In an eight-page shareholder letter released this morning, Goldman Sachs is offering its side of the controversies that have dogged the firm in 2009.
As is often the case with these letters, the best stuff is at the end. That is where CEO Lloyd Blankfein and Chief Operating Officer Gary Cohn explain the Wall Street firm’s relationship with American International Group and their role in the Mortgage Securities market.
The two key takeaways (after much painstaking details): One, Goldman’s exposure to AIG did not put the firm in danger. Therefore,Goldman didn’t need the “back door” bail out that critics claim it received from the Treasury when the government paid Goldman’s AIG credit defaults swaps contracts in full. Second, Goldman never “bet against” its clients when it began shorting the sub-prime mortgage market. It was merely hedging the firm’s and its clients’ exposure to housing.
Here are the key excerpts from the letter:
Our Relationship with AIG
Over the last year, there has been a lot of focus on Goldman Sachs’ relationship with AIG, particularly our creditexposure to the company and the direct effect the U.S. government’s decision to support AIG had or didn’t have on our firm. Here are the facts:
Since the mid-1990s, Goldman Sachs has had a trading relationship with AIG. Our business with them spanned a number of their entities, including many of their insurance subsidiaries.
And it included multiple activities, such as stock lending, foreign exchange, fixed income, futures and mortgage trading.
AIG was a AAA-rated company, one of the largest and considered one of the most sophisticated trading counterparts in the world. We established credit terms with them commensurate with those extended to other major counterparts, including a willingness to do substantial trading volumes but subject to collateral arrangements that were tightly managed.
As we do with most other counterparty relationships, we limited our overall credit exposure to AIG through a combination of collateral and market hedges in order to protect ourselves against the potential inability of AIG to make good on its commitments.
We established a pre-determined hedging program, which provided that if aggregate exposure moved above a certain threshold, credit default swaps (CDS) and other credit hedges would be obtained. This hedging was designed to keep our overall risk to manageable levels.
As part of our trading with AIG, we purchased from them protection on super-senior collateralized debt obligation (CDO) risk. This protection was designed to hedge equivalent transactions executed with clients taking the other side of the same trades. In so doing, we served as an intermediary in assisting our clients to express a defined view on the market.
The net risk we were exposed to was consistent with our role as a market intermediary rather than a proprietary market participant.
In July 2007, as the market deteriorated, we began to significantly mark down the value of our super-senior CDO positions. Our rigorous commitment to fair value accounting, coupled with our daily transactions as a market maker in these securities, prompted us to reduce our valuations on a real-time basis which we believe we did earlier than other institutions.
This resulted in collateral disputes with AIG. We believe that subsequent events in the housing market proved our marks to be correct — they reflected the realistic values markets were placing on these securities.
Over the ensuing weeks and months, we continued to make collateral calls, which were based on market values, consistent with our agreements with AIG. While we collected collateral, there still remained gaps between what we received and what we believed we were owed. These gaps were hedged in full by the purchase of CDS and other risk mitigants from third parties, such that we had no material residual risk if AIG defaulted on its obligations to us.
In mid-September 2008, prior to the government’s action to save AIG, a majority of Goldman Sachs’ exposure to AIG was collateralized and the rest was covered through various risk mitigants. Our total exposure on the securities on which we bought protection was roughly $10 billion. Against this, we held roughly $7.5 billion in collateral. The remainder was fully covered through hedges we purchased, primarily through CDS for which we received collateral from our market counterparties. Thus, if AIG had failed, we would have had the collateral from AIG and the proceeds from the CDS protection we purchased and, therefore, would not have incurred any material economic loss.
In this regard, a list of AIG’s cash flows to counterparties indicates little about each bank’s credit exposure to the company.
The figure of $12.9 billion that AIG paid to Goldman Sachs post the government’s decision to support AIG is made up as follows:
• $4.8 billion for highly marketable U.S. Government Agency securities that AIG had pledged to us in return for a loan of $4.8 billion. They gave us the cash, we gave them back the securities. If AIG hadn’t repaid the loan, we would simply have sold the securities and received the $4.8 billion of value in that way.
• An additional $2.5 billion that AIG owed us in collateral from September 16, 2008 (just after the government’s action) through December 31, 2008. This represented the additional collateral that was called as markets continued to deteriorate and was consistent with the existing agreements that we had with AIG.
• $5.6 billion associated with a financing entity called Maiden Lane III, which was established in mid-November 2008 by the Federal Reserve to purchase the securities underlying certain CDS contracts and to cancel those contracts between AIG and its counterparties. The Federal Reserve required that the counterparties deliver the cash bonds to Maiden Lane III in order to settle the CDS contracts and avoid any further collateral calls.
Consequently, the cash flow of $5.6 billion between Maiden Lane III and Goldman Sachs reflected the Federal Reserve paying Goldman Sachs the face value of the securities (approximately $14 billion) less the collateral (approximately $8.4 billion) we already held on those securities. Goldman Sachs then spent the vast majority of the money we received to buy the cash bonds from our counterparties in order to complete the settlement as required by the Federal Reserve. While our direct economic exposure to AIG was minimal, the financial markets, and, as a result, Goldman Sachs and every other financial institution and company, benefited from the continued viability of AIG. Although it is difficult to determine what the exact systemic implications would have been had AIG failed, it would have been extremely disruptive to the world’s already turbulent financial markets.
Our Activities in the Mortgage Securitization Market
Another issue that has attracted attention and speculation has been how we managed the risk we assumed as a market maker and underwriter in the mortgage securitization market. Again, we want to provide you with the facts.
As a market maker, we execute a variety of transactions each day with clients and other market participants, buying and selling financial instruments, which may result in long or short risk exposures to thousands of different instruments at any given time. This does not mean that we know or even think that prices will fall every time we sell or are short, or rise when we buy or are long.
In these cases, we are executing transactions in connection with our role of providing liquidity to markets. Clients come to us as a market maker because of our willingness and ability to commit our capital and to assume market risk. We are responding to our clients’ desire either to establish, or to increase or decrease, their exposure to a position on their own investment views.
We are not “betting against” them. As a market maker, we assume risk created through client purchases and sales. This is fundamental to our role as a financial intermediary. As part of facilitating client transactions, we generally carry an “inventory” of securities. This inventory comprises long and short positions. Its composition reflects the accumulation of customer trades and our judgments about supply and demand or market direction. If a client asks us to transact in an instrument we hold in inventory, we may be able to give the client a better price than it could find elsewhere in the market and to execute the order without potential delay and price movement.
This inventory represents a risk position that we manage continuously. In so doing, we must also manage the size of our inventory and keep exposures in line with risk limits. We believe that risk limits are an important tool in managing our firm. They are established by senior management, and scaled to be in line with our financial resources (capital, liquidity, etc.). They help ensure that regardless of the opinions of an individual or business unit about market direction, our risk must remain within prescribed levels. In addition to selling positions, we use other techniques to manage risk. These include establishing offsetting positions (“hedges”) through the same or other instruments, which serve to reduce the firm’s overall exposure.
In this way, we are able to serve our clients and to maintain a robust client franchise while prudently limiting overall risk consistent with our financial resources. Through the end of 2006, Goldman Sachs generally was long in exposure to residential mortgages and mortgage-related products, such as residential mortgage-backed securities (RMBS), CDOs backed by residential mortgages and credit default swaps referencing residential mortgage products. In late 2006, we began to experience losses in our daily residential mortgage-related products P&L as we marked down the value of our inventory of various residential, mortgage related products to reflect lower market prices. In response to those losses, we decided to reduce our overall exposure to the residential housing market, consistent with our risk protocols — given the uncertainty of the future direction of prices in the housing market and the increased market volatility. The firm did not generate enormous net revenues or profits by betting against residential mortgage-related products, as some have speculated; rather, our relativel early risk reduction resulted in our losing less money than we otherwise would have when the residential housing market began to deteriorate rapidly.
The markets for residential mortgage-related products, and subprime mortgage securities in particular, were volatile and unpredictable in the first half of 2007. Investors in these markets held very different views of the future direction of the U.S. housing market based on their outlook on factors that were equally available to all market participants, including housing prices, interest rates and personal income and indebtedness data. Some investors developed aggressively negative views on the residential mortgage market. Others believed that any weakness in the residential housing markets would be relatively mild and temporary. Investors with both sets of views came to Goldman Sachs and other financial intermediaries to establish long and short exposures to the residential housing market through RMBS, CDOs, CDS and other types of instruments or transactions.
The investors who transacted with Goldman Sachs in CDOs in 2007, as in prior years, were primarily large, global financial institutions, insurance companies and hedge funds (no pension funds invested in these products, with one exception: a corporate-related pension fund that had long been active in this area made a purchase of less than $5 million). These investors had significant resources, relationships with multiple financial intermediaries and access to extensive information and research flow, performed their own analysis of the data, formed their own views about trends, and many actively negotiated at arm’s length the structure and terms of transactions.
We certainly did not know the future of the residential housing market in the first half of 2007 any more than we can predict the future of markets today. We also did not know whether the value of the instruments we sold would increase or decrease. It was well known that housing prices were weakening in early 2007, but no one — including Goldman Sachs — knew whether they would continue to fall or to stabilize at levels where purchasers of residential mortgage-related securities would have received their full interest and principal payments.
Although Goldman Sachs held various positions in residential mortgage-related products in 2007, our short positions were not a “bet against our clients.” Rather, they served to offset our long positions. Our goal was, and is, to be in a position to make markets for our clients while managing our risk within prescribed limits.
By Michael Corkery
Tags: economy, money, mortgage, mortgage market, us economic, US Economy
