Recently, I have been running a 30,000 U strategy for about a week.

The returns in this area are relatively stable, basically maintaining a level of 1–2% daily, and occasionally catching a small surge of 5%. The drawdown is also well controlled, steady at within 1–2%. The overall rhythm is gradually being felt.

Next, I plan to continue optimizing the model parameters to further improve stability. To be honest, I'm quite looking forward to seeing what the profit curve will look like a month from now...

However, there is a major flaw right now – there is still too much manual intervention. It would be perfect if it could be fully automated, but unfortunately, reality is always harsh 😭

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.

15 Likes

Reward
15
5
Repost
Share

Comment

0/400

MEVEye

· 11-30 13:35

Is a stable return of 1-2% per week serious, or is it backtested data? Parameter tuning is never-ending; the core issue is still to solve the automation bottleneck. With 30k principal, this rhythm is achievable; don't expect too much. The amount of manual intervention depends on the market conditions; the real test is whether it can withstand extreme situations. The returns are stable, but what about the Transaction Cost? Has it been accounted for? Complete automation is a false proposition; the market changes too quickly. Choosing the right track is actually more crucial than the parameters. I want to ask what cycle the 1-2% drawdown is based on. I need to observe this stability further; it's too early to draw conclusions with such a short cycle.

View OriginalReply0

ForkTongue

· 11-27 14:05

1 week of data is too little, can backtesting and real trading be the same... Wait, is manual intervention still very frequent? Then how do we calculate this return? 1-2% seems stable, but don't be too optimistic before full automation Can 30,000 U really keep running? I'm a bit curious Daring to optimize parameters with such short data? It's easy to overfit, buddy Can you share which part requires the most manual effort? Maybe there's a way to simplify it We'll see the truth at the end of the month, it's still too early to say it's stable

View OriginalReply0

ApeEscapeArtist

· 11-27 14:05

1-2% stable in a week, this data is quite impressive, but the gap between automation and reality is indeed a dream. Too much manual intervention is really annoying, and when things get busy it's easy to mess up, so we need to find ways to reduce this vulnerability. That's how reality is; the perfect strategy is always in the next version. The profit curve a month from now is worth looking forward to, but remember to keep an eye on the risks as well. A 5% explosion is a bit exciting; how do we balance stability and profitability? If the parameters are optimized well, can we reduce some interventions? It feels like this is the key.

View OriginalReply0

gas_fee_therapy

· 11-27 13:59

A stable 1-2% weekly is indeed possible, but there's quite a bit of manual intervention; complete automation is the way to go. --- 5% surges occur occasionally; the rhythm is still okay... just worried that optimizing parameters might destroy the feel. --- To be honest, too much manual intervention is equivalent to having no strategy; automation is the ultimate form. --- Seeing a stable 1-2% daily is reassuring; after a month, compound interest really isn't bad. Looking forward to your monthly report. --- The hard truth hits home; to truly scale up, complete automation is necessary. --- 1-2% drawdown must be tightly controlled, but the question is, how long can it last... that's the real test. --- Finding the feel is just the beginning; the real bottleneck lies in automation. Fighting!

View OriginalReply0

BridgeNomad

· 11-27 13:59

ngl 1-2% daily sounds smooth until the first flash loan hits your slippage tolerance... been there, trust me. manual intervention is probably saving you rn tbh

Reply0

Trending TopicsView More
#JoinGrowthPointsDrawToWiniPhone17
235.22K Popularity
#DecemberMarketOutlook
31.66K Popularity
#GIGGLEAndPIPPINSurge
5.32K Popularity
#LINKETFToLaunch
5.2K Popularity
#SharingMy100xToken
4.8K Popularity

Hot Gate FunView More

1
MANMi Airdrop
MC:$3.53KHolders:1
0.00%
2
PIKACHUSuprised Pikachu
MC:$3.57KHolders:1
0.29%
3
DF61DF61
MC:$3.52KHolders:1
0.00%
4
DLHDr Lin Han
MC:$3.58KHolders:1
0.31%
5
BTCcBTCc
MC:$3.48KHolders:1
0.00%

Sitemap