The Death of Portfolio Theory
Posted: Wed Apr 28, 2004 8:49 pm
Now for what brought me here, Robs response to a post I wrote about the death of portfolio theory being of use in todays valuations. I will preface my previous comments with the fact I haven't studied SWR at all, but I have my own dabbling with numbers in looking at the hazards of stock market returns being used to sustain retirement. As rates have moderated, being an insurance agent, I can see and share my appreciation of insurance and annuities in relationship to this matter. I will also state that any withdrawals that invade the principal of an investment are not sustainable, even if the asset continues to go up in value. In that regard, one must have an idea as to what the principal of an investment is and that is for another discussion.
I have done my own work with Dr. Shillers figures he posted on his site. In fact, I have written him and questioned some of his techniques for valuation, including the idea about increasing dividend values by 25% to reflect stock buybacks. But, in sum I have been able to confirm some of my own suspicions by reading his stuff, articles I have read and my own finance education at a major university where they taught the formula for market valuation, but didn't tell you how to get the values. It seems it was an article in Fortune or Forbes last summer that mentioned the risk premium that allowed me to fill in the blanks for valuing the market.
Owning a portfolio of stocks and reaping a return has become an academic fact taught in every university and pushed by every financial planner without any basis of proof other than that stocks have averaged so and so over the last million years. What is neglected in this matter is that there have been long periods of overperformance and long periods of underperformance in stocks. There was what was known as the random walk theory, which in so many ways said, you don't know what to buy or when to buy it, but history show something so just buy and hold. I don't know what can be proved that can be accepted in principal, but this isn't a valid theory, though one might be able to make a case for it. Bear markets, in almost every case were either due to bad economic times or were periods when stocks went from vast overvaluations to simular undervaluations. The opposite hold true about bull markets and the public always gets out or slows down investing during the undervaluation portions of both markets and overinvests in the overvaluation periods. I think random walk came out of the idea that from 1959 to the early 1970's stocks were continually overvalued and thus, the market cannot be timed. I think if one looks at the 1959-1990 period in sum, the opposite would have to be proved out, that the market can be timed, even if for a good period of time the fools continue to pile in and make money. Something tells me the academic nonsense that has been pushed over the past 30 years will run into difficulty when we are not in a bull market and are instead in a bear market and then the proof will be stated as the opposite, that people would be foolish to get involved in stocks.
I wasn't a hard working student in college, but I was on the verge of brilliance in how I could grasp and develop what I saw to be simple ideas. I can assure you that they never told us what capitalization rate to put in the valuation formula for stocks. Nor was I told what to put in the growth portion. Once stocks got so overvalued, I began to think that instead of dividends, they must have been talking about earnings, as I knew stocks needed to be priced a couple of percent above the risk free rate of return. What they did teach me was the risk free rate of interest was 3% and that there was a difference between a capital instrument and a money instrument, something else that seems lost on the modern financial world. I also knew there was a yield curve that assigned a greater amount of risk to a greater amount of time.
Where else they touched was portfolio theory, which I didn't quite understand because I didn't understand what they were putting into the formula. I did understand the principal that you could reap a real return of maybe 5% with a 3.5% risk by combining risky assets, knowing that statistically the group would perform at the expected return. I believe the late 1970's gave birth to a load of financial factors that were used in the 1980's and became the norm in the 1990's and among them was portfolio theory. I think it was always assumed that risk could be diversified away, but I doubt there were many valid formulas for this diversification before modern times.
Now, I think portfolio theory is okay and for that matter very valid, as long as it is used with assets that have been evaluated. But, this has been a total failure, in that with portfolio theory has come a group of theories, most notably random walk and the widely accepted falsehood that stocks return 10%. Stocks return 10% when they are priced to return 10%, not when they are bought at any price and held. There is no mix of portfolio that is going to make a group of stocks 300% overvalued return the historic return.
The best view of this happens when someone buys a bond and risk or interest rates change. If one buys a bond at 10% and rates decline to 6%, the price of his bond goes up, allowing him to realize a significant capital gain or hold what it is valued now for a 6% return over its remaining life. If he makes a decision to hold the bond, he makes both 6% and 10%. 6% from that time forward and 10% from the date he bought the bond. But, since he has so far made much more than 10% by holding the bond, he might assume that holding bonds makes 17%, while instead speculating in bonds makes 17%. This confusion can go even farther as the move back to par would result in a lower percentage loss than it did a gain going up.
Though the valuation formula for stocks is a little different, they are a lot closer to bonds than one many think. There are differences, mainly that the outcome is more volatile and there is no corpus for stocks. That means a change in the return on stocks creates a much larger price impact than bonds. But, there is a well kept secret about stock returns that it takes insight to see. That is that stocks aren't guaged by inflation plus for a return, but instead are directly related to the real or risk free rate of interest by measuring the growth above inflation and dividends. Knowing stocks represent a significant market risk and are prone to bad economies and bankruptcy, then it reasons that an instrument that promises no return of profit or principal must also carry with it a significant risk premium. The risk premium of the market as a whole should show up in the price of an index such as the SPX, a return that should be reaped by managing the risk of such a large portfolio.
Well, the dividend rate is 1.61% and the historical growth rate over inflation is 1.11%. Currently, the trend growth is 1.22% above inflation, but 2 years ago, it was zero. So, the number fluctuates over time, but for 133 years, the average has been 1.11%. What this means is the SPX is priced to yield 2.72% if you use history and 2.83% if you use the current trend growth over inflation. The risk free rate on debt instruments is reputed to be 3%, so the SPX, as a group of stocks is now priced to yield under the risk free rate and for that matter is not getting a risk premium for the extreme length of time holding stocks might represent. Like the bond example I presented above, one is now buying stocks at a portfolio return of less than the rate of inflation and holding them because they don't understand that what they are holding now will hold true for evermore.
Thus, when I say that portfolio theory is dead, I don't mean the theory is invalid, but when followed as a religion, with no regard to risk and return, it cannot work. When the portfolio proves out to be priced to yield less than the risk free rate of interest, based on generations of returns, then the theory cannot work to provide a greater return. The best it can do is hedge inflation and given the risk forgone in the present market is much greater than the risk of inflation, this isn't a good hedge. The return on stocks isn't based on some long term generality, but instead in reverse is based on the risk of holding stocks over the generations. Because the theory of portfolio reduced the risk to a managable level, the risk associated with the individual assets also disappeared and the assets ceased to be priced to reflect the risk they all have as individual assets. What a 10% return for stocks really means is inflation, plus dividends plus the risk premium should equal 10% and be offset by inflation plus dividends plus growth, which should also equal 10%. Trend inflation is in the 2.5% range and despite speculation to the contrary, must be recognized as such. Historical growth is 1.1% above the long term rate of inflation. That gives 3.6%. Add the 1.61% trailing dividend yield (forward dividends from year to year don't count because they are implied in the growth rate) and you arrive at a return of 5.21% on an asset that has no call other than bankruptcy or liquidation.
In my answer to Rob on the Prudent Bear Fund Chat board, I noted that our posts were related, but not exact. The subject of this board revolves around how much can you draw from your assets and not run out of money in retirement. There is a relationship because academia and assumptions made with portfolio theory have given people the impression they can use stocks as a retirement vehicle. They likely can if they have plenty of cash to back themselves up, but in having an entire retirement portfolio based on stocks or even a 50% mix between stocks on one side and bonds and cash or near cash on the other side still presents a dilemma under this form of valuation. There is absolutely no valuation assumption that is valid that can be used on stocks other than the dividends they pay, due to the fact they no longer represent the risk and return values they have historically represented. Even in the mid 1960's, when dividends reached historically low levels at the time, we had in place a trend where growth above inflation was over 2% and dividends were over 3%. Even that led to a period where the CPI adjusted value of stocks in 1966 weren't exceeded on a permanent basis until the early 1990's. Reaping inflation as a return out of an asset represents nothing but a taxable liquidation of something that has not actually sustained a real gain.
I have seen a lot of talk about the Shiller data on this board so I will reveal something I have gathered out of this data. If you take a spreadsheet or a financial calculator, you can find the IRR of the rate of inflation from 1966 to the last date, 2002. That is done by entering the formula (P-2002/P-1966)^36, calculating the inverse 36th power. If you do that with the dividends paid in 2002, you will come up with literally the exact same figure. What this says about the market is that corporate America, as represented by the SPX, did nothing to outperform inflation. Being the SPX was around 100 in 1966, which led to a bear market where values were not CPI exceeded for 26 years, then peak valuations based on this factor should be no more than roughly 570, which bear in mind could very well lead to another bear market where after a time, the PV of SPX 570 may not be exceeded for some 26 years. Since we are not down to this value yet, one might report that I am full of nonsense and that stocks are poised for even higher values. A new paradigm, as they might say. Well, if that be the case, then the idea of holding a portfolio of stocks to achieve an above risk gain is a thing of the past and chronic low returns is all that can be expected from stocks. At SPX 575, we are looking at a market today that is priced a little cheaper than the one in 1966 and priced to yield inflation plus about 4.35. Those type returns are available on AAA bonds in the present time, assets quite superior in risk to the underlying stocks. Being there is no excess return to reap out of the portfolio as priced, the theory is all but dead.
Academia has come with some new nonsense about portfolios. Being the SPX and the Nasdaq are cap weighted indexes, they are now recommending that more stocks be added to the mix. What this means is the rest of the stocks on the market are also going to near a return that results in a risk free, before expense return out of stocks. Nonsense begats more nonsense and as long as there is a premium being paid for stocks, there will be more stocks issued to draw more money at premium prices. Mal investment will continue and the return out of capital investment will be destroyed, the asset to not be visited again until the next bull market is well under way. Academias excuse this time was the high caps were required to be bought and that brought on the overvaluation, when in fact, the random walk theory, the long term hold theory and portfolio theory just don't wash together. Portfolio theory has to be combined with intelligent financial analysis of the risk associated with a specific group of assets and nothing else. Hence, we have stocks valued today that bear no resemblance to what they represent as potential returns in the future.
Lastly, I believe there is a real hole in the long term return on stocks, as the academias only rate the surviving stocks. As recently as 1997, Woolworth and Bethleham Steel were in the Dow Industrials and the went out of the index and straight into bankruptcy. I think Johns Manville was in the index in the early 1980's and suffered a similar fate. Recently we saw AT&T taken out of the index and the yields I have seen on their bonds is an indicator they too are headed for the scrapheap. We are looking at the evidence salmon exist, in that these are the fish that made it while thousands of fry died between birth and their return to the stream of their birth. If you check the SPX data, there were massive additional investments that were required to be made to make up an SPX point on the top of the market and the losses are much greater than the index shows. But the real point is the SPX and its various entities no longer represents a long term gain potential at its current price. It won't until the flaws of the current investment intelligencia have been exposed and accepted. I don't know about you, but I don't want to be holding these assets when the world figures this out.
I have done my own work with Dr. Shillers figures he posted on his site. In fact, I have written him and questioned some of his techniques for valuation, including the idea about increasing dividend values by 25% to reflect stock buybacks. But, in sum I have been able to confirm some of my own suspicions by reading his stuff, articles I have read and my own finance education at a major university where they taught the formula for market valuation, but didn't tell you how to get the values. It seems it was an article in Fortune or Forbes last summer that mentioned the risk premium that allowed me to fill in the blanks for valuing the market.
Owning a portfolio of stocks and reaping a return has become an academic fact taught in every university and pushed by every financial planner without any basis of proof other than that stocks have averaged so and so over the last million years. What is neglected in this matter is that there have been long periods of overperformance and long periods of underperformance in stocks. There was what was known as the random walk theory, which in so many ways said, you don't know what to buy or when to buy it, but history show something so just buy and hold. I don't know what can be proved that can be accepted in principal, but this isn't a valid theory, though one might be able to make a case for it. Bear markets, in almost every case were either due to bad economic times or were periods when stocks went from vast overvaluations to simular undervaluations. The opposite hold true about bull markets and the public always gets out or slows down investing during the undervaluation portions of both markets and overinvests in the overvaluation periods. I think random walk came out of the idea that from 1959 to the early 1970's stocks were continually overvalued and thus, the market cannot be timed. I think if one looks at the 1959-1990 period in sum, the opposite would have to be proved out, that the market can be timed, even if for a good period of time the fools continue to pile in and make money. Something tells me the academic nonsense that has been pushed over the past 30 years will run into difficulty when we are not in a bull market and are instead in a bear market and then the proof will be stated as the opposite, that people would be foolish to get involved in stocks.
I wasn't a hard working student in college, but I was on the verge of brilliance in how I could grasp and develop what I saw to be simple ideas. I can assure you that they never told us what capitalization rate to put in the valuation formula for stocks. Nor was I told what to put in the growth portion. Once stocks got so overvalued, I began to think that instead of dividends, they must have been talking about earnings, as I knew stocks needed to be priced a couple of percent above the risk free rate of return. What they did teach me was the risk free rate of interest was 3% and that there was a difference between a capital instrument and a money instrument, something else that seems lost on the modern financial world. I also knew there was a yield curve that assigned a greater amount of risk to a greater amount of time.
Where else they touched was portfolio theory, which I didn't quite understand because I didn't understand what they were putting into the formula. I did understand the principal that you could reap a real return of maybe 5% with a 3.5% risk by combining risky assets, knowing that statistically the group would perform at the expected return. I believe the late 1970's gave birth to a load of financial factors that were used in the 1980's and became the norm in the 1990's and among them was portfolio theory. I think it was always assumed that risk could be diversified away, but I doubt there were many valid formulas for this diversification before modern times.
Now, I think portfolio theory is okay and for that matter very valid, as long as it is used with assets that have been evaluated. But, this has been a total failure, in that with portfolio theory has come a group of theories, most notably random walk and the widely accepted falsehood that stocks return 10%. Stocks return 10% when they are priced to return 10%, not when they are bought at any price and held. There is no mix of portfolio that is going to make a group of stocks 300% overvalued return the historic return.
The best view of this happens when someone buys a bond and risk or interest rates change. If one buys a bond at 10% and rates decline to 6%, the price of his bond goes up, allowing him to realize a significant capital gain or hold what it is valued now for a 6% return over its remaining life. If he makes a decision to hold the bond, he makes both 6% and 10%. 6% from that time forward and 10% from the date he bought the bond. But, since he has so far made much more than 10% by holding the bond, he might assume that holding bonds makes 17%, while instead speculating in bonds makes 17%. This confusion can go even farther as the move back to par would result in a lower percentage loss than it did a gain going up.
Though the valuation formula for stocks is a little different, they are a lot closer to bonds than one many think. There are differences, mainly that the outcome is more volatile and there is no corpus for stocks. That means a change in the return on stocks creates a much larger price impact than bonds. But, there is a well kept secret about stock returns that it takes insight to see. That is that stocks aren't guaged by inflation plus for a return, but instead are directly related to the real or risk free rate of interest by measuring the growth above inflation and dividends. Knowing stocks represent a significant market risk and are prone to bad economies and bankruptcy, then it reasons that an instrument that promises no return of profit or principal must also carry with it a significant risk premium. The risk premium of the market as a whole should show up in the price of an index such as the SPX, a return that should be reaped by managing the risk of such a large portfolio.
Well, the dividend rate is 1.61% and the historical growth rate over inflation is 1.11%. Currently, the trend growth is 1.22% above inflation, but 2 years ago, it was zero. So, the number fluctuates over time, but for 133 years, the average has been 1.11%. What this means is the SPX is priced to yield 2.72% if you use history and 2.83% if you use the current trend growth over inflation. The risk free rate on debt instruments is reputed to be 3%, so the SPX, as a group of stocks is now priced to yield under the risk free rate and for that matter is not getting a risk premium for the extreme length of time holding stocks might represent. Like the bond example I presented above, one is now buying stocks at a portfolio return of less than the rate of inflation and holding them because they don't understand that what they are holding now will hold true for evermore.
Thus, when I say that portfolio theory is dead, I don't mean the theory is invalid, but when followed as a religion, with no regard to risk and return, it cannot work. When the portfolio proves out to be priced to yield less than the risk free rate of interest, based on generations of returns, then the theory cannot work to provide a greater return. The best it can do is hedge inflation and given the risk forgone in the present market is much greater than the risk of inflation, this isn't a good hedge. The return on stocks isn't based on some long term generality, but instead in reverse is based on the risk of holding stocks over the generations. Because the theory of portfolio reduced the risk to a managable level, the risk associated with the individual assets also disappeared and the assets ceased to be priced to reflect the risk they all have as individual assets. What a 10% return for stocks really means is inflation, plus dividends plus the risk premium should equal 10% and be offset by inflation plus dividends plus growth, which should also equal 10%. Trend inflation is in the 2.5% range and despite speculation to the contrary, must be recognized as such. Historical growth is 1.1% above the long term rate of inflation. That gives 3.6%. Add the 1.61% trailing dividend yield (forward dividends from year to year don't count because they are implied in the growth rate) and you arrive at a return of 5.21% on an asset that has no call other than bankruptcy or liquidation.
In my answer to Rob on the Prudent Bear Fund Chat board, I noted that our posts were related, but not exact. The subject of this board revolves around how much can you draw from your assets and not run out of money in retirement. There is a relationship because academia and assumptions made with portfolio theory have given people the impression they can use stocks as a retirement vehicle. They likely can if they have plenty of cash to back themselves up, but in having an entire retirement portfolio based on stocks or even a 50% mix between stocks on one side and bonds and cash or near cash on the other side still presents a dilemma under this form of valuation. There is absolutely no valuation assumption that is valid that can be used on stocks other than the dividends they pay, due to the fact they no longer represent the risk and return values they have historically represented. Even in the mid 1960's, when dividends reached historically low levels at the time, we had in place a trend where growth above inflation was over 2% and dividends were over 3%. Even that led to a period where the CPI adjusted value of stocks in 1966 weren't exceeded on a permanent basis until the early 1990's. Reaping inflation as a return out of an asset represents nothing but a taxable liquidation of something that has not actually sustained a real gain.
I have seen a lot of talk about the Shiller data on this board so I will reveal something I have gathered out of this data. If you take a spreadsheet or a financial calculator, you can find the IRR of the rate of inflation from 1966 to the last date, 2002. That is done by entering the formula (P-2002/P-1966)^36, calculating the inverse 36th power. If you do that with the dividends paid in 2002, you will come up with literally the exact same figure. What this says about the market is that corporate America, as represented by the SPX, did nothing to outperform inflation. Being the SPX was around 100 in 1966, which led to a bear market where values were not CPI exceeded for 26 years, then peak valuations based on this factor should be no more than roughly 570, which bear in mind could very well lead to another bear market where after a time, the PV of SPX 570 may not be exceeded for some 26 years. Since we are not down to this value yet, one might report that I am full of nonsense and that stocks are poised for even higher values. A new paradigm, as they might say. Well, if that be the case, then the idea of holding a portfolio of stocks to achieve an above risk gain is a thing of the past and chronic low returns is all that can be expected from stocks. At SPX 575, we are looking at a market today that is priced a little cheaper than the one in 1966 and priced to yield inflation plus about 4.35. Those type returns are available on AAA bonds in the present time, assets quite superior in risk to the underlying stocks. Being there is no excess return to reap out of the portfolio as priced, the theory is all but dead.
Academia has come with some new nonsense about portfolios. Being the SPX and the Nasdaq are cap weighted indexes, they are now recommending that more stocks be added to the mix. What this means is the rest of the stocks on the market are also going to near a return that results in a risk free, before expense return out of stocks. Nonsense begats more nonsense and as long as there is a premium being paid for stocks, there will be more stocks issued to draw more money at premium prices. Mal investment will continue and the return out of capital investment will be destroyed, the asset to not be visited again until the next bull market is well under way. Academias excuse this time was the high caps were required to be bought and that brought on the overvaluation, when in fact, the random walk theory, the long term hold theory and portfolio theory just don't wash together. Portfolio theory has to be combined with intelligent financial analysis of the risk associated with a specific group of assets and nothing else. Hence, we have stocks valued today that bear no resemblance to what they represent as potential returns in the future.
Lastly, I believe there is a real hole in the long term return on stocks, as the academias only rate the surviving stocks. As recently as 1997, Woolworth and Bethleham Steel were in the Dow Industrials and the went out of the index and straight into bankruptcy. I think Johns Manville was in the index in the early 1980's and suffered a similar fate. Recently we saw AT&T taken out of the index and the yields I have seen on their bonds is an indicator they too are headed for the scrapheap. We are looking at the evidence salmon exist, in that these are the fish that made it while thousands of fry died between birth and their return to the stream of their birth. If you check the SPX data, there were massive additional investments that were required to be made to make up an SPX point on the top of the market and the losses are much greater than the index shows. But the real point is the SPX and its various entities no longer represents a long term gain potential at its current price. It won't until the flaws of the current investment intelligencia have been exposed and accepted. I don't know about you, but I don't want to be holding these assets when the world figures this out.