Investment Restraint, The Liquidity Glut, and Global Imbalances1Remarks by Raghuram G. Rajan, Economic Counselor and Director of Research, IMF
At the Conference on Global Imbalances organized by the Bank of Indonesia in Bali
November 16th 2006.
Good afternoon. Even while the Dow Jones Industrial Average has scaled new heights recently, the 10 year Treasury bond is yielding substantially less than the Fed Funds rate, a phenomenon typically associated with recessions. The yield spread in the Euro area also briefly turned negative, even while Europe is enjoying strong growth. Are the markets seriously out of kilter?
I think not, and my argument will rely on three global ingredients. The first is a widespread surge in productivity across the world, with the associated impact on domestic demand varying country by country based on the strength of domestic financial markets. The second is a high desired savings rate that continues unabated, particularly supported by corporations, but also by emerging market governments. The third, and perhaps least well understood, is global investment in physical assets that has yet to return to past levels despite the higher productivity and available savings. Taken together, these forces have resulted both in large global imbalances but also in the benign conditions needed to finance them. Which of these will give, how, and what the consequences will be is a matter of both great import, and unfortunately, one where we have little guidance from the past.
The Productivity Revolution
We are now in the fourth year of strong world growth, growth that has been maintained in the face of headwinds such as soaring commodities prices. In my view, productivity growth, fostered in part by the revolution in information technology, but also in part by the rationalization of production through the creation of global supply chains, has played a critical role in this expansion. While much attention has been focused on the extraordinary surge in U.S. productivity since 1995, equally impressive productivity growth in emerging markets has been little commented upon. Taken together, rapid, and largely unexpected, worldwide productivity growth can explain why the demand for commodities is so strong, how emerging markets have weathered commodity price increases without a serious slowdown in activity, why inflation is still largely contained despite the unprecedented rise in raw material costs, and why both household incomes and corporate profits are buoyant at the same time.
The Rise in Desired Savings
Even though incomes have grown (albeit unevenly within countries), the global desire to save has kept pace. We know this is not because of the behavior of households in industrial countries where, with a few notable exceptions like Germany, savings have been drifting lower.
In recent years, though, the surprising increase in industrial country corporations' excess savings — their undistributed profit less capital investment (see WEO September 2005) — has partly offset the decline in household savings. Part of the increase in corporate excess savings is because of the previously mentioned productivity growth which, because it has not been paid out to labor through higher wages, has resulted in greater corporate profits. Indeed, undistributed profits have been enhanced even more because of lower corporate taxes, and the fact that dividend payments have been steadily decreasing. But undoubtedly some of the increase in excess savings has to do with the desire of corporations to maintain large cash hoards, even while reducing investment. While academics are still puzzling over this behavior, one explanation of the mounting corporate cash hoards has to do with the increasing competitiveness of the environment for individual corporations, causing it to be individually more volatile even while the overall macroeconomic environment is calm.
In emerging markets, rising incomes have also resulted in rising desired savings. Certainly, governments have played a part by becoming more careful with their finances — many countries are running primary fiscal surpluses for the first time, some on the backs of gains in commodity exports. Households too are playing a part. In some countries like China, where citizens are increasingly experiencing the uncertainties associated with a market economy, the absence of a safety net is an important factor driving higher desired household savings. Moreover, with little ability to borrow against future incomes because of the paucity of retail credit, emerging market household also have an incentive to save to buy the durable goods such as cars and houses that they are increasingly able to afford. And finally, emerging market corporations, especially but not exclusively those in the natural resource sector, are also building excess savings — perhaps because they too face an uncertain environment with global competition and takeovers increasing. A continuing desire by households, corporations, and governments, especially but not exclusively in emerging markets, to save out of the larger income generated is the second ingredient of the story.
The Fall in Realized Investment
Given strong productivity growth and an unabated desire to save, it is therefore surprising that actual physical investment has not kept pace. After all, if productivity growth is strong as is the desire to save, investment should be both profitable and easily financed. Yet investment is only slowly returning to the levels reached in the last decade, and I would conjecture, probably below the quantities that might be warranted by the tremendous growth experienced over the last few years.
To my mind, overall investment restraint is the real macroeconomic conundrum (Bernanke (2005) offered an early discussion of the phenomenon, though based on work at the Fund, I believe the problem of the excess of desired savings over realized investment is better described as investment restraint rather than a savings glut). One explanation is simply that the world is still working off the consequences of past excessive investment. For instance, telecoms may have invested too much in capacity during the boom in the late 1990s, and may still be working off the overhang of those investments as well as the debt taken to finance them. Emerging markets too have become aware of the past inadequacies of their financial systems in allocating investment, and some of the caution displayed by their corporations now may reflect their experience of past booms and busts.
A second explanation is that the nature of investment may have changed — from hard physical assets like plant and equipment to items like training and research and development that are expensed and not as easily tracked. But if such expenses were high enough to compensate for the "missing" physical investment, profitability would be low — however, corporate profitability is high the world over.
A third explanation is that capital goods are now cheaper (think how much the cost of computers has fallen), so that lower nominal amounts have to be spent to get the same amount of real investment. There are two difficulties with this explanation. First, the kind of capital goods such as computers whose price has fallen the most are also ones with high depreciation rates. So it is not clear that lower nominal amounts have to be spent to get the same net flow of real investment. Second, if capital goods were indeed cheaper, corporations should buy more of them, so it is not obvious that this could account for the fall in overall investment.
And finally, perhaps investment is low because of economic uncertainty. For emerging markets, uncertainty about the policy environment has always been significant, even more so in the current highly competitive global markets where small policy errors can rapidly make domestic industry uncompetitive. In addition, government intervention to hold down the value of the exchange rate can make investment in non-traded goods unattractive, even while subjecting the profitability of traded goods to the vagaries of government policy.
In industrial countries, a different dynamic may be playing out. As the cost of trade falls and emerging markets become more cost-competitive, it makes sense for industrial country corporations to invest, not domestically, but primarily overseas — indeed this conjecture is supported by the fact that FDI in emerging markets is near an all time high. The key question is "Is it enough?".
Put differently, from investing in their predictable domestic environments, industrial country corporations now are looking to invest far more in non-industrial countries, where despite recent policy improvements, the environment is still substantially less predictable than at home. Not only do industrial country corporations have to worry about whether China is a better place to invest than Vietnam, they also have to forecast whether it will continue to be a better place ten years from now. No wonder then that many corporations are focused on trying to improve the efficiency of their existing production, as well as their stock of easily transportable knowledge, rather than on making really long term, hard-to-move, fixed investments in distant places. Indeed, the uncertainties surrounding industrial country FDI into emerging markets could also serve as a deterrent, limiting investment by emerging market corporations.
In other words, given the investment restraint by domestic corporations in emerging markets, and given the undoubted competitiveness of investment in emerging markets, greenfield FDI from rich countries to emerging markets ought to be higher than it is — indeed it should offset the decline in domestic investment in rich countries. That it has not may be because political and economic uncertainty are holding it back.
To summarize, I have argued the world has experienced strong productivity growth, and desired savings that continue to remain high, but actual investment, after plunging at the turn of the century, is recovering very slowly. Let me now turn to the consequences, which do not depend very much on what rationale you accept for the apparent investment restraint.
The Global Liquidity Glut
Debt securities typically need to be backed by hard assets that can be repossessed in case of default. So regardless of why the flow of hard new assets is low, the amount of debt that can be issued by corporations is likely to be constrained when investment is low.
Of course, the different rationales have different implications for the value of corporate equity. It is not clear, however, that the value of equity would necessarily rise to compensate for the low debt issuances. Indeed, if the "missing" investment is because individuals are investing in acquiring more human capital, then it is not clear that this investment can be as easily securitized. When an employee co-invests by acquiring firm-specific human capital — probably an increasing form of investment as the share of services increases the world over — she acquires a claim on the future revenues of the firm which will be paid out in the form of higher wages. The firm looks more profitable today, and invests less in hard assets, but will have to pay in the future for the soft assets its employees are acquiring. In the meantime, it is harder to create financial claims on these soft assets — how does a lender to an employee repossess specific investment in human capital given that slavery has been abolished?
Therefore, regardless of whether investment in physical assets is unnaturally low, whether it has been naturally displaced by investment in intangible assets, or whether physical goods have become cheaper so that less is spent on them, the point is that there is less in the way of incremental collateralizable assets being produced the world over. The collateral value is even lower if more of the assets are being created in emerging markets that, typically, are less well governed.
The mismatch between unabated global desired savings and lower realized investment, between the amounts available for finance and the flow of hard assets to absorb it, has led to a liquidity glut which has pushed long term real interest rates the world over lower. This has spilt over into markets for existing real and financial assets — real estate, high-risk credit, private equity, art, commodities, etc — pushing prices higher. Indeed, casual empiricism suggests that the most illiquid markets, where typically there are few transactions, and small infusions of liquidity can have substantial effect, have been pushed the highest.
The attention a market gets can be flattering. A number of commentators have noted that emerging market debt has become an accepted part of investor portfolios — it is now a well accepted asset class. While indeed emerging markets have done much to raise their creditworthiness, the achievement is less praiseworthy when we recognize that almost all financial assets have now become mainstream. Our enthusiasm should be tempered by the realization that it is the shortage of collateralizable new real assets rather than the improved creditworthiness of old borrowers which is causing financial markets to rediscover the latter.
The attention can also be self-fulfilling. If refinancing is easily available, no borrower will default, allowing lenders to believe that a "structural" change has brought down the credit risk associated with hitherto untrustworthy borrowers.
The point is that the enormous amount of excess liquidity currently sloshing round in global financial markets is, in my view, not primarily because of the accommodative policy followed by the G-3 central banks in recent years. While the historically low levels of the policy rates in industrial countries till recently may have prompted some arbitrage between the short and the long end of the interest rate spectrum, which may have compressed the long end a bit, and while the still low real policy rate in Japan may be prompting some carry trades into other currencies, I do not think accommodative monetary policy is the primary explanation for the low level of long rates. Indeed, the dramatic rise in US policy rates in the last two years has had little effect on long US rates, and monetary tightening is likely to only have moderate impact on the excess liquidity in asset markets. Nor do I think foreign central banks are directly responsible for low long rates in the US, even though their purchases may be correlated with downward pressure on long US rates (see Warnock and Warnock (2006)). Instead, I believe that foreign central bank purchases of U.S. assets reflect the savings investment imbalance in their own countries — even though the specific motivation may be oriented to maintaining the value of their currency.
The Liquidity Glut and Global Imbalances
Many of the factors that have created the liquidity glut have also widened the current account imbalances that are the subject of this conference. The reaction of domestic demand to rising productivity growth has varied across countries, differing in part based on the sophistication of their financial sector. In the United States, for example, the surge in productivity led to a boom in investment in the late 1990s, financed by deep financial markets. Not all the investment was wise, but the debris created by the bust was quickly cleared up by the financial markets. The recent expansion, since 2002, has been more focused on consumption and real estate, supported by the United States' strong arm's length financial system, which has allowed consumers to borrow against future incomes (a possibility that is easier in the financially sophisticated United States than elsewhere) and consume immediately. The expectation of higher future incomes coupled with accommodative monetary policy and low interest rates may have fueled the housing boom, which expanded consumption even more as the financial system allowed borrowing through vehicles such as home equity loans. Thus the United States' financial system helped to translate productivity growth into strong domestic demand, and a large current account deficit (also see WEO September 2006).
In emerging markets, though, the experience with investment booms and very severe busts has led to continuing circumspection in investment, with the exception of China. Moreover, because of the limited availability of housing- and retail finance, households in these countries have not been able to expand consumption through borrowing. Thus, domestic demand in these countries has been relatively muted and these countries have generated net savings or current account surpluses. Exceptions support the rule. In India, a relatively vibrant financial sector has fueled a consumption and housing boom, even while corporate investment is expanding once again after years of adapting to past excessive investment. India has moved from running a current account surplus to running an increasing deficit.
The global liquidity glut has permitted easy financing of all borrowing, including that of countries like the United States running large current account deficits. In the face of a worldwide shortage of investible assets, the innovativeness of the US financial system, which has brought many non-traditional real assets into the financial markets through the process of securitization, has helped draw world savings into the United States, and continued to support the dollar.
How will all this end?
Current conditions are unlikely to be permanent, though a quick look at history (see Catao and Mackenzie (2005), for example) would suggest that the low real interest rates of the present period are more representative than the high interest rates of the previous three decades. Given aging populations in developed countries though, one would presume that the rebalancing of worldwide investment to desired savings will have to take place primarily in non-industrial countries. Investment will increase partly through foreign direct investment, but partly mediated by the financial systems in emerging markets, which will have to develop further. Increases in consumption as safety nets improve and retail finance becomes widely available will also reduce desired savings. Certainly, the seemingly perverse pattern of net capital flows, from poor to rich countries, will have to change, if for no other reason than to accommodate demographics.
Innovative financial systems are also adapting by creating new assets — such as securitized claims on residential investment — that have helped satisfy the unmet desire to save. A danger in all this is overshooting — that too many assets of one class are created leading to a glut and abrupt correction as is currently being experienced in the US housing market. Such a danger is compounded by agency problems in financial systems, which are exacerbated in a period of cheap and easy finance. I have discussed these elsewhere (Rajan (2005, 2005)). A related danger is that small changes in excess liquidity could have large changes on the price of illiquid assets, with large attendant effects on valuations. If and when the imbalance between desired savings and realized investment is corrected, these asset markets could experience large swings.
Let me conclude. Any analysis of the global current account imbalances cannot be complete without discussing the global liquidity glut — the capital account side so to speak. I have spoken of some fundamental forces that might be driving this. My hope is that as a better balance between desired savings and investment is achieved over time, long term interest rates will move up slowly, exchange rates will adjust, and global imbalances will narrow without major blow-ups. We should, however, ensure that government policy does not impede adjustment, hence the IMF's efforts in this arena. We should also be aware that adjustments, either on the real or financial side, rarely take place as smoothly as hoped for. Continued caution is therefore warranted.
1 The following reflects my views only and are not meant to represent the views of the International Monetary Fund, its management, and its board. I thank Roberto Cardarelli, Charles Collyns, Laura Kodres, Subir Lall, and David Robinson for helpful comments.