Mark Best

I model professionally. HFT, MM, rust.

[Mark Best @MarkBestForex]

If anyone wants to read about one of the ways you can get flip flop bugs like this i wrote about it a few months ago. Its basically really easy to do and one of the fastest ways you can incinerate and account.

https://t.co/ov6nMP5Rso

[yenwod @yenwod_]

~100m gone in in 13 minutes

rough estimate assuming ~50-75 bps px impact times the 16b bybit volume

[Mark Best @MarkBestForex]

@PeterSchiff Please tell me your boat is called 21m?

[Mark Best @MarkBestForex]

@moving_charlie define need?

[Mark Best @MarkBestForex]

@josusanmartin we dont say things like this out loud

[Mark Best @MarkBestForex]

@dkosov255 Its really not

[Mark Best @MarkBestForex]

async code really is one of the leakiest abstractions.

[Mark Best @MarkBestForex]

@ukboomers There's no need to flex about how rich your children are.

[Mark Best @MarkBestForex]

@max_conradt yes

[Mark Best @MarkBestForex]

@stalequant hmmm

[Mark Best @MarkBestForex]

@annanay is it though?

[Mark Best @MarkBestForex]

Pretty much all lead lag analysis is wrong if youre using public data. Even if you collect it yourself and do the time alignment perfectly, crypto data is hot garbage when things get spicy.

Binance latency has 100 ms tails during stress and their public feeds can completely dislocate. You can "fix" this with exchange timestamps in post trade, but then you cant align timestamps with much confidence.

Actual lead-lag cross ex works completely differently to how most people think. What matches is often very different from what the public feed suggests.

So in summary, This is great analysis if you ignore reality and don't want to make money.

[Dylan O'Donnell 🦋 @erfmufn]

Two yachts have run aground at Ballina and Byron Bay recently and are both still stuck there, but last night Markus Pache 📷 took this shot of the one stuck at Flat Rock, East Ballina. Captures the dark emu at the perfect angle !

[Mark Best @MarkBestForex]

Its also worth saying beta hedging is not a perfect because estimated betas mean revert. There's methods that "deflate" betas to try to compensate for this. But you're likely to use the wrong beta of you're doing mean reversion as when you're putting on the positions the betas will be at extremes.

[Mark Best @MarkBestForex]

@real_brad_smith @ScottPh77711570 dollar neutral is fine if youre trading stuff with similar betas

[Mark Best @MarkBestForex]

@real_brad_smith @ScottPh77711570 betas imply you have some benchmark (you normally do) in equities its an index, crypto its BTC.
If you're doing delta neutral on something without a benchmark then its more like a risk parity model which is just adjusting size by vola. If you want to be fancy there is HRP.

[Mark Best @MarkBestForex]

@ScottPh77711570 @real_brad_smith I'm waiting to hear if they know delta neutral != dollar neutral. I know a lot of people use that interchangeably.

[Mark Best @MarkBestForex]

@Butterknifeopr8 Im from the UK. I had to find out where it is on a map!

At least the queues should be good!

[Mark Best @MarkBestForex]

@real_brad_smith @ScottPh77711570 Im assuming youre just talking about being dollar neutral?

[Mark Best @MarkBestForex]

@andyhyfi Can you get a PhD in losing money?

[Mark Best @MarkBestForex]

@erfmufn @astrojml Does this seem right to you? It follows my intution that even bortle 3 is pretty good.

[Mark Best @MarkBestForex]

@erfmufn @astrojml Its still amazing that cuiv manages get the results he does from tokyo. I know most people move their scopes to places like NM but i thought the main advantage was clear skys and time on target. Ive not thought before what the S/N curve looks like as a function of bortle.

[Mark Best @MarkBestForex]

@erfmufn @astrojml How much of a difference does bortel 3 vs 1 make for photography? I'm assuming it mainly helps for faint stuff. I know astrobackyard does trips.