ChemistDeFi
20 Posts
ChemistDeFi
Square VerifiedSoftware Engineer & Onchain Degen
Writing about Crypto & DeFi | AI is my slave, I'm the master
8 Following
1.0K+ Followers
313 Liked
Posts
PINNED
Still believe in $BTC.
Maybe more importantly, I still want to believe.
Some trades are about price.
Some are about conviction.
This one is both.
NFA. Always DYOR.
Maybe more importantly, I still want to believe.
Some trades are about price.
Some are about conviction.
This one is both.
NFA. Always DYOR.
Bitcoin is trying to recover.
Short term ETF flows have turned supportive.
Stablecoin supply has not.
BTC is trading around $64,000 at the time of writing.
At the time of writing, Farside shows the following US spot Bitcoin ETF flows:
→ July 13: -$424.7M
→ July 14: +$181.1M
→ July 15: +$107.7M
→ July 16: +$79.1M
→ July 17: +$132.3M
After the sharp outflow on July 13, four consecutive inflow sessions totaled $500.2M.
Across the full five session window, however, net inflows amount to only $75.5M.
I’d label the ETF rebound as encouraging, but it still needs follow through.
DefiLlama shows a different side of the liquidity picture:
→ Total stablecoin supply: $309.99B
→ 7 day change: -$1.572B (-0.50%)
→ 30 day change: -1.40%
Aggregate stablecoin supply is therefore contracting across both the weekly and monthly windows.
That metric is not a perfect measure of liquidity.
Issuance and redemptions can also reflect treasury management, chain migrations and issuer operations.
Even so, stablecoin supply remains a useful proxy for deployable onchain capital.
That leaves us with divergent signals:
→ US spot ETF demand has improved
→ Aggregate stablecoin supply is still shrinking
These signals do not contradict each other.
They measure different channels of capital and different parts of the market.
My take:
Bitcoin’s recovery now has short term support from US spot ETF flows, but broader liquidity confirmation remains incomplete.
The next signal I’d watch is whether ETF inflows persist while stablecoin supply stabilizes or begins expanding again.
Short term ETF flows have turned supportive.
Stablecoin supply has not.
BTC is trading around $64,000 at the time of writing.
At the time of writing, Farside shows the following US spot Bitcoin ETF flows:
→ July 13: -$424.7M
→ July 14: +$181.1M
→ July 15: +$107.7M
→ July 16: +$79.1M
→ July 17: +$132.3M
After the sharp outflow on July 13, four consecutive inflow sessions totaled $500.2M.
Across the full five session window, however, net inflows amount to only $75.5M.
I’d label the ETF rebound as encouraging, but it still needs follow through.
DefiLlama shows a different side of the liquidity picture:
→ Total stablecoin supply: $309.99B
→ 7 day change: -$1.572B (-0.50%)
→ 30 day change: -1.40%
Aggregate stablecoin supply is therefore contracting across both the weekly and monthly windows.
That metric is not a perfect measure of liquidity.
Issuance and redemptions can also reflect treasury management, chain migrations and issuer operations.
Even so, stablecoin supply remains a useful proxy for deployable onchain capital.
That leaves us with divergent signals:
→ US spot ETF demand has improved
→ Aggregate stablecoin supply is still shrinking
These signals do not contradict each other.
They measure different channels of capital and different parts of the market.
My take:
Bitcoin’s recovery now has short term support from US spot ETF flows, but broader liquidity confirmation remains incomplete.
The next signal I’d watch is whether ETF inflows persist while stablecoin supply stabilizes or begins expanding again.
GameFi quietly ran one of the biggest UX experiments in DeFi history.
And I still think the results are underrated.
For years, people said DeFi stayed small because it was too complex.
Seed phrases.
Gas.
Bridges.
Slippage.
Approvals.
Then play to earn happened.
People who had never touched crypto learned all of it in weeks.
The interfaces were not suddenly better.
The reason to care was.
That is the part GameFi proved.
Complexity is real, but it behaves like a price.
People will pay it when the thing behind it feels worth the effort.
Of course, the first wave had problems.
Many economies leaned too hard on rewards, emissions, and new players entering.
When incentives slowed, activity slowed too.
But that does not make the category dead.
It makes the lesson clearer.
The next version of GameFi cannot just be DeFi with a game skin.
It has to be a real game where ownership, markets, identity, and progression make the experience better.
And this is where timing gets interesting.
AI has changed the cost of building.
A simple or mid level game does not take the same time to ship as it used to.
Art, prototyping, NPC logic, world building, testing, iteration.
A lot of the slow parts are getting faster.
That does not magically create good games.
But it does make the category harder to ignore.
Because if GameFi already proved people will learn complex systems when they care about the world behind them,
I am genuinely curious what happens when we start seeing GameFi products that feel closer to AAA-quality experiences.
Or even strategy games with the depth of something like Age of Empires.
Not just rewards with better graphics.
Actual games.
With economies that support the world instead of replacing it.
And I still think the results are underrated.
For years, people said DeFi stayed small because it was too complex.
Seed phrases.
Gas.
Bridges.
Slippage.
Approvals.
Then play to earn happened.
People who had never touched crypto learned all of it in weeks.
The interfaces were not suddenly better.
The reason to care was.
That is the part GameFi proved.
Complexity is real, but it behaves like a price.
People will pay it when the thing behind it feels worth the effort.
Of course, the first wave had problems.
Many economies leaned too hard on rewards, emissions, and new players entering.
When incentives slowed, activity slowed too.
But that does not make the category dead.
It makes the lesson clearer.
The next version of GameFi cannot just be DeFi with a game skin.
It has to be a real game where ownership, markets, identity, and progression make the experience better.
And this is where timing gets interesting.
AI has changed the cost of building.
A simple or mid level game does not take the same time to ship as it used to.
Art, prototyping, NPC logic, world building, testing, iteration.
A lot of the slow parts are getting faster.
That does not magically create good games.
But it does make the category harder to ignore.
Because if GameFi already proved people will learn complex systems when they care about the world behind them,
I am genuinely curious what happens when we start seeing GameFi products that feel closer to AAA-quality experiences.
Or even strategy games with the depth of something like Age of Empires.
Not just rewards with better graphics.
Actual games.
With economies that support the world instead of replacing it.
Strategy disclosed yesterday that it sold 3,588 $BTC for $216 million.
The largest sale in the company’s history.
But the reason matters more than the size.
The proceeds went to dividends on its preferred shares.
Those dividends are paid in cash.
They are contractual.
And the software business does not cover them.
Grayscale’s research desk estimates the annual bill at around $1.5 billion.
For years, the machine only worked in one direction:
Issue paper.
Buy bitcoin.
Repeat.
The promise was simple:
Never sell.
Now, one of the preferred shares trades below its $100 par value.
New paper is harder to place.
And the bill still arrives every quarter.
So the structure built to convert dollars into bitcoin has started converting bitcoin back into dollars.
On a schedule.
At whatever the price is that day.
Strategy still holds 843,775 BTC.
This is not a collapse story.
But an asset the market priced as permanently off the market has now become a recurring seller.
And what I keep thinking about is simple:
How many other treasury companies copied this machine?
And when do their bills arrive?
The largest sale in the company’s history.
But the reason matters more than the size.
The proceeds went to dividends on its preferred shares.
Those dividends are paid in cash.
They are contractual.
And the software business does not cover them.
Grayscale’s research desk estimates the annual bill at around $1.5 billion.
For years, the machine only worked in one direction:
Issue paper.
Buy bitcoin.
Repeat.
The promise was simple:
Never sell.
Now, one of the preferred shares trades below its $100 par value.
New paper is harder to place.
And the bill still arrives every quarter.
So the structure built to convert dollars into bitcoin has started converting bitcoin back into dollars.
On a schedule.
At whatever the price is that day.
Strategy still holds 843,775 BTC.
This is not a collapse story.
But an asset the market priced as permanently off the market has now become a recurring seller.
And what I keep thinking about is simple:
How many other treasury companies copied this machine?
And when do their bills arrive?
Crypto is the only major market that does not close for the weekend.
Every market before it had rest built into the structure.
The bell rang.
The week ended.
Everyone recovered, whether they planned to or not.
Crypto removed the bell and quietly made recovery a personal responsibility.
Most people never noticed the transfer.
Here is the chemistry part.
Your stress system was built for short episodes.
React hard.
Return to baseline.
Cortisol is useful in bursts and corrosive as a background level.
An always-on market with an open position keeps that system mildly activated around the clock.
Nothing dramatic happens on any single chart check.
That is exactly why the cost stays invisible.
And weekends make the trade worse.
Order books thin out.
Moves get exaggerated, then often unwind by Monday.
Weekend price action is the least reliable information of the week, offered at full emotional price.
Peak attention spent on the market’s lowest quality signal.
What changed my behavior was a simple test:
If no decision can come out of this chart check, it is not risk management.
It is just stress wearing a productivity costume.
On most weekends, with a plan already written, that test fails every single time.
The market does not rest.
That was never in question.
The question is whether you still remember that you can.
Every market before it had rest built into the structure.
The bell rang.
The week ended.
Everyone recovered, whether they planned to or not.
Crypto removed the bell and quietly made recovery a personal responsibility.
Most people never noticed the transfer.
Here is the chemistry part.
Your stress system was built for short episodes.
React hard.
Return to baseline.
Cortisol is useful in bursts and corrosive as a background level.
An always-on market with an open position keeps that system mildly activated around the clock.
Nothing dramatic happens on any single chart check.
That is exactly why the cost stays invisible.
And weekends make the trade worse.
Order books thin out.
Moves get exaggerated, then often unwind by Monday.
Weekend price action is the least reliable information of the week, offered at full emotional price.
Peak attention spent on the market’s lowest quality signal.
What changed my behavior was a simple test:
If no decision can come out of this chart check, it is not risk management.
It is just stress wearing a productivity costume.
On most weekends, with a plan already written, that test fails every single time.
The market does not rest.
That was never in question.
The question is whether you still remember that you can.
In its July 1 weekly report, Glassnode flagged that more bitcoin was held at a loss than at a profit.
Roughly 10.8M BTC underwater against 9.2M in profit.
The latest crossover first hit in early June. Since then, price has been chopping around the flip level, moving in and out of the zone as the market tests that line.
Most feeds posted the June flip as a bottom signal the day it happened.
The history is more honest than that.
Yes, this crossover has appeared near every major low. But the time spent inside it varied wildly.
Around six months in 2019.
One month in the covid crash.
Six months again in 2022.
Close to a year in 2015.
So the metric answers one question and gets quoted as if it answers another.
It tells you which zone you are in.
It says nothing about how long you stay there.
This cycle is already proving the point. We have spent a month oscillating around the line itself.
The part I keep coming back to is that this same data made people early in every one of those windows.
Being right about the zone and wrong about the clock feels exactly like being wrong.
Maybe patience is the actual signal here.
Roughly 10.8M BTC underwater against 9.2M in profit.
The latest crossover first hit in early June. Since then, price has been chopping around the flip level, moving in and out of the zone as the market tests that line.
Most feeds posted the June flip as a bottom signal the day it happened.
The history is more honest than that.
Yes, this crossover has appeared near every major low. But the time spent inside it varied wildly.
Around six months in 2019.
One month in the covid crash.
Six months again in 2022.
Close to a year in 2015.
So the metric answers one question and gets quoted as if it answers another.
It tells you which zone you are in.
It says nothing about how long you stay there.
This cycle is already proving the point. We have spent a month oscillating around the line itself.
The part I keep coming back to is that this same data made people early in every one of those windows.
Being right about the zone and wrong about the clock feels exactly like being wrong.
Maybe patience is the actual signal here.
DYOR is the most repeated and least explained advice in crypto.
Here is the strange part. The checklist itself is not a secret.
→ who unlocks tokens, and when
→ would anyone use this without the rewards
→ how many months of runway before the team has to raise
→ who backed it, and who stayed involved after the check cleared
Anyone can write that list.
Most people reading this could.
The question that actually matters is why almost nobody runs it.
My honest answer:
the list feels like homework, and buying feels like action, hello dopamine.
A few hours of reading has a visible cost.
Skipping it has an invisible one, and invisible costs never win the argument in the moment.
The item I would keep if I could only keep one is not even on most lists.
The competition.
A good project in a crowded sector is often worth less than an average one in an empty sector.
Quality does not price in a vacuum.
Every expensive lesson I have taken from this market traces back to a question I already knew to ask and decided to skip.
The research was never the hard part.
Sitting down to do it before the position, instead of after, is.
Here is the strange part. The checklist itself is not a secret.
→ who unlocks tokens, and when
→ would anyone use this without the rewards
→ how many months of runway before the team has to raise
→ who backed it, and who stayed involved after the check cleared
Anyone can write that list.
Most people reading this could.
The question that actually matters is why almost nobody runs it.
My honest answer:
the list feels like homework, and buying feels like action, hello dopamine.
A few hours of reading has a visible cost.
Skipping it has an invisible one, and invisible costs never win the argument in the moment.
The item I would keep if I could only keep one is not even on most lists.
The competition.
A good project in a crowded sector is often worth less than an average one in an empty sector.
Quality does not price in a vacuum.
Every expensive lesson I have taken from this market traces back to a question I already knew to ask and decided to skip.
The research was never the hard part.
Sitting down to do it before the position, instead of after, is.
DefiLlama's front page is telling two stories at once, and most people only read one of them.
Total DeFi TVL sits around 70 billion, down from roughly 114 at the start of the year.
That is the story everyone reads. DeFi is shrinking.
But TVL is a strange metric. It mixes two different units inside one number.
Most of it is crypto collateral. ETH in Lido, ETH in restaking, coins sitting in lending markets. When prices fall, that TVL falls with them. Nobody has to withdraw anything. The same coins just get repriced.
A smaller slice is denominated in dollars. Tokenized treasuries and private credit. That part does not reprice when ETH falls. It only moves when someone actually deposits or actually leaves.
Open the one month column and the gradient is hard to miss.
→ Staking and restaking categories down 15 to 25 percent, tracking ETH
→ Lending down high single digits, tracking its collateral
→ RWA down low single digits, the smallest drop on the board
So the honest read is that most of this year's decline is a price echo, not an exodus. The further a category sits from crypto price exposure, the less it fell.
And inside that quietest category, the dollar credit names are still growing. Centrifuge is up double digits in a month where almost everything else bled.
To be fair, holding steady is easy when your unit does not reprice. Stability proves nothing here. Growth does.
New money choosing tokenized credit and treasuries in the middle of a drawdown is a decision someone made on purpose.
The sector's onchain value went from about 6 billion to over 25 billion in eighteen months and did not pause for any of this.
What I keep asking myself is which number I will actually care about in two years.
The TVL that breathes with the price of ETH, or the TVL that only moves when someone decides.
Total DeFi TVL sits around 70 billion, down from roughly 114 at the start of the year.
That is the story everyone reads. DeFi is shrinking.
But TVL is a strange metric. It mixes two different units inside one number.
Most of it is crypto collateral. ETH in Lido, ETH in restaking, coins sitting in lending markets. When prices fall, that TVL falls with them. Nobody has to withdraw anything. The same coins just get repriced.
A smaller slice is denominated in dollars. Tokenized treasuries and private credit. That part does not reprice when ETH falls. It only moves when someone actually deposits or actually leaves.
Open the one month column and the gradient is hard to miss.
→ Staking and restaking categories down 15 to 25 percent, tracking ETH
→ Lending down high single digits, tracking its collateral
→ RWA down low single digits, the smallest drop on the board
So the honest read is that most of this year's decline is a price echo, not an exodus. The further a category sits from crypto price exposure, the less it fell.
And inside that quietest category, the dollar credit names are still growing. Centrifuge is up double digits in a month where almost everything else bled.
To be fair, holding steady is easy when your unit does not reprice. Stability proves nothing here. Growth does.
New money choosing tokenized credit and treasuries in the middle of a drawdown is a decision someone made on purpose.
The sector's onchain value went from about 6 billion to over 25 billion in eighteen months and did not pause for any of this.
What I keep asking myself is which number I will actually care about in two years.
The TVL that breathes with the price of ETH, or the TVL that only moves when someone decides.
We are in a new month, and the July catalyst threads are already going up.
Every start of a month works this way. Someone lists the dates, the market nods, and the calendar gets treated like a forecast.
The dates are real this time too.
→ a jobs report this week
→ the Fed meets at the end of the month
→ the usual run of token unlocks, lighter than last month
None of that is what actually decides July.
BTC walks into this month roughly halved from its high, and the market around it is sitting in extreme fear.
That is the input almost nobody puts on the list.
In this state, events stop setting direction. They hand a frightened crowd permission to do what it already wanted to do.
The same jobs number lands one way in a confident market and the opposite way in a scared one.
The event does not change.
The crowd receiving it does.
And the reflex that carried two years of dip buying is quietly gone.
For most of this cycle the bull case rested on one silent assumption. Money would get cheaper. Cuts were coming, liquidity was on the way, and every dip had that floor underneath it.
The Fed spent June removing that floor. Rates held, the tone turned hawkish, and the market stopped pricing cuts for this year.
So a scared crowd is being asked to price this market without the one story it was built on.
What I keep coming back to is that none of this shows up on a catalyst list.
The dates will move price for an hour.
The fear is what gets to decide the month.
Every start of a month works this way. Someone lists the dates, the market nods, and the calendar gets treated like a forecast.
The dates are real this time too.
→ a jobs report this week
→ the Fed meets at the end of the month
→ the usual run of token unlocks, lighter than last month
None of that is what actually decides July.
BTC walks into this month roughly halved from its high, and the market around it is sitting in extreme fear.
That is the input almost nobody puts on the list.
In this state, events stop setting direction. They hand a frightened crowd permission to do what it already wanted to do.
The same jobs number lands one way in a confident market and the opposite way in a scared one.
The event does not change.
The crowd receiving it does.
And the reflex that carried two years of dip buying is quietly gone.
For most of this cycle the bull case rested on one silent assumption. Money would get cheaper. Cuts were coming, liquidity was on the way, and every dip had that floor underneath it.
The Fed spent June removing that floor. Rates held, the tone turned hawkish, and the market stopped pricing cuts for this year.
So a scared crowd is being asked to price this market without the one story it was built on.
What I keep coming back to is that none of this shows up on a catalyst list.
The dates will move price for an hour.
The fear is what gets to decide the month.
GM, new month, new platform.
First day on Binance Square.
Same chaos, better focus.
I’ll be sharing how I look at crypto, DeFi, liquidity and market narratives from here too.
Let’s cook 🧪
First day on Binance Square.
Same chaos, better focus.
I’ll be sharing how I look at crypto, DeFi, liquidity and market narratives from here too.
Let’s cook 🧪
Log in to explore more content