So I want to get into that in a second, but let's go back and explain the Exxon deal, for instance, how that worked, how it functioned. What was the utility of the credit default swap in that case?
Well, the basic concept or the original driver of credit derivatives was for banks to be able to transfer credit risk off of their balance sheet without transferring the loan itself. Say, for example, Exxon has come to JPMorgan and said, "Can we borrow X billion dollars?" JPMorgan says: "You've been a client for a long time; we certainly want to help you out with this. We want to give you this loan, but it's a big number. You need to borrow a big number. And for our risk management purposes, that's a lot of risk to take on our books and for us to hold for whatever -- five, 10 years, whatever the majority is going to be. So in order for us to give you this loan, I think we're probably going to have to transfer some of the credit risk."
Now, they probably didn't have that conversation with Exxon, but they probably had that conversation internally, which was: "We want to service the client, but this is a big risk for us to take. What can we do with it? How can we manage it?" And so the best way to manage it is to buy protection, or like buying insurance from another party.
[Let's go back to the] mid-1990s, with the creation of the credit default swap. Who came up with this idea? And how was it propagated?
The credit default swap evolved from swaps. So swaps were created in the 1980s. These are just contracts. You and I enter into an agreement. I'm going to pay you money on one leg of the swap -- it's called a leg -- and you're going to pay me money on the other leg of the swap.
So it's a contract. I have an asset and a liability. I've committed to make payments and you've committed to make payments to me.
There were lots of swaps before credit default swaps, and they were based on all kinds of different financial variables. They might be based on interest rates or foreign currency or stock.
And then, a group of people -- some of them were at JPMorgan; many of them were at Morgan Stanley, where I worked -- came with the idea of including credit in one of the legs of the swap. So I would continue to make payments on my leg of the swap, where I have a payment stream going out. I'm paying money every three months.
But instead of receiving a guaranteed payment of some kind from my counterparty to this contract, I'll receive a payment only if something called a credit event occurs, only if there's a default, only if something bad happens, if Greece defaults on its debt, if Enron defaults on its corporate obligations. Then I'll hit the jackpot and I'll be paid. It's a sort of gamble on a company going under.
But it was a swap. It was the exact same kind of technology as the previous swaps from the 1980s and the earlier 1990s. It's just that it involved credit -- that's why it's called a credit default swap -- which means the borrowing by a country or a company. And it involves default, a bet on whether or not that company or country will default on its debt. …
It's not traded on an exchange. It's a privately negotiated transaction. And because Wall Street was so effective in its lobbying activities, it's an almost entirely unregulated transaction.
In other words, on the bank's balance sheet, on JP Morgan's balance sheet, by extending that line of credit to Exxon Mobile, you were therefore-- having to hold more capital.
Equity. More capital.
More equity. That's right.
Well, more equity in--
Okay. But it's the same thing, right? It's the flipside of the same thing.
Well, so imagine a scenario where you don't have any loans on your books and the first thing you do is give out a loan for a billion dollars. Well, you're gonna need to go and borrow the billion to do it in the first place, because you need to find that cash somewhere.
The next step, very simply, is that the regulator says, "You must hold some capital against the loan." Let's say originally the minimum capital number was eight percent, let's say now it's it's higher. Let's make it ten percent. That's equity. That's shares.
We issue shares against that capital because we don't have to pay those shares back. The money we've borrowed, the billion we've borrowed to lend to Exxon in the first place, we have to pay that money back. But the shares we don't have to pay that money back.
And let's say that number's ten percent, the concept is that if I have a huge portfolio of loans that I've given out to hundreds of different borrowers, is that I'm not expecting to lose more than ten percent of that portfolio. So I can borrow 90 percent to lend 90 percent of the cash, but the other ten percent I need to issue shares because it's money I never have to pay back.
And so in the world of banking we have two different calculations of capital. The first is regulatory capital. What does the regulator say is the minimum amount of capital we have to hold against this risk? And the second is what we call economic capital. That's our determination of what the risk is. And that might be four percent. So it's four percent capital compared to the ten percent that the regulator requires, or it might be 12 percent compared to what the regulator requires.
And in 1990, you know, in the 1990s, the regulators required a flat number. They didn't look at the credit quality of your portfolio. And so if your portfolio had on average a lower credit quality than what a lot of, you know, than what the regulators considered the average, then you should, in theory, have been holding more regulatory, more economic capital than the regulatory capital. Whereas if you had a portfolio that was higher credit quality than the average.
Like a loan to Exxon.
Like a loan to Exxon. Or like, at the time, JP Morgan's portfolio, we primarily dealt with very highly rated investment grade companies. Blue chip, you know, we had a blue chip client list for the lending activities that we did. And so we looked at our loan portfolio and we looked at the minimum regulatory capital, and we felt that the economic capital was lower than the regulatory capital that we were required to hold. And so--
Now, Exxon is a big company. It's not like they're not going to pay on their line of credit. Was it really -- I mean, wasn't it more about regulatory capital and not so much about protecting against the risk?
But you can never say today who's going to be around in five years' time. You know, before 2008 we would have all said, sure, you can lend as much as you want to Lehman Brothers; you can lend as much as you want to Fannie Mae or Freddie Mac. You can lend as much as you want to. They all look like great credits. They're going to be around forever.
But that's not the way that the world of credit risk works. The fact is that today, there is always a chance that within the next five years, there's always the probability that they may not be able to pay that money back. And it's a small probability, but that's the risk that a bank manages, which is, you know, it should be a small probability.
Hopefully we're not lending money to someone that we're expecting to default, by any means, but the probability exists. So we need to transfer that risk. And, yes, we did have regulatory capital against that loan as well. Every--
In other words you had to hold less debt on the-- I'm sorry. Yes, you had to hold less debt on the balance sheet because you had a letter of credit extended to, or a line of credit extended to Exxon Mobile. The regulations required that you not be so that there be less debt. There [would] be more equity in the company.
Sorry, I don't understand the question.
And in the middle of the table you had a $14 billion loan portfolio.
And we had a $14 billion mortgage portfolio. And we also had salespeople represented, because they of course had to go and distribute the product. And we sat around, and the ideas, you know: Can we quantify the risk we're about to take? Can we model it? Can we book it? Can we manage it over its life? Do we have the operations to do it? How do we feel about the market risk? How do we feel about the credit risk? Can we distribute this? How do we feel about our reputation if we are distributing this? Those were the discussions we had.
And we had those discussions because this was a new product for us. We would have gone through what's called a new product approval process, and it involves a whole -- every element of the trade life cycle needs to be analyzed and discussed.
And, you know, the big hang-up for us really was data. Our quant guys said: "Look, you know, you the trader, you tell me what you think the price is, and I can give you a model that, when you hit F9, that's the price that pops up. But we don't have enough data to support it." So -- (laughs) -- we were nervous about the data.
And the issue was we had data about the '80s savings and loan housing crisis and what happened in a tri-state area -- I think it was Texas; I forget exactly, but it was around Texas. We had some data about some bad mortgage performance then.
We didn't have anything that was nationwide. We didn't have anything that was global, because there were a lot of mortgages from around the world in that portfolio as well. They weren't just U.S. mortgages. We didn't have anything on credit cards; we didn't have anything on auto loans; we didn't have anything on student loans. We just didn't have the data.
So it takes rocket science to -- I mean, -- (laughs) -- these are very complex deals, correct?
They are complicated. But for someone who is well versed in fixed income products or who has been looking at portfolios of fixed income products for a long time, it's not that much of a leap from what they're currently looking at. So if you're looking at individual bonds and loans, already thinking about them on a portfolio basis, and then thinking about the tranches of risk isn't that much of a leap.
OK. But by that argument, why did other banks go forward when your bank and your team decided to stop? So if it's not so complicated, why did so many others keep going, marching toward the cliff?
Look, very simply, there are certainly some investors, some banks, some borrowers who are a bit greedier than they should be. And we decided to stop because the products just got more and more risky. The risk became something that we weren't comfortable with.
And then what happened?
They started getting riskier. So, as I was saying, one of the original -- the first product that we had, this CDO of mortgages, they were AAA tranches of mortgage-backed portfolios. They were prime mortgages. There were certainly no subprime mortgages in them. We did the trade only for a three-year time period. The way we --
So the insurance was only good for three years.
It was, yes. It was a short-term trade. And we got comfortable with that. And then [there] were a number of other structural features within the transaction that we got comfortable with. And then --
So it was a profitable deal for you.
Remember, we couldn't retain the risk in our trading book; we had to distribute it. So the way that we priced it was, we were very transparent about what it cost us to distribute and what the risk that we would retain and what price we put on that risk.
I would say, from the perspective of we had no losses and as a result the money that we put on the risk that we retained ultimately became profit. But it was against risk that we had on our books. So it was a profitable transaction for us at the end of the day, but it could have just as easily been a loss, because we retained quite a bit of risk.
And so we decided to do the trade. There was a number of things we did, you know, off the back of that conversation that got us comfortable with the trade. You know, the first --
So you wrote the swap.
We did the transaction. The big risk that we were concerned about is what we call tail risk. Tail risk is that these mortgages [have], you know, 30-to-40-year lives. And while we could get comfortable with what might happen in the market for the next three to four years, 30 to 40 years was the tail. That was way so far out there, we couldn't imagine what could possibly happen in the market in that length of time.
But you did the deal.
We did the deal. ...
So -- but generally, going forward, with CDOs [collateralized debt obligations] --
We didn't do any more.
You didn't do any more?
We -- no, that's not true. We did two or three other CDOs of mortgages.
Why is it not traded on an exchange?
I think the reason it's not traded on an exchange is that people want to keep it in the dark. They don't want to be subject to an exchange.
Part of the reason is that these are customized contracts. They may have individual features. They might be four and a half years, as opposed to five years. They might have specific wording as to what a default is. And they aren't necessarily easily made into some sort of a plain vanilla transaction that you could put on to an exchange.
But I think one of the big reasons why people wanted to have these swaps off an exchange was so that they could be done privately, so that they wouldn't be disclosed, and most importantly, so that the swaps wouldn't appear as assets and liabilities on the balance sheet.
So back in the 1980s, there was a kind of showdown, where the regulators of accounting said: "We're looking at these swaps, and each leg looks like it's an asset or a liability. It looks like something that should be on the balance sheet." Just like a bank that loans money has an asset, and a bank that has deposits has a liability -- these are things that show up on the balance sheet. …
Wall Street mobilized in response to that, and they formed ISDA. At the time, this was called the International Swap Dealers Association. And the swap dealers got together and they said: "We cannot record these swaps on our balance sheet. It would introduce too much volatility. People would look at them, and they would be scared to death of the amount of risk that we're taking and the volatility of our income."
And they successfully lobbied the accounting regulators to push all of that off the balance sheet. That's where the idea of an off-balance-sheet transaction came from, from moving these swaps off balance sheets so they're not recorded as assets and liabilities.
And you might say, well, that's crazy. These are obviously assets and liabilities, just like a bank loan is an asset or a bank deposit is liability. But the word from the regulators was OK, we'll give in. We won't count them as assets and liabilities.
So originally, ISDA was called the International Swap Dealers Association, and it evolved over time and is now known as the International Swaps and Derivatives Association. So they kept the same acronym, but they have evolved in their name over time.
I should say it's a very effective marketing organization. Anyone who's interested in lobbying should really look to ISDA as one of the ultimate lobbyists of American history, unbelievably successful.
… But it seems unbelievable that a regulator would accept this, because ultimately, a balance sheet is important to understand if a bank is in trouble. And if you don't know how much risk it's taking or what its liabilities and assets are, you have no idea what the health of a bank is.
I think part of the problem was the banks didn't want anyone to know how healthy they were, how unhealthy they were, how much risk they were taking on. The banks had an incentive, starting really in the 1980s, but certainly building through the 1990s, to move into this new, complicated financial business. And that involved taking on a lot of risk. And they didn't want to have to tell the world how much risk they taking on. They didn't want to have to quantify it on their balance sheet.
They wanted to be able to push it off and hide it in these off-balance-sheet transactions. And that was why they lobbied so hard to make sure that swaps and derivatives would be treated differently from other kinds of financial products.
Let's cycle back to that. Let's go through this. In 1999 you started looking at the possibility of doing CDS [credit default swap] deals on mortgage loans. What brought that about? Tell me that story.
Well, we had a client, a bank, and they had a $14 billion portfolio of AAA mortgage-backed securities, and they wanted to release regulatory capital on that portfolio.
So they wanted to get the risk off their books.
They wanted to also get the risk off their books. But for a variety of reasons -- it was a lot of complexity around it, but they wanted to retain the assets. They didn't want to sell it into the market; they wanted to retain the assets on their portfolio.
In other words, they wanted to keep the mortgage-backed securities.
They wanted to own the mortgage-backed securities; that's right. They wanted to own them, but they wanted to distribute the majority of the risk.
Can you say who that bank was?
I don't know actually. I've always assumed that everything's absolutely a secret, and you're not supposed to say anything. It was a German institution. It was a German bank, a very big German bank.
German banks bought a lot of loans.
Mortgage-backed securities, yes.
Right. From the United States.
From the U.S., that's right.
So this bank came to you and said they want to get a credit default swap on this.
I don't know the origin of the conversation, whether we went to them and introduced the concept of risk management or whether they came to us. But the discussion transpired, and they needed to do a large transaction.
And them doing a large transaction meant that they needed to pass through my trading book this $14 billion portfolio of AAA mortgage-backed securities. And then we would turn around and sell pieces of it out of my trading book to investors, and --
You would tranche it up.
We were going to tranche it up and distribute it. And it was complicated, because prior to that, we had only done portfolios of corporate risk, entities to whom we had lent ourselves, entities that we were comfortable [with]. We knew the names; we understood them. We had teams of credit analysts looking at them. We absolutely, you know, for, as I said, 100-plus years we had been looking at credit portfolios. We knew -- we understood that risk.
So here you get a mortgage portfolio, and you're like --
And so suddenly we've got a mortgage portfolio. And we had -- we traded mortgages. We had some mortgages on our books. We certainly understood the mortgage-backed security market, but we had a lot of trouble getting comfortable with that risk.
We didn't feel that we had enough data to analyze that risk. You know, we had years and years of historical data about how corporates performed during business cycles, but we didn't have that much data about how retail mortgages borrowers, you know, individual borrowers, how they performed during different business cycles. ...
... So you're there in Boca Raton figuring out, discussing, "How can we better manage these risks?"
That's right. How can we better manage these risks, and how can other people that have the same risks better manage those risks? ...
Was there a reason for this happening at this point in time? ...
I think there were two realizations in the background that were prompting our thinking. One was that the derivative markets were growing very quickly.
And so ... you had just much more of this risk floating around, correct?
Correct. You had more of the risk, but that risk was less certain. So if you lent somebody $100, you know that you were never going to be owed more than $100 plus interest. That was as much as you could be owed.
If you entered into an interest rate derivative contract with somebody, on the first day he didn't owe you anything at all, and you didn't owe him. But a year from now, he might owe you $5; he might owe you $10; he might owe you $100 if it was a really dramatic move in whatever you were hedging. So you had more risk, and you also had more uncertainty.
So in this way, derivatives spawned the need for more derivatives in the form of credit default swaps [CDS]?
That's right, and I think you'll find as the derivate markets developed over the subsequent 20 years -- and for that matter as the securitization markets developed, the close cousin to derivatives -- a lot of the product that was developed in the second, the third and the fourth generation was only necessary because of the product that had been created in the earlier generations. ... That was one of the motivations.
There was a second motivation as well, which was that banks were getting smarter. And that may sound shocking when we look back at what happened in the subsequent 15 years, but certainly at JPMorgan we did some really quite rigorous analysis to answer the question: "When we lend money to our clients, are we earning a decent return? Are we earning an acceptable return on our capital?"
And the answer was, in a nutshell, no, that the loan business, lending money to large corporations, not the small businesses, was a loss leader business. It's a business that banks were doing at a below-market or a below-cost spread in order to curry favor with the corporation, in order to do other business.
And by doing that you were tying up capital?
... Not only tying it up, but tying it up at a poor return. So banks were looking at ways to deal with this increasing risk from derivatives, this uncertain risk from derivatives, but also ways to just get rid of the risk so that that capital could be recycled into other things. ...
... At the end of that [Boca] weekend, had you made assignments? ...
I think there were a number of areas that we had committed as a group to advance. One was to more actively manage our own portfolio of risk, which was not common in banking circles those days. I mean, the risk management approach of banks in those days was you said yes or no. And if you said yes to the right ones and no to the right ones, you did well at the end of the day. But having said yes or no, you didn't get a chance to change your mind down the road.
I think we realized both as a group but also as a bank that we needed to have the ability to manage the portfolio of credit much more accurately.
Fair enough. So this is growing and becoming a larger and larger part of JPMorgan's overall business.
Yes, and the markets in general. Every bank --
And the markets -- every bank.
Every bank has started up a credit derivatives team, and in fact, a lot of the business, I guess a lot of the banks, were educated by JPMorgan. In 1999 we had a huge conference where we mostly had banks and other asset managers and other financial institutions where we were talking about credit derivatives.
It was our first credit derivative conference, and it was held at -- I think it was 23 Wall Street. That was where the first JPMorgan building ever was. And we had this huge conference. There must have been hundreds of people, and we had several different presentations running all day.
And I gave two of the presentations. One of the presentations was on valuing credit derivatives, exotic credit derivatives, and another presentation was on counterparty credit risk, which we haven't talked about, but is a huge risk in the marketplace and one that is today still a big discussion.
But those presentations, because they talked about risk management, were the best attended presentations by far. There were 500 people in one of the presentations that I gave. And I must have given it over and over and over. I must have given it three times that day. There was so much interest in risk management. ...
By buying a credit default swap.
By investing in a credit default swap, because it was a name that they hadn't previously had access to. So there was a lot of very positive reinforcement of the market, and it just -- it grew. It grew very naturally.
Once the seed was planted, there wasn't any stopping it. Just going back to what was going on on the loan portfolios, you have to remember that prior to credit derivatives, there was really no way for a bank to manage credit risk.
So you say on day one, "Well, aren't you happy giving Exxon a loan?" Well, sure, but we would have been happy giving Enron a loan or giving Ford a loan -- (laughs) -- giving Lehman Brothers a loan, giving all of these entities loans that at that time seemed like fine and credible risk decisions to make. And yet, five years down the line, they're not all going to be there. And previously in the loan market, you were stuck with that loan.
In other words, if you made a loan to Exxon, you couldn't make one to Ford and IBM because your regulatory capital requirements were such that you were maxed out.
Well, you might have maxed out your regulatory capital requirements, but over and above regulatory capital -- forget regulatory capital for a second. Selling loans was hard. There wasn't a secondary market in loans, really. And people say, "Oh, sure, they were -- " There was no liquidity in the loan market.
So once you made the loan --
Once you made it, you were stuck.
You sat on it.
You sat on it. So if you weren't happy with it, in four years' time, there wasn't a lot you could do. And credit derivatives gave you the ability to manage that risk.
So they made a liquid market for these --
It made a liquid market for credit, far more liquid than had existed before. So forget regulatory capital and capital efficiency; it allows banks to manage risk. And if you think that fundamentally that banks are there to be doing -- they are there to intermediate the market for borrowers, they need to be able to manage that risk at the same time.
I mean, you say forget regulatory capital, but that was -- just to be clear -- that was also a factor.
That absolutely was a factor. But it wasn't the only factor, right?
What was the role of the credit default swap in this machine?
The credit default swap became an important instrument because you could feel that you were actually insured against the loss, except for one problem: The insurance company was not really a good insurance company. It was gambling, so it didn't have the resource to back the insurance which it was riding. ...
The credit default swaps also had other functions in that they allowed, and CDOs allowed, the banks to do this outside of the view of regulators, even if one was assuming that the regulators would have been on top of it. ...
Lack of transparency and lack on understanding of what was going on was absolutely essential to the blowing up of our financial system. The regulators took at face value the insurance that was being written by AIG. AIG was effectively allowing the banks to print money willy-nilly. …
Did the regulators know generally where the credit default swaps were in the system, in other words, where the counterparty risk lay?
The way these over-the-counter credit default swap markets work is they are very non-transparent. In general, market participants could not see what was going on.
So if I were buying a share of Bank America or Citibank, there'd be absolutely no way that I could ascertain what was going on. There was no sense of market discipline, which is essential for the working of a market economy, no way that market discipline could be exercised.
We have really undermined the basic principles of capitalism, of a market economy. The regulators had the authority to demand that information, so in principle, they could have looked underneath this non-transparency. They could have said: You bought this insurance, but how do we know it's insurance? You claim it's insurance, but how do we know this company that you claim you have insurance from will be able to pay off?
In other words, is the insurer going to be solvent when there's a calamity?
Exactly, and that's why every state has regulations to make sure that insurance companies can pay off the insurance. It's a very regulated industry, because there's a long history of insurance companies taking advantage of people. ...
... What kind of profits were you making off of writing these credit default swaps?
In the terms of the profits that we generated on our own business, hedging our own loans, very little. Those were really risk management transactions. But as we started to help other banks do the same thing for themselves that we were doing for ourselves but using JPMorgan as an intermediary, it was a profitable business. ...
What is the advantage [for JPMorgan] of getting that risk off the books?
That loan to Exxon consumed capital, required some capital to be set aside, and every bank has a limit to how much money they're prepared to lend to any one counterparty. And certainly at JPMorgan in those days, and I think it probably continues today, we always said to our clients: "When you really need us" -- because, for example, you've had the Exxon Valdez tragedy; if we take it in modern times the BP Gulf of Mexico disaster that required $20 billion in a hurry -- "if you really want to be sure that you can count on us, then you can't leave our balance sheet clogged up with stuff that we've accumulated through the years. Let us get that down to a relatively small amount so that we have the capacity."
By writing insurance on the line of credit --
By writing insurance, right.
-- so that if we have to pay up, we're insured.
If you cost us money because you default, we've got insurance on some of the old stuff.
So there were a lot of these sort of single deals?
There were a lot of single deals, and there were also the beginning of portfolio deals. I think that the first transaction we did in London, which was at about the same time, was what was called a first-to-default basket.
We took three European sovereigns -- and prophetic when we think about what subsequently happened 15 years later -- but you had the Kingdom of Belgium and the Republic of Italy and the Republic of Austria perhaps, and we said the first of those three names to go bankrupt triggers this insurance contract, but if the second two go bankrupt, no further payment is required.
That was a very attractive transaction for insurance companies and for some other banks that said: "European sovereigns are never going to go bankrupt. If these jerks at JPMorgan are prepared to pay me a third of a percentage point" -- actually probably even less than that then -- "to basically buy insurance against just one of these names going bankrupt, that's very good value for us and more than we could get if we just bought the bonds."
So it frees up your capital.
And so you're willing to pay a premium ... to the counterparty to insure your risk, even though you don't think that's a very big risk.
That's right. ...
Why would you write a credit default swap?
Because independent of what the regulators said -- and let's be clear: this is what kept JPMorgan out of trouble through the whole crisis -- we asked ourselves the question, "Does it make sense to have a whole lot of any risk?" ...
But in many cases not really because of the risk but because you wanted to free up capital, correct?
I would say in many cases, as this market developed across the banking sector, large proportions of the transactions that were done were motivated by regulation, no question about it. And we saw many transactions where banks very specifically chose the assets on which they sought insurance. They chose the assets that were likely to have the biggest impact on their regulatory capital requirement, independent of whether they were really releasing any risk or not. ...
To free up capital so they could have more money to lend.
And I think when you look at the firms that did relatively well through the crisis versus the firms that didn't do so well, for the most part, ... you can see that the firms that looked at their own view of the real riskiness, the real risk in a market -- and I would include JPMorgan and Goldman Sachs and a few other firms -- actually performed pretty well, and the firms that organized themselves in order to "optimize" their use of regulatory capital --
In order to end-run the regulations?
To end-run or just play by the rules. But [those] who didn't at the same time look at what they considered to be the real riskiness, they effectively let the regulators run their business for them. Those firms, almost without exception, went bust.
... When you offered this first bistro deal, you then sliced it into different, what you call "tranches." ... Why slice it?
Because different investors wanted different levels of risk. There were some investors that wanted to earn a big return on really risky stuff, and there were some investors that wanted to earn a little return on stuff that wasn't risky at all. ...
... And that party that takes the higher-risk position is really designating where they stand in line when it comes to payday?
It's all perfectly clear in the contract. So the way that we structured the bistro transactions was the very first losses were for our own account. That way nobody could ever accuse us of flogging something that we weren't standing behind. ...
Why do these if you're not covered?
Because we were shifting ... the bulk of the risk. We retained some of it, which was this first-loss tranche. The next tranche typically went to insurance companies, pension funds, perhaps other banks. The bulk of the risk of the transaction went to other people, but only behind our own loss.
Then what subsequently became known as the super-senior risk, the stuff that was statistically highly unlikely to ever incur a loss at the beginning, we retained ourselves. As we developed this market more and more, we started to shed that risk as well, because there were people that were prepared to buy it.
Then along comes the idea that packages of mortgage-backed securities could be insured with credit default swaps. Walk me through that. ...
When you think about the securitization markets and the earliest applications of a lot of this financial technology, [it] was in the U.S. mortgage market. There were collateralized mortgage obligations years before there were collateralized debt obligations, so it was a standard feature of the market, having to do a lot with the fact that the U.S. GSEs [government-sponsored enterprises] Fannie Mae and Freddie Mac were very sophisticated financially themselves in terms of developing new tools and allowing their bank counterparts and broker dealer counterparts to access leverage in different ways.
The idea that bundles of mortgages had been created, and that those bundles of mortgages were then tranched into very risky tranches and less risky tranches and really not very risky tranches, that had been around well before credit derivatives were invented. And in fact, a lot of the most esoteric interest rate-related instruments had come out of the mortgage market. ...
But what was not done at that point was writing of credit default swaps on them? ... What did that change?
I think there were two big changes. One was that the U.S. mortgage market wasn't seen as a market itself, wasn't particularly risky. In other words, people tended to pay the mortgages back, or they were guaranteed by Fannie Mae or Freddie Mac or Ginnie Mae.
It wasn't a credit-intensive market. It was much more an interest rates and prepayment-intensive market, and what changed was the idea that the real credit risk in mortgages could be bundled up and sold to investors who hadn't normally taken that risk.
It started with mortgages that were just a little bit away from the very low-risk, Fannie Mae-type mortgages. So maybe it was the jumbo mortgages, the mortgages of more than, at the time I guess it was $300,000 or something, but then slipped further and further into what would eventually become subprime mortgages that in some cases were very risky indeed.
Do you remember when this idea first was put before you, that we should perhaps consider writing credit default swaps to offload the risk on collateralized mortgage obligations?
I think the way the market developed was not so much banks deciding to write credit derivatives to protect somebody else's subprime mortgage portfolio, but rather is there a business question to the team? Is there a business to take a bunch of subprime mortgages that we either originate ourselves, so through Chase branches and across the United States, or that we buy from brokers or buy from some other broker dealer, and put those into a pool and then write credit derivatives out to the market against that pool?
... So you were looking at whether or not you could get more into this market and write off the risk to make it worthwhile to you?
Yeah, and hedge the risk. That's right. ...
... In the beginning, you took a look at this, assigned to some others the job of really crunching the numbers on this.
The respective teams took it upon themselves to analyze what could be done at any point in time. And I think in terms of when it really came to my attention as a problem or an opportunity that we might be missing, the train had already left the station. Quite a bit of business had been done by our competitors.
The consultancy firms that will come in for a small fee every year and tell you how you stack up relative to your competition, and in 2002 and 2003, 2004, 2005, 2006, these consulting firms would come in and say, "JPMorgan made $100 million last year in the whole firm in securitized products and dealing with nonconforming mortgages" -- so mortgages that don't fit inside Fannie Mae or Freddie Mac -- "but the best in class made $400 million. And the next year it was $600 million. And the next year it was $1.2 billion. ... You're missing the boat," which would prompt us to say, "Well, what are we actually missing?" And in some cases it was clear to us that we charged ourselves more for the use of our own balance sheet than other banks seemed to be charging themselves.
In other words, you had capital tied up.
We had capital tied up.
... Had it not been for credit derivatives, how serious would the financial meltdown have been?
It's very hard to say. It's a tough question to answer, because the backdrop for the whole financial crisis was a complicated set of circumstances. One is investors had become very complacent about everything.
I've talked to other of your colleagues who have agreed that that was in part driven by the idea that you could always just sort of write another insurance policy; you could insure yourself. But finally we found out that people that were holding those insurance companies had no capital, couldn't pay up.
You're right. There was real complacency all around. There was a real search for incremental yield or incremental return, and combined with complacency was a very dangerous thing.
How much of that complacency ... was driven by the fact that everybody had laid off the risk through the use of credit default swaps?
I don't think that was a central feature. I think the complacency came about because nobody could imagine house prices dropping by 50 percent or 60 percent. Nobody could imagine interest rates being increased by 3 percent in a short period of time.
But had those things happened and everybody not having been sort of tied together with all these insurance contracts, would the damage have been the same?
I think if for whatever reason derivatives had never existed -- either nobody thought of them, or some regulator at some point in the '80s said, "There will be no derivatives," a little bit like what the Chinese say today -- what would the AIGs of the world had done instead?
What they would have done instead, most likely, is lent a huge amount of money to borrowers in the form of subprime mortgages. And in fact, what they probably would have done was lent the money to owners of this mortgage pool and that 30 percent or 50 percent loan-to-value mortgage.
... A credit derivative is essentially an insurance policy?
It is an insurance policy.
... So why do you call it a credit default swap?
Because the payout on the insurance policy isn't fixed. If you buy house insurance -- you've got a $200,000 apartment and you buy fire protection and the building burns down, you're going to get $200,000, not $190,000, not $210,000.
If you enter into a credit default swap, the first thing that has to happen is the equivalent of the fire. So the borrower needs to go bust or needs to default. That in and of itself already requires a little bit of contractual fine-tuning. What is actually a default? But we'll leave that aside for the moment.
The payoff, though, is a function of the ultimate payoff from that defaulted borrower to his creditors, the real creditors, and the people that actually lent the money. And typically that's measured shortly after the event of default. ...
JPMorgan loans money to, or at least insures -- writes a letter of credit, a line of credit, if you will -- to a big corporation. They then insure that by writing a credit default swap and getting somebody to buy it. Why would somebody else buy that? What's in it for them?The same reason anybody would write an insurance contract: because they think that the value of the premium is statistically, or actuarially, or from a market perspective, worth more than the likelihood of default.
And why would Exxon or some big corporation want their bank to sell off the risk of default to somebody else?
They didn't always want that, ... because when they held a bank meeting, ... if they did get into trouble, they want to be able to call their bankers together into a room and say, "Ladies and gentlemen, we're going to work through this together." And if the banks have sold the loan to somebody else, then they don't know who's going to show up in the room.
So they were very keen not to have the banks either sell or assign or novate their loans. In the early days of credit derivatives, they may not have been aware that the bank had bought insurance against their loan exposure, or if they were aware, they may not have been terribly concerned. ...
But in some cases they were concerned, and they were concerned for a very simple reason: That the bank was effectively influencing the market value of their debt without them being in control of it. So in some cases they were concerned, and legitimately. ...
... Explain to me how corporations were using these funds and these derivatives. How is it helping? ...
When the credit default swap market started to take off, it became very much a fashionable thing to do, and there were several groups involved. One, there were the banks, and the banks were basically able to make millions.
And they were able to make loans to people who did not necessarily qualify for loans in the first place with the bank, because the banks then said, "Well, I'll admit this loan, but because I can sell off this risk to somebody else, ... I can basically take the risk, which I normally wouldn't do." ...
For corporations there were two benefits. One, people who would normally not have received funding were receiving funding, and the people who would have received funding were able to raise more money. So the volume of credit available expanded quite substantially. That was the first benefit.
But the other thing was, smaller banks around the world could actually now get access to clients, because if you're a small bank in the world wanting to lend to, say, an Exxon or somebody like that, you may not have the direct relationships. But you could now acquire these loans quite easily through these credit insurance contracts in the secondary market.
So what happens is we saw this risk gradually permeate through the financial system, and that has two effects. The first is the pool of money available for a corporate borrower vastly expanded. But the second thing that happened, which was less probably impressive and far more risky, was we were gradually tying people in very different parts of the world, who you may not necessarily know, into this web of finance.
So you might have a loan to Exxon, which everybody thinks JPMorgan has made, but then JPMorgan has distributed the risk to, say, 30 banks in different parts of the world, like in Japan, China, in the United Kingdom, in Germany, in Australia. Then they may in turn have hedged their risk with somebody else.
So you get this almost web out there, and the regulators all loved this because they thought risk was being distributed, reducing the risk and the chance of a crash. But in reality, it was introducing a new risk into the system.
And that risk was not very well understood, because remember, an insurance contract -- what was actually happening is the contract is only as good as the credit quality of the insurer. They have to pay you, and if they can't pay you for whatever reason, then this whole process of risk transfer breaks down. And people didn't really understand that the insurance market wasn't getting rid of the risk; it was just moving the risk around in a very interesting way. ...
The other thing it did ... was kind of an interesting psychological process. People changed the way they looked at credit risk. ... Once you get into a world where you think you can at a moment's notice go and insure the loan with somebody else, there are a couple of psychological changes which occur. One, you're less concerned about the repayment risk, because you think your risk is only from the moment you make the loan to when you hedge it with these instruments.
Second, you're assuming that the market for these instruments is always going to be good, and it's going to be liquid, and it's going to be functioning. And one of the strange things about financial markets is they're liquid generally when you don't need them to be liquid, and when you need liquidity, unfortunately, they tend to be illiquid.
... Explain a credit default swap. How is it structured?
A credit default swap, or CDS for short, is really a form of insurance. That's all it is. Now, think about a bank which has made a loan to ABC company. For whatever reason, it wants to avoid the risk of loss in case ABC doesn't pay them back.
What it could do is sell the loan to somebody, or the other alternative would be instead of selling the loan, they go to a third party, whether it's a bank, an investor, a hedge fund, and buy insurance from them. And the insurance would work something like the following: The bank which made the loan pays a fee to this party, and in return they say, "If ABC doesn't pay you back, we'll make good your losses." ...
It became a way that banks could make loans but reduce the risk of losing money on those loans. ...
Why do banks need this?
Well, couple of reasons. One is every time a bank makes a loan, it has a couple of issues. One, [there's] the risk that somebody doesn't pay you back, and therefore you have to manage that risk. And that limits, number one, who you lend to. And honestly, even for good clients, you don't want to lend a huge amount because you might be wrong, and they might not pay you back. So that limits the amount of credit that you can make available.
The second is effectively, you also may for a whole bunch of just logical reasons limit your risk to a particular country, a particular industry. And this is what we call concentration risk. Banks generally, because of where they're based, their geography or their client base, tend to build up concentrated risks to specific clients, so they want to manage that to some degree.
And the reason they prefer to do this using this quaint form of credit insurance rather than selling the loans is they have a relationship with the client, and they want to maintain their relationship with the client, because it's a basis for getting other forms of business.
So you don't want to go to the client and say, "Look, we love you very much, but we want to sell these loans off," because that is seen by the client as effectively a vote of no confidence in the client, ... whereas the insurance company who you're buying the actual credit insurance from, or the investor you're buying the credit insurance from, it's a completely separate contract. It never has to be disclosed to the ultimate borrower, and that became really central to this process. ...
Give another example of the Exxon Valdez. How did that work? ... JPMorgan came up with a new instrument because of this spill.
... JPMorgan used [the credit insurance market] on a scale which was quite revolutionary, and the transaction involved the settlement of the damages arising from the oil spill of the Exxon Valdez in the Alaskan Prince William Sound.
JPMorgan, as I understand it, extended money to Exxon to make the payment. But JPMorgan, in making that loan, were concerned about the risk that they were taking on, because Exxon is a very large [and] very creditworthy company, but it operates in some risky industries. And the amount involved is very, very large, and there was no guarantee that this was the final settlement of that stage either, because the litigation had been ongoing.
JPMorgan wanted to protect itself against the risk that something went wrong with its loan, so what they did was buy credit insurance on that loan from a whole bunch of players which were other banks. Some were genuine investors like pension funds and insurance companies. ...
And the fact that JPMorgan, which was a very reputable name, had actually entered the market in a large size and used this as such a pivotal transaction in managing its own risk created the environment for the growth of the credit default swap market. ...
So there was nothing bad about this?
There was nothing wrong, and there is nothing wrong with hedging, but there were two elements to the transaction. The first was obviously JPMorgan hedging its risk.
But the second element of it, which often is not concentrated on, every time a bank makes a loan, under banking regulations they're required to set aside certain reserves of capital for the loan. This is prescribed by what we call the Basel standards, ... and they actually set a framework for banking regulation.
So JPMorgan, when they made the loan to Exxon, would have had to set aside some capital. Now generally, highly rated companies like Exxon don't pay a lot for their loans, and so under those circumstances, the earnings relative to the amount of money that has to be set aside did not meet internal targets within JPMorgan.
So part of the transaction was the very notable one of managing the risk, but another part of the transaction was that by selling off this risk, we avoided making a loan which was going to give us a substandard return.
And in time, that second [element], which I call arbitrage or regulatory arbitrage, of doing things to manipulate the terms that a bank might make, ... became of greater importance and the much bigger driver of the credit default swap, or the credit insurance market.
And give me some sense of how this part of the business grew.
Well, certainly the single-name credit derivative business was becoming more and more popular. It was clearly a product that was much easier to understand, which is -- look, if you're comfortable investing in Exxon's debt, you should be comfortable selling protection on an Exxon credit default swap, because they have the same risk-return profile. They have almost the same cash flows -- not exactly, but very similar cash flows.
So the idea is that when you talk to investors about credit derivatives, although it has the term "derivative" -- and even in the '90s, we certainly had a lot of reticence to touch anything that said "derivatives." Derivatives have always had a little bit of an edge associated with them. They've always seemed a little bit more risky than what everybody's prepared to invest in.
But we talked to a number of asset managers who were comfortable with derivatives generally and as a result who became comfortable with credit derivatives. And those single-name credit derivatives, certainly they grew exponentially well beyond anyone's expectation.
Every year the BBA, the British Bankers' Association, would do a survey of what did different individuals within the credit derivative market anticipate that the growth of the credit derivatives could be the next year. And every year the actual growth far surpassed any of the numbers that we talked about. It was amazing. It was clearly a product that was in need. We had identified a need. There were banks who wanted to better manage their regulatory and risk capital. They wanted to manage their credit risk. They wanted to better service their clients by being able to give them loans, but they needed to free up some capital in order to do that, or they needed to be able to manage the risk.
And there were investors who were able to invest in some entities that they had not had access to before, entities who had previously only borrowed via the loan market. So they weren't necessarily in the bond market, so you had a bunch of investors who were very happy to be able to diversify their portfolios.
And give me some sense of -- I mean, this was pretty new stuff.
(Laughs.) This was incredibly new stuff. And now, certainly, I think, depending on who you talk to, you can always find a banker who will tell you that they were the -- you know, "I traded the first credit default swap in 1980," something. Or some other guy will tell you: "I traded the first credit default swap in 19--, you know. I have the contract in my desk."
But there were a handful of them traded. The concept existed long before 1997 or '98 when the market really -- [where] JPMorgan really stood -- started to take off. The contract certainly existed already. The concept existed And there were banks who were looking at employing it on their own books. But JPMorgan made the decision in '94 to put some effort behind it and really try to standardize the documentation, try and engage other dealers and get them to transact, to educate the sales force, to educate the investors, to build a sensible risk management framework around the product. And that was part of my job, was to come in as a trader and look at the more exotic version of the credit derivative and to build [an] exotic credit derivative trading book including all the risk management around the more exotic products. That was what I was brought in to do....
And so the deal was done. What effect does having that credit default swap deal made have on the books of JPMorgan?
What it allows JPMorgan to do is to say to the regulator, "Look, we have transferred the credit risk from our books to a third party via this contract, and as a result, against this particular loan, we should not be required to hold the capital -- whatever the number is, 8 percent, 10 percent -- that we've had to hold in the past against that loan." So if that was a billion-dollar loan, and we had 10 percent regulatory capital against it, we were able to release $100 million of regulatory capital by virtue of transferring the risk to someone else.
And so that money was therefore freed up, right?
That capital was then freed up to be able to --
Do more business.
Do more business, lend to other entities. Or it could have even been that the decision was, "Look, we actually don't have the capital to lend the full amount that Exxon needs, so in order for us to even lend this, give this loan in the first place, you know, based on our capital restrictions, we're going to need to distribute some of the risk." And we had very open and frank conversations with the regulators. There was certainly -- we were not looking, by any means, to hide the risk management practice that we were putting in place.
... How much of the motivation for these bankers to take this significant haircut comes from the fear that many banks have some sort of exposure through credit default swaps? How much is this sort of a hidden bomb driving this?
It's very hard to say for sure. I think with a deal that we're trying to close, everybody has something to gain. It is clear that if at the end of this the debt becomes sustainable, the bond prices will go up. It is very hard to say to what extent credit default swaps and the positions of the various partners of the various institutions with swaps play in their decisions. Certainly for hedge funds it makes a huge difference [as] to whether they will participate in this or not.
This is why there are number of critical issues. Will the CDSs [credit default swaps] be triggered or not in [an] event? That's something that we do not know yet. We would like to avoid it. But we will see whether that can be avoided in the end.
So that's part of the problem. The CDS market is very opaque. We don't really know who holds the positions, how much, and who is about to gain and who will lose. Some of it is washing out. Some are gaining; some are losing. But there is a clear suspicion that there is an important element that is also driving part of this discussion.
Why could all these geniuses and rocket scientists not have figured out that this could become such a problem?
That's a very good question, because I think fundamentally the financial market and the people who work in financial markets are overrated in terms of their, I suppose, ability to foresee certain problems. And there's a couple of reasons for that.
One is I think a lot of people who came into banking from the 1990s onward were relatively young, and because the business had changed, they were promoted to roles which were quite influential without having a long pedigree in banking. ... These people were almost parachuted into jobs where they did a very specific thing without necessarily understanding how a company works. ...
The second thing is a lot of the people who had quantitative skills came from the physical sciences and had very little financial experience. ... They didn't understand banking, but they understood the model. And the model, like all models, made huge simplifications in how things worked. ...
In financial markets, there are no hard and fast laws, and the models were built for a world of certainty when we worked in a world which was generally far more uncertain.
The last element is [that] in financial markets, there is a huge element of groupthink. ... Particularly when something is working and something is making money, there is a tremendous resistance to changing that.
And ... from, say, the early 1990s to, say, 2007, 2008, everything seemed to work. There were periodic setbacks, but they were minor, so everybody assumed we had found the new sort of touchstone; we could turn lead into gold. ...
Did top management at JPMorgan understand credit derivatives?
Yes, they did. Absolutely they did.
Did they at other banks?
No, not at all other banks. Certainly not. And that was part of the problem we had.
Did the buyers of credit default swaps in many cases -- I mean, how many of them actually understood the products they were buying?
I think, for the more basic product, for the simple product, the vanilla product, most of them absolutely understood, and they still do understand. You know, if they invested -- as I said, if they invest in a bond, the credit default swap in the bond is the same thing, is the same risk.
But there are some differences and some nuances that need to be understood. They're not fungible, so you do need to understand that the use of cash and the use of capital and the counterparty credit risk that comes along with it, they are different. ...
"The FRONTLINE Interviews" tell the story of history in the making. Produced in collaboration with Duke University’s Rutherfurd Living History Program. Learn more...