我姐姐的养老金资产与代理问题

GMO QUARTERLY LETTER April 2012 My Sister’s Pension Assets and Agency Problems (The Tension between Protecting Your Job or Your Clients’ Money) Jeremy Grantham The central truth of the investment business is that investment behavior is driven by career risk. In the professional investment business we are all agents, managing other peoples’ money. The prime directive, as Keynes1 knew so well, is first and last to keep your job. To do this, he explained that you must never, ever be wrong on your own. To prevent this calamity, professional investors pay ruthless attention to what other investors in general are doing. The great majority “go with the flow,” either completely or partially. This creates herding, or momentum, which drives prices far above or far below fair price. There are many other inefficiencies in market pricing, but this is by far the largest. It explains the discrepancy between a remarkably volatile stock market and a remarkably stable GDP growth, together with an equally stable growth in “fair value” for the stock market. This difference is massive – two-thirds of the time annual GDP growth and annual change in the fair value of the market is within plus or minus a tiny 1% of its long-term trend as shown in Exhibit 1. The market’s actual price – brought to us by the workings of wild and wooly individuals – is within plus or minus 19% two-thirds of the time. Thus, the market moves 19 times more than Exhibit 1 Long-Term Corporate Profits Are Very Stable and Seem to Offer Little Long-Term Risk Real S&P price vs. perfect foresight fair value*: 1882 – 2005 13.0 20 Volatility 12.5 2/3 of the Time GDP ± 1% 18 Real Price and Fair Value in Log Space 12.0 2/3 of the Time Fair Value ± 1% GDP 16 2/3 of the Time Prices ± 19% Real GDP in Log Space 11.5 11.0 Fair 10.5 Value 12 10.0 9.5 S&P Real Price 8 9.0 8.5 8.0 4 * Shiller model Source: GMO, Standard & Poor’s, Federal Reserve As of 12/31/05 1 John Maynard Keynes, The General Theory of Employment, Interest and Money, 1936.

is justified by the underlying engines! This incredible demonstration of the behavioral dominating the rational and the “efficient” was first noticed by Robert Shiller over 20 years ago and was countered by some of the most tortured logic that the rational expectations crowd could offer, which is a very high hurdle indeed. Shiller’s “fair value” for this purpose used clairvoyance. He “knew” the future flight path of all future dividends, from each starting position of 1917, 1961, and all the way forward. The resulting theoretical value was always stable (it barely twitched even in the Great Depression), but this data was widely ignored as irrelevant. And ignoring it may be the correct response on the part of most market players, for ignoring the volatile up-and-down market moves and attempting to focus on the slower burning long-term reality is simply too dangerous in career terms. Missing a big move, however unjustified it may be by fundamentals, is to take a very high risk of being fired. Career risk and the resulting herding it creates are likely to always dominate investing. The short term will always be exaggerated, and the fact that a corporation’s future value stretches far into the future will be ignored. As GMO’s Ben Inker has written,2 two-thirds of all corporate value lies out beyond 20 years. Yet the market often trades as if all value lies within the next 5 years, and sometimes 5 months. Ridiculous as our market volatility might seem to an intelligent Martian, it is our reality and everyone loves to trot out the “quote” attributed to Keynes (but never documented): “The market can stay irrational longer than the investor can stay solvent.” For us agents, he might better have said “The market can stay irrational longer than the client can stay patient.” Over the years, our estimate of “standard client patience time,” to coin a phrase, has been 3.0 years in normal conditions. Patience can be up to a year shorter than that in extreme cases where relationships and the timing of their start-ups have proven to be unfortunate. For example, 2.5 years of bad performance after 5 good ones is usually tolerable, but 2.5 bad years from start-up, even though your previous 5 good years are well-known but helped someone else, is absolutely not the same thing! With good luck on starting time, good personal relationships, and decent relative performance, a client’s patience can be a year longer than 3.0 years, or even 2 years longer in exceptional cases. I like to say that good client management is about earning your firm an incremental year of patience. The extra year is very important with any investment product, but in asset allocation, where mistakes are obvious, it is absolutely huge and usually enough.

What Keynes definitely did say in the famous chapter 12 of his General Theory is that “the long-term investor, he who most promotes the public interest … will in practice come in for the most criticism whenever investment funds are managed by committees or boards.” He, the long-term investor, will be perceived as “eccentric, unconventional and rash in the eyes of average opinion … and if in the short run he is unsuccessful, which is very likely, he will not receive much mercy.” (Emphasis added.) Reviewing our experiences of being early in several extreme outlying events makes Keynes’s actual quote look painfully accurate in that “mercy” sometimes was as limited as it was at a bad day at the Coliseum, with a sea of thumbs down. But his attribution, in contrast, has proven too severe: we appear to have survived.

You apparently can survive betting against bull market irrationality if you meet three conditions. First, you must allow a generous Ben Graham-like “margin of safety” and wait for a real outlier before you make a big bet. Second, you must try to stay reasonably diversified. Third, you must never use leverage. In my personal opinion (and with the benefit of hindsight, you might add), although we in asset allocation felt exceptionally and painfully patient at the time, we did not in the past always hold our fire long enough or be patient enough. It is the classic failing of value managers (and poker players for that matter) to get impatient and bet too hard too soon. In addition, GMO was not always optimally diversified. We are generally more cautious (or, if you prefer, “more experienced”) now than in 1998 with respect to, for example, both patience and diversification, and at least we in asset allocation always stayed away from leverage.3 The U.S. growth and technology bubble of 2000 was by far the biggest market outlier event in U.S. market history; we had previously survived the 65 P/E market in Japan, which was perhaps the greatest outlier in all important equity markets anywhere and at any time. These were the most stringent tests for managers, and we were 2 to 3 years early in our calls in both cases. Yet we survived, although not without some battle scars, with the great help that we did, in the end, win these bets and by a lot. Hypothetically, resisting the temptation to invest too 2 Ben Inker, “Valuing Equities in an Economic Crisis,” April 6, 2009.

3 Leverage can be interpreted quite broadly. For our asset allocation strategies, it means no borrowed money.

GMO 2 Quarterly Letter – My Sister's Pension Assets – April 2012 soon in 1931 may have been a tougher test of survival in bucking the market. Luckily we, and all value managers, were not around to be tempted by that one. (Although Roy Neuberger – who died in December 2010, unfortunately – was, and he could talk about it as lucidly as any investor ever.)

This exemplifies perfectly Warren Buffett’s adage that investing is simple but not easy. It is simple to see what is necessary, but not easy to be willing or able to do it. To repeat an old story: in 1998 and 1999 I got about 1100 full-time equity professionals to vote on two questions. Each and every one agreed that if the P/E on the S&P were to go back to 17 times earnings from its level then of 28 to 35 times, it would guarantee a major bear market. Much more remarkably, only 7 voted that it would not go back! Thus, more than 99% of the analysts and portfolio managers of the great, and the not so great, investment houses believed that there would indeed be “a major bear market” even as their spokespeople, with a handful of honorable exceptions, reassured clients that there was no need to worry. Career and business risk is not at all evenly spread across all investment levels. Career risk is very modest, for example, when you are picking insurance stocks; it is therefore hard to lose your job. It will usually take 4 or 5 years before it becomes reasonably clear that your selections are far from stellar and by then, with any luck, the research director will have changed once or twice and your deficiencies will have been lost in history. Picking oil, say, versus insurance is much more visible and therefore more dangerous. Picking cash or “conservatism” against a roaring bull market probably lies beyond the pain threshold of any publicly traded enterprise. It simply cannot take the risk of being seen to be “wrong” about the big picture for 2 or 3 years, along with the associated loss of business. Remember, expensive markets can continue on to become obscenely expensive 2 or 3 years later, as Japan and the tech bubble proved. Thus, because asset class selection packs a more deadly punch in the career and business risk game, the great investment opportunities are much more likely to be at the asset class level than at the stock or industry level. But even if you know this, dear professional reader, you will probably not be able to do too much about it if you value your job as did the nearly 1100 analysts in my survey. Except, perhaps, with your own assets or, say, your sister’s pension assets.

My Sister’s Pension Assets All of this brings me to my longest-lived investment: the pension of one of my sisters, which started very modestly in 1968, just after I got my first job in the investment business. Partly because the value of the assets was small and partly because I was more aggressive (or unimaginative) then, her pension assets were invested 100% in those equity mutual funds (always value and mostly small cap value) that I was involved with. However, the allocations within the portfolio changed from time to time, sometimes quite significantly. For example, as we entered the GreenspanBernanke over-stimulated market in the 1990s and onwards, she was notably underweight stocks on average, and hugely tilted to emerging markets and away from the U.S. after 1998 (as were all GMO strategies within their respective operational constraints).

Later, as GMO started to crank out 10-year forecasts in 1994 and build up a body of experience in asset allocation, my sister’s pension faithfully followed the forecasts, but I’m a little embarrassed to admit that on a risk-adjusted basis she has done a little better than our first dedicated institutional asset allocation client. There are two principal reasons for this. The first very large advantage for her is that I have only had to consider absolute return without the investment constraints some investors impose. I have felt absolutely no career risk. She is not going to fire me easily after 43 years as a mostly effective manager. In any case, she does not ask nor has she been told about investment changes or short-term performance for several years, and she is, after all, my sister. The complete lack of investment constraints and pressure from being fired gives me the greatest of all investment freedoms – the freedom to make very big asset bets when the numbers call for it as they did most notably from 1998 to 1999 and in 2007. For an institutional client, these conditions are impossible to match. (Before moving on I should tell you of my one experiment in investment communication with my sister. After a painful losing experience in emerging market equities, I felt I should mention it along with my confidence that it would eventually work out fine. On hearing of the loss and before I could provide any details, she very quickly cried out, “Sell! Sell!” right out of some 1929 movie. In spite of her pleas, I did no selling and, in the nature of emerging, it came storming back quite rapidly to brilliant new highs. So, to balance the Quarterly Letter – My Sister's Pension Assets – April 2012 3 GMO old bad news, I told her of our lightning-like gain only to be admonished with the same instantaneous exclamation, “Sell! Sell!” End of experiment.)

In dealing with clients as opposed to sisters, we have tried to adopt a range of asset allocation moves that, even when we are 2.5 years too early in extreme bull markets (bear markets tend to be much quicker), will leave the portfolio looking at least faintly normal and leave the clients’ pain just tolerable. Too big a safety margin and we are leaving too much money on the table; we are probably protecting our job rather than attempting to maximize our clients’ return. Too narrow a safety margin and clients may fire us, as some have done in the past. I believe this is not good for us or our clients, who tend to rebound into much different portfolios, often, given the circumstances of an extremely mispriced market, at a very inauspicious time. It is, of course, a central dilemma of investing. In the first 15 years of our asset allocation experience, our attempt to address this dilemma was to limit the range of our global equity shifts in our flagship Global Asset Allocation Strategy (formerly known as Global Balanced Asset Allocation Strategy) between a minimum of 50% and a maximum of 75%. We had tried to imagine what the typical investor – both individual and institutional – would consider to be at least faintly normal so that they would hang in when we would inevitably be too early in our market moves. That range – 50% to 75% – had seemed very conservative in theory but not so in practice as we learned in 1998 and 1999. It was then that P/E ratios, which had previously peaked at 20 times in 1929 and 19 times in 1964, moved up to an astonishing 33 times in early 1999. Failing to follow the crowd in this environment turned out to be uncomfortably beyond the tolerance of about 40% of our clients. Basically, they thought that we had been left behind because of our inability to see Greenspan’s new high plateau – a golden era in which he claimed that the internet and other technology would permanently change profitability and “probably” valuation levels as well. (Many of the largest asset allocation funds in the market today had the notable good sense to bypass this ultimate stress test by not existing in 1998 and 1999.) But 60% of our clients stayed. With hindsight, a philosopher might argue that if in a test of that magnitude – statistically over a 3-sigma event, or about a 1 in 1000 chance if it were a normally distributed world – you did not lose business, you were being too timid in general. That sounds reasonable, but few would volunteer to go through that bloodletting too often. For GMO, however, winning the bet attracted a flood of new business from 2003 to 2006 and, despite proving Keynes’s point about receiving “no mercy,” we seemed to have disproved the general thesis that rational investing could not survive an irrational market. Keynes himself, by the way, had not seen such irrational markets as those that occurred in the U.S. in 2000 or in Japan in 1989, either in person or in the history books. Encouraged by those two experiences and by the intensity and breadth of overpricing in the next bubble of 2007, which was truly global in nature, we pushed successfully to have the equity minimum moved down to 45% in our Global Asset Allocation Strategy, and we also made sure that our equities were, on average, more conservative. This time, our timing was better as the market moved down reasonably quickly to fair value in early 2009, resisting the temptation to repeat the crazy overpricing of 1999 and early 2000, despite the usual encouragement and over-stimulation from the Fed. Consequently, 2008 was a year of healthy outperformance against the strategy’s benchmark (+6.9% net of fees). Yet it still left us feeling quite dissatisfied, for our absolute return was -20.8% and it came in a year when we had felt extremely confident of a financial crunch and a severe market decline. By July 2008, for example, I had even thrown in the towel for my beloved emerging market equities and had advised our clients to “take as little risk as possible.” I suggested ignoring benchmark or career risk by reducing equity holdings to rock-bottom. I was, I wrote, “officially scared,” and I confessed to finding fundamentals far worse than I had expected. With such dire warnings and with real life turning out perhaps even worse, one can imagine we were a little unsatisfied with strong relative gains but painful actual losses. And salt was rubbed into my wounds by two other events. First, my sister’s pension assets, driven by exactly the same inputs as our professional accounts but carrying zero career risk and no benchmark at all – I perceived my job description for her was to make money when opportunities were good and protect money when opportunities were poor – was already down to 20% equities by late 2007. By July 2008, this allocation had ducked down to zero equities and not unreasonably so, in my opinion, because GMO’s highest 7-year forecast for any equity subset in October 2007 was a dismal 1.9% a year real and this number rose only slowly in the first few months of 2008.

GMO 4 Quarterly Letter – My Sister's Pension Assets – April 2012 The second and more important factor that increased our dissatisfaction with our 2008 “relative” win was the existence of our own professional version of an asset allocation strategy with “no benchmark,” a shorthand way for saying, I suppose, a strategy attempting to show much reduced benchmark risk, for surely every investment strategy (except perhaps my sister’s) has to make some comparisons somewhere.

In 1999 we offered a suite of asset allocation strategies that had no official benchmarks. One was deemed high risk, one medium risk, and one low risk. The original 1999 exhibit from our client conference is shown as Exhibit 2. It shows the imputed real returns (from our 10-year forecasts at that time) were in the range of 5% to 6% real, compared to the S&P 500 imputed return of 2.2%. Yet, such was the enthusiasm back then for all things bullish that we could sign no one up until 2001, and even then our approach was deemed so unusual that we were asked to match it up with a static 20% allocation to our conservative Multi-Strategy. This combined strategy was offered to institutions and was closed because of capacity concerns in 2004. But the 80% long-only component was, for accounting purposes, also run as a separate portfolio (GMO Benchmark-Free Allocation Strategy). It did well, handsomely beating our flagship Global Asset Allocation Strategy (see Exhibit 3). The main reason for this outperformance was precisely its freedom Exhibit 2 Achieving a 5% to 5.75% Real Return Using a Non-Traditional Portfolio 7% From 1999 GMO Fall Conference 5.75% 6% 5.5% Expected Real Return from Asset Class Emerging 5.0% Non-Traditional Equities 5% 19.3% Frontier REITs 15.0% 10.0% Emerging 12.3%% 10.0% 15.0% Debt 10.0% 8.1% 4% U.S.

16.9% 33.6% Bonds 27.7% Inflation 50.0% 22.1% Protected 3% Gov’t Bonds 50.0% 2% 1% Traditional Portfolio 0% 0% 2% 4% 6% 8% 10% 12% Risk (Annualized Volatility) Traditional Portfolio Non-Traditional Portfolios 5.0% 5.5% 5.75% (65% Global Equities, Real Real Real 35% U.S. Bonds) Return Return Return Return 2.0% 5.0% 5.5% 5.8% Risk 10.4% 4.7% 6.8% 8.8% Probability <0% return: over 3 year 36.9% 2.6% 7.0% 11.7% Note: Based on GMO’s 10-year asset class return forecasts. These forecasts above were, at the time they were made, forward-looking statements based upon the reasonable beliefs of GMO and were not a guarantee of future performance. Forward-looking statements speak only as of the date they are made, and GMO assumes no duty to and does not undertake to update forward-looking statements. Forward-looking statements are subject to numerous assumptions, risks and uncertainties, which change over time. Actual results could differ materially from those anticipated in forward-looking statements.

Source: GMO As of 9/30/99 Quarterly Letter – My Sister's Pension Assets – April 2012 5 GMO to be much less invested when the market was overpriced (remember, our 7-year-forecasts for equities were very low). I will point out that this last decade was a very favorable one: bumpy enough, as in 2008, to give asset allocation something to play against, but not really crazy, as the tech bubble had been. Alternatively, this strategy could not hope to do very well in a quiet decade, absent several wild over- or under-valuations, should there indeed be such a decade again. It would also have been badly beaten by aggressive portfolios during the tech bubble of 2000. You will also note the usual GMO tendency to do most of its heavy lifting when investment times are bad.

Exhibit 3 Global Asset Allocation Strategy and Benchmark-Free Allocation Strategy Performance 160% Benchmark-Free Allocation Strategy +151% 140% 120% Performance (CPI-Adjusted) 100% 80% Global Asset Allocation Strategy +72% 60% 40% +30% 20% 0% Global Asset Allocation Benchmark* -20% Dec- 01 02 03 04 05 06 07 08 09 10 11 The performance of the Benchmark-Free Allocation Strategy appearing in the chart above shows the past performance of the Benchmark-Free Allocation Composite (the “Composite”) which consists of accounts and/or mutual funds managed by Grantham, Mayo, Van Otterloo & Co. LLC (“GMO”). The Composite is comprised of those fee-paying accounts under discretionary management by GMO that have investment objectives, policies and strategies substantially similar to the other accounts included in the Composite. Prior to January 1, 2012, the accounts in the Composite served as the principal component (approximately 80%) of a broader real return strategy** (which has its own GIPS composite) pursued predominantly by separate account clients of GMO. The Benchmark-Free Allocation Strategy is not expected to differ significantly from that component of the broader real return strategy. It is expected that the strategy’s investment exposures will not differ significantly from the allocations the strategy would have had as a component of the broader real return strategy, although the strategy will likely allocate a greater percentage of its assets to the strategies that have cash-like benchmarks. Not all of the accounts included in the Composite may be mutual funds; however, all the accounts have invested their assets in other mutual funds. All of the accounts that make up the Composite have been managed by the Asset Allocation Division. Although the mutual funds and the client accounts comprising the Composite have substantially similar investment objectives and strategies, you should not assume that the mutual funds or the client accounts will achieve the same performance as the other accounts in the Composite. The client accounts in the Composite can change from time to time. The performance of each account may differ based on client specific limitations and/or restrictions and different weightings among the mutual funds. Performance data quoted represents past performance and is not predictive of future performance. Net returns for the Benchmark-Free Allocation Strategy are presented after the deduction from the composite’s gross-of-fee returns of a model management and shareholder service fee. For the Global Asset Allocation Strategy, net returns represent the weighted average of the net-of-fee returns of all accounts within the composite. Net returns include transaction costs, commissions and withholding taxes on foreign income and capital gains and include the reinvestment of dividends and other income, as applicable. A GIPS compliant presentation of composite performance has preceded this presentation in the past 12 months or accompanies this presentation, and is also available at www.gmo.com. Actual fees are disclosed in Part II of GMO’s Form ADV and are also available in each strategy’s compliant presentation. The information above is supplemental to the GIPS compliant presentation that was made available on GMO’s website in April of 2011. * The Global Asset Allocation Benchmark is comprised of a weighted average of account benchmarks. Many of the account benchmarks consist of S&P 500, MSCI ACWI ex-U.S. and Barclays U.S. Aggregate or some like proxy for each market exposure they have. For each underlying account benchmark, the weighting of each market index may vary slightly (generally 65% MSCI ACWI, 35% Barclays U.S. Aggregate). The index is internally blended by GMO and maintained on a monthly basis. ** For the broader real return strategy, as of 2/29/12: Net cumulative return = 175.18%. Net year-end performance is: 12/31/01: 0.16%; 12/31/02: 8.80%; 12/31/03: 34.20%; 12/31/04: 15.29%; 12/31/05: 13.54%; 12/31/06: 11.01%; 12/31/07: 9.99%; 12/31/08: -6.61%; 12/31/09: 13.41%; 12/31/10: 2.72%; 12/31/11: 4.22%.

Source: GMO As of 2/29/12 GMO 6 Quarterly Letter – My Sister's Pension Assets – April 2012 So, how does the Benchmark-Free Allocation Strategy compare to our original Global Asset Allocation Strategy? This is where I must take yet another detour to talk about measures of investment efficiency, or, Sharpe Ratios.

Sharpe Ratio Managing Sharpe Ratios at the portfolio level has been Portfolio Management 101 for most of my investment career, but in real life it has been used to a negligible degree. The Sharpe Ratio is a measure of how many units of price volatility an investor has received in the past per unit of return. Exhibit 3 shows that the benchmark for our Global Asset Allocation Strategy, which is 65% global equities and 35% U.S. bonds, delivered a net real return for the last 10 years of 2.7% per year and had a volatility of ±11.6% (at 1 standard deviation or two-thirds of the time). It is therefore considered to have a Sharpe Ratio of under .25 – that is to say, it delivers more than 4 units of volatility for 1 unit of return. The Global Asset Allocation Strategy, in contrast to its benchmark, returned +5.4% net of fees, with a volatility that was 20% less, or ±9.2% a year. It has had, therefore, a Sharpe Ratio of 0.6, or less than 2 units of volatility per unit of return. This means that the strategy has been more than twice as efficient, if you will, as its benchmark, delivering over twice the return per unit of volatility (often referred to, rather dangerously, in my opinion, as “risk”). With this as background, for the last 10 years the Benchmark-Free Allocation Strategy has delivered 1.7 times the yearly return of Global Asset Allocation Strategy with 10% less volatility. This gives it a Sharpe Ratio of 1.1, or nearly twice the efficiency of the Global Asset Allocation Strategy and over four times that of the balanced benchmark. I will add that I believe that clients’ enthusiasm for all long-only investment products has been geared almost entirely to the “raw” return. That part of the increased efficiency that is due to lower volatility has been, in my experience, more or less ignored. This obviously makes it commercially dangerous to offer a strategy that can be caught out twice as badly as an existing strategy on those occasions when the market rallies a lot from an already overpriced level, which it can quite easily do from time to time. A theoretical example of such pain would have been a 2008 that was up another very unexpected 30% before collapsing, say, in 2009. A real example was 1999, when the market really did rally strongly from an already record overpriced level. That said, I believe the concept of the Sharpe Ratio is one of the few aspects of “modern portfolio management” that is useful at the level of a global balanced portfolio. It is a reasonable, although short-term, measure of the chance of real loss of money. “Information Ratio” or “benchmark risk” is, in contrast, very widely used. These measure how much you deviate from the benchmark per unit of extra return. In other words, it measures career risk: the risk of embarrassing your boss and losing your job. It is no wonder, perhaps, that the Sharpe Ratio – the risk to the ultimate beneficiary, the pensioner, say – is more or less ignored.

In a nutshell, I am pleased to say that we can now offer an investment strategy that reflects little career risk. It is the Benchmark-Free Allocation Strategy, and it is now available on a stand-alone basis, independent of the real return strategy of which it has been a part for the past 10 years. Our willingness to make asset class bets has enabled this strategy in this past particularly dangerous decade to do very well because it won its big bets. To state the obvious in the interests of very full disclosure, there was clearly some risk that it would not have won its big bets, in which case, of course, performance would have been considerably worse. To make such big bets, it is vitally important to have real confidence in the very strong historical tendency for extremes in financial enthusiasm or pessimism to move back to normal. Which confidence, thankfully, we have.

At this point, a good question from any reader who has been paying close attention is: why do we think this strategy can survive the client’s standard patience test? Well, an accurate answer is that it may not. We have tried to improve the odds by branding the strategy as benchmark-free. It is intended to protect capital first and yet still make good money. But Keynes knew, as I know, that in a 1999-type frenzy it would be all too easy to imagine someone at the client end looking down the performance list and seeing the +47%, +31%, and +24% of bullish competitors and then GMO’s +12%. In the heat of the battle, his memory of longer-term performance and job descriptions fades, and the response, “Who hired that +12% guy and why is he in the portfolio?” could easily be heard. We, perhaps fondly, hope that our surviving and eventually winning previous tests might be remembered. And, who knows? But even if frequently fired on occasion, it is probably worth it. After all, the great opportunities only exist because career Quarterly Letter – My Sister's Pension Assets – April 2012 7 GMO and business risk really, really matter and they only matter because of the pain that accompanies them. So, in the end, the urge to have the best long-term record that we can have and the feeling that this is the highest and best use of our patience, forecasting ability, and, yes, willingness to lose business along the way, has won out, giving us the conviction to offer Benchmark-Free Allocation Strategy as a stand-alone strategy. This is, I believe, the first time I have written specifically and in some detail about a single GMO strategy and it is likely to be the last. My excuse is that of all of the investment issues close to my heart, this one – career risk – is in my opinion the most important. The existence of the strategy will do at least two things. First, it will substantially replace my own efforts to manage my sister’s pension assets. (Although I will retain the ability to go, on rare occasions, the extra 10% short, using futures, if only to rub it in that she has absolutely no career risk.) Second, it can, by virtue of its willingness to make occasional very big bets against markets that appear very overpriced, give clients great opportunities to fire us enthusiastically from time to time when our timing is off. Bear in mind – as such clients will not – that although we are taking enormous career or business risk (and, admittedly, passing it on to clients … ah, there’s the rub), the extra risk, I believe, is career risk, or benchmark risk, not real risk. I believe that this strategy, like roughly 90% of GMO’s long-only strategies relative to their benchmarks, takes considerably less real risk. This is reflected in its (and their) high Sharpe Ratio. Its “risk” has been that of bad underperformance in bull markets. Real risk is the risk of a permanent loss of capital as my colleague James Montier likes to call it. The cardinal rule is to not underperform in bear markets. And though it may be a cardinal rule, there are, as we all know, no useful guarantees in our business.

Investment Outlook From now on, my letter will focus on a particular issue every quarter as it has increasingly over the last few years. Sometimes I have also covered a broad investment outlook, but sometimes I have given only cursory comments on nearer-term bread and butter issues, which can be unsatisfactory for some readers. To remedy this, Ben Inker, the leader of our asset allocation group and general portfolio manager, will take up this role. So I will comment on whatever I like, and he will attempt to make sure we cover most, if not all, of the important investment issues. Some people get the good jobs and some people don’t! Ben’s comments follow as a separate section of this letter.

PSFalse Pretenses It is easy for a spokesperson to receive credit for the work of a team, and this has often been the case for me. GMO’s asset allocation process has always been a team effort; indeed, for much of our history, the ability of the individual GMO strategies to beat their benchmarks contributed more to the success of our asset allocation strategies than did our movement of the assets.

In the allocation piece itself we have always depended on the portfolio management skills, particularly sizing and risk control, of Ben Inker and he has been commander in chief of our group for more than 10 years. (My attitude toward work has always been to delegate early and often.) In the idea generation part of asset allocation, we have always tried to be a democratic, idea-driven group and as our aspirations grew, so did our need for brain cells and expertise: our asset allocation brainstorming team has grown from 4 to 25 members over the last 5 years. We want to be well-informed on almost every opportunity to make or save money by moving assets.

Performance data quoted represents past performance and is not predictive of future performance. Returns are presented after the deduction of management fees and incentive fees if applicable. Net returns include transaction costs, commissions and withholding taxes on foreign income and capital gains and include the reinvestment of dividends and other income, as applicable. A GIPS compli-ant presentation of composite performance has preceded this presentation in the past 12 months or accompanies this presentation, and is also available at www.gmo.com. Actual fees are disclosed in Part II of GMO’s Form ADV and are also available in each strategy’s compliant presentation. The performance information for the Global Balanced Asset Allocation Strategy is supplemental to the GIPS compliant presentation that was made available on GMO’s website in April of 2011.

Disclaimer: The views expressed are the views of Jeremy Grantham through the period ending April 18, 2012, and are subject to change at any time based on market and other conditions. This is not an offer or solicitation for the purchase or sale of any security and should not be construed as such. References to specific securities and issuers are for illustrative purposes only and are not intended to be, and should not be interpreted as, recommendations to purchase or sell such securities.

Copyright © 2012 by GMO LLC. All rights reserved.

GMO 8 Quarterly Letter – My Sister's Pension Assets – April 2012 GMO COMMENTARY April 2012 Force Fed Ben Inker Over the years, we at GMO have certainly done our share of Fed bashing. Most of our complaints have centered on the way in which overly accommodative monetary policy and a refusal to see the dangers of, or even the existence of, asset bubbles can lead to economic problems. We’re about to pile on the Fed again, but this time it’s personal. Our major complaint about Fed policy is not about the risks today’s ultra-loose monetary policy imposes on the global economy (which are considerable1), but rather the fact that Fed policy makes it tricky for us to know whether we are doing the right thing on behalf of our clients.

One thing that we can say about the 2000 and 2007 asset bubbles is that, while they may have done significant damage to the economy and investors’ wealth, it was at least simple for us to know what to do with our portfolios. If we avoided the overvalued assets (which in 2007 was pretty much everything risky) we knew we were doing the right thing. Of course, even when investing is simple, it isn’t necessarily easy. In both episodes, but particularly 2000, the conservative portfolios we were running underperformed until the bubbles burst, causing plenty of consternation for our clients in the process.

Today, the Fed has engineered a situation in which the really unattractive asset classes are the ones we have always thought of as low risk: government bonds and cash. And unlike the internet and housing bubbles, this time it isn’t a quasi-inadvertent side effect of Fed policies, but a basic aim of them. The Fed has repeatedly said that a central part of the goal of low rates and quantitative easing is the creation of a wealth effect by pushing up the price of risky assets. By keeping rates very low and taking government bonds out of circulation, the Fed is trying to entice investors into buying risky assets. The question we are grappling with today is whether we should take the bait. So what makes this different from 2007? We’ve got some very unattractive assets and some others that look a good deal better by comparison. The trouble is that if those unattractive assets are cash and bonds today, moving to the relatively attractive assets involves increasing portfolio risk, whereas in 2007, moving away from risky assets lowered portfolio risk. In 2007, we could hold a portfolio that, whether assets took 7 years to revert to fair value or reverted tomorrow, we would still outperform. This was reassuring, because even though we use a 7-year reversion period in our forecasts, we know that the timing of mean reversion is highly uncertain. Today, the portfolio you would want to hold if assets were going to mean revert immediately is quite different from the one you would hold if you believed it would take 7 years to get back to fair value.

Perhaps the easiest way to think about the problem is to look at the efficient frontier for long-only absolute return portfolios, which we have reviewed on a number of occasions over the years (see Exhibit 1).

By their nature, efficient frontiers are upward sloping with regard to risk. The major ways in which they change over time is the slope of the line and the level of the line. As investors, we all want the frontier to be high on the chart, which will occur when asset classes are generally priced to give strong returns. When the line is steep, it means you are getting paid a lot for taking on additional risk. Today’s line (in black) is very low, but reasonably steep. That 1 Specifically, the Fed’s policy of zero short rates and quantitative easing creates the potential for an inflation problem if they cannot remove the accommodation fast enough when the financial system is back to functioning normally. In the meantime, the extremely low cost of leverage encourages speculation, the misallocation of capital, and encourages the formation of asset price bubbles in the U.S. and around the world.

Exhibit 1 Absolute Return Portfolios Over Time Higher Risk Portfolios (more Emerging and International) 14% 12.8% 13% 2/2009 Frontier 12% 11% Expected Real Return with Alpha 10% Lower Risk Portfolios (more Fixed Income) 9% 8.1% 8% 7.0% 6.7% 7% 2/2012 Frontier 6% 5% 4.4% 6/2007 Frontier 4% 4.8% 4.3% 3.2% 4.3% 3% 2% 3% 4% 5% 6% 7% 8% 9% 10% 11% 12% 13% 14% 15% Risk (Annualized Volatility) Note: Based on GMO’s 7-year asset class return forecasts. These forecasts are forward-looking statements based upon the reasonable beliefs of GMO and are not a guarantee of future performance. Forward-looking statements speak only as of the date they are made, and GMO assumes no duty to and does not undertake to update forward-looking statements. Forward-looking statements are subject to numerous assumptions, risks and uncertainties, which change over time. Actual results could differ materially from those anticipated in forward-looking statements.

Source: GMO As of 2/29/12 means you are getting paid to take risk, but the reason is not very high returns to risky assets but very low returns to low risk assets.

Traditional quantitative analysis tends to assume that the level of the line is irrelevant. You might wish that the line was higher, but you can’t do anything about that. All you can do is decide how far out on the frontier you want to be, which is a function of your risk tolerance and the slope of the line. So, if we believe in our 7-year asset class forecasts, some reasonable forecast of volatility, and nothing else, we should be pushing out on the frontier, which is what Bernanke wants us to do.

In reality, the portfolios we are running for our clients are not particularly far out on the frontier. We are running at “normal” levels of risk or slightly lower than that. Why? You could chalk it up to a knee-jerk reluctance to do what a central banker is telling us, but when the asset allocation team is debating the appropriate level of risk to take in our portfolios, Bernanke’s name does not generally come up. The reason for our reticence to move out on the risk spectrum really comes from our idea of what drives portfolio risk. We believe strongly that the risk of an asset rises with its valuation. Stocks at fair value are less risky than stocks trading 30% above fair value because the expensive stocks give you the risk of loss associated with falling back to fair value. That risk – “valuation risk” to use my colleague James Montier’s terminology – leads to losses that should not be expected to reverse themselves anytime soon. A cheap asset can certainly go down in price, but when it does, you should expect either high compound returns from there, which make your money back steadily, or a reasonably sharp recovery when the conditions that drove prices down dissipate, which will make your money back quickly. The loss is therefore temporary, although it may GMO 2 Commentary – Force Fed – April 2012 seem unpleasant while it is occurring. When an expensive asset falls back to fair value, subsequent returns should only be assumed to be normal, which means that the loss of wealth versus expectations is permanent. Today, stocks are expensive relative to our estimate of long-term fair value. The trouble is, so are bonds and cash. If everything was guaranteed to revert to the mean over 7 years, we would hold equity-heavy portfolios, because the gap between stocks and either bonds or cash is wider than normal. But we don’t know that it will take 7 years. Because cash and (most) bonds have a shorter duration with regard to changes in their discount rate than stocks do, fast reversion would lead to smaller losses for them than for equities. Holding a portfolio where we are crossing our fingers that mean reversion will be slow is difficult to be excited about and, as a result, we are lighter on equities than the 7-year forecasts would otherwise suggest. That leaves us around 63% to 64% in equities for a portfolio managed against a 65% equities/35% bonds benchmark and 48% to 58% in equities for absolute return oriented portfolios, depending on their aggressiveness and opportunity set. On the government bond side, given the incredibly low yields around, the only bonds we have much fondness for are Australian and New Zealand government bonds, because only those countries give a combination of a decent real yield and government spending policies that are sustainable in the long run. But our appetite for even these bonds is not great, leaving us with significant holdings of cash and “other.” If our 7-year forecasts play out exactly to plan, we are leaving some money on the table by not moving more heavily into stocks. However, our positioning does leave us in a place where we need not fear the circumstance whereby asset classes revert to fair value faster than expected, and that does help us sleep better at night.

Mr. Inker is the head of asset allocation.

Disclaimer: The views expressed herein are those of Ben Inker as of April 18, 2012 and are subject to change at any time based on market and other conditions. This is not an offer or solicitation for the purchase or sale of any security and should not be construed as such.

Copyright © 2012 by GMO LLC. All rights reserved.

Commentary – Force Fed – April 2012 3 GMO