Daru (@DaruTrading) header
Daru (@DaruTrading) profile picture

Independent quantitative research organization Top 1% author on SSRN Publish it, or it didn't happen. Contact [email protected]

Joined November 2024
Photo by Daru (@DaruTrading)Photo by Daru (@DaruTrading)Photo by Daru (@DaruTrading)Photo by Daru (@DaruTrading)Photo by Daru (@DaruTrading)Photo by Daru (@DaruTrading)Photo by Daru (@DaruTrading)Photo by Daru (@DaruTrading)Photo by Daru (@DaruTrading)Photo by Daru (@DaruTrading)
Profitable web3 traders aren't better at reading charts. We tested tokens across 27 blockchains. Every strategy FAILED even a null at zero cost Rug pulls. Dev wallets. Insider information. That's where their wins are The edge is real, It's just illegal papers.ssrn.com/sol3/papers.…
579
390
35
3,872
3,972,103
You tell yourself memecoin trading isn't gambling I analyzed over 44 million positions, rebuilt from the chain. Memecoins win 75c per dollar. Roulette pays 94.7c. But what about perps and prediction markets? Much better. It's all here 👇 Full paper papers.ssrn.com/sol3/papers.…
15
30
420
623,770
How much historical data should you train on? Most of the time this question is settled by habit So I tested it with six lengths, three model families, two markets, 176K configurations Which one earns best? daru.finance/alpha/training-…
5
5
12
1,156
Nothing says a candle has to be one minute long Try one candle per fixed volume traded instead. Same data, different formula Crypto returns got 3.4x closer to a Gaussian curve. US stocks and forex had similar results But does it help you trade better? daru.finance/research-review…
9
16
134
1,382,026
Does a portfolio beat the sum of its strategies? 80 strategies on 105 coins, every signal funded on its own. Versus the same signals, but in one account. 1.2 million signals, 3 years out of sample, costs included. They differ more than you may expect daru.finance/alpha/portfolio…
1
3
17
1,027
If you want to gamble, where should you go? Memecoin "trading"? Prediction markets? Casino? I tracked 44 million memecoin trades, 21 million Polymarket bets, and the odds of the top online casino games. Do what you must with this information.
43
18
4
200
458,659
Does trading more coins actually make you more money? I ran the same strategies on 6 coins, then 15, then 30. Changing nothing else 91,277 trades, every fee, spread and funding charged One formula explains the whole picture daru.finance/alpha/universe-…
2
6
17
1,131
This is why you should not trust anyone who posts strategies on here. A strategy can look great with a high profit factor and winrate, But nobody mentions trial count, DSR, FDR or OOS performance... This is how a great backtest can fool you. Learn the basics of statistics.
4
4
13
943
How much information do you miss out by only trading OHLCV candles? One minute of a liquid US stock is thousands of trades. A candle keeps 5 numbers I kept all the data, and compared to regular charts, forecasting the same returns Is more data better? daru.finance/alpha/tick-data…
1
2
12
957
Support and resistance levels mean nothing. On a random chart, plot RANDOM lines, and it WILL look like valid levels. Why? Support and resistance are nothing but confirmation bias. All of the lines in the video are random. Test it for yourself, link in the comments.
18
5
57
427,963
Would you trade a strategy you cannot explain? Most people say no, then trade one anyway So I tested it. Strategies written from a reason against strategies found by searching, same budget, 60 coins, 18 quarters daru.finance/alpha/economic-…
7
2
1
19
1,707
Most traders lose for a reason you can calculate Run 50,000 strategies with no skill at all and the luckiest still shows a Sharpe of 2.72. Tested empirically, with every fee, spread and funding charged, all 92,500 That's the deflated sharpe ratio. daru.finance/research-review…
8
6
95
547,392
What is the best time to trade a coin after it's listed on a major exchange? Seven age buckets across the first year, each priced on volatility, funding, spread, depth and volume Measured on 832 contracts, 15.3M ledger rows. See more daru.finance/alpha/market-ma…
2
1
14
991
Is machine learning enough to make you a good trader? 30,960 models searched, each tested against a null hypothesis for statistical significance. 8,400 rule-instrument pairs, all losing strategies. Tested in crypto and US equities. daru.finance/alpha/ml-strate…
4
1
9
946
How easy is it to make money market making in crypto? Post a buy, post a sell, pocket the gap. Bots do it all day I simulated 125M fills on Bybit, Binance and Hyperliquid with their own buy/sell records It's not what the bot sellers promise Full paper papers.ssrn.com/sol3/papers.…
6
5
21
989
Your strategies have a shape. Almost nobody trades it. Turn how each pair moves together into a distance and they fall into a map: crowded middle, thin rim. The shape of your strategies might contain valid information you never considered using. daru.finance/articles/topolo…
6
5
21
1,159