Minh explains why you should take a break from trading after a big win
“If I’m up a boatload now, I take like six months off and I don’t play, that’s the key”
“If you keep playing, you’ll lose. I wish that when I made $750,000 last year I stopped playing and didn’t try to get to $1,000,000. Went to Japan for 2 months, New York for another 2 months… that’s only what maybe $150,000? I would’ve saved like $600,000”
“Sometimes doing stupid stuff that allows you to feed your dopamine and do really fun experiences and it may be expensive… that might save you a lot of money”
you're paying $560/month for something that costs $0.12.
tested on 89 accounts.
everyone keeps VPN connected forever for US targeting.
i disconnect after 7 days and still hit 85%+ US audience on 160 accounts.
here's what actually happens:
warmup phase (days 1-7):
> connect to dansvpn
> browse niche content 20min/day
> post 3-5 non-promotional reactions
> TikTok locks in audience preference
the algorithm learns during warmup. targeting gets permanently trained into the account.
automation phase (day 8+):
> disconnect VPN entirely
> load account into scheduler
> post 2x/day automated
> US targeting STAYS locked
you're not paying for ongoing targeting.
you're paying for initial training.
one week setup = permanent US targeting.
my actual workflow:
week 1: warmup 30 accounts manually with one VPS
week 2: disconnect, load into social bee, queue 120 videos
week 3: automate posting, warmup next 30
week 4: repeat
120 accounts warmed up per month on $3.50 infrastructure.
the math everyone misses:
permanent VPN model: 20 accounts max ($70/month)
warmup-only model: 400 accounts possible ($3.50/month)
20x less cost. 11x more accounts.
you can batch-warmup 30 accounts simultaneously on one $3.50 VPS.
each account costs $0.12 total for lifetime US targeting.
not $3.50/month forever.
this is how phone farms scale to 1000+ accounts.
batch-warmup, disconnect, automate.
while you're stuck at account #15 manually posting from NordVPN forever.
the sequence:
phase 1: manual warmup + VPN (days 1-7)
phase 2: automated posting + no VPN (day 8+)
VPN trains the account.
scheduler scales the output.
confusion on which phase needs which tool is why most people stuck at 20 accounts.
don’t use NordVPN for TikTok if you’re trying to reach a US audience you’ll get shadowbanned fast.
TikTok flags cheap VPNs and shared proxies because thousands of users share the same IP. You and 50,000 others hitting the US audience with NordVPN? Of course TikTok knows.
The fix: host your own private VPN on Hetzner with Outline VPN. It costs ~$3.5/month, takes 10 min to set up, and lets you create 100s of private VPN keys.
But the VPN alone isn’t enough, you also need to warm up your TikTok account properly to avoid detection and reach the right region.
Here’s how I do it (FULL BREAKDOWN)
1. Hetzner VPS with Outline VPN
Set up your own VPN (not NordVPN) in 5 min
> Follow Hetzner’s guide → link in first reply
(installs Outline via Docker automatically)
> Download Open Outline Manager → add server → create VPN keys
(each key = 1 device password).
you can create 100s of keys using only 1 VPS server
> Paste key in Outline app → connect → done
Tips:
– Use US/EU servers
– Hetzner = 2TB bandwidth/month for ~$3.5
– DigitalOcean = 1TB
– Avoid GCP/AWS = $300+ bills just for traffic
2. Reach US Audience with a New TikTok Account
Before downloading TikTok:
– Use a phone w/ no SIM
– Turn on VPN (I use DansVPN)
– Change region + phone time to target country
> Download TikTok → check if number prefix matches (e.g. +1 for US)
> Create account via email → your FYP should match your region
> Post after 2–3 days (don’t act like a bot)
> Switch to Creator Mode. check TikTok Studio analytics
Tip: turn on “Profile Views” to see if your visitors are from your target country
Keep VPN always on for that account.
3. Warmup Account
> Do not follow, like or comment anybody in the first day!!
Don't even touch your bio or photo. Just use search to find posts in your niche.
tiktok will learn from your search and start sending you more relevant content.
> Day 2. continue with natural search in your niche. change bio and photo.
> Day 3. like and follow 5 accounts
> Day 4. lile and follow 10 accounts
> Day 5-6 day follow 10 accounts in your niche. post one non promotional video. UGC reaction with hook works best for me.
guy makes $15k/month from TikTok. never posted a single video.
he runs 89 VPS servers for 340 creators who think they're just "good at the algorithm."
they're not. they're just posting from clean IPs.
the business model:
> rent Hetzner VPS servers ($3.50 each)
> install Outline VPN
> support 3-4 creator accounts per server
> charge creators $180/month for "premium VPN service"
$312/month cost. $61,200/month revenue.
the performance difference is insane.
same creator, NordVPN: 430 avg views, shadowbanned every 2-3 weeks
same creator, his VPS: 52,000 avg views, zero shadowbans in 6 months
only variable: IP trust score.
consumer VPNs share one IP with 5,000+ users. dropshippers. bot farmers. banned accounts. TikTok sees spam network.
his VPS? 3-4 vetted creators per IP. clean history. TikTok sees legitimate users.
the brutal math:
shared IP gets 200 test viewers. need 8% CTR = 16 engaged viewers. not enough data. video dies.
clean IP gets 20,000 test viewers. same 8% CTR = 1,600 engaged viewers. enough data to scale. video explodes.
100x reach difference before content quality matters.
everyone's optimizing hooks when they should optimize infrastructure.
perfect hooks on shit IP: 400 views
mediocre hooks on clean IP: 40,000 views
the game is decided before content quality matters.
I already covered the full VPS setup in a previous thread.
link below.
infrastructure > content.
Dad buys stock at $50K.
It grows to $5M.
If he sells, he owes tax on a $4.95M gain.
Instead, he puts it in a trust.
Borrows against it.
Lives tax-free.
Dies holding.
Kids inherit at a $5M basis.
IRS gets $0.
Trezor is having a 50% off sale.
You can pick up the Model One for only $24.50 plus free shipping.
Only buy from their website directly at https://t.co/uWWXIttFDO
This is what I learned from @bpkdasbaum so make sure you save these important instructions!
I’ve been seeing a lot of chatter on X about “peak cycle” and how the economy looks late-cycle. So I wanted to tackle this head on and share a few thoughts of my own...
This is from the August 21st MIT publication:
A classic late-cycle economy typically has all the following ingredients:
✅ Manufacturing sentiment is extreme (think ISM ~60)
✅ Services sentiment is extreme
✅ Homebuilder sentiment is extreme
✅ Consumer confidence is high
✅ Worker confidence is high (JOLTS quits rate rising sharply)
✅ Investor sentiment is very bullish
✅ Small business confidence is high
✅ Job openings and hiring plans are rising
✅ Wage data and surveys show accelerating pay increases
✅ CEO confidence is strong and capex is booming
Now, I could add more to this, but when you score all of these inputs and turn them into a single timeseries, here’s what you get (chart 1).
Using data from ISM, NAHB, NFIB, BLS, AAII, The Conference Board, etc., US sentiment, when viewed as a complete picture, remains very subdued. We’re just not even close to the euphoric levels we see late in the business cycle, when everything listed above is stretched to extremes.
Peak cycle is when the ISM rolls over from 60+ to sub-50, inventories unwind, and demand cools. Supply and demand reset, inflation pressures ease, and the cycle eventually recovers out of the slowdown or recession – mostly depending on the extent to which financial conditions tightened during the cycle, particularly late on as central banks hike rates and drain liquidity.
However, based on this full set of indicators, the data is pointing to something very different. This does not look like an above-trend late-cycle economy. It looks much more like an early-cycle economy trying to build momentum.
Another really important factor, and a key reason we believe both the ISM and this sentiment composite will grind higher this year and into 2026, is the sheer scale of central bank easing via rate cuts.
Right now, nearly 90% of central banks are cutting rates. That is extraordinary, and on a forward-looking basis, it is a massive tailwind for the business cycle (chart 2).
By my playbook, the time to start talking late-cycle is when the teal line rolls over and begins to drop, as central banks turn to hiking rates to slow growth. Even then, there’s usually a nine-month lag before higher rates hit the real economy.
Right now, we’re just nowhere near that... in fact, the opposite is true.
To my earlier point, slowdown or recession is largely a function of how much financial conditions tighten late in the cycle. Oil prices are a big part of this equation. When oil runs 50% above trend, that represents a massive tightening and has almost always signaled recession, looking back to the early 1970s.
However, right now, we are nearly 20% below trend and still falling, which shows this component of financial conditions is still easing (chart 3).
Also, as I’ve pointed out many times in previous reports, when you look at Temporary Help Services, it has early-cycle vibes written all over it (chart 4).
Rising growth from deeply negative levels is an early-cycle dynamic. It tells you the economy is in recovery mode, not rolling over.
Late-cycle is the opposite: positive year-on-year growth that’s slowing, which reflects an overheated economy losing steam.
Why is unemployment still rising?
Because it lags the cycle. Jobs data is a six-month look in the rear-view mirror.
Here’s the thing: full-time hires are expensive. Benefits, pensions, overhead…
So what do businesses do first?
They typically increase overtime hours and bring in temp workers. Only when they feel confident do they finally lock in full-time staff. That way, they can scale without locking themselves into long-term payroll commitments.
So, this isn’t late-cycle. It’s early-cycle (growth up + inflation down = Macro Spring), soon transitioning to mid-cycle (growth up + inflation up = Macro Summer).
That’s how I see it, anyway...
Started a group called Archive of Axis. For art, visuals and quiet ideas.
Feel free to join if it feels right. The banners are yours if you want them. 🌌
One of the most significant individual sources of ‘hidden sell pressure’ has mostly finished exiting the @RichardHeartWin ecosystem in the short-term. It’s a series of wallets you’ve probably never heard of as their ‘selling’ very rarely appears on the charts, so we thought we’d dig into these wallets to present a case study on LP management and the impact that can have on an ecosystem. Here's a 🧵