krazypepe

2.5K posts

krazypepe banner
krazypepe

krazypepe

@_krazyniko_

Always a Degen | Sometimes a Quant

Katılım Şubat 2022
2.4K Takip Edilen640 Takipçiler
krazypepe retweetledi
OfframpESP
OfframpESP@OfframpESP·
¿Estás en @ethlatam en São Paulo? ¡Participa en nuestro sorteo! ⚡️ Del 6 al 9 de noviembre, usa Offramp para pagar tus compras locales con PIX y participa por un Trezor Safe 5. 🤯 Cada dólar gastado es una chance. Más gastos = más oportunidades El ganador será anunciado el día 9 noviembre.
OfframpESP tweet media
Español
0
2
10
1.2K
El Dorado
El Dorado@eldoradoiobr·
FINALISTAS DO LEADERBOARD!!! Aqui no @ethlatam, junto com a @arbitrum finalizamos o leaderboard que possibilitou o pagamento de USDT ao principal restaurante do evento. Parabéns a todos que participaram!
El Dorado tweet media
Português
2
0
7
133
krazypepe
krazypepe@_krazyniko_·
@dappnode @ethlatam Not our nodes, not our blockchain! Real decentralization comes with people like you and me contributing to the network. Run Nodes 🫡
English
0
1
2
380
krazypepe
krazypepe@_krazyniko_·
@anthdm Tks G. The historical market data goes back until the exchanges start of the data or reduced ?
English
0
0
1
27
Anthony GG
Anthony GG@anthdm·
Market Monkey - the fastest and fully customisable trading terminal on the planet. Pre Order your lifetime access before prices are getting raised again.
English
4
0
37
3.3K
krazypepe retweetledi
pedma
pedma@pedma7·
Amazing episode. @ScottPh77711570 lays out a realistic path for a retail trader to compete, with the weight of someone that actually has done it. highly recommend both the episode and the podcast as usual. open.spotify.com/episode/1eYYt9…
English
1
6
44
4.7K
krazypepe
krazypepe@_krazyniko_·
@pedma7 Insightful analysis, pedma! Thanks for that! Quick question, what would the daily contribution to signal be ? Is it a correlation delta from previous periods ?
English
0
0
0
5
pedma
pedma@pedma7·
we can see that the correlation with forward returns is mostly significant within the first few days and decays after that. that spike later on the plot, is random spikes in momentum that are obviously not derived from our initial signal. what we care about is at the start.
pedma tweet media
English
2
0
6
1.8K
pedma
pedma@pedma7·
this is the argument for measuring signals on a continuous. if signals are most predictive when we get them, why am I holding a position with the original signal weight for multiple weeks? why am I discarding all the information within that period? x.com/ScottPh7771157…
Temu Robot James@ScottPh77711570

@moritzheiden @risk_biscuits @bylethquant Trend on crypto is most predictive 12-48hrs from signal, typically a continuous version of the same kind of signals the non-vol target systems use are better than looking at short timeframes This is correlation with next day returns at 1 day, 2 days, etc

English
1
2
62
17.9K
krazypepe
krazypepe@_krazyniko_·
@arnaud710 In fact…. another doubt came to my mind. Why not longing the perp instead of the spot? In this approach you would be receiving funding while having the same exposure with a slightly better executed average price than the spot. I mean, the alpha is in the oscillator, so why spot?
English
1
0
0
44
krazypepe
krazypepe@_krazyniko_·
@arnaud710 If I am grasping that correctly, the spot buying and perps shorting should be executed at separate moments ( ie. when big deviation down, buy spot only. when big rip, short perps only - managing inventory on the way). Is that the idea? Thanks for the explanation, boss.
English
1
0
0
93
krazypepe
krazypepe@_krazyniko_·
@arnaud710 Can you please give us an example? Im trying to do the math here… if funding rates are negative ( perps < spot ), and you buy spot and short perp, you are assuming a negative pnl on the price difference and also paying funding fee if you long spot and short perp at the same time
English
1
0
0
228
krazypepe
krazypepe@_krazyniko_·
@pedma7 It’s clear now. Thanks for the explanation, mate!
English
0
0
1
25
pedma
pedma@pedma7·
thanks mate! in this case, I use the "cash balance" to track both equity and also the "borrowed" position used for the short. so it increases as shorts grow and decreases as shorts get smaller. the position size acts as the balancing variable for the total equity. for example. I have $100 in my portfolio I short $100, I now have a "cash balance" of $200 I now have a position value of $-100 I have total equity of cash balance + position value = $100 because I consider it as a liability that I owe. maybe a different name from cash balance would be better, but works for now.
English
1
0
2
55
pedma
pedma@pedma7·
went on a side quest today. but I think this is it, I've fixed all the bugs in the continuous exposure simulation. now every day has a "state" based on our forecasts, and size is adjustable by thresholds. started as a simple project last week, took me ~30hours to complete.
pedma@pedma7

finally, started modifying my infra to model for continuous signal exposure 3 days ago, and finished it today. now I can both model for traditional binary signals and continuous signals, with just a change of a variable. let's analyze what changed and reconcile. 1/n

English
2
0
8
3.4K
HangukQuant
HangukQuant@HangukQuant·
🚨🚨🚨 celebrating my 20k account with a 40k+ quant content giveaway. The last time I did this was my bday last year. doubling it this time, with the following prizes: form is here: forms.gle/m8q8sFBsdpskXr… roughly 200~800 participants each time, so make a wise choice🤭 good luck! just rt+like/comment, boost the algos and good luck to you? let's speedrun it to 30k
HangukQuant tweet media
English
55
127
261
24.1K
Wizard Of SoHo (🍷,🍷)
Wizard Of SoHo (🍷,🍷)@wizardofsoho·
This Cuba lore is wild. The original has to be worth something A country CTO Target is their GDP
English
54
35
242
29.1K