Ketchup is Bad
Vol #181: May 30th, 2024
No bid at a nickel, tell me your size. I’ll sell any sweet red tomato based condiment all day. I was taught never to hit a cabinet bid1, but I could retire tomorrow given a long enough list of fancy ketchup contenders.
When I was out to lunch last weekend, alongside the sandwich came a tiny translucent bottle with vaguely red contents. Could this be hot sauce? Unwilling to waste even the tip of a single shoestring fry, I placed a drop of the mystery sauce on brown wax paper and pinky sampled. My wife commented on the behavior. It’s disappointing when your hedges pay.
The majority of ketchup is made by Heinz, who since 1876 has dominated the market and forced the entire category to imitate their saccharine red squiggles. I believe there could be a high quality ketchup. Evidence has proved otherwise.
A good chutney or BBQ sauce manages to blend the sweet and savory well. Forget about the variations of mouth watering relishes you can get. At some point, there will be a ketchup that passes mustard. But until then, I feel pretty confident about my naked short.
Starting fifteen years ago, the idea of “strong beliefs, weakly held” gained popularity amongst techno optimists and blogo-philes. Coined by futurist Paul Saffro, it’s a mental model that sounds venture capitalist, but also applies to price discovery.
The idea is that you should let your intuition quickly drive an opinion. Kahneman’s system I thinking, and Gladwell’s blink reaction. Our guts know quite a bit; don’t get distracted by muddying thoughts and emotions.
Once you have a strong belief, it’s important to test it quickly and be willing to change. That’s the weakly held part. Richard Feynman nailed it when he said that “we are trying to prove ourselves wrong as quickly as possible, because only in that way can we find progress.”
The biggest critique of this philosophy is that in practice it ignores the second part. Strong beliefs quash discussion, and drown out nuance. But if we’re speaking about objective truths, any serious opinion must be weakly held in the face of facts.
Markets are truth seekers. The orderbook exists to match buyers and sellers who believe the current price represents a positive trade off. With different objectives, undervalued for me might be overvalued for you. That makes a trade.
In my theoretical condiment market, if a chef comes in and starts lifting my nickels, I’m going to get scared. It’s bad practice to offer a fill without knowing the size. Who’s on the other side - Alain Ducasse or Guy Fieri? Even if it’s just a Flavortown bid, I need to move my market.
When a market maker gives a quote, it must be a strong belief because they put capital on the line. But in order to make money, you’ve got to adapt quickly. Liquidity provision is not about knowing the true price with confidence, but quickly adjusting to an intangible truth that orderflow and exogenous factors move in real time.
Breaking down a quote, it has two basic components on two sides; size and price. Both dimensions demonstrate a type of confidence and contribute to liquidity.
Size is the capital available, while the width is the price being charged to hold the inventory. They’re related, in that larger size will pay more for liquidity. The function for “I would do more at better prices (to me)” is geometric until it plummets. Too much size and I’m out all together.
Bigger is relative to the product, but it all comes down to how likely that order is to move pricing inputs. Stable markets can digest significant orderflow. A thousand contracts for a penny in SPY feels a lot safer than a ten lot for a dollar in DJT.
Sizing does not necessarily have to be balanced. While it’s most common to be “50 up” or equal bid and offer sizes, depending on where prices get set, you might skew the market to 50 x 100 or more.
Risk managers will debate whether it’s better to move your size or move your pricing. If you think the bid is too close to fair, don’t be on that market. Conversely, if the orderflow is going to trade there, it’s worth collecting a little bit of edge in smaller size. After all, that’s your job.
Mano y mano, the sizing game is fairly straightforward. One broker has an order, and one market maker provides the levels where they can execute. But sizing displayed in electronic markets with multiple participants and allocation systems requires second order game theory.
Pro-rata exchanges allocate inbound orderflow based on the size displayed by participants. If you’re showing 50 contracts on a quote with 450 others, you’ll be allocated 10% of each inbound order. (Ignoring specialist allocations, preferred routing, match percentages, etc.)
For customers, the beauty of this system is that it often brings more liquidity than the sum of individual parts. If I call up ten market makers, they might each be willing to trade 25 contracts at the given level. But because interacting with orderflow is their business, and they want to trade 25 contracts at that level, the pro-rata allocation system incentivizes them to display 50, inflating the overall market size. (Extra Credit: Multiply this by 8 other venues with the same market model.)
The competing exchange allocation system - price/time - solves for the other dimension of confidence. By incentivizing price improvement with the largest allocations to the first mover, the market structure here draws inventory pricing closer and closer to “fair” value.
Tighter markets tend to be thinner, but for most retail orderflow this is immaterial. Anything that’s less than a 250 lot will be benchmarked at the top of book. Further, the incentive of capturing that marginal flow even for less edge is enough to encourage participants to leap frog each other.
Size will ebb and flow according to confidence, but prices get downright jumpy. Positive feedback loops amplify the weakly held price opinions. Automated systems quickly react to perceived shifts from each other in rising uncertainty. Air pockets and wide spreads at short and medium time scales is because when large size comes in, prices fold rapidly until the facts are digested.
Regardless of the allocation model, pricing opinions must be weakly held. For dealers, even more important than getting a hedge off, is fading markets. Every single trade represents new information coming to market, and levels rapidly adapt. Prints are conflicting opinions that bring us closer to what that option should be worth. Price discovery is not quite like the scientific method, but it does repeatedly propose and test hypotheses.
Confidence is streaming a billion dollars worth of quotes, fragility of opinion is fine tuning the edge and retreat settings every single day. Market structure channels competing beliefs to efficiently price options and allocate capital.
Sadly, it hasn’t yet improved the market for tomato condiments. My strong belief is that there’s still no such thing as ketchup done well. But I’m willing to adjust my pricing quickly when it happens.
A cabinet trade is a mechanism that predates penny decrements, and was designed for trades to happen below the minimum price increment for “accommodation liquidations”. Traders and institutions who wanted to clear their balance sheets of worthless options could buy or sell at a $1 per contract ($0.01 in pricing terms).
After transaction and clearing costs, this was almost never worth it to be on the short side if you weren’t closing. You’d give up roughly half the edge premium just in processing the trade, all just to bear a ton of risk. But hey, a 500 lot could buy the crowd lunch?