Three Arrows Capital, Long Term Capital Management and George Soros all have something in common.
They are making massive one-directional bets, but their outcomes were dramatically different.
History is written by survivors and as such, survivorship bias colors the lens by which we view and perceive things that have happened.
In the summer of 1998, when most self-respecting hedge fund managers were playing with their yachts, the Nobel Prize-winning economists and renowned Wall Street traders at Long Term Capital Management had other things on their mind.
Instead of being out on the water, Stanford University’s Myron Scholes and Robert Merton of the famous Black-Scholes-Merton options pricing formula, were betting that the world had overpriced volatility and were buying options that their models said would never be exercised.
An option is a right to buy or sell a security at some future date at a predetermined price.
Say for instance the price of an Apple stock is $100 today, and an investor believes that it will be worth $200 in a year’s time.
What if it was possible to have the option to buy the Apple stock at $100, but in a year’s time?
If the investor was right about their bet, they would be able to exercise that option, buy the Apple stock at $100 a year from now, and sell it for $200.
Their profit would be:
Profit = $200-($100+Price of the Option)
If Apple’s stock was below $100, it wouldn’t matter anyway, and the investor who bought the option would just lose the money they invested in buying the option to begin with.
The challenge of course is, how much should the option cost?
Which is where our story returns to Merton and Scholes who used quantitative mathematics to derive a complex options pricing strategy which they believed was an infallible source of risk-free returns:
Because markets were (and still are) full of people who have no idea how to price options, Merton and Scholes, founders of Long Term Capital Management, made tons of money by selling options that would never be exercised.
In the options business, the one selling the options that never get exercised makes money because the buyers of these options guess wrongly about outcomes, while Long Term Capital Management calculated right.
Long Term Capital Management didn’t just make money by selling options, it also snapped up what Merton and Scholes had identified as mispriced securities using sophisticated mathematical models.
But in order to generate stellar returns, Long Term Capital Management (“LTCM”) had to deploy leverage, and lots of it, allowing them to bet more than just their own money.
And by August 1997, LTCM’s capital, which was just shy of US$7 billion, had assets funded by borrowing amounting to US$126 billion.
While most investors would have shuddered at the thought of carrying such massive amounts of leverage, Merton and Scholes saw no risk according to their models.
And more importantly, since LTCM was pursuing supposedly multiple uncorrelated trading strategies, around a hundred in total, with over 7,600 different positions surely not all of them would go wrong simultaneously?
Which is why the quants at LTCM saw no risk in carrying the levels of leverage that they were shouldering.
LTCM was trading in markets all over the world, but their biggest business by far was selling options in the U.S. and European stock markets, options that would be cashed in if there were big future stock price movements either up or down.
In the late summer of 1998, the high premium that options were fetching implied that markets would become highly volatile, but LTCM’s models suggested that this view was incorrect.
According to LTCM’s calculations, volatility would actually decline and that meant that the chances of options being exercised would be low.
So LTCM piled the options high and sold them cheap, calculating that the odds of the firm going bust was 1 in 10²⁴ (that’s 10 with 24 zeros behind it), or virtually zero.
However by the late summer of 1998, something happened that threatened to blow the lid right off LTCM’s otherwise failsafe money-making models.
In evolution, big extinctions tend to happen because of outside shocks, like an asteroid hitting the earth blindsided many a dinosaur.
On Monday, August 17, 1998, a giant asteroid smashed into planet finance, and it struck from the other side of the world, in an especially flaky emerging market.
Weakened by political upheaval, declining oil revenues, and a botched privatization drive, the ailing Russian financial system collapsed.
A desperate Russian government was driven to default on its sovereign debt, fueling the fires of volatility throughout the global financial system.
Stock markets plunged.
Remember all those low cost options LTCM had sold, based on their prediction of low stock market volatility?
The ones they thought that no one would exercise?
Well now they did, in droves.
LTCM had calculated that they could not lose more than US$35 million on a single day, but on Friday, August 21, 1998, the firm lost US$550 million, or around 15% of its entire capital.
By the end of the month, LTCM was down 45% and because selling options is a form of implied leverage, it required a concerted effort to bailout the firm.
Fearful that LTCM’s collapse could trigger a broader market meltdown, the U.S. Federal Reserve brokered a multi-billion dollar bailout by 14 Wall Street banks.
LTCM had placed their bets, levered up, and ultimately got it wrong.
But had they been right, today LTCM could have been celebrated as one of the most successful hedge funds in history, having mathematically solved the markets.
Similarly, Three Arrows Capital, a tiny hedge fund that cut its teeth not in cryptocurrencies, but with US$1.2 million in 2012 trading emerging market currencies, would go on to become a massive player in the nascent digital asset industry by making the right bets before leverage finally blew them up.
As digital assets started to gain more traction in 2016, following the successful initial coin offering of Ethereum, Kyle Davies and Su Zhu cofounders of Three Arrows Capital, identified the same pricing gaps in cryptocurrency markets that had become increasingly elusive in currency markets.
Having correctly called a bottom for cryptocurrencies in 2018, Three Arrows Capital (“3AC”) became a major player in the digital asset industry, and one of its “risk-free” trades was taking advantage of the “Grayscale Premium.”
The Grayscale Bitcoin Trust or GBTC, allows investors who can’t or don’t want to hold Bitcoin directly, to instead buy shares in a fund that invests in them.
Because GBTC was one of the only regulated cryptocurrency products for a long time, it cornered the market on institutional exposure to Bitcoin and became so popular that shares of GBTC traded at a premium to the value of the Bitcoin held by the Trust.
Hedge funds such as 3AC capitalized on the “Grayscale Premium” because GBTC made it possible to purchase shares in the Trust by pledging Bitcoin and receiving shares in return.
3AC would simply use Bitcoin and pledge that to GBTC, receive the shares of GBTC (that could only be sold after a lock-up of 6 months) and sell the shares of GBTC later on to realize the premium.
For instance, if the price of spot Bitcoin was US$50,000, the GBTC share price could have been US$60,000, 3AC would simply obtain spot Bitcoin, GBTC would issue shares for that Bitcoin, and 3AC would sell shares of GBTC later on to realize that premium.
Armed with this “risk-free” way to generate serious money, by December 2020, 3AC had become the largest holder of GBTC, with a position worth US$1 billion at the time.
If the price of Bitcoin kept going up, the premiums would keep growing, and 3AC would keep minting money.
But 3AC didn’t just pledge Bitcoin it owned in its own right, it also borrowed as much as it could lay its hands on, wagering that as long as the cost of leverage was less than the GBTC premium, it would still make a “risk-free” return.
But that strategy hit a snag by early 2021, as GBTC, facing stiffer competition from a bunch of similar institutional-grade Bitcoin products, including the ProShares Bitcoin Strategy ETF, a CME Bitcoin futures-backed exchange traded fund, saw the “Grayscale Premium” turn into a discount.
As the GBTC discount deepened, 3AC’s position started to grow increasingly tenuous and the 6-month lockup on GBTC shares became a major problem, but not the only one.
3AC didn’t just borrow Bitcoin to bet on the durability of the “Grayscale Premium,” they borrowed from a range of other specialty cryptocurrency lenders to wager on everything from algorithmic stablecoin TerraUSD and its sister token Luna, to staked Ether or stETH on the Lido Finance platform.
Believing that they had divined countless arbitrage opportunities that looked to be ways to collect free money, just like Merton and Scholes, Zhu and Davies piled up the leverage high and took massive bets.
Because the cryptocurrency markets are tiny compared with traditional finance, for years, prices of tokens had little correlation to macroeconomic conditions.
But as institutional investors started to trickle into the cryptocurrency space, interest rates and other macroeconomic factors, long ignored by insulated cryptocurrency traders, were becoming increasingly relevant to price movements.
By November 2021, when the U.S. Federal Reserve realized that inflation was anything but “transitory” and was forced to raise interest rates, planet finance finally collided with planet crypto, taking the latter down with it.
Almost overnight, the value of cryptocurrencies plummeted, the value of a 3AC’s investment into Luna, which was worth US$200 million at one point, spiraled to around US$600.
Remember all of that borrowing that 3AC took on, betting that the markets were in a temporary slump and would rebound sharply?
Those bets failed to consider the weight and influence of the U.S. Federal Reserve on risk appetite, especially for cryptocurrencies.
And as investors headed for the exits, many of the leveraged bets that 3AC had taken found few buyers, liquidity dried up overnight and one of cryptocurrency’s most iconic hedge funds imploded, taking down a string of other firms with it.
But had Davies and Zhu got it right, today, they would be celebrated as heroes, likely sipping champagne and plying the Mediterranean on their US$50 million superyacht, a SanLorenzo 52Steel no less.
Which brings us to our final story, that of legendary macro hedge fund investor extraordinaire, George Soros.
In 1992, Stanley Druckenmiller, a shaggy-haired Pittsburgh native who had dropped out of a PhD program in economics and who was now in charge of George Soros’s billion-dollar Quantum Fund, walked his boss’s office.
Druckenmiller wanted to slowly expand the Quantum Fund’s existing wager against the British pound.
At the time, the British pound was tightly linked to the German deutschmark, via the European Exchange Rate Mechanism.
But the Bank of England and the Bundesbank, the German central bank, were pursuing increasingly divergent monetary policies to cater to their disparate economic conditions.
Scarred by the high levels of inflation that had been a precursor to the Second World War, Druckenmiller had a strong belief that the Germans were obsessed with keeping inflation in check and would not let the mark go down.
Meanwhile, the Bank of England had to pursue a far looser monetary policy to stimulate Britain’s moribund economy and stave off a recession.
So by the late summer (as so many of these things happen to occur when hedge fund managers return from their summer vacation) of 1992, Druckenmiller told Soros that authorities in Britain were bound to break from the European Exchange Rate Mechanism and allow the pound to fall in value, which would help the country emerge from recession.
His eyes lighting up, Soros didn’t just agree with Druckenmiller, he recognized that such bets only come once in every two decades or so and wanted to pile on the leverage.
Soros had brought to global finance a brand new theory of economic behavior that underlined the fallibility of human nature and the inherent instability of financial markets.
According to Soros’s theory of reflexivity, markets can’t possibly be perfectly efficient, much less rational, for the simple reason that prices are just the reflection of the ignorance and the biases of millions of investors.
Because Soros recognized that our actions have unintended consequences that will not correspond to expectations, he implored Druckenmiller to have the Quantum Fund sell short about US$10 billion of the British pound on assets of just US$400 million, the biggest bet of Soros’s (and Druckenmiller’s) life.
But so sure was Soros that the pound would drop that he was convinced the risk was disproportionate to the reward and therefore a good speculation.
Rival hedge funds, learning what was happening at Quantum Fund, or independently arriving at the same conclusion as Soros and Druckenmiller, started piling on short positions against the pound as well, exerting pressure on the Bank of England.
By September 16, 1992, the Bank of England finally buckled, abandoning efforts to prop up the pound and sending it plummeting over 20%, earning Druckenmiller and Soros over US$1 billion in just under 24 hours.
For over a decade thereafter, Druckenmiller and Soros’s trade would be considered the greatest ever, a testament to how much money can be made with equal doses of savvy and moxie.
What ties Druckenmiller and Soros, Zhu and Davies, Merton and Scholes together is that they had equal parts savvy and moxie, but only Druckenmiller and Soros came out the opposite side as winners.
It’s easy to say with the benefit of hindsight that the collapse of algorithmic stablecoin Terra, and its sister token Luna, could have been foreseen, or that the “Grayscale Premium” would eventually evaporate.
It’s clear now that had Merton and Scholes fed 11 years worth of data into their risk models instead of 5, they could have captured the 1987 stock market crash, and had they gone back 80 years, they’d have captured the last great Russian default, after the 1917 revolution.
And if Soros hadn’t led a heavily-levered bet to short the pound, legions of traders in the multi-trillion dollar foreign exchange markets may not have noticed, nor piled into that trade that would outsize the meager reserves of the Bank of England and bring it to its knees.
Case in point is the most recent effort by hedge funds to short the Japanese yen.
With the Bank of Japan keeping interest rates low, foreign exchange traders piled into yen shorts, as the U.S. Federal Reserve raised borrowing costs and sent the dollar soaring.
For a brief few days in June, it looked like the yen might go into freefall before it suddenly staged a remarkable rally in mid-July, proof that there’s no such thing as a sure bet, no matter how sure a trader is.
It’s the job of traders to take bets and when convictions run high, moxie means that these bets are more often than not, juiced with borrowing because moments such as these can define a career.
But whether these bets end in adulation or anguish often hinge on things that are very much beyond a trader’s control.
Zhu and Davies could have been right.
Algorithmic stablecoins like TerraUSD could have ushered in a new era of decentralized currency, allegations of ponzinomics and fraud notwithstanding (all currencies are in some form or other prone to ponzinomics through inflation and debasement).
And Russia’s sovereign debt default could just as easily have been shrugged off by the market (as it has in recent times) as it could have set off the volatility that ultimately sank LTCM.
But such is the nature of trading and the only difference between domination and disaster is often only obvious in the retelling of history, tinged with a heavy dose of survivorship bias.