“主啊，使我谨慎，但不是现在。”

GMO Quarterly Letter April 2004 Jeremy Grantham, Chairman “Lord Make Me Prudent, But Not Yet.”

Skating on Thin Ice in the First Quarter: Greenspan and the Nasdaq in general did not show signs of Coughs and the Ice Creaks regrouping until the last week or so of the quarter. For So we survived the first quarter. The return to risk taking, example, the most volatile stocks (top 25%) were up 4% relative to the market, became very high 18 months ago over the market on January 17 and fell to -3% late in the and GMO took more risk than it normally does. This was quarter, and technology stocks were ahead by 6% and fell most evident in asset allocation where we had maximum to -6%. More important for us was that other measures of positions in emerging country equity and small cap risk also took a hit, but then rallied back to new highs. international, but even within specific funds we also took Emerging equities, emerging debt, small cap international more risk than usual. This was still the pattern as we came stocks – all higher risk assets – ended with strong to very into this year. We started the year with most of our funds strong quarters, and by the closing 2 weeks, all the risky moderately more aggressive than their benchmarks, but assets were outperforming again. But the real decreased the risk during the quarter. By the end of June significance of this substantial hiccup based on such we expect to be about normal for GMO, which is to say slender input from Greenspan was the reminder of how moderately more conservative than our benchmarks. And thin the ice is, how central moral hazard is to this rally of by the end of the third quarter, we should approach as the last year of the political cycle, and how central the much conservatism as our specific client mandates and Chairman of the Federal Reserve is to the issue of moral specific fund positioning make appropriate. hazard.

All in all, the first quarter was kind to our risk taking. Our Credit Problems, Housing Prices, and Mortgage two biggest bets – emerging equity and small cap Refinancing international – finished up 10.4% and 9.4%, respectively, The essence of the Fed’s moral hazard is to create an compared to the S&P 500’s +1.7%. In the funds, our environment conducive to debt expansion: the unspoken relative performance was also ahead generally in U.S. promise that there will be no negative consequence from equities, other developed country equities, fixed income, more debt . . . at least for the “foreseeable future”. The and hedge funds – where all 14 of our funds had a real problem with credit, though, is that you never know positive return. In asset allocation, global balanced was how strong the elastic band is, and in the U.S. it tends to up 4.6%, over 2% ahead of benchmark, and all of the be very strong.

Unprecedented levels of debt can always other asset allocation funds were ahead. The only get larger, and with a favorable wind all will still be well. important equity category in which we were behind the But higher debt levels reduce the margins of safety and benchmark was emerging country equity where we markets, and therefore, with higher debt, should return trailed by 2.5%, but last year’s +13% outperformance more to compensate for higher risk. More highly takes some sting out of the pain. leveraged markets should be cheaper, for only cheapness increases the returns to stock holders in the long term. (In Greenspan, however, gave us risk takers a scare in the the short term, to be sure, surges of earnings can look first quarter. On January 17, he changed his statement of impressive, but profit margins are mean reverting and moral hazard so delicately that only stock and bond upsurges are followed by downsurges whereas buying traders – all linguistic experts – noticed.

And the return to cheaply is a joy forever.) risk nose-dived. Almost every high risk investment was punished for a few weeks and several important ones The credit vulnerability that impresses us most – out of such as tech stocks, growth stocks, high volatility stocks, many good candidates – is mortgage refinancing. Easy credit and low rates were designed to offset the negative reported that the price of Manhattan apartments is up a effects on the economy of the bubble forming and then stunning 35% year over year! These are real bubble breaking – expensive excess investment and a negative numbers even if much of the rest of the country’s housing wealth effect. Probably the most effective component of is still merely in a bull market.) the program was the enormous increase in mortgage refinancings. Exhibit 1 shows that an incremental credit Our emphasis here on mortgages and housing is not equal to 6½% of GNP became available for consumer intended to minimize the obvious credit risks posed by spending in the last 3 years, probably a more powerful other consumer debt, corporate debt, government debt, stimulus than the tax cuts.

Cheap mortgages made home and U.S. foreign debts, all at or near record levels. A buying attractive relative to renting and pushed up prices. particular flash point might be the increase in the so-This, in turn, stimulated house building, but equally called ‘carry trade’, the unprecedented use of short-term importantly, by increasing the amount of housing equity borrowing to buy higher yielding or higher returning that could be used as collateral for yet more borrowing, it bonds and stocks, particularly by banks, investment reinforced the virtuous cycle, to stretch the use of the houses, and hedge funds. Another vulnerability might lie word “virtuous”. Increasing house prices also directly in the extravagant increase in hedge funds with their offset the negative stock wealth effect; in fact, for most sheer speed and high turnover. middle class families, net wealth gained from increased The consequences of these increased financial housing prices outweighed decreasing stock prices. vulnerabilities for investors is that in selecting hiding The prodigious increase in outstanding mortgages places, investors should take credit problems as seriously obviously helped keep consumer spending, and the as an overpriced market. general economy, far stronger than it would otherwise have been.

Unfortunately, it has also increased the Fallen Angels sensitivity of consumer spending to interest rates, and In the 50% decline from the March 2000 high, value mortgage refinancing has perhaps become the most stocks and small stocks cushioned most of the pain and vulnerable part of this 4-year political cycle. The usual small value stocks cushioned all of it. The very probable need to ‘clean house’ in the first and second years of the cause for this outperformance was that small stocks were next cycle of any administration will be confronted by very cheap indeed relative to the rest of the market. In dangerously extended credit on all fronts, but particularly fact, Ben Inker and I took more grief in our asset for extended house mortgages. High and still increasing allocation forecasts over favoring small caps (and home prices will consequently have a central role in how emerging equities) than anything else.

“If you’re so this cycle’s politically driven financial house cleaning gloomy on the market and particularly Nasdaq, how on plays out in the next 2 years. The whole credit system, as earth can you recommend small caps (and emerging) as a well as consumption, the strength of the economy, and the haven?” would be a close approximation of the objection. stock market, will be unusually vulnerable to a Our answer was that yes, in the past, both categories had, combination of a house price decline and higher rates. on average, underperformed in market declines, but that (For the record, on Friday, April 16, The New York Times if you divided all bear market periods into four quartiles (based on how cheap these two groups were relative to Exhibit 1: Refinancing as a Percentage of GNP the S&P) you would see that when they were in the 8% cheapest quarter, they in fact outperformed in bear 7% 7.2% periods and that in March 2000, they were at the very top of the best quartile of value.

The good news is that they 6% did indeed handsomely outperform in the bear market 5% after March 2000; the bad news is that now small stocks 4% are in the worst quartile of their range (although emerging 3% – thank heavens – is still in its top quartile).

2% Value stocks in March 2000 were off-the-scale cheap 1% relative to the market and so performed heroically on the 0% downside. This time, they are in the worst half and therefore will probably only eke out a modest reduction Q1 Q3 Q1 Q3 Q1 Q3 Q1 Q3 20 20 20 20 20 20 20 20 00 00 01 01 02 02 03 03 in market pain, even in a severe decline.

GMO Quarterly Letter – April 2004 2 Other angels that have fallen since March 2000 include opposite direction. By 1983, small stocks reached the real estate, particularly REITs that yielded 9.1% at the highest relative value that they have ever attained by market peak and, after brilliant performance, now yield tripling relative to the S&P 500! In this recent bubble, under 5.5%; all of fixed income, and especially emerging small caps, now up over 100% relative to the market, country debt and TIPS, which were also very protective have also substantially over run their relative fair value, and, in fact, attractive. Now they are all overpriced. but by substantially less. What are the prospects from here?

Small Stocks In the longer term, it is easy – small will correct back to Far from being defensive, as they were in the 2000 to trend. In the very short run, however, the fortunes of 2003 market, small caps may well turn out to have one of small caps are caught up with the same moral hazard the more unexpected and painful setbacks in a broad political issues we’ve been obsessing about. ‘Small’ market decline. Exhibit 2 shows the long-term benefits from the current speculative tone of the market performance of small caps versus large caps. There are and has prodigious momentum, but what it does not have two great rallies – 1973 to 1983 and 2000 to present – and – at least not yet – is strong relative earnings. In the 1974 one extended decline from 1983 to 2000. Exhibit 3 shows to 1983 rally, the profit margins (and hence earnings) of our composite measure of value for the small caps small companies had an unexpected surge to a unique relative to the large. It can be seen how spectacularly level, where for a minute or two in 1983, the profit cheap ‘small’ got in the Nifty Fifty market of 1973, still margins of small companies exceeded those for large their biggest distortion in history. Certainly I did not companies for the only time in history outside of World believe there was a 1 in 100 chance of this distortion War II, when profit controls were applied only to large being exceeded, but 2000 really gave it a run for its companies. Since 1983, relative profits and profit money. As has always happened, the bubble in large margins of small companies have not been strong and stocks of 1972 was followed by an over correction in the have shown little sign of improving dramatically yet. So it is unlikely that the stock surge of 1973 to 1983 will be Exhibit 2: Relative Return of Small Cap closely rivaled this time. However, it is a decent bet (over 2.5 50/50 I think) that they have a few more months of 2.3 relative outperformance. If they get their outperformance, 2.1 ‘small’ will become another particularly vulnerable 1.9 segment of the market in the next decline, and even with 1.7 no further relative gains it is likely to be vulnerable 1.5 enough.

1.3 Exhibit 4 shows the rising level of debt in small stocks.

1.1 This is frightening enough that I think today’s level can 0.9 be fairly described as a ‘possibly terminal level of debt’.

0.7 Small stocks are definitely not the place to be in a debt Dec-70 73 76 79 82 85 88 91 94 97 00 03 crisis.

Exhibit 3: Small Cap Valuation Relative to the S&P 500 Exhibit 4: Debt Leverage of Small Companies 1.30 0.55 1.0 = Fair Value 1.0 = S&P 500 1.20 0.50 1.10 0.45 1.00 0.40 0.90 0.35 0.80 0.30 0.70 0.25 Dec-70 73 76 79 82 85 88 91 94 97 00 03 Dec-70 73 76 79 82 85 88 91 94 97 00 03 3 Quarterly Letter – April 2004 GMO Volatility Equally important, today ‘quality’ as a characteristic Volatile stocks also reached a relatively cheap point in in the market is usefully cheaper than average. It also September 2002, but the most volatile 10% have had a comes with absolutely no exposure to credit spectacular rally of over 55% against the market in the problems and consequently in any credit crisis will rally since then, along with all other speculative likely accrue a scarcity premium. In a severe market measures, and are now substantially overpriced. This decline, the strategy should be a relative haven for group is always dangerous and has the added ugly those of us who have to invest in U.S. equities. characteristic of underperforming the market in the long Cash run by over 4% a year (and even the most volatile 25% have underperformed by 2.5% a year). Over valuation I have described investing in cash at -1% real return further increases the pain that volatile stocks are likely to as ‘chewing glass’ and I was grateful for Warren inflict in any market setback. Buffet’s annual report. Berkshire Hathaway’s portfolio has built up a huge reserve of unattractive Any Port in a Storm cash to prevent making even less attractive equity investments. What else is there to say? Yes, you As the year wears on, prudence should become a more might well make some more money in more normal compelling virtue, and reducing exposure to credit portfolios in the next 6 months, but only by taking a problems, small cap, and volatility should feature lot of risk. prominently. With so many factors to worry about and with the much reduced dependability of value on the Market Neutral Hedge Funds downside, what is left? For accounts that cannot short ‘Cash Plus’ in the form of market neutral hedge funds extremely overpriced asset classes like U.S. equities, the seems a lot less painful a proposition than pure cash, painfully brief list of not very satisfactory responses and might reasonably return 5% to 10% depending is: ‘quality’, cash, and market neutral hedge funds. A on the success of selecting from the huge pool of fourth, more problematical response is commodities and offerings. We warmly recommend that strategy to real assets, of which forestry (surprise, surprise!) is by far clients, although we recognize it would be difficult the most dependable in the longer run. for Mr. Buffet’s $31 billion. Ah well, the problems of size are very democratic; they reduce flexibility and The GMO U.S. Quality Equity Fund returns for everyone!

We introduced a new U.S. quality equity fund in the first quarter and took initial small positions in it in And while I’m on the subject, it was also interesting many asset allocation accounts. It owns no stocks to read Mr. Buffet’s regret at not having sold more in with material debt and no small caps. The largest 1999 and 2000. His laid-back attitude to overpricing positions are the obvious companies with global at the 2000 peak, of the ‘yes it’s a bit overpriced’ franchises. variety, undercut GMO’s then ‘shrill’ bearish position. We warmly welcome him as an ally this Both modern portfolio theory and common sense time. strongly suggest that quality stocks should return a little less than ordinary stocks in the market. A Digression on the Risks and Theoretically, in return for accepting lower returns, Advantages of Hedge Fund Investing clients should get to feel comfortable with the There is no question that hedge fund obviously successful companies in their portfolios. investing is extremely faddish and that both Should another 1932 depression come along, these the number of funds and the total amount are also the names that will survive, as many did in invested in them have shown explosive fact actually survive 1932.

Remarkably however, growth. There is also no doubt that this quality stocks have slightly outperformed the market growth of hedge fund investing has caused over at least the last 40 years. I believe this is due to some types of market risk to increase. For the willingness of the market to overpay for the sex example, the ultra low costs of borrowing appeal and rapid volatile growth of lower quality and moral hazard have increased the companies. The ‘great’ or ‘quality’ companies are vulnerability of hedge funds to rate staid in comparison. They are a little boring, and increases, and given their leverage and the ‘boring’ sells at a discount in the real world, as speed that they can move, it is clear that if opposed to the academic, theoretical world. they are collectively surprised there could GMO Quarterly Letter – April 2004 4 be a bigger rush for the door than normal.

Real Asset Fund The seed of another big problem is that To create a means for achieving exposure to most hedge funds are short liquidity: their commodities, GMO will be starting the GMO Real short positions are more liquid than their Asset Fund in the third quarter unless we hit some longs so that a liquidity squeeze has the major snags. It will own gold and precious metal possibility of being self-sustaining as one stocks, industrial metal stocks, oil and natural gas wave of hedge funds covering shorts forces stocks, and securities creating exposure to others to follow, and there is many times agricultural commodities.1 more hedge fund money today than in Long Term Capital’s day! Our legal department limits what I can say about this topic of a new fund including, strangely, the several Most fads end badly and perhaps this one caveats about what can go wrong with commodities. will too one day, but I believe this fad is currently one to follow because by sheer random good luck it has timing in its favor.

The Changing Return to Risk Taking If we are within months of a major market Sometimes the incremental return to taking another unit decline – a decline that may easily last of risk, as defined by volatility, is very high. Sometimes through 2006 – as I believe we are, then it is negligible. It is intuitively appealing that if the hedge funds, with even modest positive riskiest stocks and bonds and risky asset classes in returns, will look brilliant and will no doubt general have just had their best relative 18 months in attract even more money. Any hedge fund years, then the future return to risk taking must have liquidity or interest rate crisis that occurs dropped precipitously. This is perhaps best illustrated by before the market low will probably seem our absolute return portfolios because we have been small potatoes relative to the broad market showing an efficient frontier exhibit for that approach for pain. several years, for all levels of risk.

As shown in Exhibit 5, Exhibit 5: Absolute Return Portfolios – Efficient Frontier of Risk and Return 16% Higher Risk 14% 12.5% 9/2002 12% Absolute Return Portfolio Frontier Expected Real Return w/Alpha Medium Risk 9.7% 9.6% 6/2003 10% Frontier 7.8% 8% Lower Risk 7.4% 3/2004 6.1% 6.3% Frontier 6% 6.0% 4.5% 4% 3.3% 2% GMO Global Balanced Fund, with Alpha Global Balanced 0.2% Portfolio (Indexed) 0% 3% 5% 7% 9% 11% 13% 15% Risk (Annualized Volatility) 5 Quarterly Letter – April 2004 GMO by September 2002, the return to risk curve we showed • The weak first 2 years of the next Presidential Cycle, had become very steep. Risky investments had been with particular weakness in these years for risky stocks. badly hurt and traded at much lower prices than less risky assets, and they offered a real inducement to move to the • Unusually high credit vulnerability, going into the first right up the curve and take more risk. By March 2004 2 years of the Presidential Cycle, which are all about however, the risk curve has flattened very considerably cleaning up the credit situation. for the predictable reason that the higher risk portfolios • Very little reward for taking risk (a flat risk curve). did spectacularly well recently.

There are two points to be made here. The first is that the median risk taker should • Dramatically strong performance over the last 4 years rationally not be fixed in the middle of the risk curve. He in those asset categories and sub-categories that were should move to more risk when the curve is steep and less so useful in 2000 in cushioning pain, so that today they risk when the curve is flat. For example, when the curve will be mainly of little use. is totally flat, that is when risk taking shows no incremental return, we can all agree that both high and low risk takers should end up all in cash. The Outlook for the Next few Months In the blue corner we have over valuation of the market Traditional risk calculations that are based solely on and a fear of more terrorism, higher interest rates, higher historical volatility and ignore shifts in return miss this inflation, and all that that would mean.

In the red corner, point. Value at risk (VAR), for example, will always we have a high probability of good earnings, very high produce the same amount of risk taking for each investor, margins, great productivity, good GNP, a strong China assuming that market volatilities stay the same. Under the and India, and reasonable global economic strength. same assumption of unchanged volatility, GMO’s risk return frontier however can dramatically change, as (-) Market Over Valuation indeed it has recently. The difference is that our return Normal p/e (16x) on normal profit margins (6.0% on forecasts change dramatically with price changes and sales) produces a trend line or fair market value that since volatility by and large stays fixed, the GMO return still looks like 720 on the S&P 500 versus today’s 1120. to risk curve rises and falls with asset price changes.

With We could, of course, be in a new paradigm of VAR it does not. permanently higher average p/e’s. We know intellectually The second point concerns today. The risk curve is that new paradigms are possible despite none occurring relatively flat and I believe the odds are better than 50/50 yet at the asset class level, but hey, there’s always a first that it will flatten further in the next few months. Surely time. this justifies a move towards the lower end of the risk (-) Interest Rate Risk curve? Our current low risk absolute return portfolio, for example, has 60% bonds, 20% enhanced cash through The postponing of rate rises, which allows further our multi-strategy hedge fund, and only 20% high risk extension of record debt levels, has been likened to a assets. This contrasts to our current high risk absolute cosmic game of chicken or as St. Augustine might have return portfolio that has no bonds and 80% high risk asset said, “Lord, make me prudent, but not yet.”

The longer classes. Both portfolios reflect the unique scarcity of the raising of rates is left, the higher the price will be that attractive assets and the inability to go materially short in will eventually have to be paid. these accounts.

(+) Good Near-Term Fundamentals Unfortunate Convergence of Negatives There is a strong consensus that this year and particularly The market characteristics we face going into 2005 are as this second quarter will show higher margins, earnings, follows: GNP, and productivity. We have no reason to disagree.

Therefore, since value works slowly and so does • The broadest overpricing of all assets yet recorded: Greenspan, the probabilities favor a reasonably good global equities, global bonds, and with a few market and a continued positive return to risk for a while exceptions, global real estate. longer. So the sun continues to shine, but the ice gets • By far the most important single market, U.S. equity, is thinner and thinner. particularly badly overpriced.

GMO Quarterly Letter – April 2004 6 Conclusion I wrote in the last quarter that our skating on thin ice dependence on the use of emerging equity and strategy – our steady rather than precipitous reduction in international small cap. risk taking – was subject to “changing data and loss of nerve”. Well, I believe the data has shifted slightly for the We realize that it can be very irritating to clients, potential worse and certainly my nerve has decreased. So, even clients, and consultants when we close funds, especially though the odds favor a few more months of reasonable without several months’ warning. Unfortunately, lengthy outperformance, our new recommendation would be to warnings can be dangerous as we found out with increase ‘quality’, cash, market neutral hedge funds, and emerging equity where the size of the inflow was very any other relatively safe investments faster than high. We apologize for any inconvenience. previously recommended, that is, to move to a highly The policy of closing is central to maintaining defensive portfolio by September 30 rather than year-end. performance for you (and us) and that, in turn, is central to how we see ourselves. As I’m fond of saying, there is Postscript: nothing more supremely useless than a mediocre money manager.

Closing GMO Funds to New Money My diatribe as to why closing is integral to decent Possible Use of a New Emerging Equity Fund performance was attached to my last quarterly letter. We GMO is seriously considering the start of a new emerging had originally closed Emerging Equity, Emerging equity fund that will not have the objective of creating Country Debt, the Foreign Small Fund, and our Emerging alpha (value added), but will provide for exposure to Country Debt Hedge Fund. International Small (our emerging equities. This fund, which would be priced quant version) was recently added to the list and on April appropriately cheaper than our active fund, will provide a 7, three more funds closed: International Equity solution to the capacity constraints of our active fund. We Allocation Fund and related strategies, Global Allocation will keep you posted. Some clients might also consider Absolute Return Strategy, and the Mean Reversion Long- the use of other firms’ emerging equity funds, at our Short Hedge Fund. These recent three were all closed suggested percentage weight, together with other GMO because of strong continuing inflow and their heavy funds.1 The foregoing does not constitute an offer of any securities for sale.

7 Quarterly Letter – April 2004 GMO