The fourth and final book in the “Investing for Adults” series by William Bernstein is Rational Expectations: Asset Allocation for Investing Adults. In Book 1: The Ages of the Investor, I learned to take advantage of a lucky streak in stocks and stop when I’ve won the game. In Book 2: Skating Where the Puck Was, I learned why it’s so hard to find any “new and improved” asset classes. In Book 3: Deep Risk, I learned about the scenarios that have led to permanent capital loss.

This final book includes the most specific advice about constructing your retirement portfolio. The entire series is great (and honestly not very long even read back-to-back), but this final book is especially dense with additional practical ideas for those that are already comfortable with investing basics. This isn’t at all scientific, but upon counting my Kindle highlights, Book 4 had 75 highlighted passages vs. 33, 25, and 36 respectively for Books 1-3. I’m only going to touch on the few that directly impacted my own portfolio construction.

Stocks. Here is an excerpt regarding how much of your portfolio should be allocated to international stocks.

Deployment among stock asset classes is relatively easier. The obvious place to start is with the total world stock market, as mirrored reasonably well by the FTSE Global All Cap Index, which in early 2014 was split 48/52 between U.S. and foreign equities. From there, we make three adjustments to the foreign allocation, two down and one up. First, the downs: if you’re like most people, your retirement liabilities will be in dollars, so a 52% foreign allocation is inappropriately high. Second, foreign stocks not only are slightly more difficult and expensive to trade but also are subject to foreign tax withholding. This presents no problem in taxable accounts, since those taxes will offset your liability to the IRS, but you lose that deduction if you hold foreign stocks in a sheltered account.

The up adjustment is a temporary one, since foreign stocks, as was discussed in chapter 1, currently have higher expected returns. So at the time of this writing, a foreign stock allocation somewhere in the 30% to 45% region seems reasonable.

Simplifying all that, as of early 2014, the middle recommendation would be roughly 60/40 US/international while the world market cap weighting was roughly 50/50. A little home bias is recommended for US investors.

As of mid-2020, the world market cap weighting is 57% US and 43% International (source), which you might round to 60/40. The adjustments are mostly the same, except that foreign stocks probably have even slightly higher future expected returns as the US stocks keep climbing. If you want to maintain a slight home bias, I would speculate this might change the recommended range closer to 65/35 or 70/30?

Bonds. The recommended list includes short-term US Treasuries/TIPS, bank CDs, and investment-grade municipal bonds. Bernstein is not a fan of corporate bonds.

Sooner or later, we’re going to have an inflationary crisis, and in such an environment, long duration will be a killer. Stick to short Treasuries, CDs, and munis.

Own municipal bonds via a low-cost Vanguard open-ended mutual fund for the diversification. Own Treasury bonds and TIPS directly, as there is no need for mutual funds or ETFs since they all have the same level of risk. Own bank CDs and credit union certificates under the FDIC and NCUA insurance deposit limits.

Asset location. I found this advice about spreading your holdings across Traditional IRAs, Roth IRAs, and taxable accounts to be very useful and practical. Importantly, this may be somewhat different that what you have read elsewhere. I don’t want to summarize incorrectly, so I will just use the excerpts:

To the extent that you wish to rebalance the asset classes in your portfolio, all sales should be done within a sheltered account. If possible, you should house enough of each stock asset class in a sheltered account so that sales may be accomplished free from capital gains taxes. Next, all of the REIT allocation certainly belongs in the sheltered portfolio, since the lion’s share of their long-term returns come from nonqualified dividends.

The real difference made by location occurs at the level of overall account returns. In terms of tax liability, Traditional IRA/Defined Contribution > Taxable > Roth IRA. This means that, optimally, you’d like to arrange the expected returns of each account accordingly, with the highest returns (i.e., highest equity allocation) optimally occurring in the Roth, and the lowest returns (i.e., lowest stock allocation) in your Traditional IRA/Defined Contribution pool. To the extent that this is true, it conforms with the stocks-in-the-taxable-side argument. That said, for optimal tax-free rebalancing, unless your Roth IRA is much bigger than your traditional IRA, you’re still going to want some stock assets in the latter.

It is definitely nice to be able to rebalance and not have to worry about picking stock lots, making sure you have the right cost basis at tax time, and paying capital gains taxes.

The third book in the “Investing for Adults” series by William Bernstein is Deep Risk: How History Informs Portfolio Design. As before, I’m just trying to pull out a few practical takeaways rather than summarize the entire book. In Book 1: The Ages of the Investor, I learned to take advantage of a lucky streak in stocks and stop when I’ve won the game. In Book 2: Skating Where the Puck Was, I learned why it’s so hard to find any “new and improved” asset classes.

The main problem addressed in this book is “deep risk”, the permanent loss of real (inflation-adjusted) capital. This contrasts with “shallow risk”, in which the value of something usually rebounds within 5-7 years or less.

Here are some deep risks that you can offset by purchasing insurance (and be happy if you never have to use it!).

• Death of income earner.
• Long-term health disability.
• Legal risk – lawsuit with large judgment.
• Select types of asset loss (i.e. theft, building fire).

Unfortunately, there are other deep risks against which you can’t buy insurance.

• Hyperinflation, prolonged and severe.
• Deflation, prolonged and severe.
• Confiscation by government.
• Devastation (war).

Over an extended period of time, history has shown us that “safe” bonds are often more sensitive to deep risk than stocks. Many countries saw 100% losses for their bondholders, while partial ownership in a business survived wars and regime changes. An example given was in Germany after World War II. Bonds are also at risk for inflation, while a 30-year fixed-rate mortgage (a negative bond) can be a great inflation hedge.

A portfolio of internationally-diversified stocks is the most practical way to protect yourself from both inflation and deflation. Historically, inflation is much more likely than deflation. You might have an event in one country, but it would be very rare to have a large majority of nations experience severe inflation and low stock returns all at the same time. In such a case you’d be looking at global devastation.

As for local confiscation and local devastation, you would be looking at foreign-held assets, foreign property, perhaps the right passports, and a plan to escape in a timely manner. This sounds like something that a billionaire might pay someone else to set up, but not so sure how practical it would be for most people.

Bernstein offers his own summary:

This booklet’s primary advice regarding risky assets is loud and clear: your best long-term defense against deep risk is a globally value-tilted diversified equity portfolio, perhaps spiced up with a small amount of precious metals equity and natural resource producers, TIPS, and, if to your taste, bullion and foreign real estate.

I admit that I am somewhat fascinated by worst-case scenarios, and I recommend reading the entire book for the full discussion. But in the end, my primary takeaway is that if you have a globally-diversified stock portfolio, you’ve done most of what you can in terms of deep risk. The rest is the same advice as before: consider TIPS if you have enough money, maximize Social Security, and keep some nice safe bonds and bank CDs for short-term needs (shallow risk).

A recurring theme in investing is that you start out learning the simple basics, then you feel like you can optimize things and spend a lot of effort trying to do so, and eventually you realize that simple is probably just fine. No matter how closely you mine the past, you can’t predict the future. As the Buffett quote goes, “If past history was all there was to the game, the richest people would be librarians.” That’s what came to mind when I read William Bernstein on safe withdrawal rates in retirement:

Even the most sophisticated retirement projections contain so much uncertainty that the entire process can be summarized as follows: Below the age of 65, a 2% spending rate is bulletproof, 3% is probably safe, and 4% is taking chances. Above 5%, you’re taking an increasingly serious risk of dying poor. (For each five years above 65, add perhaps half of a percentage point to those numbers.)

Source: The Ages of the Investor: A Critical Look at Life-cycle Investing.

Something to keep in mind when you become obsessed about getting from a 98% success rate to a 99% success rate on a simple retirement calculator from Vanguard or a fancy one like FIRECalc. (Not that I’ve done that, ever, of course…)

I am re-reading a series called “Investing for Adults” by William Bernstein. By “Investing for Adults”, Bernstein means that he assumes that you already know the basics of investing and that he can skip to more advanced insights. There are four parts:

• The Ages of the Investor: A Critical Look at Life-cycle Investing
• Skating Where the Puck Was: The Correlation Game in a Flat World
• Deep Risk: How History Informs Portfolio Design
• Rational Expectations: Asset Allocation for Investing Adults

A commonly-cited part of the first book The Ages of the Investor is the question “Once you have won the game, why keep playing?”. If you have enough money to buy a set of safe assets like inflation-adjusted annuities, delayed (and thus increased) Social Security payments, and a TIPS ladder to create enough income payments for life, you should seriously considering selling your risky assets and do exactly that. (This is referred to as a liability-matching portfolio, or LMP. You can keep investing any excess funds in risky assets, if you wish.)

A wrinkle to this plan is that you won’t know exactly when the stock market will help make that happen. Before you reach your “number”, you’ll most likely be buying stocks and hoping they grow in value. Let’s say you saved 20% of your salary and invested it in the S&P 500*. How long would it take you to “win the game”?

Historically, it could be as little at 19 years or as long as 37. That’s nearly a two-decade difference in retirement dates! Same savings rate, different outcomes.

This paradigm rests on too many faulty assumptions to list, but it still illustrates a valid point: You just don’t know when you’re going to achieve your LMP, and when you do, it’s best to act.

If, at any point, a bull market pushes your portfolio over the LMP “magic number” of 20 to 25 times your annual cash-flow needs beyond Social Security and pensions, you’ve won the investing game. Why keep playing? Start bailing.

If you don’t act, the market might drop and it could take years to get back to your number again. This is one of the reasons why some people should not be holding a lot of stocks as they near retirement. Some people might need the stock exposure because the upside is better than the downside (they don’t have enough money unless stocks do well, or longevity risk), but for others the downside is worse than the upside (they DO have enough money unless stocks do poorly, or unnecessary market risk).

I find the concept of a risk-free liability-matching portfolio (LMP) much harder to apply to early retirement, as it is nearly impossible to create a truly guaranteed inflation-adjusted lifetime income stream that far into the future. Inflation-adjusted annuities are rare, expensive, and you’re betting that the insurer also lasts for another 50+ years if you’re 40 years old now. Social Security is subject to political risk and may become subject to means-testing. TIPS currently have negative real yields across the entire curve, and only go out to 30 years. (As Bernstein explores in future books, you’ll also have to avoid wars, prolonged deflation, confiscation, and other “deep risk” events.)

* Here are the details behind the chart:

As a small thought experiment, I posited imaginary annual cohorts who began work on January 1 of each calendar year, and who then on each December 31 invested 20% of their annual salary in the real return series of the S&P 500. I then measured how long it took each annual cohort, starting with the one that began work in 1925, to reach a portfolio size of 20 years of salary (which constitutes 25 years of their living expenses, since presumably they were able to live on 80% of their salary). Figure 11 shows how long it took each cohort beginning work from 1925 to 1980 to reach that retirement goal.