In this episode, David Greely welcomes Beau Taylor, J.P. Morgan’s Former Global Head of Energy Sales & Trading, to recount his experience in the U.S. Natural Gas market following Hurricane Katrina. Together they explore the risk miscalculations, market conditions, and elements of human nature that led to one of the largest hedge fund collapses on record.
Greely and Taylor revisit how a multi-strategy hedge fund transformed into one big bet on winter natural gas prices, why their collapse didn’t lead to wider systematic distress in the financial market, and what learnings we can apply to manage volatility in today’s LNG markets better.
The following Q&A is created using slightly edited excerpts from the episode transcript, optimized for readability. Download full transcript.
BT: Natural gas, at the time, was an emerging market that was in the process of attempting to mature. Natural gas trading began in earnest with a number of pipeline companies, such as Enron. They began to make markets and trade the contracts much more aggressively, but, at the time, you had a largely surplus market where you had excess supply. You ultimately had to go to a price where you incentivized power facilities to burn gas, to balance the market whenever prices would get too high. The inverse would take place, and you’d incentivize power plants to burn oil or alternative fuels. For a number of years, natural gas largely stayed in a relatively well-defined range, regardless of what the weather was, what happened with a hurricane or tropical events, or what happened in anything.
Hurricane Katrina was an interesting one because, I think, it caught people off guard. You had a scenario where you had a hurricane that largely looked like it was going to go into Florida, potentially as far north as hit land in Georgia or North Carolina. The market, not that they were sleeping on it, but the market had run up a little bit the weekend, on Friday before the weekend of Hurricane Katrina. The market sold off because it looked like a non-event. As soon as that happened, over the weekend, Hurricane Katrina hit the coast of Florida instead of dissipating. It made it to the other side of the coast of Florida, hit the warm waters of the Gulf, strengthened to an insanely strong hurricane, and started heading right towards the meat of the US natural gas market production.
At that point, anybody looking at balances for natural gas could see a credible scenario where the US simply would not have enough natural gas to get through winter. If enough production was damaged, if enough production could not come back online in the market, wasn’t able to fill storage, at the point that they’d anticipate that they needed to make it through the winter almost immediately. The price of natural gas doubled. You came home on Friday, and my natural gas is trading at $7 in MM BTU. When you woke up on Monday, natural gas was trading at $15 plus in MM BTU. The market was largely unglued because this was a scenario that the market hadn’t priced in. The market did not have a number for how you price the potential to run out of natural gas.
Natural gas was incredibly volatile for a period of time. I think what happened during that period of time, whether it be the risk models that a number of people managed to use to manage risk, whether it be a number of the training models, whether even be option pricing models, everybody kind of recognized at that time that there are these known unknowns that exist. As soon as it looked like natural gas was going to settle down, all of a sudden, Hurricane Rita came through. Right after that follows almost the identical track. It went a little bit further west. So while Katrina took out a lot of the stuff and the Central Gulf, Rita came through and took out a number of the things in the Western Gulf. I think you had a credible crisis situation that the market largely wasn’t ready for.
When you have an in-product, nobody’s prepared to turn off the heat during the wintertime. You’ve got to have some mechanism to balance that. Mechanism becomes price, but if that’s the case, the market has shown you that prices could go well beyond the realm of volatility that had ever been witnessed. Hurricane Katrina and Rita started a new era in natural gas trading, where all of a sudden, now we’re pricing for scarcity. We’re not pricing for abundance. Everybody had to utilize all the models, whether they’d be fundamental models, pricing models, or various technical models, which had to be readjusted. You were simply trading in a new era, and from that point on, the sky was the limit in terms of people’s beliefs, in terms of where prices could go, and how far away they could go.
Were they delusional? Who knows, but you began to have a number of people trading very differently than they had in the past. For instance, spread trading became very aggressively traded. People would begin to trade the front winter contracts and sell the contracts further back in the storage season by putting a price on how likely you were to run out of natural gas and, if it did happen, what would the risk parameters be. You started seeing similar things with option pricing models where calls became much more expensive than puts because, in a crisis situation, there wasn’t any known top for where the market could go. In risk models, people might have assumed that you had a certain amount of risk, short natural gas contracts, but that risk became much higher.
Even a modest amount of money going into Katrina, cutting across Florida, you lost insane amounts of money on Monday morning when you came back. Once again, a number of the traditional capital providers were forced to provide less capital simply because now we’re in a new regime, in a new era where prices were potentially going to go much higher. That was largely on the heels of the various merchants, whether it be Enron, Diana G, or some other people who had financial problems in the early 2000s. You had a scenario where you had fewer liquidity providers than you had in the past. A number of the banks had gotten into the business and were able to provide liquidity, but they were in a position where they couldn’t provide as much liquidity as they could in the past. They were looking at new risk models and riskier contracts. The market was calling for hedge funds to come in and provide additional liquidity. At that point, the only way to balance the market and liquidity was to find speculative capital willing to come in and take the other side of the commercial business. That would allow the markets to remain in balance, remain stable.
Katrina was the beginning of a new era that, in many ways, is echoing itself even today.
Traditionally spread trading commodities is largely a function of what the physical market looks like at the time. For example, if you were in deficit, you need to attract commodities out of storage, and the near-term contracts will trade above the future contracts. It’ll incentivize people to bring a physical commodity to market. Conversely, if you’re in a scenario where you have excess, you need to create an incentive for people to store that excess. So the front month contract will trade below the forward contracts because people will buy the prompt contract. They’ll put into storage, and they can maintain an arbitrage by selling those forward contracts. Generally, natural gas spread trading certainly wasn’t anything new. There were certainly lots of people who traded very actively across the curve.
The difference was that for a number of years, you had a relatively predictable curve shift largely dependent on current fundamentals. And so, whatever the market needed to incentivize, the curve would largely function as a way to incentivize what it required in the physical market. What hurricane Katrina and Rita showed was that if you have a scenario where you have extreme tightness – you could have a window where prices go substantially higher than anything you’d ever anticipated before. Therefore people became much more comfortable purchasing spread contracts beyond anything that might even have a historical basis. The market has shown that if something bad happens – you can have a payout on an event that’s dramatic. The other thing you have with spreads is there’s a lot of implicit leverage in spreads.
Whether it be the margin that you have to apply for futures, or contracts, the margin that you might have to apply for swaps because you have an offsetting buy-in sale, you can take a much larger position. Suppose you’re making a bet that gas will run out by the end of winter. In that case, you can make that bet with substantial leverage without having to put up as much capital as if you were going to make that bet. For example, buying the futures contract outright without that offset. I think for a number of reasons, a lot of people were becoming enticed to make these different bets. Plus, from a pricing standpoint, many people were looking at the model similarly, right? They’re looking at a model that had insufficient commodities to meet demand.
A number of people are pricing out this kind of hope, for lack a better term, that if something goes wrong, we will have a significant asymmetric payout. You mentioned Amaranth. They had made a significant amount of money because they’d correctly predicted that if things got tied, or if there was a tropical event, or if there was some other thing that occurred – prices could go substantially higher in one contract versus another. They were one of the first big winners on some of these spread trades. They’d been relatively ahead of the curve, specifically when you mention the risk and how it looks.
A bank doesn’t necessarily have to make all of its money and energy. They tend to be incredibly quantitative about a number of different things when they’re looking at risk, and a bank is going to be agnostic as to what the product is. I’m not saying they view everything as a widget, but they’ll certainly say that we’re going to normalize things to units of risk. For example, at a place like JP Morgan or any other big banks, they would typically say, if you’ve got significant volatility that’s occurred in the recent past, it’s going to factor heavily into that risk model. It’s going to create an impact. It will limit what people can take in terms of risk in the bank’s books and balance sheets.
The other issue with banks is that banks don’t necessarily have to be in energy. They’ve got lots and lots of different things they’ll attempt to make money on. So if they view, whether it be natural gas or any product, as being excessively risky, at any point in time, you’ll have liquidity, and that’s going to pull back out of that market. So following Rita and Katrina, investment banks had a way of calibrating risk that tended to be much more conservative. It tended to be limiting in terms of risk. It tended to take some of the liquidity out of the market, even as additional banks were getting in. The financial players, the non-bank financial players, such as the hedge funds and others, employed teams of incredibly bright people in the field of risk. A number of the banks and financial institutions felt like this could easily get over our heads. We want to make sure that we aren’t in a position to be adversely affected by these events.
They used sophisticated techniques to calibrate and determine what that risk may look like. Many of these people were very comfortable in how they looked at how they were calibrating. They were very comfortable in terms of the market’s liquidity and how they would view it going forward. So even then, you had an interesting dialectical where the banks were becoming more conservative. The non-bank participants were becoming more aggressive. The banks viewed this as excessively risky, and a number of the non-bank participants viewed it as highly opportunistic with scenarios that could throw off insanely high profits. If, for example, you had a scenario where prices did get incredibly tight, they were looking at asymmetric payouts in general.
They had profited from the Katrina and Rita debacle and were getting more aggressive as other people pulled out. They were coming in to fill the liquidity gap largely left by the non-bank participants or the bank participants, the pipeline companies, and others that had pulled out. At this point, you had teams of PhDs who’d become incredibly comfortable with risk metrics. The flaw in their risk analysis was that they assumed a traditional level of liquidity for the market, where banks aggressively grew the risk in the market. At a time, the banks were pulling back; traditionally, you had a number of the pipeline players and others, the physical players that were extremely aggressive, that had largely pulled back.
Marginal liquidity was being traded back and forth amongst the hedge funds. The belief was that the liquidity pool happened to be substantially different and deeper than you might have seen in the real world when things began to come unglued. I think that dichotomy between how the banks and financial players looked at it versus how the non-bank players looked at it. It gave rise to volatility, where most of the people passing futures back and forth between each other weren’t the people who had any intention of buying natural gas or selling natural gas in the physical market. People were largely making paper bets on the ultimate end of winter balances in natural gas.
You created a false sense of security. In some cases, you had echo chambers where the people looking at risk the same way were sharing ideas and comfort levels. There was a general belief that the market was in a new paradigm, and there was a new way to profit from the new paradigm. In contrast, there might have been a historical precedent, which may lead to more opportunity and put substantial capital behind those bets.
It’s always amorphous, attempting to figure out exactly who has what and what the ultimate scale was, but what you certainly see over time is that you’ll see a number of people who will take concentrated bets. And the general belief is that people believe they can be the market. People sometimes think they have a deep enough balance sheet to take on enough risk to wait out the other people coming in on the other side. For example, in a scenario like a March or April spread where the higher that market goes, the more physical incentives for different players who can store or withdraw from storage, to come in and make bets against. So when something gets out of whack, there’s this belief that the old rules never applied. Still, the physical players are substantially larger than any financial player.
And so, when you have metrics that create substantial incentives for people in the physical market to react and act differently than they might have, you’ll enhance liquidity coming into the market against your position. And it almost doesn’t matter how right you think you are in the short term; you’ll have substantial liquidity to balance the market from the physical players. Look at the situation like the March-April, where you had the spread up to a level where anybody that had any storage whatsoever was incentivized to sell their storage forward into that contract. They were incentivized to buy futures contracts to refill storage in the back end. So even if you’d gotten to a lower storage level, some physical players felt like they had it within their system to monetize that.
At that point, once the market becomes more balanced, even if you’ve got a substantial balance sheet, or even if you’ve got a scenario where you’ve got the fundamentals behind you; if the incentive structure changes and you get a balance of liquidity, effectively, you’re making a bet that looks right, in case of normalizing within the weather. The chances are things will collapse, go back to lower levels, and certainly look to the extent that you have another tropical event. Or if you have something that causes weather to be much colder or something of that nature, it can certainly go higher. It’s just that you might not have as much company to push prices much higher, and I think some of the things the PhDs and others would forget about when they were looking at the different models they were coming up with. In contrast, when Hurricane Katrina hit – you didn’t have time to balance your system largely because half your system was gone.
It was either underwater or wiped out by the hurricane, versus if you’re attempting to trade something a season in advance, the different participants in the physical market have time to react to those incentives and time to optimize their assets and take advantage of the incentives that the market’s providing. I sometimes think if you’ve not been in that market your entire career, the way a number of these PhDs and other people have been in the risk department, I think they are more focused on the different outcomes that you could have in a heavy tropical season or other things like that, without focusing on the fact that if you’ve got a different set of incentives and a different set of timing for how quickly people can react, you’re probably going to get very different outcomes. And I think that naive in terms of the scale and scope of the physical players and what they can do within their system to balance the market. Sometimes it can lead people to take risks that may be in excess of the risk they anticipated that they were taking.
We’re all humans; we’re moist robots where we respond to stimuli in certain ways; we’re prone to want to believe the good things. We’re prone to wanting to believe almost as a protective mechanism from psychology that these bad things can’t happen to me, and if I were wrong, I’d be able to get out, and if I’m right, I’m going to make all this money. You see it a lot in a market like natural gas, where there’s the opportunity to make a lot of money quickly. But the inverse is that there’s an opportunity to lose an insane amount of money quickly. As humans and traders, we believe we will be on the right side of the big moves. But, unfortunately, many people kind of line up on the same side of the trade, and for a variety of reasons, whatever the fundamental outlook you were looking at, it can change very rapidly.
In a market like natural gas, figuring out what to do two days in advance is not easy. So, figuring out what will be six months or a year in advance is virtually impossible. A lot of times, what people would do is make bets predicated on large statistical analysis; in a different range of outcomes and how these other things could happen. But Mike Tyson once said everybody has a plan until they get punched in the face, which is what happens in natural gas. If you look at what happened, even with the fall going into what we described as the Amaranth collapse, you had a scenario where you had a heat wave coming in August prices were starting to rally.
It looked like Brian Hunter and the other people that were aggressively in this market; they were going to be proven right and that they were going to make a lot of money, and almost overnight, the forward weather outlook changed. You had some tropical events that seemed to dissipate. You had some heat events that seemed to dissipate. Then literally, almost overnight, you went from a scenario where the market looked incredibly tight to a scenario where the market looked potentially extremely loose. And so if you’ve got a large position on, and you’ve got limited ability to defend that position, or so to say to add incremental capital, you’re largely a slave to whatever the market wants to do at the time. Even with the best analysis, even with the best team around you, even with the best insights, even with the best of everything, you can still lose an obnoxious amount of money because some unpredictable variable can change almost instantly and leave you extremely vulnerable.
This is why things such as liquidity management and various different other risk management techniques are so important. Liquidity is kind of there when you don’t need it, but it’s not always there when you do need it, and a lot of people get a false sense of confidence and self-assurance when we’re in a market, and things are going their way, and they don’t think they have to worry about getting out. They just assume that there will always be lots of bids to sell, and if you’re trying to buy lots of offers and when things turn on a dime. You’ve got an outsized position in the market; you’re just in a position where it’s really difficult to get out, and at that point, you are at the mercy of the market for where they’re going to let you out.
At that point, it’s almost impossible to figure out your true risk because your risk has nothing to do with fundamentals at that time. It simply has to do with liquidity and how you have to price this to find the liquidity you need to exit a significant position. You’ve seen it happen in natural gas. You’ve seen it happen in oil. You’ve seen it happen in virtually every market, and unfortunately, time and time and time again, traders find themselves in these positions, not because of ill intentions or anything of that nature, but because of human nature because they’re very comfortable when things are going their way, and people aren’t always prepared for what the world looks like when things
go the other way and what that liquidity profile will look like then.
Think about the size and scale of that position by any metric. It was a large position in the absence of liquidity for a hedge fund, which has limited capital, which has investors who can redeem and has a scenario where when things go poorly, there are a number of risk management constraints. The kick forced the funds to act in certain different ways, and unfortunately, when there’s a hedge fund that has a big position, and there’s no way out of it; once you’re at risk of exceeding whatever the capital is in that entity and if that risk gets transferred to a significant financial institution, like a large commercial bank or a large investment bank that has access to liquidity through various different government entities, that has a significant balance sheet that can create their liquidity in multiple different ways.
It might be a big loss. It might be something that could dent the balance sheet in one way, shape or form, but it’s not something that could take that financial institution down. And in the case of Amaranth, when that risk went from relatively weekends, which is an investor product that had limited ability to post additional equity, to a large global investment bank in a commercial bank that had a substantial $3 trillion balance sheet. The risk to the mark and the systematic risk at that point was dramatically different. I think it’s the reason you see the fed step in, and you see all these other different government entities step in to provide liquidity when things get funky and in a number of financial markets because ultimately, you need something much larger to provide the liquidity that you need to balance the market.
In the case of Amaranth, because it was isolated largely to one market that was dislocated, it was isolated largely to one firm that was participating significantly in that dislocation. Those positions could be transferred to a large global financial institution with a substantial balance sheet. Simply at that point, the systematic risk effectively disappeared. That risk wasn’t going to be large enough to take down a bank the size and the scale of JP Morgan. It could have dented earnings, it could have done a number of different things, but it wasn’t going to be able to provide the liquidity of those futures and other products needed. At that point, the systematic risk in the market was largely gone. It was simply a matter of trying to buy a little bit of time to find little liquidity, to balance out the pricing so that the market could get back to functioning in a much more normal way, which happened almost instantaneously within a few days. Market conditions were largely back to normal.
It’s a great question, and it’s quite scary because, many times, people don’t necessarily know and learn from past mistakes and past things that have happened in the marketplace. But we went through a significant transition from traditional drilling techniques to shale drilling techniques, which ushered in largely a 10-plus year period of lower prices and less volatility. All the things that are good for consumers in the economy, in general, and now for a variety of reasons, the market is much more focused on lower sources of carbon to produce electricity, lower sources of carbon to heat and provide various different forms of energy and attempting to transition that in a relatively expedited fashion while we don’t have the infrastructure in place to do that.
It’s a challenge on a number of fronts because a number of the best technologies that we have for low carbon tend to be things such as solar or wind, which largely depend on whether the wind blows, whether the sun shines, and how long the sun shines. But, you can make a case, an aggregate as to over the year, how many hours you should get. Still, you don’t know exactly what it will look like every day or even the following day, how it will look, and unfortunately, what that requires. It requires substantial backup generation, which tends to be inefficient and not particularly clean to subsidize.
I think a lot of people have very good intentions; there’s a substantial amount of capital flowing into this in various different forms, whether it be research, whether it be infrastructure, whether it be other different things. We’ve got a booming global economy that’s becoming increasingly energy intensive. The present industries require extremely highly fine-tuned power. Power interruptions create massive problems. When we’ve got a system that depends on reliability, it creates issues for everything we do. You’re attempting to change that system with things that don’t have that same level of reliability.
For example, if you look at Europe, which became largely dependent on Russian natural gas, they cut off a number of their nukes because of Fukushima to usher in larger-scale renewables across the continent. You had a scenario where for a variety of reasons, they hadn’t gotten quite the output that they might have expected from the renewables when they needed it. They’ve had trouble securing gas from Russia due to the Ukraine crisis. You’ve got a scenario where even without cold weather, you have industrial shutdowns, and other things like that are occurring because they’re struggling to get the fuel they need. Now they’re looking at making changes such as not necessarily closing nuclear facilities that they’d attempted to close otherwise.
You’ve got a very finely tuned energy system, and you’re disrupting that system. You’re disrupting that
system with newer technology; that newer technology requires a substantial amount of backup if, for whatever reason, you don’t get the output that you anticipated due to not having natural gas. Largely that backup in Europe is coal, and now you’ve got largely the worst of both worlds. You’ve got extremely high prices and a low level of reliability, and you’re going to have some of the highest carbon output they’ve had in years because they’ve been forced to depend on coal generation to fill the gaps. When you look forward and see the same things occurring in the United States, where you’re going to have a much more volatile regime where everything’s working well, there’s the potential for prices to go negative.
If there’s too much wind, sun, or anything, you can potentially have too much power for the system, and you have to find a way to sync that power one way or the other. Conversely, if you’re not getting the conditions you need, you’re likely to see substantially higher prices, and there are two trains of thought. One train of thought – we should do a little bit more of everything, and if we do that, at least we’ll have reliability and give ourselves a chance to get to whatever that endpoint is. But do it in a very responsible way. And others feel like we’re here right now; we need to do it. We need to kind of force the issue. Unfortunately, suppose you’re wrong by forcing the issue. In that case, you end up with situations like the one you currently have in Europe, where you have low-level reliability and extremely high prices. You have a scenario where you have a much worse carbon footprint than you would’ve had if you were simply smart about the mix of resources you use and your transition responsibly.
I think, unfortunately, what happens is that you have these cycles, and a number of the people that
were extremely active. Fifteen years ago, when these things occurred, they moved on to other things over time and throughout their career. Conversely, people can read about these things. Still, a lot of times, until they’ve been on the other side of a trade that they couldn’t control, it’s hard to know how painful that can be or how difficult it is to manage in that situation. And on top of that, you probably have less investor capital in the space because people have pulled back from investing in funds that focus on hydrocarbons. And the banks have largely been focused on facilitating consumer businesses.
You don’t have the same level of trading that you’ve seen out of the pipeline companies and the physical companies. And so you’ve got a market that’s much more vulnerable now than it’s been in quite some time with a backdrop. That’s potentially much more volatile than it’s been in quite some time, and I think that combinations are very dangerous. I think anybody attempting to trade in those markets and take advantage of what they view as opportunities just has to be careful; they have to be thoughtful. They have to look back to the lessons of the past. If they’re smart about it, there is massive untapped potential, and I think you will have one of the most interesting tradings ranges you’ve had in years over the course of the next few years. But I think a tremendous amount of money can potentially be made during that window. But if people aren’t careful, if they don’t learn the lessons in the past, if they don’t recognize their flaws and their human nature, they will find themselves extremely exposed. And I think history could repeat itself time and time again.