Firstly, what exactly is Shiller's diversified stock portfolio. It certainly seems to have a lot more growth in the early years than say the S&P500, so is that growth due to high dividend payouts? The 1929 crash is hardly noticeable in his portfolio, yet huge in index prices. [Should those stockbrokers not have jumped?]
If your portfolio had stocks that went bankrupt or were merged in the year of holding, how do you buy them to rebalance the portfolio? A merger has already extracted that value for the "winners". An index just replaces that stock with a new "up and comer".
We do know that stocks are riskier than treasuries, and that junk bonds are riskier than treasuries. What exactly is the acceptable risk premium? Is it just a judgement of the participants, or some sort of pareto value for the slope of the risk premium?
Not everyone has the same investment goals. Pension funds need as much predictability as possible, so bonds are the preferred vehicle. That reduces bond returns if there is competition for the bonds. Historically defined pension plans were a major buyer of such investments. That has changed since those plans have largely disappeared except for government employees.
Bottom line is that I am not confident that the claimed "excess risk premium" is real and may be an artifact of index construction and other factors. Shiller is in effect claiming the market is inefficient. I don't believe it. If it was, there would be a lot more equity investors making out like bandits on the quiet.
I'm not sure of what the puzzle is. If you think of the asset price as an estimate of the potential cash flow value, buying bonds gets you fixed cash flow, but buying shares gets you a cash flow tied to overall growth in the economy.
Suppose you have a sum of money. If you lend the money to a large portfolio of business ventures, your cash flow will be some fraction, discounting for failures, of expected interest plus the return of your capital. If you use the money to buy shares of a large portfolio of business ventures, your cash flow has no upside limit. individual businesses may fail, but future cash flow will reflect overall economic growth. An interest and return of principal stream is limited no matter how much an economy grows unlike fractional ownership.
The risk in buying a portfolio of shares is that the economy will fail to grow. This risk, however, applies to both shares and bonds. Borrowers are less likely to repay when the economy is stagnant or collapsing.
P.S. Suppose Becky Sharp has somehow have gotten her mitts on 100,000 pounds. Let's not discuss how she did this. That's Thackeray's job, and he was good at it. Becky would have been sore tempted to buy consuls paying 3% and be done with it, but a more sophisticated and avaricious Becky might have considered buying a broad portfolio of shares instead. I'm sure Thackeray would have had his Becky go for the consuls and, by machinations, lose her fortune ignominiously. The shares option would have gotten us a very different novel, something more of the impoverished slime mold to mistress of the universe genre so popular in the 1980s.
The way I understand the puzzle is: If the return on asset E is more than that on asset B, one would expect more people or more funds to be channeled toward asset E. That would raise the price of E and reduce its return until that return equals that of B. That hasn't happened over more than a century. Therefore, not enough must have been channeled to E despite its persistent excess return (the premium). Why?
That is probably true for consumables like goods and services but that doesn't seem to be how non-consumable assets work. With goods and services there is an extrinsic value. People have needs and desires, and they use prices to choose which ones they will fulfill. Non-consumable assets are purchased for resale. Their price is based on what people perceive to be the price that people will be willing to pay in the future.
How do people estimate that future price? There are all sorts of rubrics, but they all involve past pricing and price trends in some manner. Someone choosing what to eat for dinner might compare B and E, decide E is too expensive and choose B instead. Someone hoping to resell something would compare B and E, decide that E was expensive but has been rising in price for a long time compared to B and, so, buy E. (I know there is something called value investing that looks for underpriced assets, but that is best used when choosing assets within a class.)
Bonds have an intrinsic limit on their future value. Stocks don't have that limit.
"Bonds have an intrinsic limit on their future value. Stocks don't have that limit." If that is true, then the puzzle gets worse. Why haven't people channeled enough funds into stocks and out of bonds until the returns from both come closer.
Bonds are more predictable. The money is highly likely to be there when you expect it. Stocks are more volatile. That's why I invest in both stocks and bonds and even keep some cash on hand.
Pensions need the predictability of bonds to make the needed payments to pension holders. It is no use telling a pensioner that the payments have to be reduced because of poor conditions but that the pensioner will be compensated sometime in the future (after they are dead?).
Simulations in the original Mehra and Prescott paper revealed that the risk aversion parameter that would justify the persistence of that type of equity premium is almost unbelievable. Plus, there is something deeply unsatisfying about attributing matters to people's preferences alone. That way just about everything can be explained away.
IIRC, consols weren't quite like that. They were issued at 3%, but the market priced them for awhile at a relatively high risk of default. The Exchequer knew what it was doing. These bonds created a large class of people with a very strong vested interest in the politico-economic success of the British Empire, because these bonds had a big upside when they became cheaper.
Just to clarify, my understanding is that equities in other developed countries do not exhibit nearly this degree of super-high returns in the modern era since WWII. If the risk aversion etc explanations posited here are correct, one would expect to see super-high returns in other stable (since WWII) developed countries, yet to the best of my knowledge, we do not. Japan is the clearest example, but European equities have also lagged the USA. In short, this appears to be an American phenomena. So the question should be, why America only? Tax laws? The collapse in the USA of defined benefit pensions and their replacement by retail investor 401Ks? Indexing? The relative inequality of wage growth vs corporate profits? The rise of monopoly power? I would be interested in Professor Delong's thoughts on these questions.
Maynard... I have always sort of assumed that the inequality of wage growth vs corp profits which is caused by monopoly power and corporate capture of law makers was at the root of the problem. But that appears to not be true given we only a short spell when that was not the case 1940 to 1980 and it looks like that period had similar growth to other decades with a great dal more Corporate Concentration and capture of law makers.
My understanding of Professor DeLong’s explanation is that it relies heavily on excess risk aversion/market volatility, and fear of being wiped out. however, I would think those concerns would be true of investors in most developed markets, Yet it is only the US that has experienced such outsized returns since World War II. That makes me think that, unless one is willing to posit that American investors are more risk averse than those in other developed countries, that the explanation for the seemingly excess returns lies elsewhere.
Perhaps the outsized returns have been based on trust in the American empire and its financial and legal system. World War II showed that it was a safe place to do business and park one's assets. I get the impression that England had this advantage and outsized returns in the 19th century when it ran the dominant world empire. I'll bet Rome had a similar advantage in its day.
The fact that America is viewed as a very safe place to invest should in theory reduce the excess returns of US equities over safe bonds. Instead, we observe the opposite.
My guess is that it is because equity prices have a steeper demand curve than bonds. If there were infinite money, equity prices would go towards infinity. Bond prices would reach a limit determined by the value of just holding cash. If a country is a more desirable place to park one's money, then one would expect a lot more demand for its bonds and equities both from locals and outsiders and, as a consequence, the spread to be larger than in a less desirable investment venue.
How does the economic theory that argues for a narrower spread work?
When it comes to the present day, if you don't like Ole Peters answer to the "equity premium puzzle", then I have nothing to say.
When it comes to the *past*, though, things are a little different. Between 1750 and 1900, annualized inflation in the UK was 0.39%; a pound in 1750 was worth about 1.8 pounds in 1900. (https://www.officialdata.org/uk/inflation/1750?endYear=1900&amount=10) This average concealed violent episodes not only of *inflation*, but also of *deflation*. That is to say, where bonds were subject to "wealth destroying inflation", equities were subject to wealth-destroying deflation.
"There is this a very strong sense in which it is much harder to raise money on the stock market then it should be, and that American business is hobbled thereby."
If uncertain losses do indeed loom larger than gains (Prospects Theory; thanks, Kahneman, Tversky and Thaler), keeping enough people from bidding on stocks, then it is astounding that a psychological issue -- to which humans may be hardwired -- has had such a large economic consequence. (I'm not sure I got this right; please correct me).
While prospects theory applies to individuals i did not see any research from Kahneman, Tversky, Thaler claiming that it applied to trading programs driven by automation. They should have filtered that out. But it appears they did not. They don't do much better than the Indexes.
The current notion of being “fully diversified” is to hold every asset in the same percentage as the market. If Apple Inc's market value is 5% of the total market value of US stocks, it should be 5% of a well-diversified portfolio.
When Apple's share price drops, so, too, does its market value. Say it goes to 4%; it automagically drops to 4% of my portfolio. As long as I held it at the market weight to begin with. Likewise if it zooms, its share of my portfolio rises proportionately
An index portfolio only buys and sells to reflect NEW stocks entering the market. Got some dividends? Spend them on the *identical* mix that you already own. Want to spend down your portfolio by say, 10%? Sell 10% of each position. Stocks that die, eg WeWork or SmileDirect? Your friend Mr. Market sold the value for you; all you need to do is delete the entry from your database.
There are some minor intricacies like mergers, buyouts, share repurchases or secondary offerings etc. They are all very small. Almost all BUYS and SELLS are to deal w cash moving in or out of the portfolio and your broker looking for commissions will go hungry
Past performance paragraph needs work.
Firstly, what exactly is Shiller's diversified stock portfolio. It certainly seems to have a lot more growth in the early years than say the S&P500, so is that growth due to high dividend payouts? The 1929 crash is hardly noticeable in his portfolio, yet huge in index prices. [Should those stockbrokers not have jumped?]
If your portfolio had stocks that went bankrupt or were merged in the year of holding, how do you buy them to rebalance the portfolio? A merger has already extracted that value for the "winners". An index just replaces that stock with a new "up and comer".
We do know that stocks are riskier than treasuries, and that junk bonds are riskier than treasuries. What exactly is the acceptable risk premium? Is it just a judgement of the participants, or some sort of pareto value for the slope of the risk premium?
Not everyone has the same investment goals. Pension funds need as much predictability as possible, so bonds are the preferred vehicle. That reduces bond returns if there is competition for the bonds. Historically defined pension plans were a major buyer of such investments. That has changed since those plans have largely disappeared except for government employees.
Bottom line is that I am not confident that the claimed "excess risk premium" is real and may be an artifact of index construction and other factors. Shiller is in effect claiming the market is inefficient. I don't believe it. If it was, there would be a lot more equity investors making out like bandits on the quiet.
I'm not sure of what the puzzle is. If you think of the asset price as an estimate of the potential cash flow value, buying bonds gets you fixed cash flow, but buying shares gets you a cash flow tied to overall growth in the economy.
Suppose you have a sum of money. If you lend the money to a large portfolio of business ventures, your cash flow will be some fraction, discounting for failures, of expected interest plus the return of your capital. If you use the money to buy shares of a large portfolio of business ventures, your cash flow has no upside limit. individual businesses may fail, but future cash flow will reflect overall economic growth. An interest and return of principal stream is limited no matter how much an economy grows unlike fractional ownership.
The risk in buying a portfolio of shares is that the economy will fail to grow. This risk, however, applies to both shares and bonds. Borrowers are less likely to repay when the economy is stagnant or collapsing.
P.S. Suppose Becky Sharp has somehow have gotten her mitts on 100,000 pounds. Let's not discuss how she did this. That's Thackeray's job, and he was good at it. Becky would have been sore tempted to buy consuls paying 3% and be done with it, but a more sophisticated and avaricious Becky might have considered buying a broad portfolio of shares instead. I'm sure Thackeray would have had his Becky go for the consuls and, by machinations, lose her fortune ignominiously. The shares option would have gotten us a very different novel, something more of the impoverished slime mold to mistress of the universe genre so popular in the 1980s.
The way I understand the puzzle is: If the return on asset E is more than that on asset B, one would expect more people or more funds to be channeled toward asset E. That would raise the price of E and reduce its return until that return equals that of B. That hasn't happened over more than a century. Therefore, not enough must have been channeled to E despite its persistent excess return (the premium). Why?
That is probably true for consumables like goods and services but that doesn't seem to be how non-consumable assets work. With goods and services there is an extrinsic value. People have needs and desires, and they use prices to choose which ones they will fulfill. Non-consumable assets are purchased for resale. Their price is based on what people perceive to be the price that people will be willing to pay in the future.
How do people estimate that future price? There are all sorts of rubrics, but they all involve past pricing and price trends in some manner. Someone choosing what to eat for dinner might compare B and E, decide E is too expensive and choose B instead. Someone hoping to resell something would compare B and E, decide that E was expensive but has been rising in price for a long time compared to B and, so, buy E. (I know there is something called value investing that looks for underpriced assets, but that is best used when choosing assets within a class.)
Bonds have an intrinsic limit on their future value. Stocks don't have that limit.
"Bonds have an intrinsic limit on their future value. Stocks don't have that limit." If that is true, then the puzzle gets worse. Why haven't people channeled enough funds into stocks and out of bonds until the returns from both come closer.
Bonds are more predictable. The money is highly likely to be there when you expect it. Stocks are more volatile. That's why I invest in both stocks and bonds and even keep some cash on hand.
Pensions need the predictability of bonds to make the needed payments to pension holders. It is no use telling a pensioner that the payments have to be reduced because of poor conditions but that the pensioner will be compensated sometime in the future (after they are dead?).
Simulations in the original Mehra and Prescott paper revealed that the risk aversion parameter that would justify the persistence of that type of equity premium is almost unbelievable. Plus, there is something deeply unsatisfying about attributing matters to people's preferences alone. That way just about everything can be explained away.
Because stock market investors are not rational.
IIRC, consols weren't quite like that. They were issued at 3%, but the market priced them for awhile at a relatively high risk of default. The Exchequer knew what it was doing. These bonds created a large class of people with a very strong vested interest in the politico-economic success of the British Empire, because these bonds had a big upside when they became cheaper.
But your analytic point, I think, is correct.
That's interesting. If I had to choose a 19th century author to manage my portfolio, I'd go for Thackeray.
Is it a puzzle or selection bias?
Just to clarify, my understanding is that equities in other developed countries do not exhibit nearly this degree of super-high returns in the modern era since WWII. If the risk aversion etc explanations posited here are correct, one would expect to see super-high returns in other stable (since WWII) developed countries, yet to the best of my knowledge, we do not. Japan is the clearest example, but European equities have also lagged the USA. In short, this appears to be an American phenomena. So the question should be, why America only? Tax laws? The collapse in the USA of defined benefit pensions and their replacement by retail investor 401Ks? Indexing? The relative inequality of wage growth vs corporate profits? The rise of monopoly power? I would be interested in Professor Delong's thoughts on these questions.
Maynard... I have always sort of assumed that the inequality of wage growth vs corp profits which is caused by monopoly power and corporate capture of law makers was at the root of the problem. But that appears to not be true given we only a short spell when that was not the case 1940 to 1980 and it looks like that period had similar growth to other decades with a great dal more Corporate Concentration and capture of law makers.
I am stumped.
My understanding of Professor DeLong’s explanation is that it relies heavily on excess risk aversion/market volatility, and fear of being wiped out. however, I would think those concerns would be true of investors in most developed markets, Yet it is only the US that has experienced such outsized returns since World War II. That makes me think that, unless one is willing to posit that American investors are more risk averse than those in other developed countries, that the explanation for the seemingly excess returns lies elsewhere.
Perhaps the outsized returns have been based on trust in the American empire and its financial and legal system. World War II showed that it was a safe place to do business and park one's assets. I get the impression that England had this advantage and outsized returns in the 19th century when it ran the dominant world empire. I'll bet Rome had a similar advantage in its day.
The fact that America is viewed as a very safe place to invest should in theory reduce the excess returns of US equities over safe bonds. Instead, we observe the opposite.
My guess is that it is because equity prices have a steeper demand curve than bonds. If there were infinite money, equity prices would go towards infinity. Bond prices would reach a limit determined by the value of just holding cash. If a country is a more desirable place to park one's money, then one would expect a lot more demand for its bonds and equities both from locals and outsiders and, as a consequence, the spread to be larger than in a less desirable investment venue.
How does the economic theory that argues for a narrower spread work?
When it comes to the present day, if you don't like Ole Peters answer to the "equity premium puzzle", then I have nothing to say.
When it comes to the *past*, though, things are a little different. Between 1750 and 1900, annualized inflation in the UK was 0.39%; a pound in 1750 was worth about 1.8 pounds in 1900. (https://www.officialdata.org/uk/inflation/1750?endYear=1900&amount=10) This average concealed violent episodes not only of *inflation*, but also of *deflation*. That is to say, where bonds were subject to "wealth destroying inflation", equities were subject to wealth-destroying deflation.
"There is this a very strong sense in which it is much harder to raise money on the stock market then it should be, and that American business is hobbled thereby."
If uncertain losses do indeed loom larger than gains (Prospects Theory; thanks, Kahneman, Tversky and Thaler), keeping enough people from bidding on stocks, then it is astounding that a psychological issue -- to which humans may be hardwired -- has had such a large economic consequence. (I'm not sure I got this right; please correct me).
While prospects theory applies to individuals i did not see any research from Kahneman, Tversky, Thaler claiming that it applied to trading programs driven by automation. They should have filtered that out. But it appears they did not. They don't do much better than the Indexes.
Retired index fund manager here…
The current notion of being “fully diversified” is to hold every asset in the same percentage as the market. If Apple Inc's market value is 5% of the total market value of US stocks, it should be 5% of a well-diversified portfolio.
When Apple's share price drops, so, too, does its market value. Say it goes to 4%; it automagically drops to 4% of my portfolio. As long as I held it at the market weight to begin with. Likewise if it zooms, its share of my portfolio rises proportionately
An index portfolio only buys and sells to reflect NEW stocks entering the market. Got some dividends? Spend them on the *identical* mix that you already own. Want to spend down your portfolio by say, 10%? Sell 10% of each position. Stocks that die, eg WeWork or SmileDirect? Your friend Mr. Market sold the value for you; all you need to do is delete the entry from your database.
There are some minor intricacies like mergers, buyouts, share repurchases or secondary offerings etc. They are all very small. Almost all BUYS and SELLS are to deal w cash moving in or out of the portfolio and your broker looking for commissions will go hungry