The Fog of Uncertainty
Vol #123: March 30th, 2023
“There was only the confusion of a people not being able to predict the future.”
Lieutenant Dunbar’s observations of the Sioux people he has warmly befriended comes after one of the most heartbreaking scenes in “Dances With Wolves”. Dead buffalo lie everywhere. Slaughtered but for the price of their hide, even their murders’ wagon tracks scar the face of the prairie.
The film tugs our heart strings from the very moment that Kevin Costner’s character jams his boot back over a gangrenous foot. Choosing death over amputation, his suicidal mount becomes an inspirational battle cry. Confederate soldiers lie stunned after their markless barrage, and the hero gets a dispatch of his choosing.
Our protagonist is caught between two worlds. As a soldier, the lieutenant’s duty is to his country and his solo post at the edge of the frontier. As a human being, it is to the humanity he has found amongst the Sioux. The magnitude of this struggle is only matched by the breathtaking expansiveness of the American West.
If not always so stark, this dual mandate is not uncommon. The opposing forces of life are easily characterized by the angel on the left shoulder and devil on the right. If only it were as simple as white versus red. Uncertainty breeds confusion.
Trading, investing, and more broadly economics are all faced with the constant struggle of needing to make decisions about the unknowable future. The Federal Reserve tweaks interest rates with the objective of balancing unemployment and inflation in the medium term. We know not what tomorrow holds, but we must choose a path based on what we hold valuable and understand today.
Simple and predictable trading strategies resonate because they provide a guide no matter what noise du jour echoes. Dollar cost averaging is not only effective, it’s incredibly easy to implement. Covered calls might give away the right tail of returns, but many investors find FOMO-less comfort and consistency with the premiums collected.
This past week I spoke to my old high school’s finance club about the options markets, explaining some of the ways investors use these derivatives. Everyone quickly understood why one of the most popular strategies is a volatility buffering position that sells calls to buy puts or put spreads. (I certainly did not grasp that at 17 years old.)
What was surprising, was the fact that there was no complex research or modeling behind what position to put on.
“So do they use fundamental research or quant signals to adjust the position?”
“Nope, the timing and the moneyness are preordained, and that’s just how the investors like it.
The most complicated thing about hedged equity is matching the tradeoffs to the individual. Discretion drives even the most steadfast soul mad. Woulda, coulda, shoulda. The fact that I am protected on a drop between 5% and 20%, and can make up to 6% or 7% in a quarter is pretty comforting in the face of the media's font of gloom and doom. Boundaries for an unscripted next chapter.
Executing that trade in size is a delicate balancing act. The massive options position must be born by dealers, who need to manage their own directional risk, and lay off other exposures. The second order effects of greek changes leading up to the roll can create little whirlpools in market pricing.
The dance of execution becomes more interesting the greater your demands. If you have tens of thousands of options to trade, or millions of shares of stock to move, navigating this uncertainty is like a vessel in fog, balancing between speed and position.
Participants enter the market with an objective. For asset managers, this might be to adjust their holdings as part of a monthly rebalance. Statistical arb shops will be looking to put on a pairs trade to capture a dislocation between Home Depot and Lowes. Options market makers come to the underlying markets to hedge their delta exposure. Everyone wants to pay a little less, or sell a little more dearly.
The liquidity of the product will dictate how much you can move how fast. An execution trader for a mutual fund knows they need to be long a certain number of shares, but must walk the wire between price and speed. If you try to scoop too much at once, you’ll end up paying a very high liquidity premium.
There is risk in being too slow also. Waiting around with cheeky bids below the market means price could move against you. All of the sudden you’re paying up for the shares your ego prevented you from buying earlier, and you’ve sold calls underwater or sacrificed bps on an equity position.
The constant jockeying of all these participants, optimizing in real time for their constraints, is what puts together an orderbook. If the stock market is a weighing machine in the long term, and a voting machine in the short term; at the bleeding edge of microstructure, it is Schrödinger's fog.
Bid and offers fly in randomly, haphazardly arranged by circumstance and intent. There is no rhyme or reason as to why someone hits buy before or after taking a bite of their club sandwich. It’s completely unpredictable how the recursive high frequency signal reading will have butterflies flapping each other's wings in a chaotic dance.
The algorithms processing thousands of points of market data light and spitting back out orders read something in every single tick - price changes and size movements are pearls of wisdom to update their conditional probabilities. ||: How I react to what you do changes how you react to what I do :||
If someone has the magic money machine for finding the box of upticks, they’ve kept it awfully quiet. Some guys in Long Island are pretty close, but even for those chugging out 60%+ annual returns it all comes down to well sized and better timed bets. Execution algorithms place little order slices across dozens of dark and lit venues, while constantly tweaking prices as they divine which way the drunk stumbles.
Sometimes the goal is to simply “do what the market does”. This is the dollar cost average approach to execution. VWAP (Volume Weighted Average Price) algorithms track how much size is trading at what price, and try to mimic that for their own tranche of flow. Just as you can’t buy the S&P 500, and SPY has some tracking error, getting the VWAP price requires some delicate maneuvering.
For traders with a time horizon longer than a few minutes, this is a stalwart. VWAPs keep you from buying high and watching the market tumble hours later, and they provide a logical benchmark. It’s fairly defensible to report back that you paid the average price of that stock on that day. Unless you’re an HFT, your edge should be coming from somewhere else.
Other times the objective is to smash and grab with as little collateral damage as possible. This analogy spoke to traders, because every executing broker used names like Guerilla or Sniper for their algos until Compliance tightened the screws. There's an art to the quick heist, and effective algorithms will pounce on liquidity across multiple venues, ping block sizes, and simultaneously scoop offers with picosecond precision to maximize capture.
For an options trader that just took on a large delta position from a trade, or got lucky with an overnight gap, getting it done quickly is more important than the marginal gains that come with the risk of waiting. Implementation shortfall - the difference between the market when you walk in and when you get filled - is significantly higher here, but what’s a few basis points between friends when you just ingested a fat piece of edge.
Price is of primo importance, but how you trade also impacts the market. The fees are different depending on what type of order gets placed. For the certainty of a fill, you not only have to cross the bid-ask spread to buy, but you’re also likely paying a higher “take” rate. Market centers give rebates to orders that add liquidity because it creates a deeper or tighter book that makes them more likely to capture net revenue on the next matched trade.
When you’re taking liquidity, you’re also signaling to the market that there’s a participant with urgency. That means one thing, price will move away from you. If the smash and grab gets too aggressive, everyone fades quickly and you’re left empty handed.
Mechanisms like dark pools have their own rules about how orders get matched. They’re so named because the depth of liquidity is only known ex-post, and there are often unique rules about mid-point matching and minimum sizing. It can be very effective to traffic out of the light if you’re moving size, but too many indications of interest and your counterparties will smell blood in the water.
All of the data points about current trading activity and past dynamics are crunched in real time by these order handlers. Dark pool fill rates are sliced by sector, symbol, and order size. Exchange rebates and take rates are plugged into probability matrices of where the next taker order is likely to go. If someone else is repeatedly hitting the bid in MSFT stock, it’s time to get more aggressive with your AAPL sell order.
The uncertainty of any individual orders life cycle means this is a game of large numbers. Sometimes you ping into the fog and there’s a fill, other times it’s empty. It’s only on the order of millions of shares traded that the fruits of this type of execution become observable.
There is no right answer in the debate between speed and price. If a hanging delta is going to distract you from the next opportunity, you’re better off cleaning that up. If you’re steel nerved enough to sweat out those extra few basis points, there’s real money to be made in confidently pushing through the pea soup.