MrRichMill
45 Posts
MrRichMill
Report
Block User
I build autonomous trading systems and publish what breaks: backtest vs production, real costs, post-mortems. No signals. Midas Protocol @ NadirLabs.AI
Open Trade
High-Frequency Trader
5.1 Years
13 Following
5 Followers
9 Liked
Posts
Portfolio
¡
--
Your stop loss is not a risk setting. It is a noise threshold.
If your stop sits closer to entry than the asset's average daily range, you are not managing risk. You are paying to be removed from trades you were right about.
The correct order is: place the stop where the idea is invalid, then size the position so that loss is acceptable. Most people do the reverse. They pick the size they want first and put the stop wherever it fits.
Where do you actually put your stops?
Your stop loss is not a risk setting. It is a noise threshold.
If your stop sits closer to entry than the asset's average daily range, you are not managing risk. You are paying to be removed from trades you were right about.
The correct order is: place the stop where the idea is invalid, then size the position so that loss is acceptable. Most people do the reverse. They pick the size they want first and put the stop wherever it fits.
Where do you actually put your stops?
If your stop sits closer to entry than the asset's average daily range, you are not managing risk. You are paying to be removed from trades you were right about.
The correct order is: place the stop where the idea is invalid, then size the position so that loss is acceptable. Most people do the reverse. They pick the size they want first and put the stop wherever it fits.
Where do you actually put your stops?
A fixed % from entry
0%
Structure, then size to fit
0%
An ATR multiple
0%
The liquidation price
0%
0 votes ⢠Voting closed
178
¡
--
Uncomfortable truth about trading bots: the code is the easy part.
What actually kills automated accounts:
The API drops mid-order and you now hold a position you cannot see
The exchange returns a partial fill your logic never handled
Funding quietly accrues on a position you forgot was open
Your VPS reboots and the bot restarts flat while the position is live
None of that appears in a backtest. All of it appears in production. Strategy research is maybe 30% of the work. The other 70% is engineering nobody posts about.
If you run a bot: what broke first for you?
$BTC $ETH
Uncomfortable truth about trading bots: the code is the easy part.
What actually kills automated accounts:
The API drops mid-order and you now hold a position you cannot see
The exchange returns a partial fill your logic never handled
Funding quietly accrues on a position you forgot was open
Your VPS reboots and the bot restarts flat while the position is live
None of that appears in a backtest. All of it appears in production. Strategy research is maybe 30% of the work. The other 70% is engineering nobody posts about.
If you run a bot: what broke first for you?
$BTC $ETH
What actually kills automated accounts:
The API drops mid-order and you now hold a position you cannot see
The exchange returns a partial fill your logic never handled
Funding quietly accrues on a position you forgot was open
Your VPS reboots and the bot restarts flat while the position is live
None of that appears in a backtest. All of it appears in production. Strategy research is maybe 30% of the work. The other 70% is engineering nobody posts about.
If you run a bot: what broke first for you?
$BTC $ETH
106
¡
--
A 90% win rate tells you nothing on its own.
Expectancy = (win rate x average win) minus (loss rate x average loss).
At 90% wins you still lose money the moment your average loss is more than 9 times your average win. That is exactly what happens when you cut winners early and let losers run.
Most traders can quote their win rate instantly and have no idea what their payoff ratio is. One of those numbers decides whether you survive. It is not the first one.
What is yours?
$BTC
{spot}(BTCUSDT)
A 90% win rate tells you nothing on its own.
Expectancy = (win rate x average win) minus (loss rate x average loss).
At 90% wins you still lose money the moment your average loss is more than 9 times your average win. That is exactly what happens when you cut winners early and let losers run.
Most traders can quote their win rate instantly and have no idea what their payoff ratio is. One of those numbers decides whether you survive. It is not the first one.
What is yours?
$BTC
Expectancy = (win rate x average win) minus (loss rate x average loss).
At 90% wins you still lose money the moment your average loss is more than 9 times your average win. That is exactly what happens when you cut winners early and let losers run.
Most traders can quote their win rate instantly and have no idea what their payoff ratio is. One of those numbers decides whether you survive. It is not the first one.
What is yours?
$BTC
138
¡
--
Your backtest is probably lying to you. Here is the cheapest way to find out.
Run it again with 0.05% slippage per side and funding costs included.
Most strategies that look profitable are running on a 0.1 to 0.3% edge per trade. Costs that size eat the edge completely.
If your equity curve survives that, you have a strategy. If it does not, you had a spreadsheet.
What is the first thing you add to a backtest before you trust it?
$BTC $ETH
{spot}(ETHUSDT)
Your backtest is probably lying to you. Here is the cheapest way to find out.
Run it again with 0.05% slippage per side and funding costs included.
Most strategies that look profitable are running on a 0.1 to 0.3% edge per trade. Costs that size eat the edge completely.
If your equity curve survives that, you have a strategy. If it does not, you had a spreadsheet.
What is the first thing you add to a backtest before you trust it?
$BTC $ETH
Run it again with 0.05% slippage per side and funding costs included.
Most strategies that look profitable are running on a 0.1 to 0.3% edge per trade. Costs that size eat the edge completely.
If your equity curve survives that, you have a strategy. If it does not, you had a spreadsheet.
What is the first thing you add to a backtest before you trust it?
$BTC $ETH
120
¡
--
Apple just priced a phone at 1,999 dollars into a 4.8% ten year.
That is the most interesting thing about yesterday, and almost nobody in this hashtag is talking about it.
Most iPhones in the US are not bought outright. They are financed over 24 to 36 months. So a premium launch at the highest rates in nearly three years is not really a product question. It is a test of whether the high income consumer is still intact when credit is expensive.
If it sells anyway, that tells you more about risk appetite than any sentiment index will.
And it reaches $BTC and $ETH through the same channel. Higher real yields raise the hurdle rate on assets with no cash flow. Discretionary demand for a 2,000 dollar gadget and discretionary demand for a volatile asset come out of the same wallet.
Two scheduled catalysts, one question.
#AppleDebutsFoldablePhone #US10YTreasuryYieldHitsHighestSinceNov2023
Apple just priced a phone at 1,999 dollars into a 4.8% ten year.
That is the most interesting thing about yesterday, and almost nobody in this hashtag is talking about it.
Most iPhones in the US are not bought outright. They are financed over 24 to 36 months. So a premium launch at the highest rates in nearly three years is not really a product question. It is a test of whether the high income consumer is still intact when credit is expensive.
If it sells anyway, that tells you more about risk appetite than any sentiment index will.
And it reaches $BTC and $ETH through the same channel. Higher real yields raise the hurdle rate on assets with no cash flow. Discretionary demand for a 2,000 dollar gadget and discretionary demand for a volatile asset come out of the same wallet.
Two scheduled catalysts, one question.
#AppleDebutsFoldablePhone #US10YTreasuryYieldHitsHighestSinceNov2023
That is the most interesting thing about yesterday, and almost nobody in this hashtag is talking about it.
Most iPhones in the US are not bought outright. They are financed over 24 to 36 months. So a premium launch at the highest rates in nearly three years is not really a product question. It is a test of whether the high income consumer is still intact when credit is expensive.
If it sells anyway, that tells you more about risk appetite than any sentiment index will.
And it reaches $BTC and $ETH through the same channel. Higher real yields raise the hurdle rate on assets with no cash flow. Discretionary demand for a 2,000 dollar gadget and discretionary demand for a volatile asset come out of the same wallet.
Two scheduled catalysts, one question.
#AppleDebutsFoldablePhone #US10YTreasuryYieldHitsHighestSinceNov2023
127
¡
--
Most accounts do not blow up on a bad call. They blow up on position size.
A 4% adverse move is an ordinary day for $BTC . At 20x that is an 80% loss on margin. At 3x it is 12%.
Same call. Same chart. Completely different outcome. The edge did not change, only the sizing did.
Most people pick leverage by how confident the trade feels. That is the opposite of risk management.
How do you actually pick your leverage?
Most accounts do not blow up on a bad call. They blow up on position size.
A 4% adverse move is an ordinary day for $BTC . At 20x that is an 80% loss on margin. At 3x it is 12%.
Same call. Same chart. Completely different outcome. The edge did not change, only the sizing did.
Most people pick leverage by how confident the trade feels. That is the opposite of risk management.
How do you actually pick your leverage?
A 4% adverse move is an ordinary day for $BTC . At 20x that is an 80% loss on margin. At 3x it is 12%.
Same call. Same chart. Completely different outcome. The edge did not change, only the sizing did.
Most people pick leverage by how confident the trade feels. That is the opposite of risk management.
How do you actually pick your leverage?
Gut feel and conviction
100%
Same fixed number every trade
0%
From ATR or daily range
0%
I do not use leverage
0%
1 votes ⢠Voting closed
214
¡
--
The top three trending topics right now are one story, and most people are reading the causality backwards.
#BrentCrudeTops$100
#USTreasuryToBuyBackUpTo$6BLongDatedDebt
#BitcoinSurpasses$79K
Oil at 100 is the input. It feeds inflation expectations, and inflation expectations are what the long end of the curve prices. That is why the 10 year sits near 4.81% on the same day the Treasury runs its largest long-dated buyback yet.
That buyback is 6 billion against a market that turns over hundreds of billions a day, aimed at old and thinly traded issues. Whatever it does for market functioning, it is not a lever big enough to set the level of yields. Tagging it bullish or bearish for crypto is a category error.
And $BTC did not surpass 79K in any meaningful sense. It tagged 79,760 and is back at 78,313 as I write this. A level you print and immediately hand back is not a breakout. It is a liquidity sweep.
Since I keep asking other people to make their calls scoreable, here is mine, with an invalidation and a deadline.
If the buyback is what matters, the 10 year should be meaningfully below 4.75% by Friday's close. If it is still at or above 4.75%, then the buyback was plumbing and oil was the only variable that moved anything this week.
I will post the result either way, including if I am wrong.
Which of the three do you think is actually driving the other two?
$ETH
The top three trending topics right now are one story, and most people are reading the causality backwards.
#BrentCrudeTops$100
#USTreasuryToBuyBackUpTo$6BLongDatedDebt
#BitcoinSurpasses$79K
Oil at 100 is the input. It feeds inflation expectations, and inflation expectations are what the long end of the curve prices. That is why the 10 year sits near 4.81% on the same day the Treasury runs its largest long-dated buyback yet.
That buyback is 6 billion against a market that turns over hundreds of billions a day, aimed at old and thinly traded issues. Whatever it does for market functioning, it is not a lever big enough to set the level of yields. Tagging it bullish or bearish for crypto is a category error.
And $BTC did not surpass 79K in any meaningful sense. It tagged 79,760 and is back at 78,313 as I write this. A level you print and immediately hand back is not a breakout. It is a liquidity sweep.
Since I keep asking other people to make their calls scoreable, here is mine, with an invalidation and a deadline.
If the buyback is what matters, the 10 year should be meaningfully below 4.75% by Friday's close. If it is still at or above 4.75%, then the buyback was plumbing and oil was the only variable that moved anything this week.
I will post the result either way, including if I am wrong.
Which of the three do you think is actually driving the other two?
$ETH
#BrentCrudeTops$100
#USTreasuryToBuyBackUpTo$6BLongDatedDebt
#BitcoinSurpasses$79K
Oil at 100 is the input. It feeds inflation expectations, and inflation expectations are what the long end of the curve prices. That is why the 10 year sits near 4.81% on the same day the Treasury runs its largest long-dated buyback yet.
That buyback is 6 billion against a market that turns over hundreds of billions a day, aimed at old and thinly traded issues. Whatever it does for market functioning, it is not a lever big enough to set the level of yields. Tagging it bullish or bearish for crypto is a category error.
And $BTC did not surpass 79K in any meaningful sense. It tagged 79,760 and is back at 78,313 as I write this. A level you print and immediately hand back is not a breakout. It is a liquidity sweep.
Since I keep asking other people to make their calls scoreable, here is mine, with an invalidation and a deadline.
If the buyback is what matters, the 10 year should be meaningfully below 4.75% by Friday's close. If it is still at or above 4.75%, then the buyback was plumbing and oil was the only variable that moved anything this week.
I will post the result either way, including if I am wrong.
Which of the three do you think is actually driving the other two?
$ETH
Oil above 100
60%
The Treasury buyback
0%
Neither, it is Fed policy
20%
They are unrelated
20%
5 votes ⢠Voting closed
224
¡
--
$UNI short, closed at 17:55. Entry 6.407, exit 6.15. Same shape as yesterday: a 4% move, 5x, +19.57% on margin.
Two shorts, two days, same structure. That is not a good read. That is a regime.
Here is what I think actually changes in 2026, and it is not that technical analysis stops working. Regime detection, trend state, volatility state, all of that is technical analysis and it works fine.
What stops working is the human sitting between the signal and the order.
The chart said downtrend. A person still has to see it, believe it, size it, place the stop, leave the stop alone, and close at target instead of at the first flush of fear. That chain is where the edge leaks out. Not in the analysis. In the execution of it.
A system that classifies the regime, selects the logic that fits it and manages the position without asking me anything removes that chain entirely. It is not smarter than a good discretionary trader. It is just never tired, never early and never hopeful.
That is the part of 2026 worth paying attention to. This kind of autonomy used to need a desk, a team and a budget. Now one person can build it.
If your rules were executed exactly as written, with zero discretion, would your last month look better or worse?
$UNI short, closed at 17:55. Entry 6.407, exit 6.15. Same shape as yesterday: a 4% move, 5x, +19.57% on margin.
Two shorts, two days, same structure. That is not a good read. That is a regime.
Here is what I think actually changes in 2026, and it is not that technical analysis stops working. Regime detection, trend state, volatility state, all of that is technical analysis and it works fine.
What stops working is the human sitting between the signal and the order.
The chart said downtrend. A person still has to see it, believe it, size it, place the stop, leave the stop alone, and close at target instead of at the first flush of fear. That chain is where the edge leaks out. Not in the analysis. In the execution of it.
A system that classifies the regime, selects the logic that fits it and manages the position without asking me anything removes that chain entirely. It is not smarter than a good discretionary trader. It is just never tired, never early and never hopeful.
That is the part of 2026 worth paying attention to. This kind of autonomy used to need a desk, a team and a budget. Now one person can build it.
If your rules were executed exactly as written, with zero discretion, would your last month look better or worse?
Two shorts, two days, same structure. That is not a good read. That is a regime.
Here is what I think actually changes in 2026, and it is not that technical analysis stops working. Regime detection, trend state, volatility state, all of that is technical analysis and it works fine.
What stops working is the human sitting between the signal and the order.
The chart said downtrend. A person still has to see it, believe it, size it, place the stop, leave the stop alone, and close at target instead of at the first flush of fear. That chain is where the edge leaks out. Not in the analysis. In the execution of it.
A system that classifies the regime, selects the logic that fits it and manages the position without asking me anything removes that chain entirely. It is not smarter than a good discretionary trader. It is just never tired, never early and never hopeful.
That is the part of 2026 worth paying attention to. This kind of autonomy used to need a desk, a team and a budget. Now one person can build it.
If your rules were executed exactly as written, with zero discretion, would your last month look better or worse?
103
¡
--
$PENDLE short, closed by the system at 16:46. Entry 2.0981, exit 2.0137. A 4% move, 5x, +19.71% on margin.
The number is not the interesting part. One trade proves nothing, and anyone showing you a single green card is selling something.
The interesting part is that I did not take this trade. Midas Protocol marked the level, sized the position and closed at target while I was doing something else. My only job was deciding, weeks earlier, what conditions justify a short and what invalidates it.
That is the real trade-off with automation. You stop making decisions in the moment and start being judged on decisions you made in advance, when you were calm and had nothing open.
Most people want a bot so they can stop thinking. It works the other way around. You do all the thinking first, and then you have to live with it.
What is the last discretionary trade you took that your own rules would have rejected?
$PENDLE short, closed by the system at 16:46. Entry 2.0981, exit 2.0137. A 4% move, 5x, +19.71% on margin.
The number is not the interesting part. One trade proves nothing, and anyone showing you a single green card is selling something.
The interesting part is that I did not take this trade. Midas Protocol marked the level, sized the position and closed at target while I was doing something else. My only job was deciding, weeks earlier, what conditions justify a short and what invalidates it.
That is the real trade-off with automation. You stop making decisions in the moment and start being judged on decisions you made in advance, when you were calm and had nothing open.
Most people want a bot so they can stop thinking. It works the other way around. You do all the thinking first, and then you have to live with it.
What is the last discretionary trade you took that your own rules would have rejected?
The number is not the interesting part. One trade proves nothing, and anyone showing you a single green card is selling something.
The interesting part is that I did not take this trade. Midas Protocol marked the level, sized the position and closed at target while I was doing something else. My only job was deciding, weeks earlier, what conditions justify a short and what invalidates it.
That is the real trade-off with automation. You stop making decisions in the moment and start being judged on decisions you made in advance, when you were calm and had nothing open.
Most people want a bot so they can stop thinking. It works the other way around. You do all the thinking first, and then you have to live with it.
What is the last discretionary trade you took that your own rules would have rejected?
103
¡
--
$ZEC is up 154% in 30 days. Almost every retail bot on the market has been trying to short it.
Mean reversion has no natural stop. The further price runs against it, the stronger its own entry signal gets. It doubles down by design.
On September 4, when ZEC broke 1,000 USDT, CoinDesk reported 34 million dollars in short liquidations on that single move.
The fix is one line of logic: if the regime is trending, the mean reversion book goes flat. Not smaller. Zero.
Most bots do not have that line.
Do you have a regime filter, or does your bot trade the same way in a range and in a trend?
$BTC $ETH #ZECHitsANewAllTimeHigh
$ZEC is up 154% in 30 days. Almost every retail bot on the market has been trying to short it.
Mean reversion has no natural stop. The further price runs against it, the stronger its own entry signal gets. It doubles down by design.
On September 4, when ZEC broke 1,000 USDT, CoinDesk reported 34 million dollars in short liquidations on that single move.
The fix is one line of logic: if the regime is trending, the mean reversion book goes flat. Not smaller. Zero.
Most bots do not have that line.
Do you have a regime filter, or does your bot trade the same way in a range and in a trend?
$BTC $ETH #ZECHitsANewAllTimeHigh
Mean reversion has no natural stop. The further price runs against it, the stronger its own entry signal gets. It doubles down by design.
On September 4, when ZEC broke 1,000 USDT, CoinDesk reported 34 million dollars in short liquidations on that single move.
The fix is one line of logic: if the regime is trending, the mean reversion book goes flat. Not smaller. Zero.
Most bots do not have that line.
Do you have a regime filter, or does your bot trade the same way in a range and in a trend?
$BTC $ETH #ZECHitsANewAllTimeHigh
168
¡
--
Bullish
Let's GO $TAO !! Algos wins!! Midas Protocol by Nadirlabs.AI
Let's GO $TAO !! Algos wins!! Midas Protocol by Nadirlabs.AI
104
¡
--
Bullish
This is Midas Protocol, Autonomous Trading Desk for Crypto Futures Binance.
This is Midas Protocol, Autonomous Trading Desk for Crypto Futures Binance.
1
144
¡
--
Algorithmic trading by our autonomous quant desk Midas Protocol
Algorithmic trading by our autonomous quant desk Midas Protocol
155
¡
--
Another one by our autonomous trading desk Midas Protocol
Another one by our autonomous trading desk Midas Protocol
153
¡
--
Autonomous execution of our Midas Protocol algorithm
Autonomous execution of our Midas Protocol algorithm
105
¡
--
Heading toward the third TP of the day with total AUTONOMY.
Heading toward the third TP of the day with total AUTONOMY.
909
¡
--
How many of you are actually automating your trading?
Building something great here for Futures.
How many of you are actually automating your trading?
Building something great here for Futures.
Building something great here for Futures.
216
Log in to explore more content
Join global crypto users on Binance Square
âĄď¸ Get latest and useful information about crypto.
đŹ Trusted by the worldâs largest crypto exchange.
đ Discover real insights from verified creators.
Email / Phone number
Trending Topics
CPIWatch
88.5M views
507,279 Discussing
Share & Win Traffic Reward in our Trending Hashtag Campaign
â¨Topic: Will Sept CPI Shake Oil & Treasuries?
đHow to Join:
- Publish a short post or article with hashtag #CPIWatch Â
- Related tokens:Â BZUSDT, TMFUSDT, TBT, RWUSD
- How will you adjust your trades/positions with September CPI coming? Use trade sharing cards to share your Oil or Treasuries related trades or positions for a higher chance to earn traffic boost.Â
- Back up your view with data or charts, and avoid AI-generated images/views.
âď¸Create content based on the below angles:
- Middle East tensions are escalating, oil prices are surging, and US energy costs are climbing with them. Markets now expect headline CPI to rise in September. Do you think CPI will come in hotter than expected? Will oil keep climbing?
- A hot CPI could revive bets on another Fed hike before year-end. Rising rate-hike expectations put upward pressure on Treasury yields. Where do you see Treasuries heading next?
â°Campaign Period:
- 2026-10-13 12:00 â 2026-10-15 03:00 UTC
đReward:
Qualified posts that comply with the above guidelines will be reviewed and may receive a random traffic boost in views. Use trade-sharing cards to share your Oil or Treasuries related trades/positions for a higher chance to earn a traffic boost. You will receive a notification from your feed secretary if your post is boosted.
Need ideas for your post? Visit the topic page #CPIWatch or the Square Guide on How to Post for Better Reach.
Binance Square Official
¡161 Likes
¡43k views
OptimismOPRises19.98%
3,256 views
154 Discussing
OpenAIEngineerArrestedOverAllegedOfficeRecording
938 views
49 Discussing