Root Two

[Temu Robot James @ScottPh77711570]

A rare example of very good AI slop

Worth a read

[Don Wilson @drwconvexity]

1/ I encounter many misperceptions about perpetual futures (aka “perp”). A perp is simply a futures contract without an expiration date. That's it. Most of what people think they know about "perps" (first contemplated by Robert Shiller) has nothing to do with the contract itself.

[Andrew Mack @Gingfacekillah]

Look, just trade what’s in front of you.

Not the news. Not the chatter. Not the hot takes on the news and chatter.

It’s hard to hide positioning changes across the vol complex. VIX, VVIX etc.

I haven’t read a single piece of news in months.

Try it. You’ll do much better.

[Root Two @RootTwo23683]

@ScottPh77711570 @systematicls @AgustinLebron3 @macrocephalopod Hm I’m wondering what this looks like if you try this on multiple LLMs? Wonder if that would help you figure out if this is from data of other users (diff users will use diff models) vs getting signals from data the llm was originally trained on (presumably similar across llms)

[Brett Harrison @BrettHarrison]

HFT Market-Making Systems

I spent the first 13 years of my career designing algorithmic trading systems for large HFTs in options, futures, and ETFs (intl/US equities, commodities, fixed income, volatility). Here’s a summary of each component, note the role AI/ML actually plays.

Market data. A market-making system requires realtime orderbook deltas and trades, not just for the symbol that’s being quoted but also for every related instrument used in computing the theoretical fair value of the instrument. Every exchange has a different protocol specification and format, from JSON over websocket (slowest) to FIX (common) to fixed-width binary (fastest). The lowest latency systems will ingest market data on a different thread or process to avoid interfering with the core trading logic of the system.

Fair Value. Successful market making depends on a precise understanding of the theoretic fair value of an instrument, which is the indifference point between buying and selling the instrument modulo fees and other costs. The complexity spans from a several-parameter model based on one or two instruments to highly multivariate calculations across a basket of different symbols. This component is usually the firm’s most guarded intellectual property.

Order placement. Accurately predicting the price of an instrument is only half of the core logic of a trading system; deciding where and how to place orders is equally determinative of P&L. Maintaining balanced two-sided quotes, optimizing for queue position, minimizing traffic to comply with order rate limits, canceling to avoid adverse selection, and order sizing are all intricately determined by the strategy designer.

Exchange connectivity. Similar to the market data component, all trading systems must send orders and cancels using each exchange’s unique messaging protocol. JSON, FIX, and binary formats are all used, with the last most common among traditional exchanges where latency and throughput are optimized. The connectivity layer can be embedded within the order placement process or segregated as its own gateway process.

Offline training. Training is not part of realtime transformation of ticks to trades, but it is one of most important differentiators between trading firms. All parameters that define fair value calculation and order placement are set via offline compute tasks that are often guided by machine learning. A significant amount of the GPUs that large HFTs buy or rent are directed to model discovery and parameter optimization.

A handful of HFTs already use specialized GPUs for the linear algebra behind valuation, but the practice remains rare, confined to strategies where predictive power matters more than shaving the last nanoseconds off latency. As purpose-built inference silicon such as Google’s TPU and OpenAI’s Jalapeño becomes widely available, more HFTs will fold online ML and inference into their systems’ trading loops.

[Root Two @RootTwo23683]

@__nemo__omen RFQs can only be initiated in yes direction, liquidity taking is all going to be in the yes direction

[Root Two @RootTwo23683]

@Merridew__ What if you don’t find anything interesting beyond eating food, depression, and shitposting online?

[NewSHA @__nemo__omen]

Tripled my money in World Cup prediction markets while taking minimal risk.
The strategy? Exploit miscalibration on Kalshi by buying near-certain game-spread contracts. Nearly all resolved in my favor.

Recently, I did another analysis of miscalibration on moneyline and parlay contracts. Check it out: https://t.co/dwUx2sII9K

[Root Two @RootTwo23683]

@SipaAirdrop @0xd1namit @mustafap0ly This implies 85% chance of >2.5 goals conditional on both teams scoring? Is that much lower than correct probability?

[Sipa @SipaCapital]

@0xd1namit Hopefully they'll improve the RFQ system because right now it's a joke. I've been using combos from day one, but in most cases the quotes are much lower than they should be lol

@mustafap0ly

[ak0 @annanay]

I'm seeing Fable and co get better and better at trading.

It's hard to imagine a more efficient way to spend tokens for short-ish term returns than to run trading strategies.

What does this mean for SOTA model pricing? Will quant firms cannibalize top model use? Will they build their own AI labs and put their researchers under heavy noncompetes? Will the AI labs build out trading for monetization?

Just like how the price of HFT infrastructure (low-latency lines, FPGAs) has converged to the price of the best trading strategies that can run on them, soon token prices should just be the expected P&L of the strategies they can produce.

[peepeepoopoo @DeepDishEnjoyer]

this will become outdated in a year or less, but what are the most engaging resources out there for how to best setup agent workflows? especially in the context of data-sciency/researchy tasks?

like any good youtube how-tos?

[sysls @systematicls]

https://t.co/Ha4fP1gn1i

[Root Two @RootTwo23683]

@Ksidiii How do I learn more about these qis indices?

[Root Two @RootTwo23683]

@bennpeifert @bennpeifert how did you know other market participants might not have to liquidate? This kind of thing can cause a positive feedback loop no?

[Root Two @RootTwo23683]

@bennpeifert Benn, how did you time when to flip into a short position? Do it too early and you get run over when it goes from 7 points to 15; do it too late and you miss it. Do you just have to get lucky?

[Root Two @RootTwo23683]

@AgustinLebron3 @GZuckerman @aosipovich @matt_levine @TheStalwart @RobinWigg @izakaminska @ByrneHobart @johnauthers

[Agustin Lebron @AgustinLebron3]

The fraction of finance journalists who actually understand trading is small. A nearly-exhaustive list:

WSJ: @GZuckerman @aosipovich
BBG: @matt_levine @TheStalwart
FT: @RobinWigg @izakaminska

@ByrneHobart isn't exactly a journalist but close enough.

Who's missing?