tl;dr

I think I just found a list of cellar boxed companies. FTD may be fabricated. The way they fabricated it, though, means you can infer, to some extent, what companies are being cellar boxed, and are heavily shorted. I’ve included examples at the bottom. I’m sharing a link at this point, this data is still somewhat rough and will be updated, but I think this is an indicator of companies that are, or are in the process of being, cellar boxed.

A fair number of these companies are SPACs (*special purpose acquisition company)* or ETFs with billions that have lost money fast recently.

What if we could squeeze, not just one shorted company, but *all of them*.

[https://docs.google.com/spreadsheets/d/e/2PACX-1vSPEsyzOPfCr4QcRQHurde9UVsPVBE-RLaG\_pdCXJFzkMQ-Xf-PWvHRw9yTaQugp-VT7HESXoLJOsQq/pubhtml](https://docs.google.com/spreadsheets/d/e/2PACX-1vSPEsyzOPfCr4QcRQHurde9UVsPVBE-RLaG_pdCXJFzkMQ-Xf-PWvHRw9yTaQugp-VT7HESXoLJOsQq/pubhtml)

====================

Givex Information Technology Group Ltd

Take a look at some of the price movement over the past year.

https://preview.redd.it/l0odzsjkq2q91.png?width=998&format=png&auto=webp&s=26c21a7c34d8c6eb91e3105520dd70574e10e479

And their financials

https://preview.redd.it/c10f9ukdq2q91.png?width=713&format=png&auto=webp&s=ba829860fe50a262b6bc7cee7e7435ba7e27f50f

Can anyone tell me what the difference between this and fraud is?

==================

ETF had a strange peak about, what, 2 months ago then crashed by \~40%? After being almost horizontal otherwise? Really?

https://preview.redd.it/vlvlcu24r2q91.png?width=961&format=png&auto=webp&s=f13603255382497c76e576e17899f20752bb2086

===================

GRRRW looks like its in the middle of a squeeze.

​

[\\”Look at you, you baby gorilla.\\”](https://preview.redd.it/xinj19k3s2q91.png?width=800&format=png&auto=webp&s=5de4e2f50b4824f6fcb733b6373f8402f0496939)

=============================

Global Technology Acquisition (GTWAC)

https://preview.redd.it/qqp43aeps2q91.png?width=992&format=png&auto=webp&s=92cbd85904a2ee23d2fddd704dd873ae681b94b5

Hidden in plain sight.

===================

HARBOR ETF CORP CULTURE LEADER

[HAPY](https://preview.redd.it/5vir892tz2q91.png?width=1295&format=png&auto=webp&s=ec4b0fdfddca36616a12b52603712b7361df75dd)

===================

Advantage Solutions Inc

https://preview.redd.it/9t0rxq4w53q91.png?width=755&format=png&auto=webp&s=361785e068d1c5f8040e724fef05276739790fe2

===================

EXPLANATION:

When people want to fake data, there are a number of hurdles they have to jump over. Probably the most famous example is “Benford’s Law”, which gives the distribution of different end digits within a random dataset, but a number of other things need to be accounted for as well (normality, regularity, apparent randomness).

Typically, when people generate random data, they sample from some kind of bell curve that relies on exponential functions, which can be interpreted as operating in the complex space. This sounds pretty underwhelming at first, but it means that the norm that these normal distributions sample from (see: [https://www.wikiwand.com/en/Modulus\_of\_complex\_number](https://www.wikiwand.com/en/Modulus_of_complex_number)) is different from the norm that positive integers “in real life” have. Complex distributions have two axes (norm = sqrt(a\^2+b\^2)), positive integers only have 1 (norm = sqrt(x\^2)).

So think of as if you were at a roulette wheel, and a ball is spinning along that wheel, your “number system” is being “partitioned” in a very different fashion (something like on the complex plane) than it does in real life (positive reals). Since the normal distribution is representing a disc, in a sense, I took the FTD data for the whole stock market modulo pi.

Here is March 2004, which should help to give a baseline.

https://preview.redd.it/rbbxepmw1zp91.png?width=502&format=png&auto=webp&s=d92ad9a5ab655b5fb60c4372b1245cd17ecba552

Here is the July/August 2008 FTD modulo pi, which was before the Merryl Lynch bankruptcy, still uniform.

https://preview.redd.it/x9k62jzx1zp91.png?width=502&format=png&auto=webp&s=c5f2bd46bc97fd458c81fd533d6c96b93e84b7c4

​

https://preview.redd.it/zv5clywz1zp91.png?width=502&format=png&auto=webp&s=825cebf7e52494542f750e7b41fa57ff0876be3a

​

Here’s the same graph in September 2008.

https://preview.redd.it/8bp1ing22zp91.png?width=502&format=png&auto=webp&s=f53024d78da1225a923499da537f2f8073367450

Here’s the data from August 2022, the most recent available data.

https://preview.redd.it/a6b9xag82zp91.png?width=496&format=png&auto=webp&s=05d66320758e906067dc98c82adbe131b8ef18fb

It seemed odd, and since I drew from multiple points of time, and the effect seems pronounced during periods where we “suspect” FTDs are being hidden, I was wondering if anyone else knew about this. I thought this was interesting.

​

Note: All Data was pulled from the SEC site. The rough form of the code I used is below:

import pandas as pd

df = pd.read\_table(“cnsfails202208b”, sep=”|”).iloc\[:-2, :\] # remove the footer data

df\[“QUANTITY (FAILS)”\].mod(math.pi).plot.density()

plt.title(“Failure To Deliver (# FTDs modulo pi) – August 2022”)

​

===================

Hey, I was continuing to look into this, and I was specifically interested in cases where the #FTDs mod pi and the ln(# FTDs) were essentially unchanged, but not trivial. One example that I found was AFRICAN DISCOVERY GROUP INC (# FTDs in august from 8/16 to 8/24 was 3200, and the price was hovering from 0.02c to 0.03c). When you go to their website, I don’t think I’ve ever seen anything more suspect in my life.

​

[https://africandiscoverygroup.com/](https://africandiscoverygroup.com/)

AFDG’s investors, at the top, list a guy named Allen ‘Chainsaw’ Kessler (his nickname is the chainsaw), an executive at Merryl Lynch. He is more famous, however, for winning >$4M dollars from poker without ever winning the WSOP. In fact, his wikipedia page says literally nothing about the fact that he is an executive at Merryl Lynch, but if you look at the management page on AFDG’s website:

“Mr. Kessler has over 20 years of experience on Wall Street, starting his career in Investment Banking at Morgan Stanley, and subsequently Investment Research at Goldman Sachs. He founded African Discovery Group in 2017. Mr. Kessler serves as a board member of Port Energy, a Senegal-based Oil & Gas exploration company, First American Minerals, a mineral resource principal vehicle, and Ogelle, a Nigerian based social media company. Mr. Kessler is the Founder of African Discovery Foundation, a philanthropic organization focused on charitable contributions for education and health care resources in Africa. Mr. Kessler holds a Bachelor of Arts in Economics from the University of Pennsylvania (Cum Laude) and an MBA from Columbia Business School.”

On the board of directors: **Dr. Barfuor Adjei-Barwuah is Ghana’s Ambassador to the United States of America. Ambassador of the United Republic of Tanzania to the United Nations.**

I’m continuing to look into this.

===================

I think we can use a property of the collinearity of complex spaces to actually show this is the case.

​

When we represent numbers on a unit circle, what we’re doing is partitioning that unit circle by some divisor (lets call this theta\_p). This partitioning is really not obvious to us, because logarithmic spaces operating with variable and usually very high frequency are completely unintuitive. They seem chaotic, but they’re not really chaotic. Its been proven that given three points along the unit circle, one can show that another point on that unit circle is collinear to those other points. We can do this by representing the points along the circle as they relate to a rotation around this circle. On the unit circle, you can form a linear relationship between these points by

e\^(i \* theta) = cos(theta) + i \* sin(theta)

​

that means that, if we’re sampling using a particular integer divisor of the unit circle (think, rotations of theta\_initial by some integer multiple of theta\_p), then the previous three terms of these rotations can derive the last one. This can be done as:

exp(alpha\*theta\_p\*2pi\*i) \* exp(beta\*theta\_p\*2pi\*i) \* exp(gamma\*theta\_p\*2pi\*i) \* sqrt(e) = exp(kappa \* theta\_p\*2pi\*i)

​

this can be equivalently expressed as

​

exp((alpha\*theta\_p + beta\*theta\_p + gamma\*theta\_p)\*2pi\*i) = exp(kappa \* theta\_p \* 2pi\*i)

​

taking the natural logarithm of both sides yields:

((alpha\*theta\_p + beta\*theta\_p + gamma\*theta\_p)\*2pi\*i) = (kappa \* theta\_p \* 2pi\*i)

​

and canceling out like terms:

(alpha + beta + gamma) = (kappa)

alpha, beta, gamma, kappa belong to the integers

​

these greek terms are 1/some integer frequency, so given p = 1/theta\_p

e\^(1/alpha – 1/beta – 1/kappa) = e\^1/p

​

when I calculated this for the first 3 entries of the august 2022 ftd data:

[https://www.wolframalpha.com/input?i=e%5E%281%2F88879-1%2F11307-1%2F19%29](https://www.wolframalpha.com/input?i=e%5E%281%2F88879-1%2F11307-1%2F19%29)

The answer was the relation of a prime:

1006428721

and a number that cannot be represented as the sum of three squares

19094142207

​

I think I just broke something.

==========================

Looking more at this thing, I think every number in this dataset can be represented as some form of

x = (2 π n – i (log(7) + log(12697)))/(π + i log(3) + i log(7) + i log(19) + i log(3769) + i log(12697) – i log(1006428721)), n element Z

​

I’m going to take a guess that this may happen in some kind of order, too. What I do know is that this is a self-adjoint operator, as it has roots at n=x for n belong to N

​

View Reddit by PhilosopherMental395 – View Source

LEARN ABOUT: Investing During Inflation

REVEALED: Best Investment During Inflation

HOW TO INVEST IN GOLD: Gold IRA Investing

HOW TO INVEST IN SILVER: Silver IRA Investing