Is Tether and Bitcoin A Ponzi Ecosystem?
Tether and Bitcoin A Ponzi Ecosystem. As Tether corporation continues to obfuscate the identity of the companies whose commercial paper constitutes the majority of its $75 billion in reserves, yesterday the increasingly excellent Protos reporters published results of their long, deep dive into the activities of the largest tether wallets on all the blockchains carrying tether.
Their main findings are summarized in a scientific and unbiased manner, with the disclaimer: “It must be stressed that Protos is not explicitly alleging any wrongdoing on behalf of any of the entities detailed in this investigation.” Protos also “understands” that Tether issues USDT in return for overcollateralized loans in bitcoin, thus hinting at the mysterious origins of their commercial paper.
I read Protos’ Tether papers report yesterday with much interest, before it was edited down to remove previously unreported time series charts of massive in-flows and out-flows of the tether wallets of the main players; and to delete their discussion of how these flows might be coordinated with bitcoin price pumps.
The Protos report sheds further light on Tether’s collusion in (what I am not alone in calling) the bitcoin Ponzi scheme. The flow chart above depicts the process as I see it. Note, this diagram is purely the result of a logical argument applied to my knowledge of crypto market microstructure and the information in the Protos report.
Like Protos, I am not “alleging any wrongdoing on behalf of any of the entities” I use in the diagram below to explain my logic. Indeed, no actions of any party are breaking any laws. It is up to regulators to take responsibility. So I hope the policy makers for crypto markets (whoever they are) read this blog.
Okay, looking at the diagram again, first consider the companies in this ecosystem (see Key). Along the bottom (grey) we have the main users of tether. The large funds and professional crypto traders identified in the Protos report. Then we have: the Tether corporation (green); on-chain liquidity pools such as Uniswap (pink); the Binance Asia exchange, (yellow); and a regulated exchange such as Coinbase which trades bitcoin for fiat currency such as US dollars (blue).
Beginning at the lower left-hand corner, we suppose the bitcoin dollar price starts at $10,000. And that the tether user XYZ holds 100,000 bitcoins, currently worth $1 bn.
Tether Crypto Props up Toxic Crypto Carousel
- XYZ sends 100 BTC to Tether corp. as collateral for a loan but receives less than 1 million USDT in return (the loan is over-collateralised). XYZ still holds 99,900 bitcoin.
- Suppose XYZ receives 90% of its deposit, say 900,000 newly minted coins, issued by Tether. It onboards these to Binance (it can’t onboard US dollars because Binance doesn’t trade any fiat).
- Why Binance? Almost all the price and volatility leadership of BTC comes from Binance – see my previous blogs with these titles. It used to be BitMEX (the other exchange identified as a major tether user in the Protos report) but for the last 12 months, highly-leveraged trades on Binance’s BTCUSDT perpetual have taken almost all the volume and have had by far the most impact on other markets for bitcoin and other crypto (see my other blog posts about price and volatility transmissions)
- Meanwhile, Tether earns interest on its collateral by providing BTC (and any other crypto used as collateral in return for USDT issuance) to an on-chain liquidity pool such as Uniswap
- XYZ uses maximum leverage on Binance. Maximum leverage used to be 100X but this was reduced to 20X after 19 May 2021. For the purposes of illustration, suppose XYZ uses 100X leverage to place massive limit order positions, with size up to 90 million USDT on a margin rate of just 1%, designed to pump up the BTCUSDT perpetual price on Binance and therefore also the spot price of BTC on Coinbase and other regulated crypto exchanges.
- Although XYZ runs one of the best HFT algorithms in the business they are up against stiff competition, as more and more advanced professional traders start using Binance. So, by the time BTC reaches 20,000 USDT and XYZ stops pumping on the perpetual, XYZ may have used all of its 900,000 USDT in auto-liquidated margins.
- In that case, XYZ have lost their 900,000 USDT so Tether gets to keep the 100 bitcoin. Tether doesn’t call in the loan because it is earning interest on Uniswap (or any number of other liquidity pool providers, there are now so many of them).
- Now XYZ only have 99,900 bitcoins left, but these would be worth $199,800 if they wanted to stop playing and cash them in for real dollars on Coinbase (or any other regulated exchange).
- And if they want to continue XYZ go back to step 1 and repeat the game. Sending Tether another 100 BTC. But now the price of BTC has doubled so they receive 90% of 2,000,000 [1,800,000 USDT] which is onboarded to Binance and once again leveraged up to 100X on the perpetual swap. The pump position size is now 0.18 billion USDT.
- By the time the perpetual price doubles again, to 40,000 USDT. Again XYZ may have had all their margins auto-liquidated by Binance. But this doesn’t matter, because they still have 99,800 bitcoins which are now worth $3.992 bn.
The depiction above is just to illustrate how Tether could end up with the majority of its reserves in bitcoin and other crypto. And, of course, how the major users of tether could be making vast profits through highy-leveraged trades that are so massive that they move the price of bitcoin any way they choose.
And assuming these major tether users hold a lot of bitcoin (and other crypto) they could be making huge profits if they follow my logic above.
For instance, in my totally hypothetical example, XYZ forfeited just 0.2% of their bitcoin to Tether because they lost the USDT that Tether minted for them to Binance in auto-liquidations. However, after using just two lots of 100 BTC in return for tether and using these coins to pump up the bitcoin price, if XYZ were to check out the remaining 99.8% of their original BTC holdings at Coinbase now, they would make a pure profit of $2.99 bn.
Written by Professor Carol Alexander
Professor of Finance at Sussex, a visiting Professor at Peking University HSBC Business School, and Co-Editor of the Journal of Banking and Finance. I have a BSc Maths with Experimental Psychology and a PhD in Algebraic Number Theory. I took a post-doc in Amsterdam, worked as a bond analyst for Phillips and Drew. And attended the London School of Economics as a Research Assistant in game theory/labour economics also taking their MSc in Mathematical Economics and Econometrics. In 1985 I took a lectureship in Mathematics and Economics at the University of Sussex. While also designing risk, pricing, hedging and trading models for investment banks, fund managers and exchanges.
Then I dropped the Economics side at Sussex to work half-time as Academic Director for Algorithmics Inc.
My PhD thesis, entitled Integral Bases of Dihedral Number Fields was supervised by Walter Ledermann at the Universty of Sussex. After a post-doc at the Universty of Amsterdam. A rather tortuous year as a bond analyst for Phillips and Drew. In addition, a delightful Masters in Mathematical Economics and Econometrics at the London School of Economics. I returned to Sussex as a lecturer in Mathematics and Economics. That was in 1985, by which time my research interests had turned to game theory.
However, after the 1987 Black Monday crash in global financial markets my econometric skills were in greater demand. And my social conscience drew me away from game theoretic research into something more practical. I undertook various consultancy roles for investment banks and other financial institutions. Where I worked with computer programmers to implement models for risk analysis and portfolio management. This way, I became drawn to research in financial risk management. Investigating the properties of various new econometric models for market risk. Including different types of generalised autoregressive conditional heterscedasticity (GARCH) models. As well as applied research on active and passive fund management. From that time on, almost all my research has been with the wonderful PhD students that I have had the privilege to supervise.
In 1997 I left academia entirely to be a director of Nikko Global Holdings and Head of Market Risk Modeling (UK).
I briefly led a team of about a dozen PhDs. We designed and built new indexing products, but the London office closed shortly after I started. I took the opportunity to write my first book (Market Models, Wileys 2001). Then, in 1999, I became a professor of finance at the ICMA Centre at Reading University. I was also Risk Research Advisor, SAS (USA) and Chair of the Board of PRMIA (Professional Risk Manager’s International Association). In 2013 I returned to Sussex, heading the development of the Department of Business and Management. Before it split into three to become the new Business School.
Further econometric research on estimation of general discrete-time stochastic processes for financial asset returns naturally shifted my attention towards the implied measure. At which point I necessarily became a rather inefficient autodidact in various elements of mathematical finance. In this sphere I developed pricing and hedging models for various types of options, exotic and otherwise. And with two very talented PhDs we proved some classic theoretical results on scale invariance and generalised aggregation properties.
Likewise, more applied mathematical finance research converged on volatility indices. And higher moment risk premia, and on trading these premia through futures and exchange traded products.
While writing my 4-volume textbook Market Risk Analysis (Wileys, 2008). Walter Ledermann read parts of the first volume Quantitative Methods in Finance. In his early career as a young mathematician in Edinburgh Walter had done some interesting research on correlation matrices. And after reading my textbook he proved the last theorem of his life at the age of 97. By coincidence, I was supervising the PhD of his grandson at the time. Together Dan, Walter and I wrote the first paper on random orthogonal matrix (ROM) simulation. Timing the publication for the centenary of Walter’s birth. In his honour, we named the Ledermann matrix that Walter discovered. As the first in a whole class of L-matrices that Dan and I developed. I continue to work on ROM simulation. In addition, have also developed another new type of simulation model based on factor quantiles.
Currently, the main focus of my research is on the exciting new universe of crypto assets and their derivatives.
