Aspirin.
Aspirin.
我拥有工程学博士学位，不仅是一个交易员，更是一个数据科学家。在这个充满噪音和情绪的市场里，我只相信数学、对数回归带和历史周期数据。 我的交易哲学很简单：在长周期的博弈中活下来。 Engineering PhD | Data Scientist Trading signals, not emotions. Guided by mathematical modeling, Logarithmic Regression Bands, and historical cycles. Execution: Objective. Rational. Long-cycle focused.
8Following
2.6Kfollowers
Feed
Livestream
Feed
Pinned
Instead of guessing the bottom, it's better to make a checklist and verify each item one by one:
1. A large-volume bearish candle appears on the weekly chart, then the price recovers above a key level.
2. The MVRV Z-score drops to around 0 or below.
3. The price approaches or falls below the realized price of 54,000, or the on-chain risk value drops below 0.1.
4. The funding rate turns negative, while the Coinbase premium shifts from negative to positive, indicating that spot buying demand in the US has returned.
5. The total market cap of stablecoins stops declining, and on-exchange buying power begins to replenish.
6. After December 10, the market starts pricing out interest rate hike expectations for 2027.
7. Gold holds above its October low. Historically, gold tends to bottom several months before risk assets.
If three conditions are met, you can start buying in batches; if five conditions are met, you can accelerate the pace. Currently, I can confirm fewer than two.
Aspirin.
"The Last Drop: The Time Is Very Near!"
November 14, 2026, Issue 115 Aspirin Cyclical Analysis from a Data Scientist's Perspective BTC fell below 70,000, currently testing the 200-week moving average, now between 66,000 and 68,000. A particularly obvious phenomenon in the past two weeks: every time it falls to a support level, someone stands up to call the bottom. 80,000 was called, 75,000 was called, 72,000, 69,000, and 67,000 were also called. Yesterday I listened in on a live broadcast; at first, everyone said the bottom was at 71,000, but before the broadcast ended, the bottom call price had been revised down several times. This issue does not call the bottom but answers four questions: Where are we in time? Why might the price not be here? What needs to be seen to confirm the bottom? And what do I plan to do myself? 01 Time: Already entered the window First, look at when the last drop of the past three cycles started. In 2014, the last drop started in November and ended in mid-January the following year. In 2018, the last drop started in early November and ended in mid-December, just around the interest rate meeting in December that year. In 2022, the last drop also started in November; FTX collapsed, and the bottom was seen that month. All three started in November after the midterm elections. This year's midterm election was on November 3; BTC surged to 81,000 after the election, then fell all the way down to now. The timing rhythm is almost the same as the past three times. Count how many days the bear market has lasted. The high point of this round was October 6, 2025; today is day 404. Past three
Pinned
After the 2018 midterm elections, BTC dropped from $6,400 to $3,200 in less than 40 days, exactly halving.
This week's post-election movement in the US stock market is almost identical to back then: a big surge the day after the election, followed by a 5.5% drop in the Nasdaq in one day, the largest single-day decline in 18 months. BTC reversed from 81,000 back to 76,000.
In my last analysis, the scenario was: after the election, a surge to around 80,000, then a pullback after the positive news was fully priced in. That step is now complete.
But a few things this week have made me downgrade my outlook:
The rebound was pushed back before reaching the 50-day moving average;
ETH has already broken below its previous low first;
The Federal Reserve's probability of a rate hike in December remains around 75% after the US stock market plunge;
And a historical pattern after rate hikes: in two samples, BTC dropped to a price 20% lower than the rate hike day price.
So in the detailed article, I did several things:
Mapped the 2018 timeline onto this year;
Recalculated the probabilities of four possible scenarios;
Clarified why $40,000 is worth using to test position sizing but not worth waiting for as a buy point.
The conclusion is here first: the timing window for the final drop remains unchanged, from mid-November to mid-December, but the price range has shifted down a notch, prepare for $60,000 to $73,000.
The first test will be whether $75,900 holds at the weekly close at 8 AM next Monday.
The full analysis is in the referenced long article, worth sharing.
Aspirin.
"Important: Updated roadmap for the final dip after the midterm election rebound"
November 6, 2026, Issue 114 Aspirin Data Scientist's Perspective on Cycle Analysis In the previous issue, I bet on a scenario, in which the second step was written as follows: After the election results came out, BTC would surge to between 79,500 and 80,600, then the positive news would be fully priced in and it would pull back. What actually happened this week was: On the night of the election, the highest point reached near 81,000, a bit higher than expected; on Thursday, the US stock market dropped 5.5% overnight, and BTC fell back to 76,000. The second step of the scenario was completed. I have to admit that on Wednesday, when oil prices fell below 100 and the Dow Jones surged, I once raised the probability of a "direct strengthening". But just one day later, the US stock market gave the answer. This also reminded me that one day’s news is not worth changing a week’s judgment. This issue puts together the changes before and after the election, the 2018 template, several new historical experiences, and on-chain data to update the roadmap for the next five weeks. 01 What exactly happened in these three days? November 3 was the midterm election. According to multiple media reports, the Republicans lost, and the Democrats took control of both houses of Congress. On November 4, the market initially traded on the "good news". The reason was straightforward: a change in Congress means added constraints on the executive branch, reducing the likelihood of the Iran war escalating. Brent crude oil fell from the October 23 high of 116 dollars to below 100 dollars, intraday reaching around 94 dollars; US stocks surged; BTC liquidated about 378 million dollars worth of short positions within an hour, and the spot ETF also ended a three-consecutive-
Pinned
"Bearish Script Timeline"
October 26, 08:00: The second weekly candle closes below 83,000, confirming a false breakout
Following days: Retrace to the bull market support zone (just above 70,000 to 75,000), sweep the September 15 low, then rebound
October 29, 02:00: Federal Reserve meeting; if no rate hike but PCE is relatively hot that day, long-term yields may rebound, with the 10-year US Treasury returning near 5.3%
November 3: US midterm elections
Mid-November: Stock market decline drives BTC lower, long-term yields peak, forming Q4 lows
There is also a timing reference: local highs in July, August, September, and October all appeared around the 21st, except August extended to September 3. Based on this, the current peak falls between October 21 and November 3. However, with only four samples, this is for reference only.
The only condition to invalidate the entire script is if the weekly candle on October 26 closes back above 83,000.
"The Last Drop: The Time Is Very Near!"
November 14, 2026, Issue 115 Aspirin Cyclical Analysis from a Data Scientist's Perspective BTC fell below 70,000, currently testing the 200-week moving average, now between 66,000 and 68,000. A particularly obvious phenomenon in the past two weeks: every time it falls to a support level, someone stands up to call the bottom. 80,000 was called, 75,000 was called, 72,000, 69,000, and 67,000 were also called. Yesterday I listened in on a live broadcast; at first, everyone said the bottom was at 71,000, but before the broadcast ended, the bottom call price had been revised down several times. This issue does not call the bottom but answers four questions: Where are we in time? Why might the price not be here? What needs to be seen to confirm the bottom? And what do I plan to do myself? 01 Time: Already entered the window First, look at when the last drop of the past three cycles started. In 2014, the last drop started in November and ended in mid-January the following year. In 2018, the last drop started in early November and ended in mid-December, just around the interest rate meeting in December that year. In 2022, the last drop also started in November; FTX collapsed, and the bottom was seen that month. All three started in November after the midterm elections. This year's midterm election was on November 3; BTC surged to 81,000 after the election, then fell all the way down to now. The timing rhythm is almost the same as the past three times. Count how many days the bear market has lasted. The high point of this round was October 6, 2025; today is day 404. Past three
On the day Jensen Huang resigned, Nvidia first plunged then climbed: intraday it dropped as much as 14%, hitting a low of 148.8, but closed down only 4.3% at 166.19. Within half an hour of the open, retail investors net bought $110 million, setting a record.
It seems the market clearly distinguishes: Huang’s personal matters are personal matters, Nvidia’s orders are orders. CUDA won’t become invalid just because someone leaves, and TSMC’s capacity won’t resign along with him.
But don’t rush to celebrate yet. The scandalous news is just an appetizer; the main course is still ahead: export compliance, information security, board investigations, congressional testimonies, DOJ antitrust investigations. After the election, Congress has new leadership, and the "regulatory season" for AI may just be beginning. Those who remember 2018 know how tech stocks were slammed by regulators.
So Nvidia’s valuation logic needs adjustment: previously it included a "founder premium," now it must deduct some "governance discount." The leather jacket remains, but the person wearing it has changed.
Next week Anthropic goes public, Nvidia reports earnings; you could say BTC now basically shadows the Nasdaq.
My forecast position: will Nvidia close above $179 on November 13? I choose "no."
It closed at 166.19 yesterday; it needs to rise more than 7.7% today to reach that. It only rose 0.7% pre-market; 170 is the first hurdle; on the eve of earnings, with no official CEO yet, institutions won’t charge in now. Giving a 10% chance for "yes" is already generous.
This does not constitute investment advice.
On the night the CPI was released, both gold and BTC took a hit. Gold dropped 1.7%, and BTC swept down to 72,744.
But the more I look, the more I feel the stories behind these two assets are diverging.
I reviewed three historical periods:
In 2008, gold bottomed in October, while the US stock market bottomed 5 months later;
In 2018, gold bottomed in August, BTC lagged by 4 months;
In 2022, gold bottomed in September, BTC lagged by 2 months.
In all three cases, gold moved first. The reason is simple: gold trades on the expectation that "rate hikes are nearing an end," while risk assets wait for liquidity to truly ease.
Looking at the present, gold's October monthly candle broke below the bull market support zone, then recovered in November, almost identical to the 2008 scenario. If gold has truly bottomed, BTC's low will most likely fall between December and February next year, later than many expect.
In the full article, I also made a somewhat counterintuitive judgment: if BTC breaks below its July low in the next two months, 2027 could actually be better. This is because on-chain indicators will be fully cleared out, establishing a more solid bottom.
So my response is simple:
Gold moves first, but I’m halving my planned position size and will buy only after it stabilizes above $4,000;
BTC will be bought in batches as planned, stretching the pace to mid-January next year, avoiding high leverage.
The toughest period may be the next two months.
The full analysis is in the referenced long article, welcome to discuss.
Aspirin.
"Gold Bottoms Out First, How Much Longer for BTC?"
November 11, 2026, Issue 114 Aspirin Data Scientist's Perspective on Cycle Analysis On the evening of November 10, the US October CPI was released, hotter than expected: overall month-on-month increase of 0.4%, expected 0.3%; core month-on-month increase of 0.3%, expected 0.2%. Fed Governor Waller previously said that with core around 0.3%, a 50 basis point rate hike should be considered in December. Once the data was released, two assets with completely different trends this year both took a hit. Gold dropped 1.7%, back near $4,050; BTC swept down from 75,000 to 72,744, then recovered to 74,700 within a day. Both fell on the same day, but I believe the stories behind them are diverging. This issue puts together the recent two weeks' materials on gold and BTC to clarify one thing: why gold has very likely already bottomed first, while BTC's final phase is not yet over. 01 Gold: Just a deep correction, the bull market will continue First, look at the retracement magnitude. Gold has retraced about 30% from its high. This number is not outrageous historically: the 2006 retracement was about 25%, 2008 about 34%, 2022 about 22%, after which new highs were made again. Next, look at the most critical technical signal. For gold, the bull market support zone is seen on the monthly chart, specifically the 20-month moving average and the 21-month exponential moving average. October's monthly candle closed below the support zone, but November has reclaimed it. This pattern is almost identical to October 2008: back then also
"Gold Bottoms Out First, How Much Longer for BTC?"
November 11, 2026, Issue 114 Aspirin Data Scientist's Perspective on Cycle Analysis On the evening of November 10, the US October CPI was released, hotter than expected: overall month-on-month increase of 0.4%, expected 0.3%; core month-on-month increase of 0.3%, expected 0.2%. Fed Governor Waller previously said that with the core around 0.3%, a 50 basis point rate hike should be considered in December. Once the data was released, two assets with completely different trends this year both took a hit. Gold fell 1.7%, back to around $4,050; BTC swept down from 75,000 to 72,744, then recovered to 74,700 within a day. Both fell on the same day, but I believe the stories behind them are diverging. This issue puts together the recent two weeks' materials on gold and BTC to clarify one thing: why gold has very likely already bottomed first, while BTC's final phase is not yet over. 01 Gold: Just a deep correction, the bull market will continue First, look at the retracement magnitude. Gold has retraced about 30% from its high. This figure is not outrageous historically: the 2006 retracement was about 25%, 2008 about 34%, 2022 about 22%, after which new highs were made again. Next, look at the most critical technical signal. For gold, the bull market support zone is seen on the monthly chart, specifically the 20-month moving average and the 21-month exponential moving average. October's monthly candle closed below the support zone, but November has reclaimed it. This pattern is almost identical to October 2008: back then also
"Important: Updated roadmap for the final dip after the midterm election rebound"
November 6, 2026, Issue 114 Aspirin Data Scientist's Perspective on Cycle Analysis In the previous issue, I bet on a scenario, in which the second step was written as follows: After the election results come out, BTC surges to 79,500 to 80,600, then the positive news is fully priced in and it pulls back. What actually happened this week was: On the night of the election, the highest point touched around 81,000, a bit higher than expected; on Thursday, the US stock market dropped 5.5% overnight, and BTC fell back to 76,000. The second step of the scenario was completed. I have to admit, on Wednesday, the day oil prices fell below 100 and the Dow Jones surged, I once raised the probability of "direct strengthening." But just one day later, the US stock market gave the answer. This also reminds me that one day's news is not worth changing a week's judgment. This issue puts together the changes before and after the election, the 2018 template, several new historical experiences, and on-chain data to update the roadmap for the next five weeks. 01 What exactly happened in these three days? November 3 was the midterm election. According to multiple media reports, the Republican Party lost, and the Democratic Party took control of both houses of Congress. On November 4, the market first traded on the "good news." The reason was straightforward: a change in Congress means added constraints on the executive branch, reducing the likelihood of the Iran war escalating. Brent crude oil fell from the October 23 high of $116 to below $100, intraday down to around $94; US stocks surged; BTC liquidated about $378 million in shorts within an hour, and the spot ETF also ended a three-day streak
Binance was hacked for $6.3 billion, with $1.5 billion liquidated overnight, yet BTC only dropped to a low of 76,642.
It just hit the 21-week moving average, only 30 dollars short.
Honestly, I was quite surprised to see this candlestick. Such a major bearish event couldn't break through, which indicates someone is buying at the bottom. But I don't think the bottom has been reached yet; volume hasn't picked up, and panic hasn't set in.
So I reviewed all the cycles, on-chain data, moving averages, and macro factors I've been watching these past two weeks and wrote a long article.
What concerns me most is a coincidence: three completely unrelated methods all point to the same range, 70,000 to 74,000. After doing cycle analysis for so long, this is rare.
My judgment is that it will hold first, then drop. It will likely hold before the election, possibly even rebound near 80,000, but the final drop may not be over. The timeframe I’m focusing on is November 16 to 21, the week Anthropic goes public.
Of course, I could be wrong. So in the long article, I clearly explain how to respond to each possible scenario and when to admit being wrong.
The first test point is the weekly close at 8 AM next Monday; whether 75,900 holds is critical.
If interested, you can read the referenced long article, and feel free to debate.
This does not constitute investment advice.
Aspirin.
"Hold First, Then Fall: The Timing and Price of the Last Drop in This Cycle"
October 30, 2026 Issue 112 Aspirin Cycle Analysis from a Data Scientist's Perspective On Wednesday night, news broke of a Binance hack involving about $6.3 billion in funds, with approximately $3 billion frozen, and all deposits and withdrawals suspended. The entire network saw liquidations exceeding $1.5 billion overnight. Based on past experience, a negative event of this magnitude is enough to trigger a long bearish candlestick. However, Bitcoin's lowest point only dropped to 76,642, just landing on the 21-week moving average at 76,612, almost exactly. This event itself is a piece of data. It indicates that the current sell-off is not yet panic selling; there are buyers at the market bottom. But it does not prove that the bottom has appeared, because a true cycle low is usually accompanied by a concentrated release of volume, which did not happen this time. In this issue, I put together all the clues from the past two weeks: cycle timing, on-chain cost, moving average structure, macro rhythm, and finally present what I believe is the most probable path along with the corresponding strategy. 01 Where we stand in the cycle The high point of this round was 126,000 on October 6, 2025, and a low of 57,700 appeared in July this year. Starting from the bear market bottom in November 2022, today is day 1,439. The previous two cycles saw lows on days 1,432 and 1,436 respectively, so from a timing perspective, the low window has opened. Another timing pattern comes from the midterm election year. Historically, the fourth-quarter lows in midterm election years generally occur after the election. This year's election is on November 3, which means the window is very likely to fall in mid-November
"Hold First, Then Fall: The Timing and Price of the Last Drop in This Cycle"
October 30, 2026 Issue 112 Aspirin Cycle Analysis from a Data Scientist's Perspective On Wednesday night, news broke of a Binance hack involving about $6.3 billion in funds, with approximately $3 billion frozen, and all deposits and withdrawals suspended. The entire network saw liquidations exceeding $1.5 billion overnight. Based on past experience, a negative event of this magnitude is enough to trigger a long bearish candlestick. However, Bitcoin's lowest point only dropped to 76,642, just landing on the 21-week moving average at 76,612, almost exactly. This event itself is a piece of data. It indicates that the current sell-off is not yet panic selling; there are buyers at the market bottom. But it does not prove that the bottom has appeared, because a true cycle low is usually accompanied by a concentrated release of volume, which did not happen this time. In this issue, I put together all the clues from the past two weeks: cycle timing, on-chain cost, moving average structure, macro rhythm, and finally present what I believe is the most probable path along with the corresponding strategy. 01 Where we stand in the cycle The high point of this round was 126,000 on October 6, 2025, and a low of 57,700 appeared in July this year. Starting from the bear market bottom in November 2022, today is day 1,439. The previous two cycles saw lows on days 1,432 and 1,436 respectively, so from a timing perspective, the low window has opened. Another timing pattern comes from the midterm election year. Historically, the fourth-quarter lows in midterm election years generally occur after the election. This year's election is on November 3, which means the window is very likely to fall in mid-November
PCE is out, and the result is lukewarm; gold, silver, and the dollar barely moved. But looking closely at this set of data, there's quite a bit of information.
Core PCE for September year-over-year is 3.1%, slightly higher than last month's 3.0%, and month-over-month is 0.3%, all in line with expectations. Inflation is not out of control, but there's no sign of it easing either, so the case for a rate hike in December remains.
At the same time, Q3 GDP annualized growth is 3.4%, stronger than the expected 3.2%, and up from 2.2% last quarter. With such a strong economy, the Federal Reserve has no reason to rush to ease. The only somewhat dovish point is the GDP price index, which is only 2.3%, well below the expected 3.2%.
My understanding is that tonight's data doesn't vote for any particular path. A strong economy and core inflation not falling support a December rate hike and could push 10-year Treasury yields higher, which favors the bearish scenario. But with GDP this strong, it's hard to see recession fears and a major stock market drop in the short term. Plus, WTI has already fallen back to around $101, so the move to $120 oil is stalled for now, which may weaken the bearish scenario somewhat.
So, the next move is up to the market price. I'll be watching three things. First, the weekly candle at 8 a.m. on November 2: if BTC closes below 83,000, the main path continues; if it closes above, a shift is likely. Second, whether the 10-year Treasury yield will surge to 5.3% due to the strong economy. Third, whether this pre-election rally will be resisted around BTC 80,000 to 82,000.
When the market doesn't give answers, the best move is to wait.
