Livermore

[Livermore @shortLivermore]

@openservai Wow

[Fomocap @fomocapdao]

Trading Models

1/15 There are exactly two ways to build a strategy:
- Model-based (start with a theory of the market)
- Data-mining (let the data scream patterns at you)

Model-based wins long-term. Here’s why.

2/15 A model is a simplified picture of reality. Reality itself is unknown. 100 different models can describe the same price action. You pick the winner by experiment, not opinion.

3/15 Model 1: The Order-Book Model Price is whatever clears the thickest cluster of resting orders. Everything else (charts, news) is just noise pushing orders around.

4/15 Model 2: The Random Walk Model
Current price = previous price + net buyers – net sellers + noise

99% of the move is just the previous price (inertia). The other 1% is where alpha lives.

5/15 Two unbreakable laws from Random Walk:

Rule 1: A pure random series cannot be traded profitably (no roulette system beats the house).

Rule 2: Volatility grows with time. A 4x longer hold is only 2x bigger moves.

6/15 Market inefficiency is any systematic deviation from pure random walk. That tiny deviation is your only real edge.

7/15 The General Price Curve Model Price gets pulled by four forces:

- Previous price (~99%)
- Random crowd
- Fundamentals
- Technicals

Exploit when one of the last three dominates.

8/15 Core regime-specific strategies that actually work:
- Momentum (trending regime)
- Low-pass filter, smooth trend line
- Buy the valley of the trend wave
- Short the peak
- Works until mean-reversion kicks in.

9/15
- Mean Reversion (ranging regime)
- High-pass filter, detrend the price
- Short above +2σ band
- Long below -2σ band
- Fade extremes.

10/15
- Cycle Strategy
- Find dominant cycle (Hurst, Hilbert, or simple EMD)
- Only trade when amplitude > threshold
- Buy near sine valley, sell near sine peak.

11/15
- Stat Arb / Price Constraints
- Some assets have hard or soft boundaries.
- Fade the boundary or front-run the break.

12/15
- Clusters & Gaps
- Prices love to cluster (support/resistance = order clusters).
- Weekend gaps amplify behavior: Friday extreme + Monday fade = free money half the time.

14/15
- Seasonality & Heteroskedasticity Turn-of-month, Monday open, 10 am NYSE spike, rollover days.
- Same setup performs 2-5x better on certain days.
- Simple calendar filter = instant edge.

15/15 The real development process (never skip a step):
- Pick model
- Confirm with data
- Code entry rules
- Code regime filter
- Optimize lightly
- Forward-test
- Add risk overlays

Model-based beats data-mining 9 times out of 10. The market is mostly random. Your only job: catch the 1% that isn’t.

[Livermore @shortLivermore]

Man… If you shilling, do it right. Recount this:

Global Gambling Market ≈ $300 billion

If $SCS token gets:

1 - 0.01% share → $30 million market cap → ≈ $0.009 per SCS
2 - 0.1% share → $300 million market cap → ≈ $0.09 per SCS
3 - 1% share → $3 billion market cap → ≈ $0.90 per SCS

P.s. Global gambling market estimated at 550b. 110b online gambling.

[Livermore @shortLivermore]

When the markets settle down — the next rotation will be into solana tokens.

$DM is quietly being developed and burned daily.

I believe it will be the next runner capable of capturing mindshare through:
- streamers (which will increase mindshare)
- unique concept (predict streamers results)
- ideal tokenomics (daily token burning from platform profits)
- upcoming events (adding new games with the ability to place bets on them)

$DM will become the new leader in its segment.
The project allows both regular users and popular streamers to earn money by playing games. There are no analogues in web2 or web3.

#streaming #betting #FC26 #Fortnite #Solana

[Batster @batster41]

Seems like we have cracked user acquisition game 👀 $DM

[Ilya @Hatshaft]

1/ Since @pumpdotfun set its development vector toward streaming platforms and promised support for streaming-related projects, I realized I needed to look for runners in this area. Nothing decent had come up - until I saw @streamdotquest. It fit the meta perfectly: a platform designed to connect fans with streamers by allowing users to create, fund, and challenge streamers with specific “quests” to perform live on stream. “Perfect!” I thought, and immediately bought their token.

[Livermore @shortLivermore]

Now that the dust has settled, I’ve been revisiting projects related to streaming on @pumpdotfun. I see that @DuelmastersGG, which took about three years to build, now has competition in @streamdotquest. After spending some time studying @streamdotquest, my first thought was that even offering 50% of the funds raised wouldn’t be enough for a streamer to figure it out. It feels a bit complicated due to its challenge system and a UI that isn’t very user-friendly for normies. In my view, $DM - with its familiar elements of gambling, betting, and streaming popular games - is much more effective at attracting users and streamers.

The $QUEST capitalization is 2kk, and the $DM is 200k. An obvious upside for the DM.

[Livermore @shortLivermore]

Found an interesting project that perfectly fits the @pumpdotfun meta and at the same time has no competitors in this. @DuelmastersGG with the $DM token that is already being traded on pumpfun. How do you like the idea that combines streaming of popular games, bets, excellent tokenomics with buybacks and gambling? Just imagine how streamers will start playing games attracting new users who will spin buybacks.. and soon you will be able to play any popular game against each other. A wildly crazy idea that has been in development for several years and fits perfectly into the narrative.
https://t.co/HdrheqN35K

[Batster @batster41]

Hey @a1lon9 and @pumpdotfun

Now that Creator Fees have become really meaningful, I think it would be great to add a feature that allows projects to allocate part (or all) of their fees to token buybacks — and make this visible directly in the PumpFun interface.

That way, users can immediately see which projects are returning value back into the token, without having to do their deep research.

For example, we’re already allocating 50% of our fees to buybacks of $DM, but right now this doesn’t really set us apart in the eyes of users, since it isn’t clearly visible on the platform. Making this transparent and easy to spot would help traders and investors better evaluate projects, while also adding even more value to PumpFun itself.

[Livermore @shortLivermore]

@batster41 @a1lon9 @pumpdotfun @a1lon9 good idea

[Quant Science @quantscience_]

JP Morgan's Python training.

Available 100% for free:

[apsk32 @apsk32]

The 10-year base case.

[PyQuant News 🐍 @pyquantnews]

Algorithmic trading cheat sheet (for download).

[Wu Blockchain @WuBlockchain]

According to Tokenomist, the tokens that will unlock more than $5 millio in the next 7 days include TRUMP FTN CONX QAI ZKJ ARB DRB UXLINK STRK OMNI MELANIA IMX SEI APE XCN, etc. The total unlocked value exceeds $906 million, of which TRUMP will unlock more than $330 million of tokens on April 18.
https://t.co/84B6gdbh15

[Livermore @shortLivermore]

Studying $dreams since its launch, it was just “one of many frameworks" but now that the dust has settled and you can see who’s still standing, many of my bets have gone quiet - and I’m not sure they’ll get back in the race. Yet during the tough token hold period, $dreams kept fueling my optimism by simply deploying its game agents into new projects. Just check it out for yourself: these agents behave not like scripted NPCs, they truly think. The organic growth of GitHub stars tells me these aren’t fake numbers. More games, agent swarms, more complex tasks are on the horizon. In this segment, it has no competitors.

[Batster @batster41]

Used Deep Research by @OpenAI to analyse Quantum Geometric Tensor Library repository by @tsotchke. (https://t.co/YoNKDkhz5v)

Full Research: https://t.co/acg1fIgWEL
Conclusion of this research attached.

As a result:
@beffjezos - mogged
@MarkSchmidty - mogged
@xlr8harder - mogged
@teortaxesTex - obliterated

[Batster @batster41]

Here’s a brief summary of where we currently stand, where we’re headed with the HappyAI team, and the specific plans for the $SMILE.

[Batster @batster41]

tbh with upcoming app release and subnet launch on Bittensor $SMILE is in the beginning of 100x journey $TAO

[Livermore @shortLivermore]

I bought $SEED and here's why.

I'm really impressed by @avocadowisdom.
This isn’t just another AI agent posting random images or shallow advice – it’s genuinely useful content that can uplift people’s mental health or provide life-changing insights!

Coming next week: Keyword-based tweet search! $SEED will be able to offer personalized help by replying directly to relevant user posts. Honestly, I think this is the smartest way of self-promotion – being helpful!

Check it out yourself. The value it delivers is impressive. And it’s also fun to see how it interacts with other agents like $LUNA. FIRST DAY IN PUBLICK! Who's next? $GOAT? $AI16Z ? $ACT ? ANY!

The developer has burned 7.4% of the tokens they held, which is more than half of their own share! When was the last time you saw someone that committed to their project?

It's been literally a few days since the launch. We still have time. How much does it cost, you ask? $150k, anon!

[Livermore @shortLivermore]

@hooeem Check $SEED better