طرح حساب های شخصی چرخه عمر امنیت اجتماعی: ارزیابی پیشنهاد رئیس جمهور بوش

The life-cycle portfolio proposal for personal accounts within a Social Security system would have the government undertake the dynamic portfolio allocation program for individuals. This paper evaluates, using U.S. historical data 1871–2004, several versions of conventional life-cycle portfolios. The results show disappointing performance relative to the rhetoric of the promoters of the proposal. Dynamic portfolio theory suggests that the optimal life-cycle portfolio may look very different from the conventional form. Moreover, behavioral finance suggests that the design of a life-cycle portfolio for Social Security should consider the attitudes and habits of individuals and as well as their diversity.

Any plan for private accounts as part of a Social Security system has to take account of the life-cycle risk-management problem. The optimal portfolio for any worker ought generally to change as that person ages, and so it has been proposed that private accounts should be designed either to optimize over time for the worker or at least to make it convenient for worker to do so within the choices that are provided. President George W. Bush outlined, in his State of the Union speech and associated documents1 in February 2005, the world's most ambitious version of this proposal for life-cycle accounts within Social Security, and campaigned for succeeding months for his plan. Although it now appears unlikely that his plan will be implemented in the near future, it will certainly be a model for such plans in the United States and other countries in future years. In this paper, I will outline the important issues for life-cycle portfolios within Social Security. For concreteness, and to highlight associated government budget and political realities, I will put this analysis of the life-cycle proposal in the context of the specific proposal of President Bush. Much discussion of the attractiveness of the life-cycle portfolio reform for social security appears to rely on assumptions that a life-cycle portfolio will yield very high returns with little or no risk. This paper investigates the possible returns and possible risks associated with a life-cycle personal account option. I use U.S. historical returns 1871–2004 to assess the potential range of future investment outcomes under several versions of a life-cycle personal accounts plan that are suggested by recent discussions and examples. The results suggest that workers would be well advised to stay away from such life-cycle plans. This paper concludes with a discussion of issues for the better design of life-cycle accounts.

In this paper, I have concentrated mostly on analyzing the Bush proposal for life-cycle accounts, and not tried to compute the optimal life-cycle portfolios for people in relation to the particular barriers that they have to investing properly. The political reality is that governments are not likely to implement any time soon a complex array of life-cycle portfolios, each intended for a certain group of people, nor are they likely to design the optimal portfolios that have been described by some authors, such as life-cycle portfolios that are 300% stocks for young investors, or that are short the domestic stock market and long the rest of the world's stock markets. Policy modeling to be relevant has to analyze the politically feasible proposals, as well as the arguments that are made for them. The dominant argument for the Bush proposal has been simply that it will capture the historically high returns on the stock market and keep risks well under control. We have seen that, as represented by the simulations presented in Table 2 and Table 3, there is a disappointing outlook for investors relative to the rhetoric of their promoters. I think that the disappointment may be even greater than the tables indicate. The simulations depended on the historical experience of either the United States or the rest of the world for over a century. While a century may seem like a long enough sample period to prove any point, in fact even with a century of data we do not know the true probability distribution of future returns. The 21st century may differ fundamentally from the 20th. Moreover, whenever we look at long historical data on stock markets, we are of course looking at survivors, stock markets that made it, and ignoring countries where conditions turned out so badly that we get little or no stock market data.15 Obviously, Russia and China were not in Dimson et al. for their century-long data set. While it is highly unlikely that the particular upheavals that hit Russia and China in the 20th century will repeat themselves in the U.S. in the 21st century, it should be remembered that nobody predicted in 1900 the kind of upheavals that were to follow in Russia or China. Similarly, it is difficult for us to imagine all the different kinds of things that might disrupt stock market performance in the future. Thus, there is additional uncertainty, uncertainty of regime change or model uncertainty or survivorship bias that should ideally be taken into account. Unfortunately, there does not seem to be any objective way of quantifying this additional uncertainty about the future. We are living in the aftermath of the spectacular stock market boom of the 1990s, a boom that generated irrational exuberance about the outlook for the stock market. A public attitude of exaggerated expectations about the stock market is still with us, and probably accounts for a good portion of the support for personal accounts. Those who rely heavily on the past U.S. experience for their expectations for the future might find attractive a life-cycle account that is even more heavily into the stock market in the younger years, such as the aggressive life-cycle account considered here or even more aggressive than that. But we have seen that the aggressive life-cycle portfolio entails serious risks too; moreover the correlation of these risks to the labor income risks will have to be a central consideration. In thinking about the redesign of Social Security, it is important not to focus on chasing past returns and instead focus more on the theoretical fundamentals that underlie Social Security. From this perspective, the government's offering margin credit to the general public, thereby duplicating a service that is already available from the private sector, should not be the focal point of discussions. There are some important theoretical justifications for the government's providing a Social Security system, if not margin credit. Notably, a justification for a pay-as-you-go Social Security system is that the private sector cannot allow people to share risks across generations effectively, since children, and unborn future children, cannot make financial contracts. We could redesign the existing Social Security system to do a better job of managing such intergenerational risk than it does today.16 But this justification for Social Security would not appear to offer a reason for the government to get involved in margin lending. The intergenerational-risk-sharing justification for Social Security is so strong and palpable that it is likely to trump any effort to design it along different lines. That is, a design that imposes risks rather than reduces risks will ultimately be discarded. In fact, people likely know this and will likely use this information in their choice of personal accounts. With the Bush plan, many people will probably choose the 100% stocks portfolio, thinking that it might do very well, and that if it does poorly, the government will probably change the program and bail out the losers. In any event, the 3% real offset rate of the Bush proposal appears to be too high, and if the program is instituted, it should be done with a lower rate. If the offset rate were lowered from 3% per annum to 2%, it would shift the net values of the accounts at retirement shown in Table 2 and Table 3 up by $30,322 the difference between the offset cumulated at 3% and the offset cumulated at 2%, generally making the median net values positive. Better yet, the offset could be cumulated at a market rate. The offset could be calculated as the terminal value of the contributions brought to the final date using actual U.S. Treasury Inflation Protected Security (TIPS) yields of the appropriate maturity. Reducing the offset rate would move the distribution of the returns to the right, but of course would do nothing to reduce the uncertainty of the returns that we saw in Table 2 and Table 3. Even with an offset rate of 2%, the baseline personal account net value at retirement as simulated with adjusted historical returns (shown in column 1 of Table 3) would still be negative over 25% of the time. Note that the personal account plan does nothing directly to improve the saving rate, or to make the economy grow faster. It might possibly increase the saving rate, if the experience of being invested in the financial markets and having made choices among investment accounts stimulates people to think more favorably about investing. But, the personal accounts might also have the perverse effect of lowering saving rates if people's perception that they have money in the stock market that is going to make them rich some day discourages them from saving on their own outside Social Security. Psychologists have documented a “wishful thinking bias” that encourages people to think that their own team will win the game or their own candidate will win the election; the same tendency might encourage personal account holders towards great expectations for their account. The experience of crashing saving rates in the U.S. at the same time as the stock market boomed in the last two decades suggests that the latter outcome might be more significant. If the personal accounts do not raise saving rates and do not make the economy grow faster, then the accounts will not increase opportunities for aggregate consumption. Any increase in consumption that is provided to some people must come one way or another from others. The argument is sometimes made that Bush's plan cannot be a bad idea because it merely expands people's choice set and so anyone can choose not to participate. But this is not valid. Those who choose not to participate in the program are still affected by it. The program will affect them through prices in the stock market; the personal accounts program could have the effect of causing speculative demand to bid up the stock market in the short run, followed by a crash in the longer run, an outcome that would help and hurt people both in and out of the personal accounts depending on the timing of their exposure to stock market risk that they chose. Bush's Social Security plan might be described as a plan to allow small investors to use their expected future Social Security benefits as collateral for margin loans, offering a plan at such a large scale that economies of scale reduce administrative costs, and streamlining the program so that workers do not impose costs by expecting the kind of services they get from brokers, thereby making it possible for the government to charge a lower rate for margin loans than is available in the private market today. If the government does this it will not be met with unmitigated enthusiasm from the brokerage community: providing a lower rate puts the government program in competition with private margin lenders, who do not offer such attractive margin rates to small accounts. More importantly, perhaps, if the government offers a lower rate on the offset balance then the government will be losing an opportunity to lower its long-term budget deficit. The present plan saves the government money by lending to individuals at a rate that is higher than the real rate at which the government can borrow money. We have seen that Bush's plan might well help some people (the example given was people who have savings but, because of psychological obstacles, are too conservatively invested) but may hurt other people (the example given was people who have no savings and who might invest their personal accounts in too risky a manner.) Thus, the personal account is a blunt tool. It would seem to be a better plan if the government merely subsidized personal financial advisors, making it cheaper for people to get financial advice that is really appropriate for their own economic circumstances. Today, most people get no significant professional help with their investments. With proper financial advice people could also make use of some better choices among their personal accounts, such as choices designed to hedge their labor income or home value. Much of the design of Social Security also depends on the need to design around human foibles, and to provide public goods that might help them out of their errors. The failure of many people to save at all for their retirement is one such foible, and ultimately our society feels a need to protect such people. More broadly, a failure of many people to embrace capitalist institutions, to learn about investing and about the benefits of ownership, may also be a factor justifying actions from a government that provides an educational that will not be provided privately because it is a public good. One might justify government intervention to help people deal with these psychological failures, and this is indeed part of the justification of Bush's ownership society. The personal accounts may indeed help achieve this, even if they are probably not the ideal vehicle for this purpose. But to say that there is a money machine in the stock market, that it can be tapped to yield great wealth without significant risk if one uses life-cycle investment methods, is a big mistake. The stock market is an uncertain place, and even if the risk is effectively managed over the life cycle, important uncertainty remains.