Anatomy Of A Crisis

Quantitative Investment Strategies September 30, 2008

Anatomy of a Crisis Causes and implications of the current financial disruption Executive Summary • After a long period of stable economic growth and rising financial prices, the market has entered a period of extreme disruption that has shaken investor confidence to its foundation. Uncertainty about the future is in turn driving a pronounced reduction in risk taking by individual investors, institutional investors and financial intermediaries. • We believe a key ingredient of this disruption is moral hazard in the housing market. That is, both home buyers and mortgage lenders were encouraged to undertake risky investments without adequate appreciation of the consequences, since they could easily walk away from the investment if things didn’t work out – homeowners, by defaulting; lenders, by securitizing the loans in pools with an implied federal guarantee. • This resulted in too much money flowing into housing, which became overvalued relative to income and alternative investment opportunities. • When housing prices declined, and mortgage default rates rose, creditors were left holding assets (mortgage-backed and related securities) that were illiquid and hard to value, thereby limiting their ability to make loans or raise additional capital. • As uncertainty and paralysis spread, many assets, including higher-quality corporate bonds and commercial mortgages, began to sell at extremely distressed levels, putting further pressure on intermediaries’ ability to lend and invest. • Faced with a budding financial meltdown that could send the whole economy into a severe recession, the Treasury has proposed a distressed asset purchase plan that would provide a backstop to pricing for certain illiquid assets, freeing up liquidity to be used in the broader economy. • We believe the Treasury proposal, however, does create a potential new moral hazard issue. Some of the questions you may ask are: if the government will always be there to bail out investors when risks don’t work out, why should investors have any reason to assess risk carefully or to assure that they earn adequate compensation for taking risk? If capital costs don’t accurately reflect risks, how can markets efficiently allocate capital to the best riskadjusted opportunities? • We believe this new moral hazard risk is smaller and less certain than the real economic risks associated with a financial crisis. We also believe increased disclosure and modest regulatory improvements can offset these new moral hazards. Clearly, the market believes that some type of backstop plan is necessary as illustrated by the events seen on Monday, September 29, 2008. • From an investment standpoint, periods of disruption often provide extraordinary opportunities. We are pursuing these opportunities by positioning our portfolios to hedge against further disruption while being ready for a return to normalcy. From a research standpoint, we are also using disruption indicators to modify our risk targets over time and incorporating additional proprietary factors that are less susceptible to disruption.

This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions and should not be construed as research or investment advice. Please see additional disclosures.

Anatomy of a Crisis 쐍 September 30, 2008

I. Introduction After many years of a booming world economy and calm financial markets, the unprecedented events of the past few weeks have clearly shaken both investors and casual observers of the financial markets. At the beginning of 2007, risks to the economy appeared minimal; the prices of insuring against default on corporate debt, high quality mortgages, and even lower-quality subprime mortgages were all at historically low levels. Since that time, however, we have seen a dramatic increase in the perceived risks to the US and global economies. As the economy and markets lurched from one crisis to the next, some measures of economic and financial risk reached their highest levels in decades. Currently, risk aversion is so high that the credit markets have seized up: lenders now require huge spreads (by historical standards) to extend credit, with some intermediaries unable to borrow or lend at all because they can’t assess their liquidity positions due to the lack of bids for their mortgage-backed securities (MBS) portfolios.1 In an effort to calm markets, Congress and the Administration are working on a proposal to create a new

government authority which could buy up to $700 billion in distressed securities, essentially providing a backstop bid that would allow lenders to assess their liquidity levels and resume their normal business of providing credit to the economy. In addition, Goldman Sachs and Morgan Stanley, the last two remaining bulge-bracket investment banks, applied to become bank holdings companies, placing themselves under the regulatory supervision of the Federal Reserve. In this paper, we discuss how the economy and markets have reached this precarious point. We begin with an abbreviated chronology of selected events over the past several months. Next, we examine how various market prices have reacted to these events. We also discuss how economic theory can help us understand what’s happened and the implications for financial markets and the broader economy. Finally, we examine the predictive power and implications of financial disruption with respect to our macro and equity strategies, and also our efforts to manage portfolio risk in this environment.

Complete definitions of select terms used in this paper are given in the Glossary (see page 16). This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions and should not be construed as research or investment advice. Please see additional disclosures.

1

Goldman Sachs Asset Management | 2

Anatomy of a Crisis 쐍 September 30, 2008

II. Chronology of Selected Events Exhibit 1: S&P 500 Implied Volatility Index (VIX), Jan. 2, 2007 – Sep. 29, 2008

I-P See below September 08

50

P

40

Annualized volatility (%)

C. Quant Liquidity Crunch (August, 2007) B. Investment-grade corporate spreads widen; SEC eliminates uptick rule (July, 2007)

30

D. Bank write-downs due to subprime mortgages (over $15 billion) (November, 2007)

F. Establishment of Fed’s lending facility; Corporate Credit Crunch II; Bear Stearns collapse; JPMorgan Chase announces plans to acquire Bear Stearns (March, 2008) O

N D

C

F E

M

K

G. CMBX spreads widen (June-July, 2008)

L H

J I

A. Subprime mortgage spreads widen (February-March, 2007)

G

A

20

B

E. Fed cuts rates by 3/4 of a point, the biggest cut in nearly 24 years (January, 2008) 10

H. Concerns over Fannie Mae/Freddie Mac collapsing; SEC bans naked shorting of 19 financial companies (July, 2008)

0 Jan 07 Feb 07 Mar 07 Apr 07 May 07 Jun 07

Jul 07 Aug 07 Sep 07 Oct 07 Nov 07 Dec 07 Jan 08 Feb 08 Mar 08 Apr 08 May 08 Jun 08

Jul 08 Aug 08 Sep 08

I. Fannie Mae/Freddie Mac placed into a conservatorship run by the Federal Housing Finance Agency (FHFA) (September 7, 2008) 50

J. Bank of America announces plans to acquire Merrill Lynch (September, 14, 2008)

Annualized volatility (%)

P

K. Lehman Brothers files for Chapter 11 protection, marking the largest bankruptcy in US history (September 15, 2008) L. American International Group (AIG) share price falls due to news of valuation of subprime MBS Federal Reserve announces creation of credit facility of up to $85 billion (September 16, 2008)

40

N K 30

J

O

M L

M. Bush administration requests Congress to authorize the Treasury Department to buy up to $700 billion of mortgage-related assets, requests raising the national debt ceiling to $11.3 trillion In the US & UK, the SEC and FSA apply short selling restrictions Intraday prices of Morgan Stanley and Goldman Sachs fall to 84% and 66%, respectively, below their all time highs (September 18, 2008) N. Goldman Sachs and Morgan Stanley become bank holding companies NY Fed immediately provides access of bank’s broker dealer arms to Fed’s Primary Dealer Credit facility (September 21, 2008)

I

O. JPMorgan Chase announces plans to acquire Washington Mutual (September 25, 2008) 20

2

3

4

5

8

9

10

11 12 15 16 17 18 19 22 23 24 25 26 29 September 2008

P. House of Representatives fails to pass the asset purchase legislation Citigroup announces plans to acquire Wachovia’s banking operations VIX hits an all-time high (September 29, 2008)

Source: Yahoo!

Goldman Sachs Asset Management | 3

Anatomy of a Crisis 쐍 September 30, 2008

III. Evolution of the current crisis In January 2007, market prices were indicating steady economic growth with few risks. Spreads on corporate debt, commercial and residential mortgage-backed securities, and even subprime mortgage-backed securities were close to alltime lows. The VIX was just above 10%, close to the all-time low of 9.31% set in 1993, and well below the previous highs of 44–45% seen in September 1998 and September 2001 (see Exhibit 2). The Federal Reserve was more concerned about inflation than recession. Accordingly, the Fed Funds target rate had been raised to 5.25% on June 29, 2006, and would not move again for another 14 months.

13.32% per year. By August 2007, the spread had jumped to approximately 45% per year, a full ten-fold increase, and prices on BBB- MBS had fallen by 40%. By mid-September 2008, these prices had fallen by another 85% to under 9 cents on the dollar. Essentially, the market was implicitly forecasting that the vast majority of subprime mortgages would default, and that lenders would recover little value for the underlying property. Exhibit 3: The sharp increase on the ABX BBB- spread illustrates investors deepening concerns about default risk ABX BBB- spread, Jul. 19, 2006 – Aug. 15, 2007 5000

Exhibit 2: Volatility increased sharply in 2008 4000 ABX spread (bps)

Daily VIX level, Jan. 2, 1990 – Sep. 29, 2008 50

Annualized volatility (%)

40

3000

2000

30

1000

20

0 Aug 06

Oct 06

Dec 06

Feb 07

Apr 07

Jun 07

Aug 07

10

Source: Morgan Markets, JPMorgan Chase 0 1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

Source: Yahoo!

Within a few months, the first clear indications of the current financial crisis began to emerge. Concerns about decreasing home values and increasing mortgage delinquency rates caused the spreads on subprime mortgages to rise in a dramatic fashion. Exhibit 3 plots the implied ABX BBB- spread. On January 2, 2007, the spread stood at 411 bps, meaning that investors who sold protection on these securities were willing to accept 4.11% of the loan value per year as an insurance premium. However, by February 27, 2007, the premium demanded had jumped almost four-fold, peaking at a level of

Mortgage-backed securities In order to understand the current financial situation, we must first discuss the securitization of mortgages. Mortgage securitization can provide many economic benefits: it promotes sharing of mortgage risk across broad segments of the population, allows more investors to earn some of the premium associated with bearing this risk, and lowers the cost of borrowing to home owners. However, in the mortgage securitization process there is an asymmetric information problem. In other words, the loan originators (i.e., the banks actually making loans to homeowners) understand default risks better than the firm or individual purchasing the mortgages or, ultimately, the MBS that are backed by the loans.

This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions and should not be construed as research or investment advice. Please see additional disclosures.

Goldman Sachs Asset Management | 4

Anatomy of a Crisis 쐍 September 30, 2008

This issue became even more pronounced as the subprime mortgage market developed. Subprime mortgage-backed securities do not carry a guarantee against default, as do those on conforming mortgages. In hindsight, we believe this should have led investors in subprime MBS to insist on tighter loan standards and better monitoring. However, perhaps because of the long, steady increase in housing prices leading up to 2007, there was little fear in early 2007 that subprime MBS would experience major losses – or rather, investors felt that a small interest rate premium was enough to compensate for the modest risk of default in a diversified subprime mortgage pool. Another factor that may have contributed to the lax loan standards was that Fannie and Freddie were purchasing large quantities of subprime securities. In particular, Freddie purchased $158 billion worth of subprime and Alt-A securities in 2006-07, which was 13% of the total amount issued over this period.2 In early 2008, as the price of subprime mortgage debt continued to fall, investors began to more fully appreciate the extent and quality of Fannie and Freddie’s subprime holdings. As a result, Fannie and Freddie found it extremely difficult to secure additional debt or equity financing. On September 7, 2008, because of the concern that a bankruptcy of one or both of these firms would lock up the home mortgage market and potentially exacerbate the decline in home values, the Treasury stepped in to assume the risk, as well as the control, of Fannie and Freddie – essentially making good on the implicit guarantee.

The crisis spreads to other markets At first glance, it is tempting to conclude that the string of events detailed in Exhibit 1 arose solely as a consequence of declining real estate prices and the resulting losses in subprime mortgages. While there are commonalities between the widening default and credit spreads witnessed over the last few months, it would be wrong to ascribe them entirely to falling home prices. Indeed, prior to July 2007, the market perceived that the fall in the value of subprime mortgages was an isolated event which would not adversely affect the rest of the economy. This is seen in Exhibit 4 which plots the spread from March 6, 2006 through August 15, 2007 on three portfolios representing different levels of credit risk in investment-grade corporate debt and commercial mortgages, the CDX, the AAA CMBX and the super-senior tranche of the CDX.3 Until June 2007, the market did not perceive that the sharp fall in the value of subprime MBS would result in an increased probability of default for either high-grade commercial mortgage-backed securities (CMBS) or investment-grade corporate debt. However, beginning in late July 2007, spreads in these instruments spiked dramatically: the CDX spread jumped from 40 to 80 bps, the AAA CMBX from 10 to over 40 bps, and the spread on the super-senior tranche of the CDX by a factor of 10: from 1.6 bps on June 29 to 16.2 bps on August 1. At this time, concerns about a contagion effect arising from the subprime mortgage debacle surfaced. Exhibit 4: Spreads spiked significantly in July of 2007 as sentiment shifted Credit derivative spreads, Mar. 6, 2006 – Aug. 15, 2007 100 CDX spread AAA CMBX spread CDX super-senior tranche spread

80 Spread (bps)

For many years, the mortgage-backed securities market was primarily based on conforming mortgage loans, which were generally issued by Fannie Mae and Freddie Mac, who also guaranteed these mortgage-backed securities against default. In fact, most MBS investors – which include pension funds, insurance companies, hedge funds, mutual funds, foreign central banks, and individuals – assumed these securities were implicitly guaranteed by the federal government as well, since it had sponsored both organizations. However, this implicit backing, coupled with the asymmetric information problem, created a moral hazard issue. More specifically, it could be argued that Fannie and Freddie (and the ultimate investor) didn’t need to worry too much about the quality of the underlying mortgages because the government would bear most of the risk of a mortgage default.

60

40

20

0

Apr 06

Jun 06

Aug 06

Oct 06

Dec 06

Feb 07

Apr 07

Jun 07

Aug 07

Source: Morgan Markets, JPMorgan Chase 2 3

From “Fannie, Freddie subprime spree may add to bailout.” By Jody Shenn, Bloomberg.com, September 23, 2008. We refer to the five-year maturity instruments for all three indices.

This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions and should not be construed as research or investment advice. Please see additional disclosures.

Goldman Sachs Asset Management | 5

Anatomy of a Crisis 쐍 September 30, 2008

Additionally, evidence from money markets also revealed increased concern about the broader economy. Exhibit 5 plots 3-month T-bill rates, Eurodollar rates (LIBOR), and the TED spread (the difference between LIBOR and T-bills) since January 2006. As LIBOR is an inter-bank lending rate and the T-bill rate is effectively the risk-free rate, the spread between the two is an indicator of credit risk in inter-bank loans. The TED spread increases when the perceived risk of bank default rises. As shown in Exhibit 5, the TED spread increased slightly in mid-June 2007 as a result of declining T-bill rates, and then jumped dramatically in early August.4 Exhibit 5: Investor fear drove rates down, borrowing premium hit record highs Treasury, LIBOR, and TED spread, Jan. 1, 2006 – Sep. 29, 2008 6

Rate (% per year)

5 4 3

3-Month T-bills 3-Month LIBOR TED spread

2 1

documented in our “Quant Liquidity Crunch” paper from last year.5 The deleveraging began in the US, then quickly became severe and spread to non-US regions. This deleveraging resulted in unprecedented negative returns to all standard quant factors in the US and overseas during the first two weeks of August. In the process, it became clear that many popular quantitative signals had become over-crowded, resulting in either reduced effectiveness and/or increased risk. After a harrowing week for quant strategies, Goldman Sachs and a number of other institutional investors announced significant investments in quantitative portfolios. Coincident with these investments, many quantitative factors flipped directions on August 10, 2007 and moderated to more normal levels. Around the same time, as a result of an increased demand for liquidity in the financial system, the European Central Bank and the Federal Reserve jointly injected roughly $150 and $100 billion of credit into the financial system on August 9 and 10, respectively.6 In the US, the increased liquidity drove down Treasury rates, but the increased demand for investment capital and increased concerns about risk drove up LIBOR, resulting in a strong jump in the TED spread.

The crisis widens

0 Mar 06 Jun 06 Sep 06 Dec 06 Mar 07 Jun 07 Sep 07 Dec 07 Mar 08 Jun 08 Sep 08

Source: Bloomberg

The large moves in July and August 2007 were likely related to two factors: new information about financial institutions’ exposures to credit instruments, and a broader demand for liquidity. On July 30, 2007, Deutsche Industriebank (IKB), a German bank, warned that it faced losses as a result of holding US subprime mortgage-related instruments in a structured vehicle and was subsequently bailed out by a consortium of German banks. This was the first major bank failure that resulted from mortgage-related distress. Also in early August 2007, the equity markets saw a dramatic deleveraging of quant trades, something we previously

Since August of last year, the shock of increased capital costs has propagated through multiple markets in a way that few could have predicted. Exhibit 6 extends the time period of the credit spread plot in Exhibit 4 up through September 29, 2008. To facilitate comparison, the blue area in Exhibit 6’s plot is that previously shown in Exhibit 4. Note that the seemingly large spread of 80 bps for the CDX in early August 2007 widened to over 200 bps per year in March 2008. The 45 bps CMBX spread increased to over 250 bps in early March as Bear Stearns collapsed following a massive decline in the values of its mortgage-related securities; it was subsequently acquired by JPMorgan Chase. Both the CDX spread and the CDX super-senior tranche spread indicate that the implied probability of major defaults on corporate debt reached extraordinary levels in March 2008, and again over the week of September 15 – in both cases to

4 The 3-month TED spread widened to 354 bps on September 29, 2008, the highest level since September 18, 2008, when it stood at 313 bps. The September 18, 2008 spread broke the previous record, which was set after the “Black Monday” crash of 1987 (300 bps on October 20, 1987). 5 Please see GSAM papers “The Quant Liquidity Crunch” and “Quantcentration,” released in August 2007 and March 2008, respectively. 6 On August 9 and 10, the Federal Reserve pumped $24 and $38 billion, respectively, into the financial system. This move was presumably in response to a perceived lack of liquidity in the financial system and, more specifically, to a (somewhat more remarkable) increase in overnight rates early Friday morning to more than 6% – well above the Fed target rate of 5.25%. At the same time, the European central bank took even stronger action, injecting $130.6 and $83.6 billion on August 9 and 10. As a result, the Fed Funds rate dropped to below 1% on Friday afternoon, also coinciding with the sharp positive move in US quant factors on Friday afternoon.

This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions and should not be construed as research or investment advice. Please see additional disclosures.

Goldman Sachs Asset Management | 6

Anatomy of a Crisis 쐍 September 30, 2008

Exhibit 6: Derivative spreads indicate increased probability for default Credit derivative spreads, Mar. 6, 2006 – Sep. 29, 2008 300 250

CDX spread AAA CMBX spread CDX super-senior tranche spread

Spread (bps)

200 150 100 50 0 Mar 06 Jun 06 Sep 06 Dec 06 Mar 07 Jun 07 Sep 07 Dec 07 Mar 08 Jun 08 Sep 08

Source: Morgan Markets, JPMorgan Chase

levels far higher than those experienced during the Great Depression. In our view, these levels were the result of investor fear and panic – i.e., market inefficiency – rather than a rational assessment of default probabilities. Similarly, the “too thin” spreads in early 2007 were probably based on overly optimistic (i.e., irrational) expectations. The magnitudes of these spreads are remarkable. If an investor had sold protection on a dollar’s worth of the 5-year investment-grade CDX on September 18, he would receive 10% over the coming five years, and would pay any losses due to defaults on the debt. However, the worst loss rate ever on a basket of debt of this quality was only about 140 bps per year (during 1932-1936).7 So even if losses on bonds over the next five years turn out to be as bad as the worst five years of the Great Depression, an investor will still clear 3% over five years. If losses are not that bad, the investor will earn more. An alternative trade at this point would have been to sell the CDX super-senior tranche. On this trade, an investor would have received 75 bps per dollar of loss insured on September 18. And with this security, investors only had to cover losses that exceeded 30% over the 5-year period – that is, if losses were more than four times as bad as they were during the Great Depression. While it is not hard to argue, with the benefit of hindsight, that the spread of 1.3 bps per year on the super-senior tranche in January 2007 was “too low,” it is certainly also the case that the spread is remarkably high now.

Of course, if these CDX and CDX-tranche prices are so high right now, why aren’t more investors selling protection, or otherwise jumping in to earn this extra premium? There are several plausible answers. One possibility is that investors really believe that the economy will move into a downturn much worse than the Great Depression of the 1930s. In our discussions with investors, however, no one admits to being this pessimistic. Another possibility is that these high premia are the result of a “credit crunch.” A credit crunch is described by the popular press as a situation where “there isn’t enough capital to go around.” This interpretation of a crunch is a bit too simplistic. There is clearly still capital available for investment today – in fact, a number of investors are currently tying up their capital in T-bills which currently yield almost zero – but the marginal investor is unwilling to take on seemingly small risks even when the rewards are large, as current credit prices suggest. One interpretation of this is that, among the investors who still have capital, there are relatively few who have both (1) gone through an analysis to try to determine the “correct” prices for these credit instruments, and (2) set up the infrastructure to allow them to trade these instruments. This further limits the amount of investment capital available to take advantage of these dislocations. The government’s asset purchase plan is a direct response to these mortgage-related asset dislocations. By proposing to buy these distressed assets, we believe the government is arguing that the true values of these securities are misunderstood, and, consequently, that they are trading at prices that are “too low.” Moreover, if the government is willing to invest taxpayer capital in these undervalued securities, we believe they should be able to acquire them at a favorable price. Finally note that, compared to August 2007, the recent crisis was not confined to quant strategies and factors. In fact, quant strategies saw less volatility and dislocation than many other investments, including corporate bonds and commercial paper. As with quant strategies in 2007, however, it is difficult to explain the dislocation in these other markets via a direct link to mortgages – most corporations have limited exposure to mortgages or real property markets. Rather, the heightened implied risk of default seems more driven by broad deleveraging and panic than by a rational assessment of default risk.

7 Historically, the worst 5-year period for investment-grade corporate defaults was 1932-1936, when the default rate on a comparable basket was 13.5%. However, the recovery rate given default was almost certainly above 50% for these bonds, giving a cumulative 5-year loss of 6.75%, and a loss rate of 135 bps/year.

This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions and should not be construed as research or investment advice. Please see additional disclosures.

Goldman Sachs Asset Management | 7

Anatomy of a Crisis 쐍 September 30, 2008

Investment banking firms and bank runs In the latest manifestation of the recent financial crisis, a number of investment banks (including Goldman Sachs) have seen their stock prices fall precipitously. Lehman Brothers filed for Chapter 11 in the largest bankruptcy ever on September 15 (Lehman had $600 billion in assets and 25,000 employees). Two others have been acquired or are due to be acquired: Bear Stearns by JPMorgan Chase and Merrill Lynch by Bank of America. Most recently the last two remaining large investment banks – Goldman Sachs and Morgan Stanley – have filed to become commercial banks. This occurred after a fall in their equity prices of 66% and 84%, respectively, from their all-time highs, indicating investor concerns that Goldman Sachs and Morgan Stanley would not be able to continue to fund their investing and financing activities. One striking aspect of these events is that they resulted more from a positive feedback loop in asset pricing than from a dramatic change in expected cash flows from the underlying investments. That is, as some firms got into liquidity problems and were forced to sell illiquid assets to raise cash, healthier firms were forced to write down their own holdings in these same assets to reflect the fire-sale prices that less healthy firms were obliged to accept. This produced new liquidity problems for the originally healthy firms, forcing them to raise capital by selling assets, and the process snow-balled. The point is that the decline in values for financial firms were driven more by a forced fire-sale in illiquid assets than by a true decline in the expected cash flows from those investments. Some economists have interpreted this latest episode as a classical bank run, with some modifications. Bank runs were relatively common in the United States up until 1933, when the Federal Deposit Insurance Corporation (FDIC) and Federal Savings and Loan Insurance Corporation (FSLIC) were established in response to a string of bank failures in the early 1930’s. Now, the deposits of individuals in commercial banks are insured up to $100,000. Economists generally believe that, prior to the establishment of deposit insurance, a number of otherwise solvent banks were taken under by bank runs.

A very simplistic explanation of the way a commercial bank operates is that it takes in demand deposits and makes loans or investments in illiquid assets, for example home or business loans, or investments in real estate. Unfortunately, the assets in which the bank invests may be hard to value; therefore, investors and depositors will very rarely know exactly what the bank’s assets are worth. A run on a commercial bank occurs when a significant fraction of the bank’s depositors fear that the bank’s asset values may have fallen to very low levels, and they subsequently rush to withdraw their deposits. The bank has a certain amount of reserves, which means that the first depositors to withdraw their money will get paid in full. However, if enough depositors attempt to withdraw their money at the same time, the bank may not have sufficient reserves to pay them all. At this point, for the bank to remain open, it will have to liquidate some of its investments. Since the bank has to do this quickly, the prices that it will obtain for these assets will likely be less than the true value of the underlying assets. This means that the bank may not be able to pay off all of its depositors in full. The FDIC and FSLIC have substantially reduced bank runs in the past 70 years because depositors know they will be able to get their savings out (up to the insured amount) quickly if a bank fails. Modern investment banks are like commercial banks in that a lot of their activities rely on raising liquid, short-term capital from investors. In turn, they use this short-term capital to make investments in longer-term, illiquid investments. The investment bank, as long as it is viewed as having a strong capital base, will be able to continue to roll over short-term financing and continue to fund its investment activities. It is this short-term financing that makes modern investment banks look much like the commercial banks of old. Once investors begin to fear that the bank may not be able to pay them off in full, they may be unwilling to extend credit to the bank. Thus, like commercial banks in the early 1930’s, investment banks today may be forced to liquidate some of their longer-term assets at fire-sale prices. Moreover, as this short-term financing dries up, customers of other divisions

This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions and should not be construed as research or investment advice. Please see additional disclosures.

Goldman Sachs Asset Management | 8

Anatomy of a Crisis 쐍 September 30, 2008

Exhibit 7: Volatility on September 18, 2008 was driven by investor concern over Goldman Sachs and Morgan Stanley Cumulative intraday changes, Sep. 18, 2008 0.3

0.25 VIX Goldman Sachs Morgan Stanley

Cumulative return (normalized)

0.2

Market prices over this recent period suggest that investors feared there could be major consequences associated with failures of investment banks and the financial sector more broadly. This makes sense, as investment banks are key intermediaries between the holders of capital and the users of capital. This was particularly apparent on the day of September 18, when Morgan Stanley and Goldman Sachs saw their largest declines.

0.20

0.1

0.15

0.0 0.10 -0.1 0.05 -0.2 -0.00

-0.3

-0.05

-0.4 -0.5

The price behavior of Goldman Sachs and Morgan Stanley, and the behavior of the VIX equity volatility index on September 18 were strongly reminiscent of a bank run. By mid-day Thursday, September 18 the VIX had achieved its highest level since 2002 at 42.4%, as the share prices of Morgan Stanley and Goldman Sachs fell dramatically following the announced acquisition of Merrill Lynch by Bank of America, Lehman’s filing for bankruptcy and the near-collapse of American International Group (AIG) over the prior four days. Exhibit 7 plots cumulative intraday changes in the VIX and in share prices of Goldman Sachs and Morgan Stanley over the September 18 trading day. Shortly after 10AM the share prices of Morgan Stanley and Goldman Sachs began to fall, and at 12:55PM hit levels 45% and 23% below the previous day’s close. As the prices of the largest two remaining US investment banks fell, implied volatilities for the overall market soared, pushing the VIX from 34.7% at the open, to 42.2% midday. The yield on the 3-month T-bills fell to 1 bp (0.01%), as investors fled to safety. Later, as the Treasury plan began to take shape, market panic abated, and the prices of Goldman Sachs and Morgan Stanley recovered. The VIX also recovered, falling back to a level of 33.1% at the close. Exhibit 7 clearly illustrates just how

Annualized volatility (%)

within the bank, including their brokerage and investment banking groups, may pull away from doing business with these investment banks. However, for investment banks, the insurance and orderly liquidation process does not exist, as it does with commercial banks through the FDIC.

-0.10 9:40

10:19

11:31

12:43

13:55

15:07

16:00

Source: Yahoo!

closely the three series moved over this day. The correlations of the 15-minute changes in the VIX with Goldman Sachs and Morgan Stanley returns were -71% and -72%, respectively. The market apparently believed that a default by Morgan Stanley or Goldman Sachs would lead to dramatically higher market volatility, and by extension, significant disruption in the proper functioning of markets globally. In the week since the events of September 18, Goldman Sachs and Morgan Stanley have filed to become commercial banks, both to secure financing and to avoid “runs” like that experienced on September 18. Goldman Sachs has secured $10 billion, and Morgan Stanley $9 billion, of new long-term financing with this same goal. Additionally, Congress is evaluating a proposal to set aside up to $700 billion to purchase distressed securities. Of course, with all of this activity, financial market volatility has remained high. While we in QIS are students of history, over the past year we have increased our focus on measures of unusual market disruption as well as on strategies for reacting to these types of infrequent events. In the next section, we describe our efforts in this area and the implications for our portfolios.

This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions. It also refers to specific securities which pertains to past performance or is the basis for previously made discretionary investment decisions. It should not be construed as research or investment advice, or recommendation to buy or sell investments in the strategy or any other investments mentioned in this report or to follow any investment strategy. Please see additional disclosures.

Goldman Sachs Asset Management | 9

Anatomy of a Crisis 쐍 September 30, 2008

IV: How market disruption affects our investment process

Exhibit 8: The FDI can warn us about changes in risk FDI, Jan. 1, 1995 – Sep. 29, 2008

Level of disruption in financial markets

4

As discussed earlier, financial markets have been characterized by a period of unusually high uncertainty. In the QIS group, we believe that measures of financial disruption may be helpful in forecasting the efficacy of our risk, return and liquidity models. Factors in our equity and macro models are inspired by economic intuition and supported by empirical evidence over long periods of time. Unfortunately, during periods of extreme uncertainty, the motivation for trading often shifts from fundamental information to short-term liquidity considerations. For example, the crisis that affected many quantitative managers in the summer of 2007 was sparked by concerns in credit markets, but quickly led to heavy deleveraging of a wide variety of investment strategies due to the immediate need for liquidity. In the short term, the main driver behind most price movements became the need to liquidate positions, regardless of fundamental information.

3 Asian crisis

2

Subprime crisis & broad deleveraging

LTCM collapse & Russian default

September 11th Accounting scandals

1 0 -1 In sample Out of sample

-2 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

Source: GSAM The Financial Disruption Indicator, which starts in January 1995, is a proprietary composite of various indicators of financial disruption, including volatility measures, credit spreads and other signals. In an effort to make the underlying data sets comparable, each component was first normalized (this means that we subtracted the average return from the realized return and then divided by the standard deviation).

With this in mind, late in the summer of 2007 we created an index which seeks to quantify the level of disruption in financial markets. This index, which we call the Financial Disruption Indicator (FDI, see Exhibit 8), is a composite of metrics that gauges the level of overall turbulence in financial markets – including measures of yield spreads, volatility, return and correlation. We combine the various measures of disruption into a single index and then scale those measures so that the index has an expected mean of zero and a volatility of one. If observations of the index were drawn from a (standardized) normal distribution, we should expect 68% of the index values to fall between 1 and -1, and 95% of the index values to fall between 2 and -2. Although we do not believe that a normal distribution is a realistic way of describing the behavior of most financial data, the bounds described above still provide a useful way of identifying unusual events.

The very high levels of the FDI in recent months provide a good perspective on the extreme nature of the current financial crisis relative to other disruptive events in the past 15 years.

The history of the FDI reveals that the Index spikes during periods of obvious turmoil, such as: the Asian crisis of 1997; the Long Term Capital Management collapse and the Russian financial crisis in the summer of 1998; the September 11 attack in 2001; the accounting scandals in the late summer of 2002; and the most recent credit crisis associated with broad deleveraging in the summer of 2007.

The values of the FDI during 2008 indicate that most of this year has been characterized by high disruption, with the FDI at levels greater than 1.0. Early in the year, the FDI reached levels close to 3.0 during the heated debate over the methods used by credit rating agencies. After a brief decline, the FDI spiked again surrounding the collapse of Bear Stearns. After these events, financial markets appeared to revert to a less turbulent

Because we originally created the FDI at the end of August 2007, in Exhibit 8 we use different colors to display the values of the FDI before and after we completed our research. We recognize that the index was built to fit historical data and, therefore, may not fully capture future disruptive events. As quant managers, we are always conscious of this type of potential bias in our models. Clearly, the FDI has been fluctuating at a high level over the past twelve months. In Exhibit 9, we provide a magnified version of the fluctuations of the FDI since January 2008. The evidence suggests that the FDI has reached several peaks during this period.

This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions and should not be construed as research or investment advice. Please see additional disclosures.

Goldman Sachs Asset Management | 10

Anatomy of a Crisis 쐍 September 30, 2008

Level of disruption in financial markets

4 3

Rating Agencies in the news

Bear Stearns collapse

Government steps in to save Fannie Mae & Freddie Mac, Lehman Bros. bankruptcy, Federal loan to AIG

2 1 0 -1 -2 Jan 08

Feb 08

Mar 08

Apr 08

May 08

Jun 08

Jul 08

Aug 08

Sep 08

Source: GSAM

equilibrium, but not for long. In late August, concerns about the financial stability of Fannie Mae and Freddie Mac sparked the most recent increase in the FDI, which has stayed near 2.0 as the market was inundated with additional news concerning Lehman Brothers, Merrill Lynch, Bank of America, AIG, Morgan Stanley, Goldman Sachs and Washington Mutual.

FDI and predictive power of our models Given that the FDI seems to accurately capture the level of disruption in financial markets, can this information be used to anticipate periods when our models might have weaker predictive power or more uncertainty? There are at least two reasons to suspect that strategies inspired by economic intuition and fundamental information may suffer from deterioration in expected alpha during periods of financial disruption. First, during times of disruption, rising volatility in the markets tends to swamp much of the underlying fundamental information from our forecasts. This is especially true for factors driven by medium- to long-term economic considerations. Second, some of the most wellknown factors tend to result in correlated positions among investors. During financial crises, the need for immediate liquidity often leads to quick and aggressive deleveraging of those positions. As a consequence, strategies exposed to crowded signals are likely to suffer considerable losses when the level of uncertainty increases.

In Exhibit 10, we overlay the chart of the FDI from Exhibit 9 with the cumulative performance of a currency portfolio that implements the carry trade. The carry portfolio is updated daily based on the relative yield of a basket of ten developed currencies. As we can see, the two lines display a negative correlation. The carry trade appears to gain popularity (and performance) every time the FDI moves to lower values, but once disruption picks up, the carry portfolio suffers significant losses rather quickly thereafter. Exhibit 10: There is a negative correlation between the FDI and the returns of the currency carry trade FDI vs. cumulative returns on currency carry trade, Jan. 1, 2008 – Sep. 29, 2008 1.2

4.0 FDI Carry trade cumulative return

3.0

1.0

2.0

0.8

1.0

0.6

0.0

Cumulative return on carry trade (%)

FDI, Jan. 1, 2008 – Sep. 29, 2008

The “carry trade” in currency markets is a textbook example of a well known (and often crowded) trade. In the carry trade, investors borrow in low-yielding (funding) currencies to invest in high-yielding (investment) currencies. As long as spot exchange rates are relatively stable, investors expect to gain from the interest rate differential between the two currencies. Given its popularity, the financial press focuses on the appetite for the carry trade as an indicator of market sentiment concerning the expected levels of relative stability.

Level of disruption in financial markets

Exhibit 9: There have been significant fluctuations in the FDI since the beginning of the year

0.4 Jan 08

Feb 08

Mar 08

Apr 08

May 08

Jun 08

Jul 08

Aug 08

Sep 08

Source: GSAM For the carry trade return illustrated in the chart above, after ranking each currency based on its three month yield, we created a portfolio by going long the top third currencies, those with high yields, and short the bottom third currencies, those with low yields. We then calculated the return to the portfolio and repeated the process on a daily frequency. Disruptions in the subprime mortgage market, widening of credit spreads and broad de-levering may have led to the unwinding of the worldwide carry trade and a general flight to quality over this period.

This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions and should not be construed as research or investment advice. Please see additional disclosures.

Goldman Sachs Asset Management | 11

Anatomy of a Crisis 쐍 September 30, 2008

It is hard to interpret the relationship between future performance and current values of the FDI by simply looking at the points in the graph. For this reason, we also show the fitted linear relationship. As expected, the evidence suggests that periods of financial disruption are associated, on average, with future periods of poor performance, at least on a relative basis. On the other hand, another implication of the fitted line is that when the disruption subsides, performance tends to be better than average. This is consistent with the intuition that extreme disruption can be costly in the short run, but can also create good fundamental investment opportunities for the long run. As frustrating as disruptions may be, it is important to be well-positioned for the return to normalcy.

Implications of financial disruption for our portfolios Having documented a negative relationship between financial disruption and expected performance, the next question to ask is: What should an investor do to mitigate the effects of disruption and crowding on expected portfolio performance? Our response has been to focus our effort on two goals: (1) search for new factors that are less likely to be crowded; and (2) evaluate the potential benefits of modulating the risk target of our factors and portfolios based on the overall level of turbulence in financial markets. To achieve the first goal, we have invested in new data that are hard to access and, therefore, have not been extensively studied by others. As one should expect, the most potentially profitable strategies are likely also the hardest ones to uncover. For the second goal, we have performed a thorough empirical analysis to evaluate the expected benefits and the costs of modulating risk over time. Clearly, if performance does vary with the level of market disruption, then in the absence of transaction costs, one would immediately adjust the risk target of the portfolio to reflect different levels of market disruption. Once transaction costs come into play, however, the answer may not always be so simple. In fact, the cost of reducing active position sizes may be too high to justify unwinding

Exhibit 11: Periods of disruption can create investment opportunities Normalized macro strategy returns against FDI 4 3 Normalized simulated selected macro strategy returns

For a more general perspective on forecasting power and market disruption, in Exhibit 11, we plot the monthly returns from our selected macro strategy against the values of the FDI that would have been observable at the beginning of the month. This allows us to assess whether the FDI has any forecasting power on future performance.

2 1 -2

-1

-1

1

2

3

4

-2 -3 -4 Risk timing signal

Source: GSAM The risk timing signal is a composite of various indicators of market uncertainty, including equity market returns, implied volatility measures, etc. Normalization means that we subtracted the average return from the realized return and then divided by the standard deviation. In the above analysis, we use historical figures and there are sub-periods of the analysis prior to the implementation of the FDI. The figures used in the above analysis are sourced from the Quantitative Investment Strategies Group (the manager of the Fund) and not the Risk Management & Analytics Group (the keeper of official performance numbers). The performance published here is estimated, unaudited, and subject to subsequent adjustments. Simulated performance results do not reflect actual trading and have inherent limitations. Please see additional disclosures. Past performance is not indicative of future results, which may vary. The returns are gross and do not reflect the deduction of investment advisory fees, which will reduce returns.

positions during periods of high disruption, especially when we expect we will have to reinstate them again after a short time. So far, the evidence for our strategies suggests that introducing an opportunistic component to our risk targets is beneficial in many of our macro strategies, but less successful for our equity strategies. There are two reasons for this difference: first, while the profitability of many of our macro strategies’ varies measurably with the FDI, the profitability of individual equity strategies is less predictable. Second, individual equities are far more expensive to trade than the very liquid futures and forward contracts that we use in most of our macro strategies. Given the limited predictability and higher trading costs, our current research indicates that we are better off maintaining steady exposures to our equity return factors than to modulate them. Our approach for equity strategies has been to focus our efforts on developing more proprietary investment factors that we believe are less crowded and therefore less vulnerable to disruption. By adding these factors, which are also uncorrelated to existing factors in the model, we increase diversification and improve our ability to manage portfolio risk, without incurring excessive transaction costs.

This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions and should not be construed as research or investment advice. Please see additional disclosures.

Goldman Sachs Asset Management | 12

Anatomy of a Crisis 쐍 September 30, 2008

Exhibit 12: Our new factors have contributed more to our returns than our old factors

Exhibit 13: Valuation spreads are nearing the peaks reached during the internet bubble A. Value spreads in the US, Dec. 31, 1976 – Sep. 29, 2008 1.0 0.9 Percentile rank of the interquartile range of value signal

In the recent period of heightened disruption, our newer, proprietary factors have fared much better than our older ones. In Exhibit 12, we illustrate the cumulative return to our new and old factors in the US, since February 2008. It is clear that our new factors have fared better than our old factors over this time period. It is also interesting to see how these factor groupings performed during the recent crisis period. Not surprisingly, our older factors were hit relatively hard when disruption peaked on September 18. This is also true in our models outside the US.

0.7 0.6 0.5 0.4 0.3 0.2 0.1

Cumulative return to investment factors in the US, Feb. 15, 2008 – Sep. 29, 2008

0.0 1977 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006 2008

2.5 New Factors Old Factors

B. Value spreads in Japan, Jul. 31, 1987 – Sep. 29, 2008

2.0

1.0 1.5

0.9

1.0 0.5 0.0 -0.5 Feb 08

Mar 08

Apr 08

May 08

Jun 08

Jul 08

Aug 08

Sep 08

Percentile rank of the interquartile range of value signal

Cumulative return (%)

0.8

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1

Source: GSAM

While we have been successful in finding and incorporating new factors in our models, we continue to hold exposure to many of our older factors. Although in this particular crisis many of our older factors have not performed well, in general we believe that some market inefficiency remains in many of the better known factors. For example, we continue to include valuation measures in our models because we believe that securities will continue to become over- and under-valued over time. The degree of mispricing can vary, however, as investors become more and less optimistic. Moreover, we believe that the recent underperformance of value has resulted in unusual opportunities today. As Exhibits 13 (A & B) show, in both the US and Japan, the spread between the cheapest and most expensive stocks are near the peaks reached during the Internet bubble.

0.0 1988

1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

Source: PACE and Compustat. We arrive at the interquartile range by taking the difference between the 25th percentile and 75th percentile book-to-price values. The above uses Global Industry Classification Standard (GICS) industries. The returns above represent cumulative performance of the value factor on a rolling one-year basis. These performance results are backtested based on an analysis of past market data with the benefit of hindsight, do not reflect the performance of any GSAM product and are being shown for informational purposes only. Please see additional disclosures.

This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions and should not be construed as research or investment advice. Please see additional disclosures.

Goldman Sachs Asset Management | 13

Anatomy of a Crisis 쐍 September 30, 2008

Equity Risk Management

Exhibit 14: Volatility has increased since mid-2007

Our equity risk model is a key input to our portfolio construction process. We use a factor risk model, with factors categorized into return, industry and control factors. Our process attempts to take risks in return factors, while neutralizing risk from industry and control factors.

Predicted volatility across industries, Jan. 2, 1998 – Sept. 29, 2008 Industry average Commercial banks

40 Annualized volatility (%)

Return factors represent risks that we expect will earn a premium going forward. For example, factors related to Valuation, Sentiment or Momentum are all return factors. Our portfolio construction process strives to achieve exposures to return factors that are roughly commensurate with the riskadjusted expected returns offered by these factors.

50

30

20

10

0 1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Source: GSAM

Exhibit 15: The risk of certain factors has increased dramatically Predicted volatilities within leverage and volatility, Jan. 2, 1998 – Sep. 29, 2008 14 Volatility Leverage

12 Annualized volatility (%)

Industry and control factors represent risks that generally offer no expected premium. We include them in our risk models to make sure we measure their risks adequately, and so we can assess the risk impact from incidental exposures. Firm leverage is an example of a control factor: being significantly over- or underweight firms that are all highly levered is risky because these highly-levered firms tend to move together in response to variations in financing costs and expected economic activity. Our goal in portfolio construction is generally to minimize exposures to both control factors and industries. However, because our portfolios are optimized taking into account transaction costs, penalties and constraints, zero exposure to each of these factors is not always optimal.

10 8 6 4 2 0 1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Source: GSAM

Exhibit 16: Volatility within our return factors has not risen significantly Predicted volatility within the optimal tilt portfolio, Jan. 2, 1998 – Sep. 29, 2008 1.4 1.2 Annualized volatility (%)

The recent crisis has differentially impacted these various sources of risk. We have seen modest increases in the volatility of some of our less proprietary return factors (e.g., price momentum), but only minimal increases in the volatility of our newer, proprietary factors. In contrast, the volatility associated with industry and control factors has increased substantially; for some of these factors, risk has spiked in a truly dramatic fashion. Exhibit 14 plots predicted volatilities for the average industry and for the commercial banking industry over the period from January 1995 through September 29, 2008. The plot clearly shows the recent sharp increase in industry risk. Over the period from July 1, 2007 through September 29, 2008, average industry volatility increased by 81%, while the volatility of the commercial banking industry increased by a whopping 221%. Exhibit 15 plots estimated volatilities for

1.0 0.8 0.6 0.4 0.2

The OTP is an optimized, unconstrained long-short portfolio which does not take transactions costs into account. It is not meant to reflect performance achieved in any realized accounts, but rather to measure the success of the model's predictions. The returns presented herein are gross and do not reflect the deduction of investment advisory fees, which will reduce returns. These performance results for optimal tilt portfolios are backtested based on an analysis of past market data with the benefit of hindsight, do not reflect the performance of any GSAM product and are being shown for informational purposes only. Please see additional disclosures. This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions and should not be construed as research or investment advice. Please see additional disclosures.

0.0 1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

Source: GSAM With regard to our simulation methodology, we generate risk and return estimates for all assets in our research database using our proprietary model. Simulated performance results do not reflect actual trading and have inherent limitations. Please see additional disclosures.

Goldman Sachs Asset Management | 14

Anatomy of a Crisis 쐍 September 30, 2008

two of our control factors: firm leverage and firm-specific volatility. While both have exhibited large increases over the last month, the volatility of our leverage factor in particular has increased dramatically since the onset of the crisis, rising 447% since July 1, 2007. For comparison, Exhibit 16 shows the volatility of our “optimal tilt portfolio” (or OTP), which has “optimal” exposure to our return factors, but almost no exposure to industries or control factors. This plot shows that, although we did see a large increase in return factor volatility during the quant liquidity crunch of August 2007 – especially for more popular/common factors – the increase since that time has been minimal in comparison to the volatility increases observed in the industry and control factors. These volatility shifts have implications for our portfolios. In particular, the cost (in terms of risk) of incidental exposure to industry and control factors has gone up; therefore, it now makes sense to forego some expected alpha and/or pay additional transaction costs to further minimize exposure to these factors. In addition, while overall exposure to our return factors has remained stable, recent changes in the relative risks within our return factors has resulted in changes in our relative exposures to factors. In general, our exposures to common/popular factors have gone down as their relative risks have gone up, while our exposures to newer proprietary factors have gone up as their relative risks have declined. The one exception is value – although its risk has gone up, its expected return has also increased due to further widening of valuation spreads. On a net basis, our exposure to this popular factor has remained fairly stable. The overall effect of the recent market disruption on the composition of our equity portfolios has been a modest but intentional shift in risk away from uncompensated risks that have risen significantly into newer, less crowded and higher expected payoff sources of return.

Conclusion Recent events have produced a crisis of confidence in financial markets, resulting in many dislocations that could potentially derail the economy for many years to come. In particular, many investors are consumed with fear and uncertainty about the

value of their illiquid assets, and have become unwilling to extend credit or take much risk despite record-high risk premiums. This extreme risk aversion has the potential to paralyze the economy as companies are unable to efficiently finance their operations or fund additional investments. It has also forced many financial institutions, which rely extensively on short-term credit, into dire circumstances, including bankruptcy (Lehman Brothers and Washington Mutual), loss of control (Wachovia, Merrill Lynch and Bear Stearns), or outright takeover by the federal government (Fannie Mae, Freddie Mac and AIG). We believe the roots of our current crisis can be traced to issues of moral hazard. That is, everyone from homeowners to issuers of mortgage-backed securities faced asymmetric risks and payoffs in the mortgage market. As long as housing values kept climbing, they would make money. When the music stopped (i.e., home prices declined), they could forfeit their leveraged investments back to the bank (homeowners) or the federal government (MBS issuers). Thus, they took risks without adequate appreciation of the consequences of default. As a result, relative capital costs were distorted and capital was misallocated, resulting in excess capital flowing into housing and inflating real estate prices. Currently, however, the problem has come full circle. Instead of demanding down payments and interest rates that are too low for marginal homeowners (thereby discouraging them from buying), investors are now demanding excessive risk premiums for even low-risk AAA corporate bonds – essentially estimating their risk of default at multiples of what it was during the Great Depression. In response, the government has proposed a plan to backstop the price of distressed assets at levels closer to those justified by the actual risks and expected cash flows. As investors, however, our job is to be prepared for disruption, while also being ready to take advantage of the opportunities that such disruption provides. We believe that recent market movements have been driven more by short-term liquidity considerations than by long-term economic fundamentals. This has produced opportunities for above-average returns when normal conditions return – which we hope will be soon. Accordingly, we have been focusing our efforts on identifying less crowded strategies that can withstand a liquidity crisis and also provide consistent alpha under normal market conditions.

This information discusses general market activity, industry or sector trends, or other broad-based economic, market or political conditions. It also refers to specific securities which pertains to past performance or is the basis for previously made discretionary investment decisions. It should not be construed as research or investment advice, or recommendation to buy or sell investments in the strategy or any other investments mentioned in this report or to follow any investment strategy. Please see additional disclosures.

Goldman Sachs Asset Management | 15

Anatomy of a Crisis 쐍 September 30, 2008

Glossary AAA CMBX spread: The CMBX spread is analogous to the annualized cost of insuring against losses due to defaults and interest writedowns on a basket of CMBS. The index is constituted of CDS on the AAA tranches of 25 different CMBS deals.

Commercial mortgage-backed security (CMBS): A type of mortgage-backed security that is secured by pools of mortgages on commercial property. These loans are typically secured by such commercial property as apartment buildings, shopping malls, warehouse facilities, hotels and office buildings.

ABX BBB- spread: This spread represents the annual cost of insuring a basket of subprime (BBB- rated) MBS against losses in case of default.

Credit default swaps (CDS): A financial instrument designed to transfer the credit exposure of fixed income securities between parties. It is essentially an insurance contract that enables a seller to protect against the risk of default on debt obligations for a specific issuer.

ALT-A: A type of US mortgage that, for various reasons is considered riskier than “prime” but less risky than “subprime,” the riskiest category. A mortgage may be classified as Alt-A for reasons including limited documentation or because the loan is subordinated such as a home equity line of credit. Alt-A interest rates, which are determined by credit risk, tend to be between that of prime and subprime home loans. Carry trade: A strategy where an investor sells (i.e., borrow in) a currency with a relatively low interest rate and uses the funds to buy a different currency yielding a higher interest rate. CDX spread: The CDX spread is effectively the annualized cost of insuring against losses from defaults in a basket of investment-grade corporate bonds. The index contains 125 investment-grade corporations and is rebalanced every six months. CDX super-senior tranche spread: This is the annualized cost of providing insurance against the most severe losses on investments grade credit – specifically, no protection is provided until losses on the CDX basket exceed 30% over the specified period. This tranche can effectively be thought of as the “safest” class of CDX debt. Conforming mortgage: A residential mortgage that conforms to the loan purchasing guidelines set forth by the Government Sponsored Enterprises (GSE), e.g., FNMA/FHLMC. Criteria include debt-to-income ratio limits and documentation requirements. There is also a maximum loan amount, which changes based on the mean home price (above which a mortgage is considered a “jumbo loan”).

LIBOR (London Interbank Offered Rate): Daily reference rate based on the interest rate at which banks offer to lend unsecured funds to other banks in the London Interbank market. Moral hazard: Moral hazard is the prospect that an individual insulated from risk may behave differently from the way she would behave if she were fully exposed to the risk. Moral hazard arises because an individual does not bear the entire consequences of her actions, and therefore has a tendency to act less carefully, leaving another party to bear some responsibility for the consequences of those actions. Mortgage-backed security (MBS): A security created by pooling a group of individual mortgage loans with similar characteristics. Monthly payments from the mortgage pool are then passed through to investors in the mortgage-backed security. Subprime: Borrowers who generally exhibit lower credit scores and higher leverage (i.e., higher debt-to-income ratio). TED spread: Measures the premium high quality firms and banks have to pay for short-term borrowing above government rates. It is measured by the difference between LIBOR and T-bills. VIX: The Chicago Board Options Exchange’s (CBOE) Volatility index (ticker VIX), which is a key measure of market expectations of near-term volatility conveyed by S&P 500 stock index option prices.

Source: GSAM

Goldman Sachs Asset Management | 16

Anatomy of a Crisis 쐍 September 30, 2008 Additional Disclosures These examples are for illustrative purposes only and are not actual results. If any assumptions used do not prove to be true, results may vary substantially. Backtested performance results do not represent the results of actual trading using client assets. They do not reflect the reinvestment of dividends, the deduction of any fees, commissions or any other expenses a client would have to pay. If GSAM had managed your account during the period, it is highly improbable that your account would have been managed in a similar fashion due to differences in economic and market conditions. This material is provided for educational purposes only and should not be construed as investment advice or an offer or solicitation to buy or sell securities. THIS MATERIAL DOES NOT CONSTITUTE AN OFFER OR SOLICITATION IN ANY JURISDICTION WHERE OR TO ANY PERSON TO WHOM IT WOULD BE UNAUTHORIZED OR UNLAWFUL TO DO SO. Opinions expressed are current opinions as of the date appearing in this material only. No part of this material may, without GSAM’s prior written consent, be (i) copied, photocopied or duplicated in any form, by any means, or (ii) distributed to any person that is not an employee, officer, director, or authorized agent of the recipient. This material has been prepared by GSAM and is not a product of the Goldman Sachs Global Investment Research (GIR) Department. The views and opinions expressed may differ from the views and opinions expressed by the GIR Department or other departments or divisions of Goldman Sachs and its affiliates. Investors are urged to consult with their financial advisors before buying or selling any securities. This information should not be relied upon in making an investment decision. GSAM has no obligation to provide any updates or changes. Holdings may change by the time you receive this report. The securities discussed do not represent all of the portfolio's holdings and may represent only a small percentage of the strategy’s portfolio holdings. A complete list of holdings is available upon request. Future portfolio holdings may not be profitable. The information should not be deemed representative of future characteristics for the strategy. Effect of Fees: The following table provides a simplified example of the effect of management fees on portfolio returns. Assume a portfolio has a steady investment return, gross of fees, of 0.5% per month and total management fees of 0.05% per month of the market value of the portfolio on the last day of the month. Management fees are deducted from the market value of the portfolio on that day. There are no cash flows during the period. The table shows that, assuming all other factors remain constant, the difference increases due to the compounding effect over time. Of course, the magnitude of the difference between gross-of-fee and net-of-fee returns will depend on a variety of factors, and this example is purposely simplified. Period 1 year 2 years 10 years

Gross Return 6.17% 12.72 81.94

Net Return 5.54% 11.38 71.39

Differential 0.63% 1.34 10.55

CORESM is a registered service mark of Goldman, Sachs & Co. Indices are unmanaged and reflect the reinvestment of dividends but do not reflect the deduction of any fees or expenses, which would reduce returns. Investors cannot invest directly in indices. This presentation has been communicated in Canada by GSAM LP, which is registered as a non-resident adviser under securities legislation in certain provinces of Canada and as a non-resident commodity trading manager under the commodity futures legislation of Ontario. In other provinces, GSAM LP conducts its activities under exemptions from the adviser registration requirements. In certain provinces GSAM LP is not registered to provide investment advisory or portfolio management services in respect of exchange-traded futures or options contracts and is not offering to provide such investment advisory or portfolio management services in such provinces by delivery of this material. This presentation has been communicated in the United Kingdom by Goldman Sachs Asset Management International which is authorised and regulated by the Financial Services Authority (FSA). This presentation has been issued or approved for use in or from Hong Kong by Goldman Sachs (Asia) L.L.C. This presentation has been issued or approved for use in or from Singapore by Goldman Sachs (Singapore) Pte. (Company Number: 198602165W). With specific regard to the distribution of this document in Asia ex-Japan, please note that this material can only be provided to GSAM’s third party distributors (on the basis they will not distribute it to third parties), prospects in Hong Kong and Singapore and existing clients in the referenced strategy in the Asia ex-Japan region. This material is distributed in Australia and New Zealand by Goldman Sachs JBWere Investment Management Pty Ltd ABN 41 006 099 681, AFSL 228948 trading as Goldman Sachs JBWere Asset Management (GSJBWAM). Investment in any herein mentioned capability is only open to wholesale clients for the purposes of section 761G of the Corporations Act 2001 (Cth) and to investors in New Zealand to who fall within the category of investors set out in sub-sections 3(2)(a) and 5(2cc) of the Securities Act 1978 (NZ) by private arrangement with GSJBWAM and this document is intended for viewing only by such clients. Tracking Error (TE) is one possible measurement of the dispersion of a portfolio’s returns from its stated benchmark. More specifically, it is the standard deviation of such excess returns. TE exhibits are representations of statistical expectations falling within “normal” distributions of return patterns. Normal statistical distributions of returns suggests that approximately two thirds of the time the annual gross returns of the accounts will lie in a range equal to the benchmark return plus or minus the TE if the market behaves in a manner suggested by historical returns. Targeted TE therefore applies statistical probabilities (and the language of uncertainty) and so cannot be predictive of actual results. In addition, past tracking error is not indicative of future TE and there can be no assurance that the TE actually reflected in your accounts will be at levels either specified in the investment objectives or suggested by our forecasts. Simulated performance is hypothetical and may not take into account material economic and market factors that would impact the adviser’s decision-making. Simulated results are achieved by retroactively applying a model with the benefit of hindsight. The results reflect the reinvestment of dividends and other earnings, but do not reflect fees, transaction costs, and other expenses, which would reduce returns. Actual results will vary. These performance results are backtested based on an analysis of past market data with the benefit of hindsight, do not reflect the performance of any GSAM product and are being shown for informational purposes only.

Copyright © 2008, Goldman, Sachs & Co. All rights reserved. (13959.OTHER)

QECrisis/09-08