Just moved SF —> NYC (life reasons) and trying to cope with leaving OBSF. Living in Brooklyn, looking at renting a board/wetsuit locker and then just commuting when the waves are good
Any recs ?
Got it as a gift, been having a tough time figuring out what it is. Even took the controller apart, just generic Chinese components
Apologies for the beginner question, somewhat stuck, figure someone might know
I am trying to compute PnL for a user given their full trading history (one row per fill). To my current understanding (and please correct me if I am wrong)
To compute PnL we need to know how much a user spent minting tokens. If we have an unambiguous set of trades for a token we can then deduce that the user must have minted at LEAST the amount we see sold that we didn’t see bought(sorted by time)
However, market data is structured as [price, amount, maker, taker, makerDirection, takerDirection, token] so for a single trade [0.3, 1, alice, bob, BUY, SELL, “YES”] EITHER:
- alice bought 1 YES from bob at 0.3, bob sold 1 YES at 0.3 to alice
- alice bought 1 YES from bob at 0.3, bob bought NO at 0.7 and was matched against alice
What is the standard way to distinguish between the cases (or more generally compute PNL without insight into minting)
Interested in everyones thoughts — I have done the standard Leetcode grind but from what I am seeing in the current market a fair amount of finance-flavored places are switching towards more design/implementation heavy questions (over algorithmic) and increasing the weight placed on clean, clear, concise code.
Anyone have any suggestions for how to build this particular skill? Last couple of interviews have gone badly for me not because my code was wrong, but because it wasn’t as clean as it could have been
In my standard flow (during normal work, not interviews) I frequently write things out, then look for opportunities to refactor and clean things up so I don’t get a lot of practice oneshotting beautiful code
Any advice appreciated