> However, my partner, a real-life HFT guy comfortable with generating millions of dollars in returns with sub-bps opportunities, began salivating.
I couldn't roll my eyes faster..
Oh why did I continue reading..
> our most recent bot required substantial modifications to the validator software, constructing and evaluating a graph of thousands of tokens, math accurate to 18 decimal places, and robust reliability and error catching, all written in super optimized Rust code.
This genuinely almost reads like a parody. My BS-o-meter is through the roof!
I believe it. On Ethereum for instance before lots of competition came whoever was behind this address made literally hundreds of thousands of dollars a day in 2021 doing atomic arbitrage. You can see this graph of their balance here: the declines are due to them withdrawing money from the contract.
> This genuinely almost reads like a parody. My BS-o-meter is through the roof!
Can you explain why? Nothing really jumps out at me as being total nonsense. A lot of things are handwavey, but for obvious reasons such as not revealing the money printer.
The man is not hiding who he is either, so it could be a troll post, but from looking at other things, doesn't seem likely.
Math generally looks good, and is similar to things I learned decades ago in chaos theory and stochastic processes, especially the comment/proof towards the middle. I have never been a professional trader, but I am familiar with a lot of the jargon, generally translates well to crypto.
Clearly the guy isn’t going to open source his work so there’s going to be a great deal of hand waving but are you just unfamiliar with this area of work, because there’s nothing eye roll worthy in what I read.
Wow, so much negativity in the early comments. I've been in the space decades and there's at least 2-3 years of education to be gained here. Great piece, thank you!
It's definitely a lot of work to build these systems (but also fun if you are competitive and like math/cs puzzles). While the margins are thin in stat arb, the absolute profits are not and risk-adjusted (but not time-adjusted) returns quite good.
One of my biggest fears is that I spend time building things that are useless and have no value for anyone else. I absolutely hate when corp reorgs and other major events redefine priorities and make everyone waste their time into something nonproductive. This person did that for 6 years and "would not have done anything differently".
> And in retrospect, from a purely EV perspective, I do think that this little crypto side quest may have hurt compounding value into my primary occupation
What exactly does "stat arb" mean in this context? I assumed it meant "statistical arbitrage", so I was expecting the article to talk about a pairs trading strategy or something in that vein, i.e. a strategy that relies on a statistical relationship. But this seems to be more what I would call "geographical" or "locational" arbitrage where they're trading price differentials in the same instrument between multiple venues.
I defined stat arb in the glossary at the end but probably should have included a more obvious link.
TL;DR: this is a pairs trading strategy that relies on the (very strong) statistical assumption that the price of a token on a centralized and decentralized exchange will converge.
My longer definition below:
Stat arbs: in finance, statistical arbitrage generally refers to any trade where a pair of assets should statistically move in a certain way. However, there are degrees of should. TradFi traders might reason that Meta and Google are both in the ads business, so if Meta is relatively expensive and Google is relatively cheap, they should short the former and long the latter. However, this is a weak argument. Perhaps Meta is just a better business or Google has structural problems. A stronger stat arb thesis is that Royal Dutch Shell used to be traded on both American and European exchanges. If the shares were trading at different prices on each, nearly risk-free profits are available to those who close the spread. This is what stat arb means in a crypto setting. AVAX may be trading at slightly different prices on Binance and on various blockchains.
Isn’t AVAX transferable between blockchains and Binance? If so, this isn’t “stat arb”, it’s just regular arbitrage. When the instruments are fungible, it doesn’t matter whether there’s a statistical relationship; your profit is guaranteed once you complete the trade, regardless of how the prices move later.
The Royal Dutch/Shell case is different because the two stocks were not actually fungible, so the trade would only have been profitable if the two prices eventually converged.
PS I enjoyed the article and it’s clear you’ve spent a lot of time thinking about this space. I just can’t resist chiming in when it comes to well-established terminology.
In theory, yes, the tokens should be transferable between the blockchain and the CEX, but this link breaks all the time and for various reasons (see the crypto problems section). Lots of things that “should” happen in regular, efficient markets don’t in crypto.
That said, call it whatever you want it. Most people in crypto call CEX/DEX stat arb to differentiate from true risk-free arbitrage, but I agree that people coming from a traditional trading perspective would call this pure arb.
Good stuff sharing! Yeah, the higher tiers are necessary but you can get access to at least VIP6 on Binance as a trial if you’ll do volume and the higher VIP tiers are accessible with “colo” access on some other exchanges. We’re almost exclusively perps.
25 comments
[ 5.7 ms ] story [ 71.5 ms ] threadI couldn't roll my eyes faster..
Oh why did I continue reading..
> our most recent bot required substantial modifications to the validator software, constructing and evaluating a graph of thousands of tokens, math accurate to 18 decimal places, and robust reliability and error catching, all written in super optimized Rust code.
This genuinely almost reads like a parody. My BS-o-meter is through the roof!
https://etherscan.io/address/0x0000000000007f150bd6f54c40a34...
Can you explain why? Nothing really jumps out at me as being total nonsense. A lot of things are handwavey, but for obvious reasons such as not revealing the money printer.
The man is not hiding who he is either, so it could be a troll post, but from looking at other things, doesn't seem likely.
Math generally looks good, and is similar to things I learned decades ago in chaos theory and stochastic processes, especially the comment/proof towards the middle. I have never been a professional trader, but I am familiar with a lot of the jargon, generally translates well to crypto.
> And in retrospect, from a purely EV perspective, I do think that this little crypto side quest may have hurt compounding value into my primary occupation
TL;DR: this is a pairs trading strategy that relies on the (very strong) statistical assumption that the price of a token on a centralized and decentralized exchange will converge.
My longer definition below:
Stat arbs: in finance, statistical arbitrage generally refers to any trade where a pair of assets should statistically move in a certain way. However, there are degrees of should. TradFi traders might reason that Meta and Google are both in the ads business, so if Meta is relatively expensive and Google is relatively cheap, they should short the former and long the latter. However, this is a weak argument. Perhaps Meta is just a better business or Google has structural problems. A stronger stat arb thesis is that Royal Dutch Shell used to be traded on both American and European exchanges. If the shares were trading at different prices on each, nearly risk-free profits are available to those who close the spread. This is what stat arb means in a crypto setting. AVAX may be trading at slightly different prices on Binance and on various blockchains.
The Royal Dutch/Shell case is different because the two stocks were not actually fungible, so the trade would only have been profitable if the two prices eventually converged.
PS I enjoyed the article and it’s clear you’ve spent a lot of time thinking about this space. I just can’t resist chiming in when it comes to well-established terminology.
That said, call it whatever you want it. Most people in crypto call CEX/DEX stat arb to differentiate from true risk-free arbitrage, but I agree that people coming from a traditional trading perspective would call this pure arb.