Published: March 13, 2023
6
4
34

1/ "They wouldn't announce it if they hadn't bought already" is a common fallacy on CT. Similar people who say "evil HFT just frontruns and provides no value." If you want to execute large size, often the best thing to do is announce it. A thread on efficient markets 🧵

2/ If you are CZ and want to buy $1 billion of bitcoin what are your options? You could start blasting the public books. Your execution is going to suck, it's gonna take you forever, and your short term price impact will be massive.

3/ Or you tell people what you're going to do. The quants do the hard work of competing to provide you the liquidity. A $1 billion buy always moves the price but this way you get the correct price impact plus a premium, which depends e.g. on probability you are lying.

4/ In a perfectly efficient market, if you commit the $1 billion to a smart contract that will market buy outside of your control at a specific time, announcing it should get you close to optimal execution. The real world is messy, but the intuition here still works.

5/ Don't just take my word for it. For example, look at tradfi where people really care about execution quality. Most levered ETFs predictably rebalance on the close to minimize tracking error. Basically a daily "announcement" for people to "frontrun" and works like a charm.

@chameleon_jeff make a bunch of wallets and blast it all at GMX and its clones

@skyquake_1 Limiting OI per account without KYC never made sense to me

@chameleon_jeff so if I understand it correctly, marketmaker A's ask price is far away from the best ask they could provide, so B steps in and offers a bit better (closer to the best price) and so on, which leads to a good ask price instead of actually just highball offers?

@fuckyourputs Yeah could be that, or when CZ lifts the offers some MMs who saw the announcement could be faster to replenish or fade less knowing the story behind the flow. In general the market will better understand what the fair price is and give CZ a better average price

@chameleon_jeff What I saw today is that most of the liquidity on spot was on the bid side, and (for example for ETH) the first really big sell walls appeared between 1650-1700.

@ltrd_ Like a concrete example maybe CZ is placing fat limit orders at the top of the book until he gets his size. The quants frontrun until they have enough, and then start hitting the bids until CZ is filled. Simplified picture and no one knows exactly, but the market as a whole knows

@chameleon_jeff Don't you think that during such time, most of the liquidity will be on the side of the news? And then you will be frontrun by others and no market makers will appear in order to take your market orders? From time to time you will see some sell walls but not that size.

@ltrd_ Liquidity provision at such large scales don't happen as discrete sell walls, but rather in the market movement as a whole. Basically all quant actors are detecting flow and responding to it, information is transmitted over time and execution is continuous in a complex dance

@chameleon_jeff A bit out of theme question. Are you hedging somehow stablecoin exposure for your futures strats?

@kdcd22802757 Nope, there are always risks and cost to hedging. We do 80/20 on the important stuff, then focus on EV

Share this thread

Read on Twitter

View original thread

Navigate thread

1/15