|
About IndexInvestor.com |
Privacy Policy |
Transaction Policy |
Legal Disclaimers |
Contact Us |
My Account |
Home
|
|||||
|
|
|
|
|
||
Tom Coyne, Editor
tcoyne@indexinvestor.com
The month of September is always an important one when it comes to understanding what is happening in the world economy. It is not only the month when the IMF publishes its semi-annual World Economic Outlook (WEO), and annual Global Financial Stability Report (GFSR), but also one that usually sees important analysis from other institutions (e.g., the Bank for International Settlements, the Centre for Economic Policy Research, the Levy Economics Institute, the Institute for International Finance) and global publications (e.g., The Economist and the Financial Times).
This analysis will begin with a brief summary of our two baseline economic scenarios, as they existed at the beginning of September. We will then review this past month's developments, and discuss their impact on our scenarios. We will then conclude with a discussion of the implications of our scenarios for asset allocation decisions.
Both of our scenarios begin with the major problem facing the world economy today: the large current account deficit in the United States, and its counterpart current account surpluses in other countries (principally in Asia and more recently in OPEC). By definition, the current account balance equals the difference between domestic savings and investment in a country. Where domestic savings exceed domestic investment, a current account surplus will result, as surplus savings are sold to buy the foreign currency needed to make investments abroad. This depresses the exchange rate, and makes exports from the surplus country more competitive in the global economy, while making imports more expensive. The result is a current account surplus. Many Asian countries have been in this situation in recent years.
In contrast, where domestic investment exceeds domestic savings, a current account deficit results, as foreign savings are used to buy the domestic currency needed for domestic investment. The resulting appreciation of the savings-short country's exchange rate makes its exports less competitive, while making imports cheaper, leading to a current account deficit. Many Anglo Saxon countries, and above all the United States, have been in this situation in recent years.
The balance between domestic savings and investment can be further divided into the net balances in the public sector and the private sector (households and business). For example, the IMF projects a 2005 current account deficit in the United States equal to an astounding (6.1%) of its Gross Domestic Product (GDP). In turn, this reflects the combination of a public sector deficit equal to (3.7%) of GDP, and a private sector deficit equal to (2.4%) of GDP. The public sector deficit reflects the combination of tax cuts with heavy increases in defense and social spending, made worse by the cost of recovering from hurricanes Katrina and Rita. The private sector deficit primarily reflects the continued borrowing by American consumers (often secured against the rising value of their houses) to finance the spending upon which the world economy has become dangerously dependent.
The imbalances that exist in the world economy today can be summed up in the following table, which summarizes the state of play in regions that account for about two thirds of the world's Gross Domestic Product.
| Country | GDP % of World | GDPDomestic Demand Growth in 2005 | Public Sector Balance in 2005 (% of GDP) | Private Sector Balance in 2005 (% of GDP) | Current Account Balance in 2005 (% of GDP) |
| United States | 20.9% | 3.5% | (3.7%) | (2.4%) | (6.1%) |
| Eurozone | 15.3% | 1.4% | (3.0%) | 3.4% | 0.4% |
| China | 13.2% | 3.5% | (1.7%) 1 | 7.8% | 6.1% |
| Japan | 6.9% | 2.0% | (6.7%) | 10.0% | 3.3% |
| Asian NICs 2 | 3.5% | 3.1% | (1.5%) | 7.0% | 5.5% |
| United Kingdom | 3.1% | 1.6% | (3.2%) | 1.3% | (1.9%) |
| Canada | 1.9% | 4.6% | 0.5% | 1.0% | 1.5% |
| Australia | 1.6% | 3.2% | 1.3% | (7.0%) | (5.7%) |
As you can see, the world economy has become very dependent for its demand growth on the United States and China. However, domestic demand growth in the United States has been based on a combination of consumer and government borrowing (along with relatively loose monetary policy) that is, in the long run, unsustainable. Quite simply, neither U.S. consumers, nor government, and certainly not both together can forever spend beyond their means by borrowing from the rest of the world.
Unfortunately, domestic demand growth in China looks even less stable. Behind the 7.8% private sector balance lie enormous internal imbalances. Over the past few years, domestic investment in China has been growing at a torrid pace. However, much of this has been financed by state-owned banks, whose primary goals have been more political (maintain fast economic growth, which is the only remaining basis for the Communist Party's claim to leadership legitimacy) than economic (make good loans at a profit). The result has been an accumulation of uncompetitive investments, and a matching mountain of non-performing loans on the books of China's state owned banks. Those with even moderately long memories (which, sadly, rules out too many hedge fund managers) recall that they've seen this movie before: say, in Japan in the run up to its crash and depression, and in the run up to the late nineties crashes in the Asian NICs. The dynamic at work isn't new: excessive lending leads to overcapacity, which leads to falling prices, recession, debt deflaton, and depression. Unfortunately, in China this process will only be made worse by the household sector's already very high savings levels, which reflect the rising costs of education, healthcare, and housing, along with a rapidly aging population and only rudimentary government old age pensions. Unfortunately, given the apparent commitment of the Communist Party leadership to retain power at any cost, the triggering of a vicious downward economic cycle could easily lead to an attack on Taiwan (it's also an old story: when threatened, focus the people's attention on an external enemy).
The essence of the problem confronting the world economy -- and investors -- is this: we all know the current system must come to an end. What we don't know is how or when this will happen (although the sharp rise in oil prices seems to be hastening that day's arrival). In the back of everyone's mind is an optimistic scenario one might call "the grand bargain." In it, the governments of the world come together, and agree to the following: (a) a reduction in the U.S. public and private deficits; (b) a fall in the U.S. exchange rate; (c) an increase in exchange rates and domestic demand in the Eurozone, Japan, and NICs; and (d) hope that this keeps the wheels from coming off in China. We don't put too much stock in the "grand bargain" scenario, for the simple reason that too many things have to go right for it to work -- and the joint probability of all those things happening appears to be quite low. This is especially true given the political pain that many of these measures would entail. If it happens at all, we think it will take a severe crisis to get leaders to seriously consider it, and by then events may have gotten too far out of control for a grand bargain to work. That leaves us with our two principal scenarios.
In our "conventional wisdom" scenario, the global economy continues to "muddle through" for a while longer without a major crisis. Logically, this requires continued demand growth in the United States and China, which fundamentally depends on the continued willingness of foreign investors (especially Asian central banks) to keep accumulating dollar denominated assets to finance the United States current account deficit.
Our most dangerous scenario is characterized by a sudden rush out of dollar investments. There is no telling for sure what could cause this to happen. It might be a sharp drop in U.S. consumer spending, triggered by the rise in oil prices or a slowdown in the increase in housing crises (for an example of the latter, one need look no further than what is happening in Australia or the Netherlands today). Or it might be caused by foreign (and some domestic) investors' declining confidence in the ability of U.S. political leaders to resolve that country's fiscal problems (particularly those related to the exploding costs of Social Security and Medicare). There has been some debate over the likely results of a rush out of U.S. dollar assets. The dominant view seems to be that is would lead to a sharp rise in nominal U.S. interest rates (as bonds are sold, their prices fall, which causes their yields to rise). Domestically, this would either trigger or reinforce a fall in housing values, leading to a wave of bankruptcies by heavily leveraged consumers and a crash stop in their spending, which would plunge the U.S. into a deep, and quite possibly deflationary recession. Internationally, this would trigger a sharp drop in export demand in all those other regions whose growth is heavily dependent on the United States. Most dangerously, this would include China, where the sharp slowdown in growth could easily trigger the bursting of domestic property bubbles and increased social unrest, with very unpredictable consequences.
A minority view was recently put forth in a paper by Joseph Gagnon of the U.S. Federal Reserve ("Currency Crashes and Bond Yields in Industrial Countries"). He notes that a sharp rise in interest rates doesn't necessarily have to accompany a currency crash, provided that investors have sufficient confidence in the Central Bank's commitment to keeping inflation under control. Unfortunately, upon closer examination, his argument seems to weaken. First, numerous analysts have pointed out that even a sharp depreciation of the U.S. dollar cannot, on its own, significantly reduce the current imbalances. You simply cannot escape the need for increased domestic demand growth in the Eurozone and Asia to offset a reduction in U.S. domestic demand. Absent the former (and the increase in U.S. exports it would generate), the necessary reduction in U.S. domestic demand would be so deep as to practically guarantee the development of a global crisis (for two good recent analyses of this issue, see "The End of Large Current Account Deficits" by Sebastian Edwards of UCLA, and "The U.S. and Her Creditors: Can the Symbiosis Last?" by Godley, Papadimitriou, Dos Santos and Zezza).
We think the most interesting question is how the "end game" of that crisis would play out. Suppose you are a U.S. political leader, and you are faced with this choice. One way to get out of the deep recession (and quite possibly deflation) I'm facing is to crank up the money supply and inflation. This will hurt bondholders (at least those who own nominal return bonds), but will help all those voters who bought their houses with fixed rate mortgages. The other alternative is to maintain monetary discipline (and the wealth of bondholders), even if it means the economy must endure a prolonged recession (which will cause a lot of voters to lose their homes when they lose their jobs and can't make payments on their mortgages). If push comes to shove, our view has been that the vast majority of political leaders will choose the inflationary route out of the crisis.
Finally, in addition to our main scenarios, we have been increasingly concerned with the actions of microscopic actors who could also have a sharp to severe negative impact on the world economy in the months ahead. We refer, of course, to the H5N1 strain of influenza virus, more commonly known as "bird flu." Recent developments in this area are mixed. On the one hand, more and more governments seem to be taking quite seriously the threat of a global flu pandemic. On the other hand, there are many questions as to whether government actions add up to "too little, too late." And in the meantime, H5N1 seems to be progressing (or, more accurately, evolving) towards the pandemic stage (for a good summary of recent developments, see www.recombinomics.com).
As a means of identifying in advance the scenario that may be developing, we have developed this list of early warning indicators:
| Indicator to Watch | Dangerous Trend |
| Real Interest Rates | Falling trend (sign of worry about long term economic growth slowdown) |
| Oil Prices | Remain high and/or rise higher (raises probability of global growth slowdown) |
| U.S. Ten Year Treasury Bond Nominal Yield | Rising trend (raises probability of housing and demand slowdown) |
| U.S. Dollar Exchange Rate | Falling trend (weakening dollar, which should trigger higher U.S. interest rates and slowing economy) |
| Price Instability in China | Excessive inflation or deflation is a leading indicator of rising economic pressures with high potential for crisis. |
| Political Instability in China | Reports of growing political unrest, which could also lead to sharp economic slowdown and/or more external tension with United States and/or Japan |
| Real Domestic Demand Growth in the Eurozone | Falling trend (no sign of region replacing U.S. as global growth engine) |
| Real Domestic Demand Growth in Japan | Falling trend (no sign of region replacing U.S. as global growth engine) |
| H5N1 Pandemic Influenza | Signs of increased communicability between humans, with no reduction in currently high mortality rates. Pandemic could result in sharp, and possibly prolonged reduction in global economic growth. |
Now let's look at what has happened in the past few weeks. On the good news front, China has made its exchange rate slightly more flexible, which should, if nothing else, moderate the rising political voices in the United States calling for protectionist measures. The election in Japan seemed to give Prime Minister Koizumi a clear mandate to aggressively move forward with reforms that may lead to increased domestic demand. The continued restructuring of Japan's financial sector will also help in this regard. Another piece of good news was the release of another Federal Reserve Board report ("Assessing High House Prices: Bubbles, Fundamentals, and Misperceptions" by Himmelberg, Mayer and Sinai) making the case that, for the most part, housing prices in the United States had not yet reached "bubble" territory.
Unfortunately, that was pretty much it on the good news front (unless you count rising government interest in flu preparedness). There were also more than a few pieces of bad news. The German elections seemed to guarantee a slowdown in that nation's structural reform efforts, and hence made a recovery in domestic demand less likely. Oil prices continued to rise, which cannot help but lead to reductions in domestic demand at some point in the future. Rising oil prices will also put upward pressure on inflation, which, along with rising government deficits in the Eurozone, will probably make an interest rate cut by the European Central Bank even less likely than it was before. Oh, and did I mention that migratory birds are bringing H5N1 to Europe? To put it mildly, the signs do not point to a rise in domestic demand in that region. In Japan, Koizumi's re-election triumph was balanced by the release of a new survey by the Bank of Japan showing that deep pessimism among the Japanese public that their nation's economic recovery could be sustained. In China, Xinhua news agency ran a number of stories noting a slowdown in domestic demand, and growing signs of deflation caused by overcapacity in many industries. Rising oil prices have caused increases in inflation in Southeast Asian nations, which has led some Central Banks in the region to raise interest rates. That makes increased domestic demand growth in this region less likely than before. And if any place has an H5N1 problem today, it is Southeast Asia. On balance, we'd say that there was more bad than good news over the past month.
Now let's take a look at what the IMF had to say about the condition of the global economy today. The September World Economic Outlook begins by noting that while "the [global] expansion remains broadly on track…risks are still slanted to the downside…Global current account imbalances -- a key medium term risk to the outlook -- have increased yet again…Associated with this, growth divergences across regions remain wide. The expansion has continued to be led by the United States and China, where the growth momentum has remained robust. Growth projections for 2005 in most other regions have been marked downward, with the important exceptions of Japan and India. The renewed weakness in the euro area is of particular concern…Oil prices have continued their ascent, hitting a new nominal high of $65 a barrel in late August…Markets remain concerned that the current very low spare production capacity will be insufficient to meet demand growth next winter…High and volatile oil prices remain a significant global risk…The tentative expansion in the euro area has faltered once again…[with] domestic demand growth slowing considerably…The consistent overestimation of the strength of the expansion in the euro area in recent eyares, as well as the fact that corporations as yet show little sign of investing their now substantial profits, underscore the risks of a more extended period of weakness…Progress with structural reforms to boost domestic demand in Japan and the euro area has been limited…Among the major industrial countries…fiscal pressures from aging populations remain a serious concern, especially for health care." In sum, the IMF seems to have the same overall view as everyone else: while the problems are clear, it remains uncertain when and how they will eventually be resolved.
Elsewhere in the World Economic Outlook, the IMF presents a very interesting analysis of an issue near and dear to our hearts: why, in many countries, the real interest rate is so far below its long term average. The dominant view of the problem is known as the "savings glut" theory. It posits that the cause of real interest rates has been an increase in the supply of savings, primarily in Asia, reflecting the ongoing reaction of that region to the sharp financial crisis it experienced at the end of the 1990s. The IMF's analysis reaches a different conclusion. In addition to an increase in Asian savings, it finds that two other forces were also at work. The first was a fall in the savings rate in many industrial countries. This was due to a fall in both public savings (i.e., a rise in public sector deficits), and in private savings. The former reflects these governments continuing political inability to balance pressures to spend with their capacity to generate tax revenue. The fall in private savings seems to have been significantly affected by the increased availability of credit and the rise in housing values (with rising asset value substituting for savings out of income).
However, more important than changes in savings levels has been a fall in the overall level of investment. As the IMF notes, "investment rates have fallen across virtually all industrial country regions, although this has been most noticeable in Japan and the euro area countries, where they reached historic lows in 2002…Investment rates in the United States are broadly unchanged from their levels in 1997, although they remain below the peak in 2000." To some extent, this "decline in investment rates over time reflects the fact that capital goods have become relatively less expensive -- mainly owing to the extensive…productivity growth in the capital goods sector." Outside the industrial countries, "with the exception of China and a handful of other countries, investment rates have fallen in emerging market economies since the Asian financial crisis -- in east Asia, by more than ten percent of GDP since their peak in the mid 1990s -- and have not rebounded." The IMF concludes that "investment appears to be below the levels that would usually be associated with this stage of the economic cycle, and this may be an important factor in explaining both the current low level of real long-term interest rates, and the shift of emerging market economies from being net importers to net exporters of savings."
In our view, you have to go back to the late 19th century to find a situation that is analogous to the one we face today (with the clear caveat that history never exactly repeats itself). That was the age of rapid industrialization, that saw the U.S. economy faced with a supply shock. This triggered a glut of goods, falls in prices and investment, and eventually a depression brought on by debt deflation (the process by which deflation raises the real burden of debt, causing more people and businesses to go bankrupt, which further decreases demand). In our era, this process has been building for years, and reached its zenith with the explosion of new investment in China, which resulted in dramatic productivity gains as reasonably well-educated and organized workers suddenly had a lot more capital to use. The consequence was an explosion of supply, price pressures, and declining investment in a growing number of industries. To be sure, the actions of western governments (especially the United States' aggressive fiscal and monetary policies) have delayed the day of reckoning brought on by the relatively sudden entry of China into the world economy. But they have not eliminated its necessity, whatever form it will take.
This brings us to another fascinating research paper, that provides food for thought about financial markets' reaction to our current global situation. In "Rare Events and the Equity Premium", Robert Barro from Harvard finds that, if investors take into consideration events that have very low probabilities but very serious negative consequences, this can explain the simultaneous existence of very low real rates of return and a quite high equity market risk premium. As the probability for an extreme event increases, the value of risk free assets increases. People will pay more to own them, raising their price and depressing their yield. By the same logic, they will demand a higher risk premium to hold equities. This raises an obvious question: we know that real bond yields are very low today. But how high are required equity risk premiums?
To answer this question, we turn to another of our favorite topics, the global market portfolio. In September, the IMF also publishes the Global Financial Market Stability Report, which contains its estimate of the market value of different segments of the global capital market. We have combined the IMF data with information from other sources (e.g., the BIS and Venture Economics) to produce the following estimate of asset class weights in the global market portfolio at the end of 2004:
| December, 2004 | U.S. $ Billions | Percent |
| Fixed Income 1 | ||
| Australia | $ 479 | 0.43% |
| Canada | $ 1,055 | 0.94% |
| Eurozone | $ 15,816 | 14.08% |
| Japan | $ 9,163 | 8.16% |
| Switzerland | $ 488 | 0.43% |
| United Kingdom | $ 2,444 | 2.18% |
| United States | $ 22,540 | 20.07% |
| Other Developed Mkts | $ 2,319 | 2.06% |
| Emerging Mkts | $ 3,659 | 3.26% |
| Total Fixed Income | $ 57,963 | 51.60% |
| space | ||
| Commercial Property | ||
| North America | $ 5,300 | 4.72% |
| Europe | $ 5,200 | 4.63% |
| Asia Pacific | $ 3,100 | 2.76% |
| Latin America | $ 400 | 0.36% |
| $ 14,000 | 12.46% | |
| space | ||
| Commodities (Notional Value of Principal O/S) 2 | $ 2,000 | 1.78% |
| space | ||
| Timberland (Investable) | $ 500 | 0.45% |
| space | ||
| Public Equity | ||
| Australia | $ 825 | 0.73% |
| Canada | $ 933 | 0.83% |
| Eurozone | $ 5,356 | 4.77% |
| Japan | $ 3,208 | 2.86% |
| Switzerland | $ 996 | 0.89% |
| United Kingdom | $ 3,683 | 3.28% |
| United States | $ 18,792 | 16.73% |
| Other Developed Mkts | $ 1,167 | 1.04% |
| Emerging Mkts | $ 2,208 | 1.97% |
| Total Equity | $ 37,168 | 33.09% |
| space | ||
| Private Equity | ||
| Buyout | $ 490 | 0.44% |
| Venture Capital | $ 210 | 0.19% |
| Total Private Equity | $ 700 | 0.62% |
| space | ||
| Global Market Portfolio | $ 112,331 | 100.00% |
It is important to understand what this represents -- and what it does not. The first important point is that, because of data measurement problems and differences in the timing of when different data are collected, this is at best a rough estimate of the "true" value (and weights within) of the global market portfolio.
Second, it is not the case that the global market portfolio necessarily represents an "optimum" portfolio for any given investor. While it reflects the aggregation of millions of investor's decisions, it does not follow that all those investors are the same. As Professor William Sharpe noted in his Princeton Lectures in Finance last year, investors differ along three fundamental dimensions: positions, preferences, and predictions. Positions refer to differences between investors' goals, timing, capital, and expected future income and savings. Differences in each of these areas can result in two investors wanting to hold portfolios with different asset allocations. Preferences refer to factors such as an investor's willingness to take risk, aversion to regret, and desire to substitute current for future consumption. Once again, differences in these areas will also give rise to differences in investors' allocations to different asset classes, even when the only available investments are index funds. Finally, predictions refer to differences between investors' forecasts. For example, one investor may believe it is impossible to identify skilled active managers who will consistently outperform an index fund in the future. Another investor may believe he possesses that skill, but only in a certain asset class (e.g., buyout funds). Hence, the second investor may choose to allocate a larger portion of her portfolio to the asset class in which she believes she possesses distinctive forecasting skill.
Third, if you know the asset class weights in the global market portfolio, and have an estimate of the covariance of their respective returns, you can make inferences about the future asset class returns that "the market" (i.e., investors in the aggregate) expects. To be sure, this analysis rests on a number of assumptions, any one of which can be challenged. It assumes that the market is efficient, and no asset class is significantly over or undervalued. It also assumes that the historical covariance matrix (the covariance of two securities equals the product of the standard deviations of their respective historical returns, times the correlation between those returns) is accurate. And it assumes that one can accurately estimate the risk aversion coefficient of the "average investor." As I said, all of these can be challenged, so the analysis that follows is necessarily an uncertain estimate. Still, it provides interesting food for thought.
The following table shows, in local current terms, the estimated future real returns on different asset classes that we derived from the global market portfolio and our historical 1989-2004 covariance matrices (for the technically inclined, we assumed a risk aversion coefficient of 2, which, while at the low end of the normal range of assumptions, seemed appropriate given conditions -- e.g., very tight credit spreads -- at the end of 2004).
| Real Return Bonds (current yield) | Domestic Nominal Return Investment Grade Bonds | Domestic Commercial Property | Domestic Equity | |
| AUD | 2.55% | 11.35% | 12.57% | 14.21% |
| CAD | 1.60% | 4.33% | 6.45% | 7.09% |
| EUR | 1.14% | 7.50% | 7.87% | 8.23% |
| JPY | 0.79% | 9.65% | 13.33% | 11.90% |
| GBP | 1.42% | 6.12% | 8.02% | 7.35% |
| USD | 1.75% | 3.21% | 4.60% | 6.51% |
This table makes some interesting points. Consistent with our view of the balance of risks in the global economy (i.e., a rising probability of an extreme downside event) and Professor Barro's paper, our estimates of the expected future returns implicit in the global market portfolio shows quite high equity market risk premia (i.e., the difference between the implied real return on domestic equity and the current yield on real return bonds).
That being said, while many of the implicit returns seem intuitively reasonable, some of them do not (e.g., Australian bonds, although one can hope). Whether this reflects estimation errors in our covariance matrix, inefficiencies (over or undervaluations) in certain segments of the global capital market, or some other problem (e.g., inaccurately measured data), we cannot say. Along with other inputs (e.g., historical results, and the predictions of a theoretical asset pricing model), these inferences are but one of the inputs we will consider when making the assumptions we will use in our asset allocation model for this year's rebalancing of our model portfolios.
This brings us to another difficult question: what are the implications of our analysis for the asset allocation in your portfolio? Experience has taught us the folly of being ideologues when it comes to market timing. Rather, we distinguish between the goal one is pursuing in trying to time markets, and the way one goes about it.
Conceptually, some people try to time markets because they are trying to generate higher returns (alpha) than their long-term strategic asset allocation policy is expected to produce. Other people time markets because they are trying to limit their risk. The former approach is heavily dependent on foresight about the relative returns an investor expects different asset classes to earn in the future. And as we have just shown, accurately forecasting future returns is not something many of us have the skill to do consistently well. In theory, if one expects to earn superior returns from market timing, one's forecast should be superior to the market's conventional wisdom, in either its magnitude or timing. It one's views align with those of the market, it will be hard to earn superior returns other than through luck.
On the other hand, people who attempt to time markets to limit their risk typically base their decisions on hindsight -- the current valuation of an asset class relative to the historical record. In our experience, hindsight tends to be more accurate than foresight; hence, we believe that, in the absence of superior forecasting skill, investors should limit their episodic market timing to avoidance of extreme overvaluations.
I say "episodic" because there are two broad ways to implement a market timing strategy. Episodic market timing refers to conscious decisions, often involving significant portfolio re-allocations or hedging decisions, made on the basis of specific valuation conclusions (based on either foresight or hindsight). In contrast, systematic market timing decisions are made automatically. Let's look at two examples this approach. A decision to invest in index funds contains an element of momentum, in so far as an index fund that uses market capitalization weighting will increase the weight of assets whose relative value is appreciating at the fastest rate. In contrast, a regular rebalancing strategy systematically sells recent winners and reallocates the proceeds to other asset classes that have not performed as well (i.e., it reduces holdings in asset classes above their target weights, while adding to those below their target weights). Similarly, employing dollar cost averaging when adding funds to a portfolio is another systematic approach to risk reduction. It may also be possible to systematically earn slightly higher returns. If you assume (as we do) that markets tend not to be in equilibrium, then they will have a tendency to overshoot and undershoot their fair value, and then correct. This suggests that, at the margin, it may be possible to generate small amounts of alpha by rebalancing overweight asset classes to a level somewhat below their target weight (e.g., from 15% to 5% instead of 10%), and underweight asset classes to slightly over their target weight (e.g., from 5% to 15% instead of 10%).
However, the situation before us today is clearly one of episodic market timing based on hindsight-based conclusions about the reasonableness of current valuations in one or more asset classes. As noted above, the problem we face is that while we know what must eventually occur (a reduction in the imbalances in the global economy today), we don't know exactly how or when this will happen. In this way, the current situation is somewhat similar to the U.S. equity market in the late 1990s. Unfortunately, we don't know whether today is more like 1998 (imbalances will persist, and we'll muddle through for a while longer) or 2000.
Given the heavy dependence of all bubbles on human emotion, and the rapidity with which they can reverse and herd into a crash, we are suspicious of anyone who believes they will be able to "get out at the top". This suggests that it would be wise under the current circumstances to consider putting some downside protection in place. However, this raises the question of the form that protection should take. While one could buy put options, they are not available for all asset classes, and they cost money, which adds up if the expected downturn arrives late (as anyone who started buying them in 1998 and then threw in the towel in 2000 can tell you). One could also switch some assets into cash. However, again as anyone who did this in 1998 can tell you, the emotional impact of opportunity costs (watching an asset class rise in value while you are in cash) can also be severe, depending on the internal trade-off one makes between fear and greed. This leaves stop-loss orders. Since they cost nothing until executed, they are more attractive than put options for hedging an existing position. And unlike a move into cash, they allow an investor to continue to earn positive returns, in the case that the downturn arrives later than expected. However, there remains the key issue of how low below the current price of an investment to place a stop loss order. In essence, you want to set it low enough that you avoid having it triggered by "normal volatility" (i.e., the regular ups and downs that characterize a market). For example, the following table shows the standard deviation of monthly nominal returns on different U.S. dollar asset classes between 1989 and 2004:
| Asset Class | Standard Deviation of Monthly Returns |
| Real Return Bonds | 1.60% |
| Investment Grade Bonds | 1.15% |
| Foreign Currency Bonds | 2.61% |
| Domestic Commercial Property | 3.66% |
| Foreign Commercial Property | 5.57% |
| Commodities | 5.43% |
| Timberland | 4.34% |
| Domestic Equity | 4.22% |
| Foreign Equity | 4.84% |
| Emerging Equity | 6.81% |
These give a rough idea of how big a downside move would have to be in order to be statistically significant. Assuming normally distributed returns, 95% of the time monthly returns should be less than two standard deviations below the mean; hence a quick and dirty estimate of a big downside move would be two standard deviations or more in a month. However, this estimate is subject to three important limitations. First, unlike statistics that describe the results of an unchanging physical process, those that describe asset class returns (which result from a process that tends to change over time) contain a lot of estimation error. Second, most asset class returns aren't normally distributed. Close, typically, but not quite. Rather, they often have slightly "fatter tails" than a normal distribution, which means that there will typically be more big downside moves than would be the case if the returns were normally distributed. Third, on a daily (or, more accurately, intraday) basis (which is the relevant one for triggering a stop loss order), returns have even fatter tails than they do on a monthly one. All of these considerations argue for setting the trigger price on a stop loss order well below the current market price on an investment -- say 15% to 20% below as a rough guide -- to avoid the risk of incurring unnecessary transaction costs.
Finally, we'll repeat what we wrote in our March 2005 issue. With trouble on the horizon, now would be a good time to rebalance one's portfolio to levels below one's target weights for asset classes that seem overvalued (e.g., many equity markets and arguably commodities except perhaps for gold) and to above target weights for those assets that would probably perform well under either a U.S. dollar decline and/or a global flu pandemic scenario (e.g., real return bonds, non-U.S. dollar bonds, short maturity U.S. government bonds, timber, and commercial property). However, having said that, we will also note that we will be concluding our biennial asset allocation review and announcing our rebalanced model portfolios at the beginning of November. So, assuming you believe (as we do), that the world economy will probably keep muddling along with out any major shocks for at least another six months (if not longer), it makes sense to hold off on this rebalancing until the end of the year.
| Equity Market Valuation Update | Economic Update: Semi-Annual September 2005 | This Month's Issue: Key Points | Global Asset Class Returns | Product and Strategy Notes: Oil ETF vs. Commodity Indexes, Reader Comments: Everbank Minimums, USDX, VSTOXX vs. VIX, IPOX-30; and Canada (Active vs. Passive) | This Month's Letter to the Editor: Convertible Bonds, Defined Pension/Social Security and Siegel's Paradox |