New Book on the Life Cycle Model - "The Missing Billionaires: A Guide to Better Financial Decisions"

[GoWithTheCashFlow]

Amazon Link

Google Preview

Rational Reminder Podcast Episode

Don't let the odd title fool you. This is probably the best book on the life cycle model of consumption and investment written for the (intelligent) layperson.

The book is laid out in 4 sections:

1. Investment Sizing

Also known as the section Sam Bankman-Fried most needed to read, in this section the authors demonstrate why strategies based on maximizing expected wealth tend to lead to ruin and introduce (and defend) the expected utility framework as a more appropriate guide for decision making. They begin with the simple example of betting on the toss of an unfair coin (60/40 in favor of landing on heads) and establish that, though betting your entire budget or net worth on heads maximizes expected wealth, a constant fractional bet sizing strategy leads to more desirable outcomes. They extend this logic to the decision of how much to invest in stocks and finally formalize the decision strategy under the expected utility framework.

2. Lifetime Spending and Investing

With the expected utility framework firmly grasped, particularly in relation to how much risk to take in investment decisions, the reader is ready to tackle the joint lifetime consumption and portfolio choice framework. The authors guide the reader through the model and compare it to other spending rules. Finally, they present the annuity puzzle.

3. Where the Rubber Meets the Road

The devil is in the details, which are explored in this third section. Topics include how one might go about estimating their own risk aversion, human capital, characteristics of various assets classes, investing without access to the risk free asset, options, taxes, and risk vs uncertainty.

4. Puzzles

This last section explores different puzzles which either arise as a consequence of the life cycle model (observations which do not match the predictions) or which can be illuminated by it (unintuitive solutions to seemingly simple problems). Puzzles include why a lottery with a positive expected value might not be such a great bet, the equity premium puzzle, how a higher expected return can sometimes lead to a lower optimal allocation to the asset, and the long and short of leveraged ETFs.


With this book some of the most important and practical contributions of academic finance have finally been made accessible to the non-academic. Though challenging at times, it was a highly enjoyable read and easily worth the effort. I cannot recommend it enough. I have already gifted several copies to friends and family and anticipate referring back to my own copy for years to come.

[ScubaHogg]

I listened to the podcast yesterday and immediately downloaded the book. It sounds very interesting but with a silly title

[asset_chaos]

Doubly unfortunate title. When I search for the book, missing billionare is also the main title of a series of romance novels, which coincidentally might also be considered books on a particular, but different, life cycle model of consumption. In any event the finance book sounds worth checking out. Thanks for the review.

[Random Walker]

I just started the book based on a recommendation in a Larry Swedroe tweet. Book looks outstanding so far.

Dave

[OuterBanks]

I started reading the book. I love the 60/40 chance of heads probability scenario being used throughout the book so far. I was always the type to double down with any slight odds in my favor and the book has already made me realize I was taking too much risk when I really just needed to be 60-70% in equities to come out ahead over the long term.

[Sandtrap]

Who is this book for:
(excerpt)

[alpenglow]

I agree that the title is terrible, but if it a finance book recommended by Bogleheads, it is going on the reading list. Thank you.

[Random Walker]

I’m about 35 pages into the book. Using the concept of biased coin flips to bring up one of my favorite investing issues: volatility drag.

Dave

[GoWithTheCashFlow]

I’m about 35 pages into the book. Using the concept of biased coin flips to bring up one of my favorite investing issues: volatility drag.

Dave

Then you'll love the Costanza Trade chapter

[Harry Livermore]

When I search for the book, missing billionare is also the main title of a series of romance novels, which coincidentally might also be considered books on a particular, but different, life cycle model of consumption.

Maybe I should order one of each! Mrs. Livermore is a big fan of romance novels.
Cheers

[Horton]

I started reading the book. I love the 60/40 chance of heads probability scenario being used throughout the book so far. I was always the type to double down with any slight odds in my favor and the book has already made me realize I was taking too much risk when I really just needed to be 60-70% in equities to come out ahead over the long term.

I’m only about 30 pages into the book, but I also really like this example. I put the book down and created a spreadsheet version of the game so that I could test my strategy. My two initial strategies were: (1) a low fixed dollar amount ($1) each flip and (2) a low fixed percentage (4%) each flip. I quickly realized that the former would take thousands of flips to reach the max, so I switched to the latter. Looks like I caught on fairly well, although the relative amounts were very conservative - possibly because the authors led on that many went bust but also because I tend to be more conservative than most my age. I’m looking forward to the rest of the book.

[sc9182]

Thanks but no-thanks. Many a references to LTCM, Black-Scholes formula -- which created a failed monstrosity, the LTCM.
What is "unbeknown" upfront is "Risk" nor "variance" -- without knowing those apriori., their formula promptly fails. Yes, assumptions can be made, or ball-park those -- then again, betting big, LTCM way ? No-Thanks!

Sequence of Returns or Black-Swan events and such ilk - are hard to predict, let alone base one's model (life-savings) on.

There is a decent "hole" in their missing billionaires calculations. If avg of those missing 120,000 billionaires (of 4000 millionaires' lineage) is 1.5 Billion a pop -- we are talking a net of 180 Trillion USD. The only problem with that math is - the "missing pie" is sure lot-bigger/chunk compared to current NetWorth of the nation/world. I understand lot more monies could be literally created in the market due to bubble'ed up valuations of assets. But, if one wants to claim they have a bubble-billion, maybe they are referring to billion Liras - may not be billion USD of wealth !?

Still listening thru podcast - looking for more nuggets

[GoWithTheCashFlow]

I started reading the book. I love the 60/40 chance of heads probability scenario being used throughout the book so far. I was always the type to double down with any slight odds in my favor and the book has already made me realize I was taking too much risk when I really just needed to be 60-70% in equities to come out ahead over the long term.

I’m only about 30 pages into the book, but I also really like this example. I put the book down and created a spreadsheet version of the game so that I could test my strategy. My two initial strategies were: (1) a low fixed dollar amount ($1) each flip and (2) a low fixed percentage (4%) each flip. I quickly realized that the former would take thousands of flips to reach the max, so I switched to the latter. Looks like I caught on fairly well, although the relative amounts were very conservative - possibly because the authors led on that many went bust but also because I tend to be more conservative than most my age. I’m looking forward to the rest of the book.

Very nice! That's some impressive engagement with the material.

For those that want to run the coin flip experiment themselves, the authors created a tool on their website.

[Ben Mathew]

Thanks but no-thanks. Many a references to LTCM, Black-Scholes formula -- which created a failed monstrosity, the LTCM.
What is "unbeknown" upfront is "Risk" nor "variance" -- without knowing those apriori., their formula promptly fails. Yes, assumptions can be made, or ball-park those -- then again, betting big, LTCM way ? No-Thanks!

Sequence of Returns or Black-Swan events and such ilk - are hard to predict, let alone base one's model (life-savings) on.

The book is not about "betting big, LTCM way" or the Black-Scholes option pricing formula. It's not about bond trading or options trading. It's not about picking stocks. It does not involve leverage or derivatives.

The book is about financial planning using the lifecycle model, which is the standard model of saving, investing, and withdrawing in the academic literature. The lifecycle model has a number of advantages over the ad hoc SWR strategy.

When it comes to asset allocation, the basic prescription of the model is to maintain a consistent level of risk across time. In a simple setting with constant risk/reward, that would translate to a fixed asset allocation. The model does not say to take a lot of risk or little risk—that depends on how risk averse you are. But whatever the level of risk you want to take, the model says to be consistent about it across time. Don't bet big in June and small in July. Consistent risk exposure across time does not increase your exposure to tail risk or sequence of return risk (SORR)—it reduces it due to time diversification.

On the spending side, the model says to amortize the portfolio to calculate withdrawals. This is Amortization Based Withdrawal (ABW). It's a variable spending strategy that adjusts to portfolio performance. This again reduces your exposure to tail risk and SORR.

See this post for an illustration of how SORR is managed in the lifecycle model. A bad sequence of returns prior to the start of retirement does not reduce spending because risk is being held constant across time and the portfolio will recover fully when the market recovers (plus pick up a rebalancing bonus). A bad sequence of returns after retirement has begun will reduce spending during the downturn. But the portfolio and spending will recover fully when the market recovers (and again, pick up a rebalancing bonus on top). There is no permanent damage to the portfolio, which is often a concern with SORR.

[Finridge]

Thanks to this thread, I got the book and started reading it. So far it is very good.

[bobcat2]

The Buttonwood columnist at The Economist discusses "The Missing Billionaires".

How to avoid a common investment mistake
Think less about what to buy, and more about how much

“The Missing Billionaires” is not a discussion of the minutiae of ltcm’s bond-arbitrage trades. Instead, it examines what its authors argue is a much more important—and neglected—question than picking the right investments to buy or sell: not “what” but “how much”. People tend to answer this question badly. ...

In practical terms, the book’s crowning achievement is its explanation of the “Merton share”. This is a simple rule of thumb for determining asset allocation, which says that allocations should rise in proportion to expected returns, fall in proportion to the investor’s risk aversion and fall a lot in proportion to volatility.

Link to article - https://www.economist.com/finance-and-e ... id=1791361

BobK

[Horton]

I thought that this was a really intriguing quote on page 233. With real yields at 20+ years around 2.3%, I wonder if the respondents are backing up the truck or moving the posts?

In early 2016, we asked 60 friends and colleagues from the finance industry, most of whom were high-net-worth taxable US investors, the following question:

What risk-free, inflation-protected, after-tax return would you be willing to accept on the totality of your wealth for the rest of your life in order to completely and forever forgo any other investment opportunity?

The answers we received were almost entirely within a range of 1-4% per annum, with the lowest required return at 0%, and the highest, which was quite an outlier, at 8%. The average was about 2.5%.

[Bill Bernstein]

When it comes to asset allocation, the basic prescription of the model is to maintain a consistent level of risk across time. In a simple setting with constant risk/reward, that would translate to a fixed asset allocation.

Except that that's not quite what the authors advocate in the book. They do indeed start out with the Merton share, but then they endorse changing allocations--radically--according to the ERP calculated from the CAPE and TIPS yield. (See pages 53-65).

The LTCM criticism is valid; if the peak of your skill set is that you can solve differential equations as easily as most people brush their teeth, then it's easy to believe that answer for better finance is more math.

This would be true if the financial markets were an airfoil or an electrical circuit. But they're not, a discordance evident in the book and, one might add, in the LTCM episode. A classic example of that is the book's Merton assumption of a constant relative risk aversion (RRA). This is nonsense, of course. At the top of the market, everyone's a long-term investor with an RRA = 1. At the bottom, RRA is much larger, an increase that afflicts even academic finance's best and brightest, as evidenced, say, by how the U. Chi. endowment board blew a gasket in 2009.

That said, I agree it's a superb book for those in need of the basics of utility-driven portfolio design. I enjoyed it and learned more than a few things from it.

[DiploInvestor]

About 80% done with the book - reading it in small chunks to try to absorb its lessons. It bridges the gap between the too-easy and ubiquitous investment advice and a dense Econ text. Some of the math is challenging, but I suspect it’s just the Greek letters that can be off-putting. The overriding theme throughout the book is how much risk to take/dynamic asset allocation. I’m no expert, but I find this book fills a need between the stuff we hear and read all the time and the too-advanced material I don’t need. Overall I think it’s excellent, though a more thorough treatment of crypto in that particular chapter would have been welcome.

[GoWithTheCashFlow]

When it comes to asset allocation, the basic prescription of the model is to maintain a consistent level of risk across time. In a simple setting with constant risk/reward, that would translate to a fixed asset allocation.

Except that that's not quite what the authors advocate in the book. They do indeed start out with the Merton share, but then they endorse changing allocations--radically--according to the ERP calculated from the CAPE and TIPS yield. (See pages 53-65).

I agree their actual strategy is over-engineered. However, the fundamental principle behind it, that a change in the price of a claim on future earnings should, ceteris paribus, lead to a change in the quantity demanded, is sound. The trouble with their strategy is, of course, that ceteris is never paribus.

The LTCM criticism is valid; if the peak of your skill set is that you can solve differential equations as easily as most people brush their teeth, then it's easy to believe that answer for better finance is more math.

This would be true if the financial markets were an airfoil or an electrical circuit. But they're not, a discordance evident in the book and, one might add, in the LTCM episode.

Is this a claim that smart (but not wise) people with fancy models often become overconfident causing them to take on too much risk? If so, I agree, but I don't understand how it's a criticism of a book which encourages the reader to put risk front and center when thinking about financial decisions. Do you think that if people followed the book's advice it would lead them to make poorer financial decisions? Or is this more a comment on their dynamic asset allocation strategy?

A classic example of that is the book's Merton assumption of a constant relative risk aversion (RRA). This is nonsense, of course. At the top of the market, everyone's a long-term investor with an RRA = 1. At the bottom, RRA is much larger, an increase that afflicts even academic finance's best and brightest, as evidenced, say, by how the U. Chi. endowment board blew a gasket in 2009.

Is their RRA actually changing? Or has the risk they're facing increased (in the case of U. Chi., the risk of ruin) and, in particular, has the lack of consideration of the risk they were taking at the top caused them to be in an even more dire situation at the bottom than they would have been had they taken on risk more in line with their underlying preferences?

A response might be that there exist only revealed preferences and so it's nonsensical to describe the actions taken at the top of the market as being out of line with their preferences. My response to that would be that the true revealed preference came later once the risk could no longer be ignored and excess risk taking at the top was due to a behavioral error of not keeping risk top of mind. In other words, a complete understanding and cognizance of the risk they were taking (as the book is trying to get readers to do) would have led them to reduce their risk, even at the top of the market. Unfortunately, that understanding didn't come to them until the bottom.

[GoWithTheCashFlow]

I thought that this was a really intriguing quote on page 233. With real yields at 20+ years around 2.3%, I wonder if the respondents are backing up the truck or moving the posts?

In early 2016, we asked 60 friends and colleagues from the finance industry, most of whom were high-net-worth taxable US investors, the following question:

What risk-free, inflation-protected, after-tax return would you be willing to accept on the totality of your wealth for the rest of your life in order to completely and forever forgo any other investment opportunity?

The answers we received were almost entirely within a range of 1-4% per annum, with the lowest required return at 0%, and the highest, which was quite an outlier, at 8%. The average was about 2.5%.

We at least know some Bogleheads are backing up the truck.

Rationally, the goal posts should move, but your observation is interesting and leads me to rethink the "completely and forever" aspect of the hypothetical. Is that really the right question to ask?

[Ben Mathew]

When it comes to asset allocation, the basic prescription of the model is to maintain a consistent level of risk across time. In a simple setting with constant risk/reward, that would translate to a fixed asset allocation.

Except that that's not quite what the authors advocate in the book. They do indeed start out with the Merton share, but then they endorse changing allocations--radically--according to the ERP calculated from the CAPE and TIPS yield. (See pages 53-65).

It's still the Merton share. The equity risk premium (ERP) is an input into the formula for Merton's share, and so when that changes, the Merton share also changes. If your estimate of the ERP and stock risk is constant, you get constant risk/reward for investing in stocks and a constant Merton share. If you don't think that the ERP and stock risk is constant (Haghani and White don't), then the risk/reward for investing in stocks changes over time and the Merton share also changes. Changing one's allocation to keep up with a changing Merton share is something to be done with transaction costs in mind, and they discuss that on page 62.

The LTCM criticism is valid; if the peak of your skill set is that you can solve differential equations as easily as most people brush their teeth, then it's easy to believe that answer for better finance is more math.

It's also easy to believe that not knowing how to solve differential equations confers sage wisdom. It doesn't. Math is a tool. If you understand how to use the tool better, you can understand some aspects of the problem better. Financial planning involves a number of complex issues that can be clarified with math. It's very easy to trip up if you don't think about it systematically—something that happens all the time. The practice of financial planning could use more math and better models, not less math and simplistic ad hoc models like SWR which have led so many people astray.

That said, I agree it's a superb book for those in need of the basics of utility-driven portfolio design.

Glad you think it's a superb book. But who is not in need of utility-driven portfolio design? What should they use instead? Rules of thumb? From whom? Sage advice from certified experts who don't know math? Pronouncements from pop psychology? Models like SWR that use pass/fail metrics that effectively assume a binary utility function which is clearly worse than CRRA?

Utility is simply a statement about an individual's preferences. So portfolio design that is not "utility-driven" would be portfolio design that does not take into account the individual's preferences. How is that supposed to work?

I enjoyed it and learned more than a few things from it.

Great to hear this!

[WilliamOfOckham]

Amazon Link

Google Preview

Rational Reminder Podcast Episode

Don't let the odd title fool you. This is probably the best book on the life cycle model of consumption and investment written for the (intelligent) layperson.

The book is laid out in 4 sections:

1. Investment Sizing

Also known as the section Sam Bankman-Fried most needed to read, in this section the authors demonstrate why strategies based on maximizing expected wealth tend to lead to ruin and introduce (and defend) the expected utility framework as a more appropriate guide for decision making. They begin with the simple example of betting on the toss of an unfair coin (60/40 in favor of landing on heads) and establish that, though betting your entire budget or net worth on heads maximizes expected wealth, a constant fractional bet sizing strategy leads to more desirable outcomes. They extend this logic to the decision of how much to invest in stocks and finally formalize the decision strategy under the expected utility framework.

2. Lifetime Spending and Investing

With the expected utility framework firmly grasped, particularly in relation to how much risk to take in investment decisions, the reader is ready to tackle the joint lifetime consumption and portfolio choice framework. The authors guide the reader through the model and compare it to other spending rules. Finally, they present the annuity puzzle.

3. Where the Rubber Meets the Road

The devil is in the details, which are explored in this third section. Topics include how one might go about estimating their own risk aversion, human capital, characteristics of various assets classes, investing without access to the risk free asset, options, taxes, and risk vs uncertainty.

4. Puzzles

This last section explores different puzzles which either arise as a consequence of the life cycle model (observations which do not match the predictions) or which can be illuminated by it (unintuitive solutions to seemingly simple problems). Puzzles include why a lottery with a positive expected value might not be such a great bet, the equity premium puzzle, how a higher expected return can sometimes lead to a lower optimal allocation to the asset, and the long and short of leveraged ETFs.


With this book some of the most important and practical contributions of academic finance have finally been made accessible to the non-academic. Though challenging at times, it was a highly enjoyable read and easily worth the effort. I cannot recommend it enough. I have already gifted several copies to friends and family and anticipate referring back to my own copy for years to come.

+1. And thanks for the heads-up!

[bobcat2]

I believe one has to be cautious about taking the Merton share too literally.

I just looked up the current CAPE. It's 30.77 so one over CAPE is 0.0325. The current yield on 10 year TIPS is 0.0216. That makes mu in the numerator of the Merton share (the equity risk premium) a tiny 1.09%. If we assume the sd of equities is 20% and the investor has risk aversion of 2 then the Merton share currently is

k= 1.09/2*(.04) =1.09/.08 = 0.136 or 13.6% of the portfolio should be invested in stocks currently if one has average risk aversion. That doesn't seem close to reasonable to me.

BobK

[exodusing]

When it comes to asset allocation, the basic prescription of the model is to maintain a consistent level of risk across time. In a simple setting with constant risk/reward, that would translate to a fixed asset allocation.

Except that that's not quite what the authors advocate in the book. They do indeed start out with the Merton share, but then they endorse changing allocations--radically--according to the ERP calculated from the CAPE and TIPS yield. (See pages 53-65).

That's about where I stopped reading. Especially when they acknowledged problems using history to validate their model, then went on the say it's a good idea anyway since it worked in the past.

The LTCM criticism is valid; if the peak of your skill set is that you can solve differential equations as easily as most people brush their teeth, then it's easy to believe that answer for better finance is more math.

This would be true if the financial markets were an airfoil or an electrical circuit. But they're not, a discordance evident in the book and, one might add, in the LTCM episode. A classic example of that is the book's Merton assumption of a constant relative risk aversion (RRA). This is nonsense, of course. At the top of the market, everyone's a long-term investor with an RRA = 1. At the bottom, RRA is much larger, an increase that afflicts even academic finance's best and brightest, as evidenced, say, by how the U. Chi. endowment board blew a gasket in 2009.

Yes! Finance is not math (or at least not just math). A point which I believe you have emphasized.

That said, I agree it's a superb book for those in need of the basics of utility-driven portfolio design. I enjoyed it and learned more than a few things from it.

I guess I should at least browse the rest of the book.

[Ben Mathew]

k= 1.09/2*(.04) =1.09/.08 = 0.136 or 13.6% of the portfolio should be invested in stocks currently if one has average risk aversion. That doesn't seem close to reasonable to me.

Why does it not seem close to reasonable to you? I'm not saying you're wrong, but I'm interested in understanding your reasoning.

[bobcat2]

I wouldn't tell a 23-40 year old with average risk aversion to invest only 14% of their retirement portfolio in stocks on account of a particular forecast of equity returns. I would instead suspect that that particular forecast of equity returns is faulty.

BobK

[GoWithTheCashFlow]

I believe one has to be cautious about taking the Merton share too literally.

I just looked up the current CAPE. It's 30.77 so one over CAPE is 0.0325. The current yield on 10 year TIPS is 0.0216. That makes mu in the numerator of the Merton share (the equity risk premium) a tiny 1.09%. If we assume the sd of equities is 20% and the investor has risk aversion of 2 then the Merton share currently is

k= 1.09/2*(.04) =1.09/.08 = 0.136 or 13.6% of the portfolio should be invested in stocks currently if one has average risk aversion. That doesn't seem close to reasonable to me.

BobK

I think it can be taken literally, but with the understanding that there will be significant error bars around it. I also think 1/CAPE is not the right measure since it only accounts for dividends but no other forms of earnings distribution. Plus, there is no forward looking component. I prefer something like Damodaran's estimates, again understanding there are error bars.

One other note, it should be 13.6% of the present value of the total lifetime portfolio (not simply the retirement portfolio), which makes it much more reasonable to me (though still low for the reasons above).

[Ben Mathew]

I wouldn't tell a 23-40 year old with average risk aversion to invest only 14% of their retirement portfolio in stocks on account of a particular forecast of equity returns. I would instead suspect that that particular forecast of equity returns is faulty.

BobK

First, the Merton share applies to total wealth, not current savings. So it would tell a 23-40 year old with a lot of future savings coming in to hold a much higher stock allocation than 14%.

Second, you're saying that you think the estimated equity premium of 1.09% is wrong. That's not a problem with the Merton share. It's a problem with the equity premium you're using to calculate it. Use an equity premium that sounds reasonable to you. What do you think is a reasonable estimate of the current equity premium?

[bobcat2]

I believe one has to be cautious about taking the Merton share too literally.

I just looked up the current CAPE. It's 30.77 so one over CAPE is 0.0325. The current yield on 10 year TIPS is 0.0216. That makes mu in the numerator of the Merton share (the equity risk premium) a tiny 1.09%. If we assume the sd of equities is 20% and the investor has risk aversion of 2 then the Merton share currently is

k= 1.09/2*(.04) =1.09/.08 = 0.136 or 13.6% of the portfolio should be invested in stocks currently if one has average risk aversion. That doesn't seem close to reasonable to me.

I also think 1/CAPE is not the right measure since it only accounts for dividends but no other forms of earnings distribution.

Well in the book they use 1/CAPE. If they thought other measures were better, I assume they would have used the better measures. So it seems you believe the authors used the wrong measure for expected stock returns.

BobK

[Ben Mathew]

I also think 1/CAPE is not the right measure since it only accounts for dividends but no other forms of earnings distribution.

1/CAPE is based on earnings, not dividends. It's an estimate of the earnings yield.

The earnings yield would be the realized real return if earnings are constant in real dollars and valuations don't change. Dividend policy can be ignored assuming dividend irrelevance.

[exodusing]

If you use e/p you'd get about 4% equity returns. Whether one year p/e or CAPE is a better basis is not entirely clear. As noted, the authors went with CAPE.

[SRM007]

I wouldn't tell a 23-40 year old with average risk aversion to invest only 14% of their retirement portfolio in stocks on account of a particular forecast of equity returns. I would instead suspect that that particular forecast of equity returns is faulty.

BobK

Thanks Bob - I understand and agree with your position above... Would anyone on this forum advice their 25 y.o. with an average risk aversion to invest ~15% of their portfolio today and any ongoing contriutions at the same level - if you would, please share your insight

[GoWithTheCashFlow]

Well in the book they use 1/CAPE. If they thought other measures were better, I assume they would have used the better measures. So it seems you believe the authors used the wrong measure for expected stock returns.

BobK

I do. By the way, for anyone looking for a CAPE adjusted for retained earnings, as well as for time lags and changing corporate tax rates, see ERN's "Building a Better CAPE Ratio.

[bobcat2]

You're saying that you think the estimated equity premium of 1.09% is wrong. That's not a problem with the Merton share. It's a problem with the equity premium you're using to calculate it. Use an equity premium that sounds reasonable to you. What do you think is a reasonable estimate of the current equity premium?

It's not my problem because I wouldn't use it. But it's a problem with the book which advocates using 1/CAPE. I was using the book method and found that method unreasonable.

I was wrong not to realize they are using wealth not portfolio size. This still seems unreasonable. That a 65 year old with average risk aversion and nearly all their wealth in portfolio wealth should today hold 14% in equities seems unreasonable to me. I would think at a minimum such a person would be at least 25% equities.

BobK

[Ben Mathew]

I wouldn't tell a 23-40 year old with average risk aversion to invest only 14% of their retirement portfolio in stocks on account of a particular forecast of equity returns. I would instead suspect that that particular forecast of equity returns is faulty.

BobK

Thanks Bob - I understand and agree with your position above... Would anyone on this forum advice their 25 y.o. with an average risk aversion to invest ~15% of their portfolio today and any ongoing contriutions at the same level - if you would, please share your insight

As noted upthread, the Merton share does not advise this for a 25 year old.

[GoWithTheCashFlow]

I also think 1/CAPE is not the right measure since it only accounts for dividends but no other forms of earnings distribution.

1/CAPE is based on earnings, not dividends. It's an estimate of the earnings yield.

The earnings yield would be the realized real return if earnings are constant in real dollars and valuations don't change. Dividend policy can be ignored assuming dividend irrelevance.

Yes, but the choice between dividends and buybacks certainly affects the price part of CAPE. A company who only engages in buybacks and pays no dividends will have a much higher price (and thus a higher CAPE) than in the counterfactual world where they only paid dividends.

[Ben Mathew]

I would think at a minimum such a person would be at least 25% equities.

Why do you think this? Does it relate in any way to an expectation of an equity premium?

[exodusing]

I would think at a minimum such a person would be at least 25% equities.

Why do you think this? Does it relate in any way to an expectation of an equity premium?

If memory serves, Ben Graham said don't go outside 75/25 and 25/75. Seems a perfectly reasonable rule of thumb.

[Ben Mathew]

I also think 1/CAPE is not the right measure since it only accounts for dividends but no other forms of earnings distribution.

1/CAPE is based on earnings, not dividends. It's an estimate of the earnings yield.

The earnings yield would be the realized real return if earnings are constant in real dollars and valuations don't change. Dividend policy can be ignored assuming dividend irrelevance.

Yes, but the choice between dividends and buybacks certainly affects the price part of CAPE. A company who only engages in buybacks and pays no dividends will have a much higher price (and thus a higher CAPE) than in the counterfactual world where they only paid dividends.

I misunderstood what you meant by "accounts for dividends."

Yes, share buybacks vs dividends, and dividend policy in general, would affect the CAPE measure because it changes the amount of assets owned by a share, making the earnings per share not directly comparable across time (and CAPE averages earnings across ten years.)

Under these assumptions, the current earnings yield would still be the real return of the company's stock. But due to changes in the size of the assets underlying a share, we cannot do a simple average of earnings across time to get the current estimated earnings of the share. This can be corrected for, but the usual CAPE measure does not do this correction.

[Ben Mathew]

I would think at a minimum such a person would be at least 25% equities.

Why do you think this? Does it relate in any way to an expectation of an equity premium?

If memory serves, Ben Graham said don't go outside 75/25 and 25/75. Seems a perfectly reasonable rule of thumb.

Sounds like Ben Graham felt that the equity premium would never go that low. And maybe he's right.

The point I'm trying to make is that the optimal stock allocation has to depend on the equity premium. If there is no equity premium, the optimal stock allocation would be 0%. There is no reward for the risk, so why take it? So when people are stating that 20% or 50% or 80% is right, there is an implicit assumption there about the size of the equity premium (and stock risk and risk aversion) without which the recommended allocation does not make any sense.

[Horton]

I believe one has to be cautious about taking the Merton share too literally.

I just looked up the current CAPE. It's 30.77 so one over CAPE is 0.0325. The current yield on 10 year TIPS is 0.0216. That makes mu in the numerator of the Merton share (the equity risk premium) a tiny 1.09%. If we assume the sd of equities is 20% and the investor has risk aversion of 2 then the Merton share currently is

k= 1.09/2*(.04) =1.09/.08 = 0.136 or 13.6% of the portfolio should be invested in stocks currently if one has average risk aversion. That doesn't seem close to reasonable to me.

BobK

If, on the other hand, the ERP is 5%, risk aversion is 2, and standard deviation of equities is 16%, then about 100% of the portfolio should be in equities!

5% / (2 x 16% ^ 2) = 98%

[bobcat2]

I would think at a minimum such a person would be at least 25% equities.

Why do you think this? Does it relate in any way to an expectation of an equity premium?

I would think that for a 65 year old to hold less than 25% equities they would have risk aversion above average, say at least nearly 3. I don't think it's that unusual for people 60 & up to have less than 25% in equities, but that's mainly because they have high risk aversion. For a person at age 65 to hold less than 25% equities and have average risk aversion strikes me as unusual.

BobK

[exodusing]

I would think at a minimum such a person would be at least 25% equities.

Why do you think this? Does it relate in any way to an expectation of an equity premium?

If memory serves, Ben Graham said don't go outside 75/25 and 25/75. Seems a perfectly reasonable rule of thumb.

Sounds like Ben Graham felt that the equity premium would never go that low. And maybe he's right.

The point I'm trying to make is that the optimal stock allocation has to depend on the equity premium. If there is no equity premium, the optimal stock allocation would be 0%. There is no reward for the risk, so why take it? So when people are stating that 20% or 50% or 80% is right, there is an implicit assumption there about the size of the equity premium (and stock risk and risk aversion) without which the recommended allocation does not make any sense.

The problem is we don't know the equity premium in advance; we just have estimates. I don't believe we have estimation methodologies that are nearly good enough to make allocation decisions. I'm fond of the Ken French quote "Chance dominates realized returns", as well as the entire https://www.dimensional.com/us-en/insig ... -investing.

You could take Ben Graham's advice and then pick a number in his range based on taste. If you wanted to appear a bit more rigorous, you could look at the 60/40 global stock/bond allocation and consider whether you consider yourself more aggressive or conservative than the representative investor (such as the average investor based on portfolio size or trading volume). See the Bill Sharpe preferred portfolio thread here or the Ken French article linked above. You could end up the same either way.

[bobcat2]

I believe one has to be cautious about taking the Merton share too literally.

I just looked up the current CAPE. It's 30.77 so one over CAPE is 0.0325. The current yield on 10 year TIPS is 0.0216. That makes mu in the numerator of the Merton share (the equity risk premium) a tiny 1.09%. If we assume the sd of equities is 20% and the investor has risk aversion of 2 then the Merton share currently is

k= 1.09/2*(.04) =1.09/.08 = 0.136 or 13.6% of the portfolio should be invested in stocks currently if one has average risk aversion. That doesn't seem close to reasonable to me.

BobK

If, on the other hand, the ERP is 5%, risk aversion is 2, and standard deviation of equities is 16%, then about 100% of the portfolio should be in equities!

5% / (2 x 16% ^ 2) = 98%

But it's the percentage of their wealth that needs to be nearly 50%, not the percentage of their portfolio. So the percentage of a 25 year old's portfolio in equities would be extremely high. Our 25 year old believes they will retire in 40 years at age 65 and calculates the value of their human capital to be $1,970,000. The current size of their portfolio is $30,000. So they need to leverage up to get their stock exposure up another $460,000 or so. This strikes me as unreasonable advice.

BobK

[bobcat2]

The problem is we don't know the equity premium in advance; we just have estimates. I don't believe we have estimation methodologies that are nearly good enough to make allocation decisions. I'm fond of the Ken French quote "Chance dominates realized returns", as well as the entire https://www.dimensional.com/us-en/insig ... -investing.

Yes. Any estimated value of the equity risk premium is not to be expected.

BobK

[Ben Mathew]

I would think at a minimum such a person would be at least 25% equities.

Why do you think this? Does it relate in any way to an expectation of an equity premium?

I would think that for a 65 year old to hold less than 25% equities they would have risk aversion above average, say at least nearly 3. I don't think it's that unusual for people 60 & up to have less than 25% in equities, but that's mainly because they have high risk aversion. For a person at age 65 to hold less than 25% equities and have average risk aversion strikes me as unusual.

BobK

So only risk aversion matters? The equity premium doesn't feature in any way?

That doesn't make any sense to me. Why would someone with whatever risk aversion hold 25% or more stocks if they don't have any conception of an equity premium? Why hold stocks? What's the reward for holding stocks?

[Ben Mathew]

I would think at a minimum such a person would be at least 25% equities.

Why do you think this? Does it relate in any way to an expectation of an equity premium?

If memory serves, Ben Graham said don't go outside 75/25 and 25/75. Seems a perfectly reasonable rule of thumb.

Sounds like Ben Graham felt that the equity premium would never go that low. And maybe he's right.

The point I'm trying to make is that the optimal stock allocation has to depend on the equity premium. If there is no equity premium, the optimal stock allocation would be 0%. There is no reward for the risk, so why take it? So when people are stating that 20% or 50% or 80% is right, there is an implicit assumption there about the size of the equity premium (and stock risk and risk aversion) without which the recommended allocation does not make any sense.

The problem is we don't know the equity premium in advance; we just have estimates. I don't believe we have estimation methodologies that are nearly good enough to make allocation decisions. I'm fond of the Ken French quote "Chance dominates realized returns", as well as the entire https://www.dimensional.com/us-en/insig ... -investing.

You could take Ben Graham's advice and then pick a number in his range based on taste. If you wanted to appear a bit more rigorous, you could look at the 60/40 global stock/bond allocation and consider whether you consider yourself more aggressive or conservative than the representative investor (such as the average investor based on portfolio size or trading volume). See the Bill Sharpe preferred portfolio thread here or the Ken French article linked above. You could end up the same either way.

So you find the equity premium estimates too unreliable. But you are able to settle on an asset allocation without this. You are saying tilt the portfolio towards stocks or bonds relative to the average portfolio based on how much more aggressive or conservative I am relative to the average investor.

Two questions in trying to put this into practice:

1. How do I decide how much more aggressive or conservative I am relative to the average investor?
2. Once I know the answer to question 1, how much should I tilt my portfolio away from the average portfolio? 5%? 10%? And how does that change over time?

[bobcat2]

[
So only risk aversion matters? The equity premium doesn't feature in any way?

I didn't write that only risk aversion matters. Here's what I did write.
" I don't think it's that unusual for people 60 & up to have less than 25% in equities, but that's mainly because they have high risk aversion."

BobK

[Ben Mathew]

[
So only risk aversion matters? The equity premium doesn't feature in any way?

I didn't write that only risk aversion matters. Here's what I did write.
" I don't think it's that unusual for people 60 & up to have less than 25% in equities, but that's mainly because they have high risk aversion."

BobK

So how does the equity premium affect the stock allocation?