1. What is TSP Advisory?
TSP Advisory is a web-based information service advising participants in the government's Thrift Savings Plan (TSP) how best to allocate their current account balance among the various TSP funds. TSP Advisory helps TSP participants decide how and when to make interfund transfers, i.e., when to reallocate their account balances among the various TSP funds. The premise of TSP Advisory is that -- while stocks are clearly the best long-term investment -- timely shifts or reallocations among TSP funds can significantly improve the overall return on TSP investments.
Back to Top
2. Is TSP Advisory connected with the Federal government?
No. TSP Advisory is a private organization. The Thrift Savings Plan (TSP) itself is administered by the Federal Retirement Thrift Investment Board (FRTIB), which is an independent government agency. To go to the TSP web site, click here.
Back to Top
3. How did TSP Advisory get started?
TSP Advisory began in 1996 as a monthly print newsletter to help government employees under the Federal Employee Retirement System (FERS) make better informed choices about allocating their TSP contributions among the various TSP funds. The development of the TSP Advisory models, however, was not completed until mid-2002.
The need to balance risk and return to achieve retirement goals without undue worry is not unique to government employees; employees of private firms offering 401(k) plans face the same set of issues. Both 401(k) plans and the TSP are defined contribution plans versus defined benefit plans. The key difference is that under a defined contribution plan, the decision about how much and where to invest is made by you, the employee. It’s a significant responsibility, since the quality of your investments may well affect the quality of your retirement.
TSP participants previously were allowed to make interfund transfers just once each month, generally initiating any transfer prior to midnight (CST) on the 15th of the month for the transfer to take effect by the 1st of the next month. As of 16 June 2003, interfund transfers will be processed on the next business day. The widespread availability of internet access and the decided speed and presentation advantages of a web-based information source required that TSP Advisory become an on-line information provider in order to continue to better serve TSP participants.
Back to Top
4. What's the best investment for TSP participants?
Over the long run, the best investment has been common stocks. Until May 2000, stock investments for TSP participants were limited to U.S. "large cap" stocks in the C-Fund, which closely tracks the S&P 500 index. The FRTIB has expanded the number of stock funds to three, including a U.S. "small cap" stock fund (the S-Fund) and an international stock fund (the I-fund). The addition of the S-Fund and the I-Fund increases the possibilities for allocation and risk diversification. One of the advantages of the TSP is the flexibility to reallocate among various funds, to diversify your investments and reduce risk, and to do so without incurring transaction costs.
TSP Advisory can help you make tactical asset allocation choices (versus strategic asset allocation choices, discussed in the next question) among the various TSP stock funds, the F-Fund and G-Fund. TSP Advisory uses its proprietary models to help you improve the return on your TSP investments over a "buy-and-hold" strategy, with no or little additional risk. (Sticking with the stock market through its ups and downs is called "buy-and-hold"). Despite the recent (2000-2002) stock market performance, investments in stocks have historically provided the greatest long-term promise of maximizing your retirement returns.
| TABLE 1 | TOTAL REAL RETURN, % | ||||
| PERIOD | Stocks | Gold | Bonds | Bills | |
| 1802-1992 | 6.7 |
0.1 |
3.4 |
2.9 |
|
| 1966-1992 | 4.2 |
1.9 |
1.6 |
1.3 |
|
| 1982-1992 | 11.1 |
-7.4 | 10.6 |
3.3 |
|
| *Adapted from Stocks For the Long Run, Jeremy J. Siegel, p. 15, Table 1-2. | |||||
The case for investing in the stock market is convincing. We frequently hear from financial columnists and advisors that the stock market is the best long-run investment. Table 1 (which appeared in Stocks for the Long Run, Jeremy J. Siegel, 1994) indicates that investing in common stocks has proven very effective. Even going back nearly 200-years in the record, stocks are best. For the period between 1966-1992, stocks returned 4.2% total real return (i.e., with dividends reinvested and net of inflation), or 10.1% nominal. From 1982-1992 (the first 10 years of the bull market that ended in 2000), the figures are 11.1% and 15.4%.
Implementing a "buy-and-hold" strategy for the TSP participant means directing most of your bi-weekly TSP contribution to the C-Fund/I-Fund/S-Fund and leaving the money there, regardless of market or economic conditions.
But "buy and hold" requires that you be prepared to suffer through difficult times. Stocks are volatile; the market fluctuates, often violently and apparently without forewarning. Even the words associated with the stock market -- "boom" and "bust", "The Great Crash", "Black Monday"-- suggest instability. And losses can be substantial, such as the >40 % drop in 1973-74, and the >20 % drop in 1982. And most recently, the stock market has suffered three straight losing years (2000-2002). The 2000-2002 three-year string of negative stock returns certainly ranks as one of the worst periods for stocks in the last 50 years.
TSP Advisory provides an alternative to "buy-and-hold." Our services are detailed in An Introduction to TSP Advisory's Premium Services.
Back to Top
5. How does the TSP Advisory asset allocation model work?For its subscribers, TSP Advisory provides recommendations for portfolios of TSP funds built around two basic model portfolios, and subscribers are free to choose between them. With the advent of daily valuation, TSP Advisory posts recommended fund allocations each week for each of the model portfolios, or indicates that previously-recommended allocations should remain unchanged.
The first model can be defined as a "tactical asset allocation" (or TAA) model. Under this model, TSP Advisory maintains 4 portfolios allocated as follows among stock and bonds: 80%-20%, 70%-30%, 60%-40% and 50%-50%. These stock-bond ratios are "strategic asset allocation" decisions that you make to establish a long-term balance between stocks and bonds. That balance should be tailored to your financial objectives and your risk tolerance. You should review your strategic asset allocation decisions as personal circumstances change, or at least every 5 years. (Assistance with establishing your strategic asset allocation is available on many web sites. TSP Advisory offers some general advice on our web site in our "Quick Primer.")
TSP Advisory's TAA model allows adjustments to that strategic asset allocation up to 15 percentage points on either side. For example, an 80%-20% strategic asset allocation may range up to 95% stocks and 5% bonds, to as low as 65% stocks and 35% bonds. The TAA model also indicates what percentage of the overall stock allocation is to be placed in the C-Fund, the S-Fund, and the I-Fund. The TAA model will also allocate the remaining funds between the F-Fund and the G-Fund. In general, the G-Fund is a riskless investment that serves as a shelter when the TAA model indicates reduced allocations to the risky funds.
The second basic model is the "dynamic asset allocation" (or DAA) model. Under this model, TSP Advisory maintains only one portfolio that can range from 100% stocks allocated among the C-Fund, the S-Fund, and the I-Fund, to as low as 0% stocks. The DAA model uses certain relatively persistent "environment" variables (based on specific economic and monetary measures) and certain "event" variables (transitory events) to define a basic allocation among stocks and bonds, and then superimposes the TAA model over the results from these "environment" and "event" variables.
TSP Advisory's models are computer models based on over 30 years of historical data that have been used to construct a mechanical system to help relate market movements with existing economic, monetary, and market conditions.
Back to Top
6. Don’t the TSP Advisory models go against the premise of the Efficient Markets hypothesis?
Yes. In particular, the TSP Advisory Dynamic Asset Allocation (DAA) model runs contrary to the theory underlying the Efficient Markets Hypothesis.
The Efficient Market Hypothesis came about from the research in the mid-1960s of a group of academic economists, including Eugene Fama and Nobel laureate Paul Samuelson. Here is a quick take on the theory. Market price will reflect all relevant information, and investors will not be able to "beat the market", that is, to get better than market returns on their investments. Why not? Financial markets like Wall Street are made up of vast numbers of talented people and investors all engaged in a zero-sum game seeking profits. Breaking news that bears on the value of stocks will be incorporated instantaneously into the price of the stocks, swiftly driving share prices up or down. Thus, the market efficiently incorporates new information into stock prices. Obvious profit opportunities tend to vanish before they can be seized, and "beating the market" becomes a pipedream.
It also follows that if today's stock prices reflect all available information, then tomorrow's price movements must be unforeseeable, since any information that might be used to forecast them will have already been incorporated by traders into today's prices. Thus, in each succeeding moment, the stock prices in an efficient market lurch about like a drunk on a bender, meandering down the street in what economists call a "random walk."
The Efficient Market Hypothesis even has its trademark joke: An economist is strolling down the street with a companion when they come across a $100 bill lying on the sidewalk. As the companion bends over to pick it up, the economist says: "Don't bother -- if it were a real $100 bill, someone would have picked it up already."
Index funds, such as the four index funds on which the TSP funds are based, were developed largely in response to the Efficient Market Hypothesis. And these passively-managed funds (wherein the fund manager does nothing more than buy the stocks that are in the index) do well. The Vanguard Index 500 fund, has in recent years generally topped actively-managed U.S. large-cap mutual funds.
If you believe in the Efficient Market Hypothesis, then expressions such as the "market is over-priced" or "this stock is undervalued" are necessarily wrong. Saying the "market is overpriced" presumes that the speaker has information about the market that is not reflected in its current price, and so contradicts the Efficient Market Hypothesis, which says that all information about stocks and bonds are reflected in their prices.
But such judgments are made all the time. Here’s one example, quoting two prominent economists about the nature of the overvalued market in 1998:
"I thinks there is a good deal of comparison between the market in 1929 and the market of today. I think both of them are bubbles….. If anything, I suspect I think there is more of a bubble in today’s market than there was in 1929."
(Milton Friedman, Nobel Prize winner in economics)*
"I define a bubble as a situation in which the level of stock prices is high, and the rate of growth of stock prices is high, because of a self-fulfilling prophecy in which people believe that the market is going to go up. On that basis, I think there has been an element of a bubble in the market for at least two years, possibly longer."
* (quoted from "Pricking the Bubble", John Cassidy, The New Yorker, August 17, 1998)
(Paul Samuelson, Nobel Prize winner in economics)*
Although both Friedman and Samuelson generally hold opposing views on macroeconomic policy, they were both in agreement that the market in 1998 was not correctly priced. And that was a full two years before the S&P 500 peaked.
The point is not that they both thought the market was over-valued. Rather, the point is that two prominent academic economists, when asked to comment about the real world, abandoned the Efficient Market Hypothesis.
In short, we accept the critique that the TSP Advisory models contradict the Efficient Market Hypothesis and the "random walk" model of financial asset prices. But so do the opinions of some Nobel prize-winning economists.
On the other hand, no critique of the Efficient Market Hypothesis lends direct or indirect support to the effectiveness of TSP Advisory’s models. The only thing that counts is whether the models work.
| More
on the Efficient Market Hypothesis
The Efficient Markets hypothesis has come under considerable pressure since its inception. It is not universally accepted. Obviously, stockbrokers and many Wall Street investment advisors cannot be expected to embrace a theory that dismisses their role, and the investment advice and esoteric charts that are the tools of their trade. But many others have issues with the theory. Here’s a sampling: ***********************
TSP Advisory has adopted the parameters of the tactical asset allocation scheme suggested by Mr. Bogle (plus or minus 15% movement in the percentage around a basic strategic asset allocation, as explained above). Mr. Bogle bases his TAA scheme around the current dividend yield, which has proved a fairly reliable predictor of stock market returns for the over the following 10 years. We should note that the TAA model used by TSP Advisory is not based on predicting future stock returns on current dividend yield. Mr. Bogle is the former Chairman of the Vanguard Group of Investment Companies, which pioneered the use of index funds in response to the "random walk" theory of the Efficient Market Hypothesis. It is telling that he advocates the use of TAA, which must be considered generally inconsistent with the Efficient Market Hypothesis. There is also evidence from statistical treatments in financial theory that demonstrate that the market is not completely random. For example, in the short run, broad market returns are positively correlated, like the weather. Just as a fair day is more likely to be followed by another fair day rather than a rainy day, a positive return are more likely to be followed by other positive returns. (Andrew W. Lo and A. Craig MacKinlay, authors of A Non-Random Walk Down Wall Street (1999, Princeton University Press, NJ.) Finally, Fama himself, working with Kenneth French, found that low price-to-book (P/B) and low price-to-earnings (P/E) stocks outperform the general market and, in particular, growth stocks (June 1992, Journal of Finance). There is no obvious reason why this "free lunch" should continue to exist over time in an efficient market, especially since documentation of the pattern goes back to the grand old man of security analysis, Benjamin Graham. But the patterns persist. Please note that none of these critiques of the Efficient Market Hypothesis lend direct or indirect support to the effectiveness of TSP Advisory’s models. That’s not the intent. The point is that stock prices -- and stock indexes -- do not necessarily follow a "random walk." |
Back to Top
7. Are TSP Advisory’s models "market timing" models?
Most financial professionals would consider TSP Advisory’s "Dynamic Asset Allocation" (DAA) model to be a variant of "market timing," that is, attempting to move money into or out of the market to catch the movements of the market, either up or down. In other words, sell high and buy low.
Most practitioners of "market timing" use "technical analysis", which generally refers to certain signals generated from the movements of the market being studied. TSP Advisory models are more generally based on the economic and market data associated with the different advances and declines of the stock market from 1970 on.
Most financial advisors recommend "buy-and-hold." In fact, some would say trying to time the market is a futile search for the Holy Grail of investing. In other words, it cannot be done. Why not?
You have to be right twice. You need to get out of an over-valued market before it goes down, and into an under-valued market as it heads up. It’s tough enough to get one out of two correctly. And, if you believe in the Efficient Market Hypothesis, the market is always correctly priced, so the question of valuation of the market "does not compute."
Some market experts point out that if you were out of the market for the five best days of each year, your returns would be considerably reduced. The argument is generally used as a caution against those who would try to time the market. Here’s a sample of the argument: A dollar invested in 1966 under a "buy-and-hold" strategy would be worth $11.71 by October 2001. But had the investor been out of the market in the best five days of each year since 1966, the dollar would be worth 15 cents (an 85% loss). This reaffirms the typical argument supporting the "buy-and-hold" theory.
On the other hand, if the investor were in for all but the five worse days each year since 1966, the dollar would have grown to $987.12. Although there is a reasonable case to be made for "buy-and-hold", arguing that you need to stay in to get the "best five days" of each year also means you get the five worse days. Not much of an argument for "buy-and-hold." (Based on a study by Birinyi Associates, as reported in Barrons, The Truth About Timing, 5 Nov 2001).
In summary, we believe the economic and market data associated with the different advances and declines of the stock market since 1970 provide markers that can be used to help chart the future course of the market. The TSP Advisory models are based on that economic and market data.
Back to Top
8. Why should TSP Advisory's system work?
TSP Advisory’s recommendations are based on quantitative models. We are not "market experts", "market gurus" or even market professionals. We believe that any effort to improve on the results of "buy-and-hold" should be based on the practice of following a good quantitative model, and not on the judgment of "market experts."
TSP Advisory has spent much time and effort in gathering and analyzing data, and developing predictive models. We follow these quantitative predictive models without superimposing any personal judgments. For that same reason, we offer no personal market commentary on our web site.
Consider that there are two ways of making predictions. The first method uses clinical, or intuitive judgments, and relies on individual expertise. The other way is "actuarial" or quantitative. In the first method, the practitioner relies on judgment, and may run through scenarios, evaluating likely outcomes, and applying knowledge, experience and common sense to reach a decision. The second method uses no subjective judgments, but instead consistently uses empirical relationship established between outcomes and data to reach decisions. In other words, a model.
James P. O'Shaughnessy, author of What Works on Wall Street: A guide to the Best Performing Investment Strategies of All Time (1997, McGraw-Hill, NY,NY) explains in his initial chapters why 80% of actively managed mutual fund portfolios cannot beat the passively managed S&P 500 index fund. His conclusion is that the actively managed funds under-perform because they are not managed by models but by "market experts" -- and the experts get it wrong. O'Shaughnessy uses a wide variety of sources to show that quantitative, empirical models beat the judgments of "experts." Models beat human forecasters because they reliably and consistently apply the same criteria time after time. He points out that even the humans who developed the models cannot out-predict their own models. As the ever-wise Pogo said: "We have met the enemy and he is us."
TSP Advisory has over the years developed and refined models with reliable predictive capability. Having invested that time and effort, we are still not "market gurus." We have learned just enough about the market to know it would be foolish to try to out-predict our own models. And we don’t try.
The reason why models beat human expertise lies in human nature. Human reason is limited by human nature; ego, fear, greed and various other emotions get in the way of reason. On the other hand, models are never bored, have no ego, no fear, and no memory. Or as O'Shaughnessy says, "If they were people, they'd be the death of any party."
Finally, consider the story of Sir Isaac Newton and the South Sea Company. Like the tulip bulb mania in Holland, the South Sea Company is a classic example of an investing mania. In the spring of 1720, Sir Isaac, then Master of the Mint, recognized that investment in the South Sea Company had become a speculative bubble, and incomprehensible to his rational, scientific nature: "I can calculate the motions of the heavenly bodies, but not the madness of people." In April 1720, he sold all his shares of the South Sea Company for a 100% profit of 7000 British pounds. But as the mania continued on into the late spring and summer, Sir Isaac reentered the market, buying an even larger stake of the South Sea Company. Unfortunately, he entered at the top, and lost 20,000 British pounds in the ensuing panic that punctured the South Sea bubble.
Sir Isaac Newton, the inventor of the calculus, the discoverer of the principles of gravitation, and arguably the most intelligent man who ever lived, could not use his human judgment to effectively time the market.
Unlike Sir Isaac, however, we heed Pogo’s advice. TSP Advisory relies on models, not subjective judgment. But given the necessary data and today’s computing capability, we suspect Sir Isaac could have developed some powerful quantitative models to aid his investment performance.
Back to Top
9. Does TSP Advisory use Modern Portfolio Theory?No. Although the theory is elegant, implementing the theory has some practical problems.
Modern portfolio theory (it dates from the 1950s) was intended to minimize risk in a portfolio of financial assets by taking advantage of statistical relationships among individual assets to obtain a given rate of return with minimum risk. Let's use stocks, for example. If you picked your stocks so they were inversely correlated (say, half went down when the other half went up, and vice-versa), you could make almost as much, but with much less risk. To most people, it seems obvious that you’re safer if you spread your bets around, if you’re willing to give up the chance of a big win. But the mathematics of portfolio theory let you diversify in the most efficient way: smaller wins, yes, but with much less risk. Or, as economist James Tobin reputedly said when asked to summarize his work upon winning the 1981 Nobel prize in economics: "don't put all your eggs in one basket."
Portfolio theory can be used to allocate among index funds, but is less useful among funds which are highly positively correlated (move together), such as the S&P 500 (C-Fund) and fixed-income investments (F-Fund). In theory, the addition of the "small cap" index fund (S-Fund) and international index fund (I-Fund) should provide greater possibilities for asset allocation that make sense under modern portfolio theory, because of weak or inverse correlations between the returns among these different funds.
But there are problems in trying to exploit the correlations among the three TSP stock funds. First of all, there is no reliable way to assess the correlation between foreign and domestic markets. Correlation coefficients rely on years of historical data, and they shift over time. You only know in retrospect what the real correlation coefficients are, which is not helpful; what you want to know are what those correlation coefficients will be. Such historical data are increasingly without meaning in an age of global capital shifts, various hedging techniques, and wide-spread privatization of formerly state-run industries.
As an aside, not everyone believes in modern portfolio theory, even when it works. Warren Buffett, a world-class investor, believes that "diversification is a protection against ignorance." And his partner, Charlie Munger, calls modern portfolio theory "twaddle." Get a couple billion ahead, and you can run against the tide.10. Which TSP participants will benefit from TSP Advisory?
TSP Advisory is for TSP participants who wish to have the opportunity to better the returns offered through a simple "buy-and-hold" strategy. For TSP Advisory to be useful to TSP participants, they should be willing to use the TSP’s interfund transfer provisions to shift according to the signals generated by TSP Advisory models.
11. Is TSP Advisory worth the money?
The longer you intend to stay invested in the TSP, and the more you are investing into your TSP account, the more money TSP Advisory can save you. The table below illustrates that point. Assuming a 7% return for the "buy-and-hold" investor, and a 9% return following TSP Advisory recommendations (both returns are real, i.e., corrected for inflation), a person saving $2000 a year (including agency contributions) would net $500 in extra retirement over 5 years. By "net", we mean that the person's retirement savings, after deducting for the cost of subscription to TSP Advisory (at $45 per year, also invested at 7%) would contain an extra $500 at the end of 5 years.
At the other extreme, following TSP Advisory recommendations would have netted over $560,000 extra for a person saving $12,000 a year (including agency contributions) over 30 years. (Just to make sure you understand that the table below is describing differences, the value of that TSP Advisory subscriber's account would be $1.76 million after 30 years, versus $1.20 million for the "buy-and-hold" investor after 30 years. The difference of $560,000 is the TSP Advisory difference).
Although we think these are reasonable estimates of the value of subscribing to TSP Advisory, please note the following assumptions. First, as one must always keep in mind for all financial comparisons, past performance is no guarantee of future return. Thus, neither the 7% real returns for the "buy-and-hold" investor nor the 9% for the TSP Advisory subscriber may hold in the future. The key point, however, is that the annual return of TSP Advisory's Dynamic Asset Allocation model was over two (2) percentage points greater than ""buy-and-hold" in back-tested modeling between 1981-2002, and four (4) percentage points greater than "buy-and-hold" in back-tested modeling between 1995-2002. We used the more conservative estimate (2% difference) to produce the table you see below.
Annual
TSP Contribution |
$2,000 | $4,000 | $6,000 | $8,000 | $10,000 | $12,000 |
Years to Retirement |
Figures in columns below are extra retirement savings from following TSP Advisory recommendations | |||||
| 5 | $500 | $1,200 | $1,900 | $2,700 | $3,400 | $4,200 |
| 10 | $2,900 | $6,400 | $10,000 | $13,500 | $17,100 | $20,600 |
| 15 | $9,000 | $19,200 | $29,500 | $39,700 | $49,900 | $60,200 |
| 20 | $21,800 | $45,600 | $69,400 | $93,200 | $117,000 | $140,800 |
| 25 | $46,200 | $95,500 | $144,800 | $194,100 | $243,400 | $292,700 |
| 30 | $90,400 | $185,400 | $280,400 | $375,400 | $470,400 | $565,400 |