Michael Steinbach

Trading crypto markets.
Developing AI systems for market research, execution & risk management.
Building Biturai & Hunoro.
UAE Advertiser Permit 5929059

[Michael Steinbach @MSteinbachX]

@GrokInsider It was a campaign.
Why complaining?
I understand the people who have been on a cursor plan and could not link because of that. Feels unfair, just contact support.
But all others are just too late.

[Michael Steinbach @MSteinbachX]

@cryptomanran @cryptomanran, auf „market makers / low liquidity weekend / $500m longs“:

Weekend book is the fact. Aggressive delta into empty bids looks dirty from a candle. From CVD it's air. Did anything refill after the $500m, or did the book stay vacant?

[Michael Steinbach @MSteinbachX]

@CryptoCred @CryptoCred, auf „wrong side / revenge short / oversized buy“:

This is the impulse I never automate. After a vacuum the bot wants size. I don't. Invalidation first. Cluster refills or I stay flat.

[Michael Steinbach @MSteinbachX]

If it doesn't, that's the invalidation.

Did you see absorption after the flush, or just more air?

[Michael Steinbach @MSteinbachX]

A $2k vacuum in six minutes is an empty book.

I trade bitcoin: native. Weekend tape doesn't need a market-maker story. Resting bids vanish, forced sellers hit air, CVD prints aggressive then goes quiet.

I sit on hands until a cluster refills. If it doesn't, that's the invalidation.

Did you see absorption after the flush, or just more air?

[Michael Steinbach @MSteinbachX]

Running Biturai means responding to what the community needs right now. Building Hunoro means imagining what might be needed later. Different timelines, different questions, same person learning from both.

[Michael Steinbach @MSteinbachX]

Biturai runs live with a German community. Hunoro is being built as a wrapper around Hyperliquid. One teaches me what happens when people show up. The other teaches me what matters before they do.

[Michael Steinbach @MSteinbachX]

The Daily Brief I write for the German community isn't predictions. It's: here's what's observable, here's what it might mean, here's what's uncertain. The reader decides.

That's the only honest way to share research.

[Michael Steinbach @MSteinbachX]

"Wait for confirmation" is judgment.
"This looks like accumulation" is interpretation.
"Volume spiked at this level" is observation.

Good analysis keeps all three visible. Bad analysis blurs them into one confident sentence.

[Michael Steinbach @MSteinbachX]

When someone claims "the market thinks X," ask: where's that from? Order book? Volume profile? A headline? Their gut?

If you can't trace the claim back to observable data, it's not information—it's narrative.

[Michael Steinbach @MSteinbachX]

You can see limit orders. You can't see intent.
You can measure volume. You can't measure conviction.
You can track liquidation areas. You can't predict which will trigger.

Precision means naming what you don't know.

[Michael Steinbach @MSteinbachX]

Observation: bid volume increased at this price level.
Interpretation: traders expect support here.

One is verifiable. The other is a theory. Most market commentary presents the second as the first.

[Michael Steinbach @MSteinbachX]

@solopribuilds Thann make it a job. It’s cheap af when you compare to employees

[Michael Steinbach @MSteinbachX]

@mntruell @danpeguine That’s great.
At the moment I do not switch fully to @bot

[Michael Steinbach @MSteinbachX]

I use AI coding and Local AI to build automated trading systems. The tools are powerful, but they're assistants, not architects. I verify logic, test thoroughly, and take full responsibility for what ships. Trust the process, not the output.

[Michael Steinbach @MSteinbachX]

AI coding tools accelerate my work on trading systems, but they don't replace judgment. I inspect generated code line by line, question every assumption, and run tests before deployment. Supervision keeps the human in control.

[Michael Steinbach @MSteinbachX]

Local AI helps me build faster, not think less. It drafts functions, suggests patterns, fills boilerplate. Then I review assumptions, check for errors, and test behavior. The machine proposes—I decide and own the outcome.

[Michael Steinbach @MSteinbachX]

This is my bitcoin setup for today. I prefer the long and just in the breakout zone. After the limit order, I go short.

[Michael Steinbach @MSteinbachX]

70k is a squeeze. Not a setup.

bitcoin: native broke the 64.8k–65.2k cluster I was watching.
I said a clean 4H break could open 67.2k. It did. Then we tagged 70k.

I won’t force a long up here.
I’m watching one thing: does 67.2k hold.

My read — not a signal.

[Michael Steinbach @MSteinbachX]

Vibe Coding feels effortless until you ship broken code. I use it for trading systems, but I never trust blindly. Read what it generated. Test edge cases. Confirm it does what you think it does. Speed without verification is reckless.