Since most of the time customers don’t interact with each other, this will be a market maker who is trading to capture edge and manage risk.
Whether it’s in an auction mechanism or outright on the screens, liquidity providers are incentivized to compete for that flow. They won’t trade for negative edge or fair value, but they’re not going to let the price get so juicy the competition takes it down first.
I like to set expectations around halfway between the midpoint and the opposite side. If the markets $1.00 - $1.20, a midpoint sale (or buy) of $1.10 is unlikely, but a sell order should usually get filled somewhere north of $1.05. Not only does that mean a few pennies of edge to the dealer, it’s roughly the level wholesalers set when they purchase orderflow from brokers. They have bigger picture incentives to make sure your order gets some price improvement.
The more liquid the name, the more likely fair value is around the midpoint. But for a variety of reasons, this might not be the case in a more thinly traded stock. Here it’s even more important to float around to see where you can get filled.
There’s a delicate balance between floating and spooking, and this is where the art of trading gets nuanced. If you’re too aggressive or rote about stepping down a sell order, a watchful market maker or clever algo will sense that and let you keep going. This is balanced by the fact that there are a half dozen competitors, but everyone can’t be watching everything all the time.
How much is too much varies for every single name. Even with the wider options far out of the money or long duration in SPX, someone’s going to get you for pretty close to fair. But if you’re regularly trading a lower volume name, pay attention to where the fills tend to happen. If that tends to be closer to the bids and offers every time, start out more aggressive so you aren’t constantly stepping down and making matters worse. Conversely, ask for a little bit more each time you get quickly filled.
When you don’t have the experience or data to tell you where to price something, you want to keep track and be nimble. Whether new to options or new to this name or strategy, it’s important to quickly adapt to every bit of information.
The good news for the bottle shop is that we can always order more. And every cork we pop gives just a little bit more information. Track the orders, watch the fills.