Decided to run the numbers on @VestExchange again.
This post (teapawt.net/flank2/status/21148272…) goes into coin flips, and why influencers buying tons of accounts are not typically outperforming a coin flip (s/o again to @MAGNUS_FUTURES )
I wanted to see if any of Vest's other accounts worked better.
One in particular does.
The $90 two-step returns +137% on the exact same coin flips. Same 850 accounts, a quarter of the spend, 2.5x the money out.
In the prior example, $319k spend, $42k out
In this example, $76,500 spend, $105k out
And its not just a bigger average. The 95% band on the one-step runs from a $21k loss to a $110k gain. On the two-step it runs +$64k to $149k. Its worst year in forty runs still beats the expensive account's average year.
Same model as last time, same assumptions, same fees. Only the account type and parameters change.
Parameters
Target $2500 then $1250
Drawdown $1500 static
Max risk/trade $650 (avoiding daily limit touch)
Funded phase: identical - the account's value is the drawdown. Bet the same max risk per trade as the eval. Withdraw everything above the floor the moment you're above it; take 80% profit split, keep going until the account dies.
And I suppose for my second post in 18 years we'll discuss geometry.
Over the years I've seen many fintwit influencers and their following (I was one) promoting the idea that all that matters at the end of the day is overall PNL on spend vs payouts. True...unless you want to identify if you have any genuine skill...
Most here understand props. Most understand trading. What few have understood because they haven't been shown aside from a couple folks in the industry is what explains the big pnl screenshots.
This is a distillation of content @MAGNUS_FUTURES and JJ Simon (youtube.com/@itsjjsimon) have put out. Give them a follow.
An eval is a bet with fixed odds. You can calculate them before you place a trade.
The math and example here is premised off @VestExchange one step $25k, no discount code... $375 (no DLL, 80% split)
$2500 to pass
$1500 drawdown
Assume the trader has zero edge. Pure coin flip. Assumptions:
- trader risks full $1500 every trade
- trader uses 75 point stops (for fee calcs)
- trader withdraws everything after every win
If you're a coin clip, the odds you pass are a distance:
$1500 / ($1500 + $2500) = 37.5%
Entry model doesn't change the math. Fees drag it to 36.5%
What is the account worth?
Risk the whole $1500 to make gross of $1500. Half the time you win, pull $1200 after the profit split, go again. Half the time it's gone. Run taht forward and the average account gives you one payout before it dies.
So a funded account is worth its drawdown times the split. 1500 x 80% = $1200. Fees take it to $1165
Since you pass 36.5% of the time, an account is worth $425 before you even buy it. It costs $375. So you make $49 per account knowing nothing about the market.
13% return on a coin flip, after fees. Yes it's a pricing mismatch, yes most props have it, and yes they know (because most traders aren't buying 100+ accounts)
$49 per account is average. Averages only show up if the action is repeated enough times. A single account pays nothing 82% of the time.
3 accounts: 55% chance you get zero payouts and lose the whole spend.
850 accounts: 0.0%. You'd need 153 failures in a row which doesn't happen.
At 850 accounts you're looking at roughly +$42k expected and a 90% chance of finishing green. Not because you can trade. Because you bought enough accounts that the average shows up in the bottom line.
This is why the screenshots of big payouts exist.
Influencers (most, not 1, 3 or 10 in particular; I'm speaking generally about the long-time influencers in the space) running hundreds (thousands?) of accounts is going to produce a big number at some point. It's arithmetic.
Statistically, 697 of those 850 accounts return nothing and it doesn't matter, because the 153 that pay cover all of them (and when luck favors them, a few of those accounts could catch big runs at just the right time to inflate more).
Then the big number becomes the content. And what gets sold is the entry mode, the risk management, and the system. The thing that produced the result, owning 850 accounts, is never the product or what's taught.
Something worth thinking on
At that scale you can't tell if the person can trade. The spread from luck alone is about $33k either way , and a small real edge is worth $15k across the whole portfolio. The luck is bigger than the skill.
To be clear, 95% band bracket where there is no statistical difference from a coin flip sits between a loss of $21,150 and a gain of $109,650.
Anywhere in that bracket is ordinary luck. Clearing it takes an edge above +.046R per trade (expectation of $118,602 across the portfolio). Under that, 850 accounts cannot separate a skilled trader from a coin flip.
The results of influencers, especially with the obfuscation of underlying metrics like account quantity, spend, and yoloing accounts before hitting a payment can't and doesn't prove skill.
Again credit to Magnus and JJ for bringing light to the math behind this. I think most influencers are unintentionally following this math, but in a much more disorganized way. I think most genuinely believe the edge is in the chart and their approach.
The math doesn't give a shit. The bankroll is the strategy at this level of the game.
Now the part that should bother people.
Take 850 one-step accounts and a trader with a genuine, real, hard-won 52% win rate at 1:1. Their median year is +$100,050.
A trader with no skill whatsoever lands between a $21,150 loss and a $109,650 gain, 95% of the time.
The skilled trader is inside the coin-flip band. You cannot tell them apart. Not at 850 accounts, not from a payout screenshot, not from a year of them.
Luck is worth plus or minus $65,400. A full point of real win rate is worth $28,800. Luck is more than twice the size of the edge.
You need 53% before the result itself says anything.
Here is what I think is actually happening.
Most of the people posting these numbers are not lying. They found this by survivorship, not by deriving it: buy a lot of accounts, full port, take the payout, repeat, and the ones for whom it worked kept going and started selling the entry model.
The behavior is the zero-edge optimum. They arrived at it sideways and credit it to the strat.
The tell is which account they buy. Anyone who had actually run this math would be on the $90 2-step, not the $375.
The tell is which account they buy. Anyone who had actually run this math would be on the $90 two-step, not the $375. They are paying four times the price for the weakest version of the trade, and calling what comes out of it skill. They do not know that is what they are doing. That is sort of the point.
What you give up (because this isn't free):
It's slow. About 24 trades to clear both phases instead of 4. Four to five weeks a cycle instead of one. Running 850 of them is 10,887 trades, not 2,198 (on 850 accounts).
It is "lumpier" per account. 87% pay you nothing, vs 82%. A single one is a worse ticket at the carnival. They are a quarter the price, so a budget buys four times as many, and that's the entire reason the band tightens.
It's unforgiving. The daily limit is measured on unrealized equity and closes the account the moment you touch it. No warning. A second trade on a losing day is how that happens, which is why the rule is simply to stop once you are red.
Note: Under about five accounts the expensive one is the better bet: at three accounts you are 45% to finish green on the $375 and 34% on the $90. The cheap account only wins on volume. By ten accounts it is 63% against 42%, and by a hundred it is 97% against 64%.
tldr: The bankroll is still the strategy; it's always been the strategy for the vast majority of influencers. This - 2 step - just makes the bankroll go four times further.
And I suppose for my second post in 18 years we'll discuss geometry.
Over the years I've seen many fintwit influencers and their following (I was one) promoting the idea that all that matters at the end of the day is overall PNL on spend vs payouts. True...unless you want to identify if you have any genuine skill...
Most here understand props. Most understand trading. What few have understood because they haven't been shown aside from a couple folks in the industry is what explains the big pnl screenshots.
This is a distillation of content @MAGNUS_FUTURES and JJ Simon (youtube.com/@itsjjsimon) have put out. Give them a follow.
An eval is a bet with fixed odds. You can calculate them before you place a trade.
The math and example here is premised off @VestExchange one step $25k, no discount code... $375 (no DLL, 80% split)
$2500 to pass
$1500 drawdown
Assume the trader has zero edge. Pure coin flip. Assumptions:
- trader risks full $1500 every trade
- trader uses 75 point stops (for fee calcs)
- trader withdraws everything after every win
If you're a coin clip, the odds you pass are a distance:
$1500 / ($1500 + $2500) = 37.5%
Entry model doesn't change the math. Fees drag it to 36.5%
What is the account worth?
Risk the whole $1500 to make gross of $1500. Half the time you win, pull $1200 after the profit split, go again. Half the time it's gone. Run taht forward and the average account gives you one payout before it dies.
So a funded account is worth its drawdown times the split. 1500 x 80% = $1200. Fees take it to $1165
Since you pass 36.5% of the time, an account is worth $425 before you even buy it. It costs $375. So you make $49 per account knowing nothing about the market.
13% return on a coin flip, after fees. Yes it's a pricing mismatch, yes most props have it, and yes they know (because most traders aren't buying 100+ accounts)
$49 per account is average. Averages only show up if the action is repeated enough times. A single account pays nothing 82% of the time.
3 accounts: 55% chance you get zero payouts and lose the whole spend.
850 accounts: 0.0%. You'd need 153 failures in a row which doesn't happen.
At 850 accounts you're looking at roughly +$42k expected and a 90% chance of finishing green. Not because you can trade. Because you bought enough accounts that the average shows up in the bottom line.
This is why the screenshots of big payouts exist.
Influencers (most, not 1, 3 or 10 in particular; I'm speaking generally about the long-time influencers in the space) running hundreds (thousands?) of accounts is going to produce a big number at some point. It's arithmetic.
Statistically, 697 of those 850 accounts return nothing and it doesn't matter, because the 153 that pay cover all of them (and when luck favors them, a few of those accounts could catch big runs at just the right time to inflate more).
Then the big number becomes the content. And what gets sold is the entry mode, the risk management, and the system. The thing that produced the result, owning 850 accounts, is never the product or what's taught.
Something worth thinking on
At that scale you can't tell if the person can trade. The spread from luck alone is about $33k either way , and a small real edge is worth $15k across the whole portfolio. The luck is bigger than the skill.
To be clear, 95% band bracket where there is no statistical difference from a coin flip sits between a loss of $21,150 and a gain of $109,650.
Anywhere in that bracket is ordinary luck. Clearing it takes an edge above +.046R per trade (expectation of $118,602 across the portfolio). Under that, 850 accounts cannot separate a skilled trader from a coin flip.
The results of influencers, especially with the obfuscation of underlying metrics like account quantity, spend, and yoloing accounts before hitting a payment can't and doesn't prove skill.
Again credit to Magnus and JJ for bringing light to the math behind this. I think most influencers are unintentionally following this math, but in a much more disorganized way. I think most genuinely believe the edge is in the chart and their approach.
The math doesn't give a shit. The bankroll is the strategy at this level of the game.
I had been in Lanto’s Discord from early 2023 until earlier this year. I defended him from naysayers in the group and on X, distilled his content into guides early on, and built the chatbot that new members used. This weekend’s fiasco confirmed conclusively what OGs had been flagging for some time.
Fintwit prop firm influencers are “successful” through survivorship bias and volume, not edge.
Buy enough independent attempts and pure luck eventually produces large winners, even when the underlying approach has little or no positive expected value once net on the portfolio is established. Those big runs are the ones that the public gets to see. The average guy in the audience is running 2, 4, maybe 5 accounts on average while paying discord fees to "learn how to trade"
Recent public examples among many in the fintwit affiliate and influencer space:
@dhesi_trades - 170+ accounts
@thebrianstonk - 240+ accounts
@lanto_trades - 819+ accounts
5-6 in eval or S2F fees on just one firm.
Eventually something hits. The run gets screenshotted and turned into content that's used for the next 10 months of promotions for the group, or pushed in promo videos encouraging use of an affiliate code. The costs typically don't see the light of day. Screenshots get cropped (usually), account numbers are hidden, the actual PNL on the machine is kept private.
The machine:
1. subscription fees, course sales, and affiliate commissions fund more accounts
2. more accounts make a big winner close to inevitable
3. the big runs, sell more subscriptions, courses, and garner more aff code use
Repeat.
The prop firm space is inherently dirty, relies on ponzinomics, and is unfixable; it's by design.
But the affiliate and furu space - that props implicitly encourage - is dirtier.
Fresh cash from courses, subs, and affiliate payouts keeps the account volume flowing so the occasional big payouts can be marketed as a skill taught to their followers.
The trading can lose money, but the machine still wins, because the audience pays for step 2 of the machine.
Now look at what's sold. Courses, discord subs, mentorships, and an implied promise to teach a skill to escape the rat race.
Learn this IFVG. Take this unicorn. Manage your risk. Trade the system. Don't forget to like, subscribe, and use code DUMB.
The average member can't play the volume game and nobody tells them that is in fact their mentor's game.
I left for three reasons: the quality of members in the group declined after the TPT run-up and subsequent repeat promotions leveraging that run, OGs who asked questions were treated poorly, and some kept raising exactly the disclosure issues that surfaced this weekend.
Buy enough attempts.
Hide or minimize the cost.
Market the winners as skill.
Accept money from people who cannot realistically replicate the process.
You can pick whatever word or phrase you want for this machine that is pervasive in this "industry."
Mine is "fraud."
