Friday, November 10, 2023

Stocks for the Long Run, Sixth Edition

Jeremy Siegel recently wrote, with Jeremy Schwartz, the sixth edition of his popular book, Stocks for the long Run: The Definitive Guide to Financial Market Returns and Long-Term Investment Strategies.  I read the fifth edition nearly a decade ago, and because the book is good enough to reread, this sixth edition gave me the perfect opportunity to read it again.

I won’t repeat comments from my first review.  I’ll stick to material that either I chose not to comment on earlier, or is new in this edition.

Bonds and Inflation

“Yale economist Irving Fisher” has had a “long-held belief that bonds were overrated as safe investments in a world with uncertain inflation.”  Investors learned this lesson the hard way recently as interest rates spiked at a time when long-term bonds paid ultra-low returns.  This created double-digit losses in bond investments, despite the perception that bonds are safe.  Siegel adds “because of the uncertainty of inflation, bonds can be quite risky for long-term investors.”

The lesson here is that inflation-protected bonds offer lower risk, and long-term bonds are riskier than short-term bonds.

Mean Reversion

While stock returns look like a random walk in the short term, Figure 3.2 in the book shows that the long-term volatility of stocks and bonds refutes the random-walk hypothesis.  Over two or three decades, stocks are less risky than the random walk hypothesis would predict, and bonds are riskier.

Professors Robert Stombaugh and Luboš Pástor disagree with this conclusion, claiming that factors such as parameter and model uncertainty make stocks look riskier a priori than they look ex post.  Siegel disagrees with “their analysis because they assume there is a certain, after inflation (i.e., real) risk-free financial instrument that investors can buy to guarantee purchasing power for any date in the future.”  Siegel says that existing securities based on the Consumer Price Index (CPI) have flaws.  CPI is an imperfect measure of inflation, and there is the possibility that future governments will manipulate CPI.

Siegel continues: “Additionally, the same caution about the interpretation of historical risk that applies to stocks also applies to every asset.  All assets are subject to extreme outcomes called tail risks or black swan events.”

Rating Agencies’ Role in the Great Financial Crisis of 2008-2009

Siegel offered a partial defense of rating agencies who failed to see that mortgage-related securities were risky:

“Standard & Poor’s, as well as Moody’s and other ratings agencies, analyzed these historical home price series and performed the standard statistical tests that measure the risk and return of these securities.  Based on these studies, they reported that the probability that collateral behind a nationally diversified portfolio of home mortgages would be violated was virtually zero.  The risk management departments of many investment banks agreed with this conclusion.”

Standard statistical tests are notoriously unreliable when it comes to extreme events.  Flawed math might say an event has probability one in a trillion trillion when its true probability is one in ten thousand.  The world is full of people who use statistical methods they don’t understand, and there are others who use statistics to get the answer they want for personal gain.


I still agree with my conclusion in reviewing the previous edition: This book is very clearly written and offers powerful evidence for the advantages of investing in stocks. I highly recommend it to investors.


  1. So, admission is the first step, and I was caught with my pants down during this last 1.5yrs. My portfolio is more or less a 65/35 Equity/Bond split, and (true passive indexing) the bond portion is just a big chunk of VAB. Obviously, in hindsight, I should have been in VSB or something shorter, and I feel silly for not seeing the obvious.

    Anyway, I am only 40, time left to go, but I admit I'm really struggling now with how to position the bonds & "safe" part of my portfolio. Especially, watching the last few days as the "risk off" comes back and long bonds have shot up again.

    Having 35% sitting in Cash seems like not the best idea?
    I could allocate entirely to shorter bonds like VSB, and just recognize that the "safe" portion is meant to be boring and just sit there?
    Or stick with an aggregate like VAB, because, who knows, we could go back the other way?

    1. I'm definitely a passive investor when it comes to stocks, but the more I think about bonds, the more 'active' thoughts I have. Long-term bond yields through most of 2020 were so low that the outcome from holding them was guaranteed to be terrible. Will that happen again? I don't know. What I can say is that long-term bonds no longer look terrible. They may or may not disappoint, but they're not obviously terrible. So, I wouldn't blame anyone for seeing what happened to bonds recently, and deciding that was the past and it's OK to own all durations of bonds going forward.

    2. Yeah I think this is what I am choking on: I was so committed to passive investing that I am also now realizing that maybe this does not work in Bonds (ironically, meaning you need some active engagement in the "safe" part of your portfolio - and said human involvement probably goes a long way toward making it UNsafe).

      If I am going to need to more actively manage it, I feel the need for some sort of rule or algorithm, so that I am not just doing this based on emotion & articles I am reading at the time.

    3. I agree. With stocks, I would only intervene in extreme situations, say the Shiller CAPE going higher than it's ever been. I've made some attempt to automate even such extreme events with stocks. I'm still thinking about how to handle extreme situation with bonds, or if the more active interventions are even limited to extreme situations.

    4. Well, I eagerly await learning about whatever algorithm or approach you devise. There's surprisingly little out there about this - everything I Google or read is either purely passive (ie, just sit in XBB, which is what I just did - to my obvious detriment) or overly active which still makes me uncomfortable, comes with higher fees, etc.

      Proposal: Could a 'manage based on exceptions" approach work? Stay in XBB & remain fully passive, until/unless bonds reach some obvious & mathematical milestone where they are no longer likely to do their job as "safety"?

    5. Your 'manage based on exceptions' sounds roughly right to me. The challenge is to define the exceptions. I prefer not to have hard switches. 'Sell everything if rates reach X' is too harsh for me. I prefer a gradual formula-based approach, but I don't know what this should be right now.

  2. An argument can be made to just hold stocks and bonds passively but I would tend towards a much higher stock to bond allocation for example 75% to 90% stocks. The bond component should be broad based with more weighting on shorter term maturities, after all this is the "safe" component of your portofolio. This approach is probably more optimal and on average it should do well. You won't squeeze out very last dollar but you'll also avoid costly mistakes and many possible wrong assumptions about the market

    1. When I can see that long-term bonds must give poor outcomes, I couldn't own them. I get that many people can't understand the reasoning about bond returns or might be tempted to make other active choices based on hunches instead of solid reasoning. So, I don't blame others for sticking to a strategy through think and thin. I just can't own something that must disappoint.

  3. bonds are essentially a toxic asset class, of a complexity that few retail investors truly understand. avoid them entirely and prosper.

    1. There is complexity in corporate bonds, but government bonds are quite simple. However, I see a lot of danger in long-term bonds because of uncertainty about inflation. If government mismanagement drives up inflation, long-term bonds could get crushed.