To the people he met along the way, Ken Griffin may have seemed like a man in a hurry. Also, a boy in a hurry.
In 1980, for a school assignment, the 11-year old Griffin proposed to research the stock market. In loopy, neatly spaced handwriting, he described his methods – he would speak with a broker, refer to encyclopedias and other books – and summed up his aims: “to explain how it works and how to read about the stock market in the newspaper.”
Seven years later, Griffin was at Harvard, where he famously persuaded the school to let him install a satellite dish on his dormitory, so he could trade from his room. Soon afterwards, he was running a convertible bond arbitrage fund – while studying.
Where did this zeal come from? In investor letters and occasional interviews, he has cited his grandmother as an inspiration. Genevieve Gratz took on the family business after the death of her husband, batting away a group of bankers who had hoped to buy it cheaply, and she later provided capital Griffin used to set up his first institutional trading account. But he had no connection to the world of finance; he was born in Florida, and none of his family worked in the markets.
Lloyd Blankfein – the former Goldman Sachs chief executive, and postal worker’s son – knows something about being a self-starter. He describes Griffin as “a person who kind of created himself. He lifted himself by his own bootstraps. Most of the people I know, who have developed in the markets, I can see where their training was, I can see the mentorship – how they developed into who they are. But Ken started his business in a dorm room in college. So, not after he graduated – in college, in his dorm room, he started doing this. Who was thinking about that stuff then? He was.”
Griffin himself simply says, “I always had an interest in the stock market”.
As evidence, he cites that fifth-grade research proposal, which he keeps by his desk. “It’s a work in progress,” he jokes. “I still don’t fully understand the stock market.”
Maybe not. And in the aftermath of the GameStop drama (more on that later) who does? But the path the 11-year old Griffin started down has since led him to a unique place in finance. Citadel, the hedge fund he founded three decades ago, is among the most successful of all time. Citadel Securities – the market-maker that sits alongside it – makes the institution one of a kind.
The story of LTCM is one I think every market practitioner really needs to studyKen Griffin, Citadel
Although Griffin rarely gives interviews, many of the firm’s landmarks are well-known. As Citadel has grown in size and prestige, commentators have picked over its best and worst years, and its boldest trades – such as the acquisition of stricken hedge fund Amaranth’s energy portfolio in 2006.
But trades and strategies come and go. The more interesting question is the secret of Citadel’s long-term success. “That’s the remarkable thing,” says Blankfein. “To succeed over the long term, virtually uninterrupted.”
The firm would not comment on performance figures, but reports put Citadel’s net gains at $41.8 billion since inception, with its flagship fund generating an annualised 19% return.
Lessons from LTCM
Arguably the best-known chapter in Citadel’s history is also the worst-understood. In 2008, the firm was rumoured to be spiralling towards liquidation – it delivered a loss of 55%, and gated its funds to prevent redemptions.
Griffin has repeatedly accepted the blame for this. In this interview, he refers to it as the firm’s “day of reckoning”, and admits he did not understand “how precarious the banking system was”. It was his “low point”.
By the end of 2009, though, Citadel had bounced back, delivering a 62% net return to its investors and removing the gate.
It’s often presented as a redemption story – and it is – but that places all the emphasis on Griffin’s determination to make good, on the firm’s grit, its desire to shake off what was almost a knock-out blow and get back on its feet.
In fact, there’s a less-hackneyed tale to tell about how Citadel survived that blow, which is rooted in the hard work, and hard lessons, that followed a trio of earlier crises – the Black Monday stock market rout in 1987, the rate-hiking cycle of 1994 and, somewhat weirdly, the bail-out of Long-Term Capital Management (LTCM).
Weird, because Griffin lauds the failed fund for its resilience.
“The story of LTCM is one I think every market practitioner really needs to study,” he says. “One of the observations we took away was their ability to withstand unimaginable losses. They lost 90% of their equity capital before they were effectively seized by the banks. That staying power was a real testament to their risk management, to their treasury function, their counterparty management, the documentation they used for their derivatives contracts – they did an incredible number of things right.”
Griffin didn’t learn these things by reading books. He did it by bringing in more than a dozen of LTCM’s staff for interview, following its near-demise. This was a deliberate tactic.
“I wanted – I desperately wanted – to know their story,” he says. “I wanted to know how they did it. I also wanted to understand what didn’t work.”
They did it by paying for term funding, in a variety of forms. LTCM was also “incredibly thoughtful on their contractual terms”, says Griffin.
He is referring to the master agreements and collateral agreements published by the International Swaps and Derivatives Association, but also repo agreements and prime brokerage agreements – all of which can be a minefield for the unwary.
To illustrate, Griffin points to the treatment of a failure to deliver a US Treasury in a repo contract, a common event that can spike in times of stress – bonds worth $110 billion failed to be delivered on March 12 last year, for example. Ordinarily, the root cause is operational and the transaction is settled after a short delay, but weak docs can allow a fail to be treated as an event of default. Similarly, Isda contracts can allow a dealer to terminate its derivatives with a hedge fund if its net asset value falls by a certain amount over a certain period of time, potentially forcing the fund to pay the current mark-to-market on the portfolio.
Taking his learnings from LTCM, Griffin tried to ensure Citadel’s counterparties would not be able to strangle the firm in a crisis.
“Once we understood how much focus LTCM put into the terms of their derivatives and funding documentation, we absorbed that focus into what we were doing. If we’d been sloppy, and used the market-standard documentation terms used by many of our contemporaries in 2008, then you and I wouldn’t be having this conversation today,” he says.
“We lost a third of our capital”
Other reinforcements date back to 1994, when the Federal Reserve hiked rates six times, starting in February. Two months later, Askin Capital was out of business as its mortgage-backed securities collapsed in value. Heavy losses followed for other funds, and Citadel was caught in the crossfire.
Bryan White was one of Citadel’s first investors – and for some time its largest – at fund-of-funds house Quellos Group. He recalls what the crisis was like from that vantage point: “Most of the capital at that time came from fund-of-funds that promised their clients monthly liquidity. When you had a negative performance surprise, people ran for the fences. They all scrambled to put in redemptions. It wasn’t a matter of optimising the portfolio – it was a matter of indiscriminately taking money from whatever source was available to take money from. That was ’94,” he says.
Citadel was one of those sources, and it was a scary moment for Griffin’s four-year-old firm.
“We lost a third of our capital base to withdrawals at a moment in time when we were actually flat for the year from a trading point of view,” Griffin says. “Really, I saw my career almost coming to an end, not because we’d had significant losses, but because our contemporaries had significant losses.”
This posed a particular threat to Citadel, which was expanding from its convertible arbitrage roots into a range of other spread, convergence and relative value-type strategies – trades that may have an effective duration of months or even years, and can’t be rapidly liquidated without wrecking the strategy. Citadel had an asset/liability mismatch.
Everybody loves a David-and-Goliath story. But we have better predictive analytics, better risk management techniques, and we believe that – on an open architecture in transparent and fair markets – we can be a very successful competitorKen Griffin
In principle, there was a simple way to manage that particular risk: Citadel could alter the profile of its liabilities by obtaining longer-term commitments from its investors. In practice, the firm needed its investors to agree, and they were unlikely to do so in the aftermath of 1994.
So Griffin waited. Four years later, with more positive years under the firm’s belt, he went to investors with the same choice they get today. They could either commit to a minimum of two years, or have quarterly liquidity – but with a charge payable to the fund, if they chose to redeem.
Changing those terms came at a short-term cost to Citadel, Griffin recalls. Some investors told him to “go pound sand”, as he puts it today – they took their money elsewhere – but Citadel stuck to its guns.
“We’d made a strategic decision that we would rather shrink the firm to a smaller base of longer-term capital – more stable capital – than be at risk of hot money flying out when markets really have a moment of turmoil. Opportunities arise in such moments,” he says.
Griffin didn’t spend long wondering whether he’d made the right call. The exercise was completed just weeks before the LTCM crisis.
Throughout these early years, Griffin was guided by a conviction formed when he was trading from his dorm room on October 19, 1987. He still remembers the innocuous stories that led the news that morning – Nancy Reagan’s recovery from a cancer operation; tension in the Persian Gulf. There was no warning the US equity market was about to lose more than a fifth of its value.
“To this day, that underpins much of how we think about risk management – it’s what we do before the event, because once it starts it’s like a bolt of lightning,” says Griffin.
In essence, Griffin and Citadel spent more than a decade preparing for the bolt of lightning that hit the firm in 2008 – locking down funding, closing out documentation weaknesses, bringing in more long-term investors that matched his own beliefs about where opportunity lay in the financial markets. That’s what enabled the firm to survive the storm, and to rebound afterwards. Griffin’s redemption story is best-understood as a triumph of long-term vision and planning.
It also gives the lie to the idea that risk management is about living in a defensive crouch, worried the sky might cave in. An arbitrageur’s biggest opportunities are when markets are most distressed – the cruel twist is that those opportunities can be placed out of reach by expensive funding, edgy investors or wary counterparties. Griffin wanted Citadel to be more resilient not to avoid losing money, but to make it.
Dark side of the moon
That attitude can be seen throughout Griffin’s career. It lies behind the Amaranth deal, Citadel’s acquisition of the Sowood Capital credit portfolio in 2007 and – most recently of all – its $2 billion investment in Melvin Capital, the most public casualty of the GameStop squeeze.
Risk.net interviewed Griffin in December, when GameStop stock was still trading at less than $20. On January 25, it closed at $76.79 as an army of retail traders sought to “send it to the moon”. The rapid rise inflicted heavy losses on short-sellers including Melvin, prompting Citadel and Point72 to stabilise the wounded fund.
Three days later, all hell broke loose. Robinhood and other brokers that had been the conduit for the squeeze abruptly limited their customers’ ability to add long positions. The internet pointed an angry finger at Citadel Securities, which competes for – and wins – a large share of Robinhood’s flow. It also pays the broker for every trade it receives. Was Griffin’s market-maker using its clout to give his hedge fund’s new investment some breathing space? In fact, Melvin is reported to have closed out its positions two days earlier, on January 26. But politicians took up the issue, with Maxine Waters – chair of the House Financial Services Committee – naming Citadel on a wishlist of firms she would like to testify about the drama. A hearing has been scheduled for February 18.
I was in my early 20s and one of my partners said to me, “You’re already climbing one great mountain – what do you do when you reach the peak?” And I thought to myself: “I’ll climb another mountain!”Ken Griffin
Citadel and Citadel Securities have both denied they had any role in the brokers’ decision to suspend or limit trading. For this article, Griffin emphasises the investment rationale for buying into Melvin – launched by Gabriel Plotkin in 2014. Like so many other Citadel trades, it was a chance to take a position cheaply in an asset that might be expected to rebound – in this case, a portfolio of dislocated stocks, and a fund that had grown rapidly prior to its tangle with the day traders.
“Gabe Plotkin has been one of the most successful equities managers of his generation. He and his team have delivered exceptional results over the history of Melvin,” Griffin says.
Robinhood’s own explanation for its decision is simple: it didn’t have enough collateral to allow unfettered trading to continue. In a February 10 affidavit filed in one of the many lawsuits it now faces, Robinhood’s chief operating offer, Jim Swartwout, says the firm imposed its limits after receiving a $3 billion daily margin call from the clearing house that handles most US equity trades – this was “an order of magnitude above typical levels”. The broker was able to reduce the total by limiting trading of GameStop and other high-volatility stocks.
The affidavit also describes speculation about Griffin’s firms as “completely false”. No third party was involved in the decision, Swartwout’s filing states.
The GameStop drama focused the world’s attention on the twin arms of Griffin’s institution. Citadel Securities may be legally and operationally distinct from the hedge fund, but the two outfits share a front-foot ethos.
In recent years, Citadel Securities has become the first non-bank to build an interest rate swaps business; it is also present in foreign exchange and off-the-run US Treasury trading. It has the biggest designated market-making franchise at the New York Stock Exchange, and has taken a string of companies public, bypassing the traditional IPO process. It dominates the market for stock trading and stock options in the world’s biggest equities market.
In many of these businesses it competes with the traditional incumbents – the banks – and Griffin’s upstart market-maker is often presented as a gadfly. Griffin doesn’t see it that way.
“Everybody loves a David-and-Goliath story. But we have better predictive analytics, better risk management techniques, and we believe that – on an open architecture in transparent and fair markets – we can be a very successful competitor. Certain institutions that are, generally speaking, distribution-led, and have tended to be very good at opaque products, don’t want to have to compete in that world,” he says.
But that critique of the banks is also less accurate than it used to be, he notes, citing a generational shift within the industry.
“The younger generation that is now running the banks is much more culturally aligned with where we were 15 years ago – and, frankly, some of them are phenomenally good competitors. And who has won? The buy side has won,” he says.
Making black boxes white
Griffin has won too, of course. In his 2010 letter to investors, Griffin started by describing his vision for the firm. He wanted “to create an institution unlike any other – a firm that harnessed technology and quantitative methods to uncover unique opportunities.”
That large, ambitious item at the top of the Griffin to-do list can be scratched off. It’s done. What does he still want to achieve?
The question elicits a long pause. “This brings back memories,” Griffin says. “I was in my early 20s and one of my partners said to me, “You’re already climbing one great mountain – what do you do when you reach the peak?” And I thought to myself: “I’ll climb another mountain!” I don’t think I have a list of what I have left to accomplish. The question is, what motivates an individual, and for me – I love to learn.”
There is plenty to learn about. The forces Griffin sought to harness 30 years ago – technology and quant finance – have not been standing still. Cloud computing provides on-demand access to vast reserves of number-crunching power; electronification is creating new, and larger, datasets through which firms can search for hidden rules and relationships; and machine learning algorithms can, in some cases, do a better job of spotting those relationships, and deviations from them.
These models are literally black boxes, even to the developer of the model – so you do lose, if you’re not careful, actual comprehension of what is taking place in the marketKen Griffin
Citadel has embraced these developments – over the past 10 years, the firm has doubled the number of engineers it employs, from roughly 300 in 2010, to more than 600 today. The number of engineers embedded directly in investment teams has increased fivefold in the past five years.
As an example of what this means in practice, Griffin says his new, more tech-adept teams might use machine learning to spot interactions between different return forecasts – these could be missed when using traditional techniques, such as linear regression, to model a relationship.
Using algorithms in this way might help Citadel avoid making a mistake, but it also bothers Griffin – he sees it as “dissipating the acquisition of true knowledge”. Many of us are content to take our cars to a mechanic when there’s a problem, but fund managers can’t do that with a busted forecast – they need to understand why it was wrong.
“These models are literally black boxes, even to the developer of the model – so you do lose, if you’re not careful, actual comprehension of what is taking place in the market,” he says.
Citadel tries to be careful. If an ML algorithm is able to show the firm something interesting, Griffin says “we really try to get our heads around what’s taking place. Let’s take a step back: why do we think this model is showing this prediction? And – this has been a huge area of research recently – how much of this is literally just data-mining, in the sense that you’ve ended up with a prediction that has no out-of-sample application?”
On the face of it, the spread of machine learning, big data and the cloud could be bad news for Citadel – acting as a levelling force, making it easier for new entrants to make good decisions, quickly.
Griffin recognises the danger, but has a ready answer: “It’s very simple. As long as we can sustain our advantage in predictive analytics and risk management, and continue to extend those advantages, we will be prosperous. But if we rest on our laurels, if we take a victory lap – then somebody else will come from behind and they will take our position.”
Quellos founder White would see that response as classic Citadel. In his experience, the best hedge funds are never comfortable, never complacent: “In order to survive, they have to be paranoid that they are about to get eaten by a bigger animal. It’s a common characteristic.”
The next 30 years
There is one moment in the interview when Griffin briefly falters. It’s a question about succession at Citadel. He pauses. His tone changes.
“You know, I don’t look forward to succession in the context that it means I’ve – bluntly – gotten older and maybe it’s my time to pursue other interests and other passions. I don’t relish that day. But I am confident, when I look across the team that is here, my colleagues will unquestionably do a great job of running Citadel when that day comes,” he says.
At its senior levels, Citadel is a mix of seasoned executives the firm has poached from around the industry, and those who have grown with the firm – such as Peng Zhao, the Citadel Securities chief executive, who joined Citadel from grad school, helped build the market-making business, and took the reins at the age of 32.
This is no accident. Citadel promotes from within – and promotes young – because it’s a way of retaining talent and creating stability.
Griffin talks about the “long runways” these individuals have, once they’re in those roles. The same kind of runway Griffin himself has enjoyed, of course.
Blankfein has known him for much of that three-decade run. First, their respective firms were counterparties; as Citadel grew, it became a more important customer to Goldman. As Citadel Securities pushed into derivatives and bond trading, it became a rival.
When he was leading Goldman Sachs, Blankfein would speak with Griffin several times a year – they speak just as often today. And in 2014, Blankfein approached Griffin with the suggestion that he make a donation to Harvard. The $150 million gift became the largest in the university’s history, primarily to give smart kids from poorer backgrounds access to top-tier education (Blankfein says “his arm didn’t need to be twisted; he gave more than I asked for”).
Blankfein describes Griffin as a friend and – as a friend would – he takes umbrage at the suggestion that the trends Griffin has ridden might be dissipating. He makes two points. First, that Griffin and Citadel didn’t just ride those trends, they also anticipated them, and made them real – “he wasn’t just a cork bobbing on a stormy sea; he was involved, he influenced it.”
Second, he says, markets have always changed. Griffin and Citadel will change, as well.
“Five thousand years ago, there were traders – people who traded spices, textiles. A thousand years ago, the Vikings were trading all kinds of things. And I’m telling you – 2,000 years from now, there will be people who are good at this and those who are not so good. I don’t think Ken’s strength is freeze-framed in this moment of time; I think his strength is his capacity to affect the market, to adapt to changes,” he says.
Griffin isn’t looking quite that far ahead, and he trusts the current crop of Citadel employees to carry the firm forward.
“The organisation is very blessed to have had a strong 30 years of existence but is more fortunate to have individuals who will be here 30 years from today, continuing to drive one of the pre-eminent firms in finance,” he says.
Don’t count out the possibility that Griffin will still be driving it, too.
Only users who have a paid subscription or are part of a corporate subscription are able to print or copy content.
You are currently unable to print this content. Please contact [email protected] to find out more.
You are currently unable to copy this content. Please contact [email protected] to find out more.
Copyright Infopro Digital Limited. All rights reserved.
You may share this content using our article tools. Printing this content is for the sole use of the Authorised User (named subscriber), as outlined in our terms and conditions - https://www.infopro-insight.com/terms-conditions/insight-subscriptions/
If you would like to purchase additional rights please email [email protected]
Copyright Infopro Digital Limited. All rights reserved.
You may share this content using our article tools. Copying this content is for the sole use of the Authorised User (named subscriber), as outlined in our terms and conditions - https://www.infopro-insight.com/terms-conditions/insight-subscriptions/
If you would like to purchase additional rights please email [email protected]
‘Parallel stacks’ proposal unlikely to appear in first draft of CRR III, due next monthReceive this by email