Which funds will produce the most long-term growth? Unsurprisingly, even the most informed investors and researchers do not completely agree. Before proceeding with my recommendations, let me be clear that by proposing a portfolio of multiple funds, I am not suggesting that a more straightforward approach is a bad idea. If daunting complexity might cause you to delay investing, feel free to simplify. Some ways of doing so are described in the next section, which proposes various allocations. Waiting to invest will cost you a lot more than not implementing someone’s idea of the perfect portfolio.
We should start with background on portfolio characteristics. With regard to market cap, a company can be small cap, mid cap, or large cap. The terms micro cap and mega cap can be used to describe the extremities. Standard and Poor's creates the S&P 500 (large cap), S&P 400 (mid cap), and S&P 600 (small cap) indices. Their market cap boundaries for small cap stocks are $850M and $5.2B; for mid cap stocks they are $5.2B and $14.5B; and for large cap stocks the lower bound is $14.5B. By this definition, the total US market distributes over 80% of its weight to large cap stocks. Of course there are many small and mid cap companies, but the sum of their market value is a small fraction of the market value of all public US companies.
Stocks can be characterized on a spectrum of relative price, where the halves of the spectrum are called value and growth. Value stocks have relatively low share prices compared to fundamental metrics like earnings. In contrast, growth stocks have relatively high share prices compared to their fundamentals. In general, growth stocks have expensive prices relative to current fundamentals because investors have high expectations for growth of the company.
Many people conflate growth of a company and growth of its stock price. Not knowing the distinction, some people even think that a fund with "growth" in its title implies the expectation of greater long-term returns (e.g., Vanguard's small cap growth ETF). As we'll see below, the opposite is true: value stocks have higher risk and higher expected return. Value and growth are not dichotomous descriptors. As an analogy, someone may be described as extraverted or introverted, but the trait extraversion is a continuous dimension. So there are important differences between two extraverts who score at the 70th and 99th percentiles of extraversion, respectively; and some people around the 50th percentile are not especially introverted or extraverted. Because a cap-weighted total US index fund mixes growth and value stocks, it is characterized as a large blend fund (where large refers to the dominance of large cap stocks).
One division among foreign markets is important to understand. There are developed markets outside the US, led in size by Japan, the UK, Canada, France, Switzerland, Australia, and Germany. Then there are emerging markets, led by China, India, Taiwan, Brazil, Saudi Arabia, South Africa, Russia, Thailand, and Mexico. South Korea is grouped with developed markets by FTSE indices and with emerging markets by MSCI. When some people say "international markets", they mean developed markets. The titles of the funds SCHF, SWISX, and FSPSX indicate that they invest in "international" equity. We can check each fund's composition and find that they invest only in developed markets. If you're noticing that this convention leads to ambiguity in the word "international", you are right. The funds VXUS and IXUS resolve this ambiguity with the phrase "total international". In this guide, if I'm referring only to developed or emerging countries, I will specify, and otherwise "international" refers to all markets outside the US. EM is short for "emerging markets", and DM is short for "developed markets" outside the US.
Cap-weighted funds that invest in the total global market are simple, great options for many reasons.[1] Many people would be well-advised to buy VT for their stock allocation and put their feet up. However, those willing to add moderate complexity to their portfolios can be rewarded for their efforts.
Unfortunately, the method most Americans seem to employ when selecting stock funds is to check which funds have performed well recently in their domestic market. Perhaps they'll add 10 or 20% to an international fund as a cursory measure of diversification. But there are many decades of of data and careful analysis we can draw on to build a portfolio with the characteristics we desire. Here I will offer advice on how to apply the academic research that led to the Fama-French five-factor model, which is explained in this video. These findings indicate that while investing in a cap-weighted index fund is a successful approach, diversification in terms of risk factors and geography improves long-term returns as well as the reliability of returns.
Not all stocks carry the same expected return — a "risk factor" refers to a characteristic which accounts for systematic differences in the risk and returns of diversified portfolios. Deliberately taking the risks associated with certain categories of stocks carries a greater expected reward. In any asset class, risk and expected return are evaluated in relation to the risk-free rate of return. The risk-free asset is considered a one-month US Treasury bill. Risk factors influence the returns of bonds as well, but those factors are quite obvious. As we discussed earlier, bonds can be riskier than one-month T-bills along two dimensions: they can have longer duration or they can have lower credit quality. Bonds with more risk along either of these dimensions have greater expected return. What are the risk factors for stocks?
The first factor is simply the market factor: a cap-weighted stock market index tends to outperform the risk-free rate of return. This factor describes the baseline risk of stocks and how much the expected return exceeds that of one-month T-bills. A cap-weighted index fund like VT has exposure to only the market factor. The second factor is size: small cap stocks tend to outperform large cap stocks. The third factor is value (or relative price): value stocks tend to outperform growth stocks. The fourth factor is profitability: stocks with robust profitability tend to outperform stocks with weak profitability. And the fifth factor is investment: stocks with conservative investment into growth of company assets tend to outperform those with aggressive investment. The stocks which tend to outperform, like value stocks, deliver a higher expected return because of their greater risk.[2] This implies that an individual can choose to invest in less risky stocks if it suits their circumstances and risk tolerance.
A common misunderstanding is that buying a fund with value stocks means that you're investing in the value factor. However, the value factor is actually a portfolio of value stocks minus a portfolio of growth stocks. If value stocks return 9% and growth stocks return 5%, then the value factor portfolio returned 4%. This is synonymous with saying that the "value premium" was 4%. However, if value stocks return 12% and growth stocks return 20%, then the value premium was -8%. The value factor can have a negative return even if value stocks have a positive return. The opposite can occur as well, as in 2022 when value stocks fell by 5% and growth stocks fell by 31%. Using these figures, there was a positive US value premium of roughly 26 percentage points in 2022.
So when we buy value funds, we're not investing in a pure value factor portfolio. Instead, we're tilting our portfolio toward the value factor by overweighting value stocks and underweighting growth stocks. The same applies to the other factors. The size premium is the return of the size factor portfolio: small cap returns minus large cap returns.
We've established that there are premiums for these five risk factors. But while the premiums are positive in the long run, they are sometimes reversed on shorter time scales. Factor-tilted portfolios cannot provide constant outperformance, because their higher returns are the result of a different risk profile. That risk must entail underperformance during some periods, or it wouldn't actually be risk. 2020 was an especially poor year for value investing, not because returns were negative but because the returns of growth stocks dwarfed those of value stocks. Stock market trends are very difficult to predict in the short term, and those who embrace any investing approach need to understand that they will experience periods of underperformance relative to other approaches. Naturally, radical outperformance is also possible. For US value this occurred in 2000-2004 after the tech bubble popped, and in 2021-2022.
Since 1926 in the US, the value premium has been positive in 66% of three-year periods, 84% of ten-year periods, and 96% of 20-year periods. The average premium has been 3.1%. In the plot below, each data point indicates the value premium over the prior three years (red) or the prior 20 years (purple). The value factor is volatile over short timespans, and has been negative over many three-year periods. However, it has been positive over nearly all 20-year periods. More details about the analysis are in this footnote.[3]
The worst 20-year and 3-year periods for the US value premium occurred recently: see value vs. growth portfolios in 2001-2020 and 2002-2021. In particular, a recent period from 2017-2020 caused many people to question the value premium. There will always be people who think trends in the last five or ten years upturn everything we know about investing. But we can observe data on the impressive historical consistency of factor premiums, and find that someone with a long time horizon is very likely to reap these risk premiums. Someone with a short time horizon is not guaranteed to benefit from factor-tilting, but the probability of increasing their return is always greater than 50%.
The small cap premium is notably less reliable than the others. See this video and these papers for evidence. This implies that you shouldn't invest in pure small cap funds with a strong expectation of higher returns. However, some factor premiums are stronger within small cap stocks, so it still makes sense to overweight certain small caps, like those in value funds. The average magnitude of the value premium among US small caps has been 4.8%, compared to 2.3% among large caps. Among US small caps, the value premium has been positive in 73% of three-year periods, 91% of ten-year periods, and 99% of 20-year periods. [4]
The market factor is the return of the cap-weighted market minus the risk-free rate of return (the return of one-month US Treasury bills). This is the basic compensation an investor receives for accepting the risk of stocks instead of a guaranteed return. The average magnitude of the US market factor has been 6.5%, and it's been positive in 78% of three-year periods, 86% of ten-year periods, and 100% of 20-year periods. Even in the unusually successful US market, there have been periods as long as 18.8 years(!) in which one-month Treasury bills outperformed the stock market (the longest was Oct 1963 to Jul 1982).
Notice in the plot below that the other factors — except size, as we discussed — are nearly as consistent over long periods as the market factor. These factors are persistent in other countries too. If you believe that investing in cap-weighted stocks is a worthwhile risk, you should consider exposing your portfolio to these premia as well.
One fascinating research result is illuminated by the value premium. A widespread assumption is that the positive returns of the cap-weighted stock market depend on the emergence of new, revolutionary companies. Presumably an index that never updated, and kept only old companies would fall further and further behind an index that represented the evolving broad market. Jeremy Siegel and Jeremy Schwartz showed that this is untrue with an intriguing question and finding. The question: what would the returns of the S&P 500 be if the index had never updated, keeping the same companies for decades? The answer: the original S&P 500 companies from 1957 delivered far better returns than the actual S&P 500 index! The average outperformance was more than 1% per year from 1957 through 2012. One explanation for this is familiar to those who understand the value premium. Siegel wrote: "Although the earnings and sales of many of the new firms grew faster than those of the older firms, the price that investors paid for these stocks was simply too high to generate good returns." The original S&P 500 was much more heavily allocated to companies that tended to be value stocks, and had higher returns than the growth-heavy real S&P 500.
For a longer explanation of what factors are at a beginner level, see this podcast in which Ben Felix explains factors to his mom. For a more advanced explanation, see this footnote in the next section. It will help resolve questions caused by a superficial understanding of factors, such as: how could more profitable stocks be riskier? The bottom line is that factors do not travel alone: selecting stocks with low relative price is good, but any multifactor model indicates a risk premium for stocks with low relative price while also controlling for the other factors (most notably profitability). This motivates selection of stocks for low relative price and high profitability simultaneously, rather than in separate parts of a portfolio. I'll conclude in the next section that the best management from this perspective is provided by Dimensional Fund Advisors (DFA) and Avantis Investors.
Let's sing praises of diversification for a moment.
Increasing risk-adjusted return is one of the central goals of rational investing. Risk is often operationalized as volatility. If someone changes a portfolio to achieve higher return while keeping volatility constant, or to reduce volatility while keeping return constant, they have increased the risk-adjusted return. Diversification is the main mechanism for doing so. Investing in assets that have low or moderate correlations with one another enhances the risk-adjusted return of the overall portfolio. In particular, investors seek assets with positive returns when stocks are crashing. High-quality bonds are the most common potential hedge against crashes in the equity market.
Ideally, a portfolio will contain diversified assets that succeed and fail at somewhat different times, making routine volatility as well as the largest crashes more tolerable. Volatility is undesirable not only because it's uncomfortable: it increases the uncertainty of your wealth at any given future point, and therefore your uncertainty about lifetime consumption. These concepts lead us to an important insight: the expected return and volatility of an asset are not what we care about. We care only about how an asset will affect the expected return and volatility of our entire portfolio. Short-term US Treasury bonds are far less volatile than long-term US Treasury bonds. But high-quality bonds often perform well when stocks are crashing. Potentially large positive returns from bonds with longer duration can partly counteract the negative stock returns and reduce volatility — much more than short-term bonds would.
Recall the section on rebalancing, where I focused on the importance of maintaining your intended allocation of stocks and bonds. We'll explore another way in which disciplined rebalancing helps you, in combination with diversification. Let's start with an example: between January 1970 and June 2009, in Canadian dollars, the S&P 500 had an average annual return of 9.66%, a Canadian index returned 9.42%, and an international index excluding the US and Canada returned 9.28%. Superficially, you might think that this means an investor would've received the highest return by investing in the S&P 500 and ignoring the other two. But a periodically rebalanced portfolio with one-third allocated to each index returned 9.84%, a higher return than any of the funds with lower volatility than any of the funds! Over a period as long as four decades, these differences are more significant than they may seem.[5]
So diversification among stock funds and rebalancing can provide your portfolio with higher returns than any of its constituents. This exact result does not usually occur — it requires that the different funds have pretty similar returns — but the example illustrates the advantages of this approach. Rebalancing into the funds that have recently underperformed is a counterintuitive but effective practice — it makes a portfolio into more than the sum of its parts. This is true not only among stocks, but between asset classes. Rebalancing also serves to restore the intended portfolio, with its original risk and expected return.
From a defensive point of view, lack of diversification threatens your financial well-being. The financial advisor Larry Swedroe often tells the story of a retired couple he met in March 2003 on a book tour.[6] Three years earlier they had been invested in a concentrated portfolio of growth stocks. They were chasing the extraordinary recent returns of tech stocks enabled by the internet, and when the tech bubble burst in 2000, they took it on the chin. Their retirement portfolio was reduced from $13 million to $3 million. This is an extreme example, but not an anomaly. These are the kind of real financial consequences to which you make yourself vulnerable if you concentrate your portfolio in a small number of similar stocks and fail to diversify broadly across the market.
See, for instance, how the stock prices of Oracle, Intel, Cisco and Qualcomm behaved in 2000 and imagine buying shares as they approached their peaks. Apple and Amazon, now two of the most valuable companies, drew down by 82% and 94%. For Apple, much of this happened on a single day in September 2000 when its shares dropped 52% following an announcement about low earnings. Hopefully this illustrates how wild investing can be for those without diversified portfolios, even though I selected examples from companies that have survived to the present. Wikipedia has a good roundup of companies with multibillion dollar valuations that declared bankruptcy or fell drastically in valuation in the early 2000s.
Another growth stock frenzy deflated recently in 2021 and 2022. Many well-known, large cap stocks took major blows. The total US market suffered a drawdown of 25%. Meanwhile, there were drawdowns of: 56% for Amazon, 59% for Salesforce, 74% for Tesla, 77% for Meta (formerly Facebook), 76% for Netflix, 66% for Nvidia, 65% for AMD, 78% for Paypal, 82% for Block (formerly Square), 68% for Uber, 85% for Shopify, 87% for Zillow, 81% for Spotify, 85% for Palantir, 76% for Moderna, 91% for Coinbase, 82% for DoorDash, 92% for Roku, 90% for Rivian, 83% for Cloudflare, 91% for Snap (formerly Snapchat), 88% for Zoom, 96% for Peloton, and 99% for Carvana. For those who thought these companies seemed like exciting investments, 2022 was a hard year.
Take the stock performance of Snap, known for Snapchat, during the wild pandemic market:
The bottom line: individual stocks are hazardous to your wealth. Most people should isolate them in a small fraction of their portfolio if they feel the urge to dabble in trading stocks.
Some of the factor funds, like small cap value, will exhibit higher volatility than the cap-weighted total market. But this doesn't imply that factor investing needs to deliver a more volatile portfolio. In other words, although factor investing involves overweighting riskier stocks, it is not necessarily a riskier approach. To the contrary, investing in a variety of funds that are not perfectly correlated means that each risk factor acts as a somewhat independent source of returns, and those returns tend to deliver at different times.
A prominent example is the entire decade of 2000-2009, in which total US market and S&P 500 funds had negative average annual returns (-0.15% and -1.01%, respectively). Below is the total return chart for SPY, the first S&P 500 ETF and the security with the highest trading volume in the world. Two deep bear markets made the 2000s a bad time to invest in large cap US stocks.
This is often called the lost decade because stock returns in the US and some other countries were so poor. But it was not lost for value stocks: a US large cap value fund returned an average annual 4.39% and a US small cap value fund returned an average 9.13%. The 2000s also illustrated the value of geographic diversification: Vanguard's emerging markets fund returned an average 9.82%, an EM value fund returned an average 13.88%, a DM value fund returned an average 6.67%, and a DM small cap value fund returned an average 11.27%. Many naive investors, who base all of their investment knowledge on the last decade and think the S&P 500 can do no wrong, should peer a little further back in history.
How does factor investing fare when stocks plummet? Because correlations are high across the market during major crashes, don't count on risk factor diversification to protect against losses. As I explained in this footnote, there are two broad types of stock drawdowns: those caused by economic downturns and those in which a growth stock bubble bursts. During recessions like those in 2008 and 2020, value stocks tend to fall more than growth stocks. But when a growth stock frenzy collapses like in the early 2000s and 2022, value stocks tend to experience smaller drawdowns or even gains. Most people are at their greatest points of economic vulnerability during recessions like the 2008 financial crisis. They might be more likely to lose their job and take longer to find a new one, or make less money in commissioned roles. The main drawback of overweighting value stocks is that your stocks experience greater losses during these periods.
Even when the movement of funds with varied exposure to different risk factors or different countries is the same directionally, one market subset may gain more or lose less than another. The smoother returns of a global portfolio with multiple risk factors are thanks to the free lunch of diversification.
See this section's last footnote for more on geographic diversification.
How can we apply these ideas to a real portfolio? First we can consider geographic allocation. An investor purely concerned with global market cap weighting would allocate about 60% to the US, 29% to other developed markets, and 11% to emerging markets (assuming you count South Korea, 1.5% of the total, as a developed market). Many investors have a home country bias, and US investors in particular think their chauvinism is justified by the domestic returns from 2012 to 2021 (which were 16.3%, on average). One could debate this topic forever, and some people do. I think a large home country bias is not sensible regardless of where you live, and there are good reasons to avoid allocating 100% to the US market or any other country's market.[7] In the section on psychology, I elaborate on why recent past returns are a poor guide for selecting investments. I recommend 50-65% allocation to the US; in other words, allocate as if you were an alien and did not acknowledge divisions between countries. 60% is allocated to the US because that's where 60% of market value is domiciled. There are some evidence-based ways of deviating from geographic market caps, but most of them would suggest underweighting rather than overweighting the US.
We should, however, depart from global cap weights by overweighting subsets of the market with positive long-term premiums. These are stocks with low relative price (value stocks) and high profitability. As discussed above, size is an unreliable risk factor on its own. But because some factor premiums are stronger within small cap stocks, we should overweight small cap value stocks, especially those with high profitability. These stocks arguably have the highest expected return of any category.
This may cause some people to wonder why it isn't advisable to invest only in small cap value stocks. There is indeed a strong case that great returns would be expected for someone with this portfolio if they had a time horizon of at least 20 years and an iron will. But over the course of those 20+ years, there would undoubtedly be periods of significant underperformance compared to a more typical portfolio, like a cap-weighted global fund. For many people, these periods would be so psychologically punishing that it's doubtful they would be able to avoid changing their allocation or bailing out of stocks altogether. And of course, it's not certain that even over 20 years, small cap value stocks will continue to outperform. For augmented returns that are more reliable, we should invest in globally diversified stocks across all market capitalizations and with different relative price and profitability characteristics. We can overweight some stocks without reaching the extreme of omitting the underweighted stocks. The reliability of our returns will be the gift of risk factor diversification and geographic diversification.
Which ETFs and mutual funds can be used to build a portfolio with these characteristics? Various suggestions are organized in tables in the next section, and the webpage for each fund is hyperlinked.
Click here for the next section — Fund proposals
All sections:
- Cover page
- Introduction to index funds
- Thinking about risk
- Tax-advantaged accounts
- Your psychology
- Investing for retirement
- Guidelines for personal finance
- Building a stock portfolio
- Fund proposals
- Advice
- Practical information for execution
- Taxes
- Vocabulary and further resources
- Advanced topics
Footnotes:
1 Total market index funds have a lot going for them. First, because stock returns are positively skewed — a small fraction of stocks drive the market's growth — a poorly diversified investor is more likely to randomly underperform than randomly overperform. Total market index funds provide the broad diversification needed to capture skewed returns. Second, management fees are low because their holdings are based on straightforward rules. Third, they have low turnover, which implies chiefly long-term holdings, leading to tax efficiency and lower explicit trading costs such as the bid-ask spread.
Fourth, investing according to market cap weight means funds are distributed in proportion to the available opportunity set. This has several advantages. It means most trades occur with large cap stocks, which are highly liquid and have minimal trading costs. It maximizes the strategy's capacity. Capacity for any investment approach would be reached if everyone piled on the same strategy, eliminating the risk premium after costs. Because a total market index fund distributes assets according to the opportunity set, reaching capacity is not a concern. The strategy's capacity is vast. Moreover, the predominance of large cap stocks means that implicit trading costs are low. Buying shares puts upward pressure on the price, and selling exerts downward pressure. This is because, if you are selling shares (for example), the number of people willing to buy from you at a given price is limited, and after exhausting those offers, you must move your sell price lower in order to keep selling. Trading many shares thus causes meaningful price movement in a way that costs the trader. Stocks with larger market caps are less susceptible to the implicit cost of price movement, because any given trade taps a smaller fraction of the available liquidity. ↩
2 Value stocks are not just generically riskier and more volatile than growth stocks, moving up and down in greater amounts. Specifically, value stocks are economically vulnerable and tend to perform poorly during economic slowdowns and contractions. This is exactly when investors are most adverse to large losses, partly because their portfolios are already suffering painful drawdowns, and partly because their own financial security is often at greater risk during an economic recession. On the other hand, value stocks thrive when economic recovery and expansion is expected.
We can observe this in the most recent stock market crashes. In February and March 2020, VTI, Vanguard's total US stock fund, drew down by 35.0%, while DFLVX, a US large cap value fund, drew down by 41.0% and AVUV/DFSVX, US small cap value funds, drew down by 46.9%/47.6%. The greater losses for value stocks occurred in the context of widespread anxiety about the economy and public health; many people became unemployed. Value investors need to understand that the additional risk they're choosing to bear will manifest as greater losses precisely when loss is most unpleasant. If you lose your job and need to sell your stocks at the bottom of a crash, you won't reap the value premium. Value stocks also underperformed during the global financial crisis: between the peak in October 2007 and the trough in March 2009, total return was -55.4% for VTI, -63.7% for DFLVX, and -64.1% for DFSVX. However, the compensation for bearing the risk of value stocks emerges most powerfully during economic recovery.
Let me be clear that value stocks do not always underperform growth stocks during periods of negative return for the total market. The period 2000-2002 was a notable demonstration of this in the US. In 2022 the US experienced negative returns that were led by growth stocks. They were due to concerns related to inflation and rising interest rates, and exacerbated by the Russian invasion of Ukraine. Between their respective peaks and troughs, the total US market fund VTI drew down by 25.4%, the US large cap growth fund VUG by 35.6%, the US small cap growth VBK fund by 38.4%, the US large cap value funds AVLV/DFLVX by 19.3%/19.8%, and the US small cap value funds AVUV/DFSVX by 20.6%/18.0%.
Value stocks are not simply riskier than growth stocks, but exposed to different risks. ↩
3 The academic value factor is conventionally defined as a portfolio of the 30% of stocks with the highest book-to-price minus a portfolio of the 30% of stocks with the lowest book-to-price. This leads to the abbreviation "high minus low" or HML. In addition, each portfolio is split evenly between large cap and small cap stocks, so the academic value factor overweights small caps. In the plot of the US value factor, the factor is calculated for each lookback period with a one-month step. So the three-year lookback starts with Jul 1926 - Jun 1929, the second data point is Aug 1926 - Jul 1929, and so on. The first data points for the 20-year lookback are Jul 1926 - Jun 1946, Aug 1926 - Jul 1946, and so on. Ben Felix wrote an excellent whitepaper about these data, but since it is more than two years out of date, I did my own analysis. See the table below for some numerical details.
To do this yourself, go to Ken French's library and download the CSV files for univariate sorts on size, book-to-market, operating profitability, and investment. For market factor data, download the CSV file under the header "Changes in CRSP Data" next to "Fama/French 3 Factors" (or 5 factors). For small value data, download the CSV file under the bivariate sorts header next to "6 Portfolios Formed on Size and Book-to-Market (2 x 3)". ↩
US five-factor premia through Dec 2022
| Market | Value | Size | Profitability | Investment | |
|---|---|---|---|---|---|
| Premium Magnitude | 6.5% | 3.1% | 2.5% | 2.7% | 2.4% |
| 3 Year Consistency | 78% | 66% | 55% | 68% | 59% |
| 10 Year Consistency | 86% | 84% | 74% | 86% | 80% |
| 20 Year Consistency | 100% | 96% | 89% | 98% | 99.6% |
| Data start date | Jul 1926 | Jul 1926 | Jul 1926 | Jul 1963 | Jul 1963 |
4 Jack Vogel of Alpha Architect recently published a paper showing that if a large cap value portfolio is equal-weighted rather than market cap-weighted, its returns are about as strong as those of a small cap value portfolio. All other things equal, large caps are better because they have lower trading costs (see footnote 1 above). Equal-weighted value portfolios are available through Alpha Architect and are discussed in the next section.
The paper linked in the main text on the size factor is from authors at AQR Capital. For the interested reader, they have also written papers on the value and momentum factors, the latter of which is not part of the Fama-French five-factor model discussed above. The AQR papers are accessibly written and provide a window into some of the disagreements surrounding risk factors. This paper is an interesting commentary on how slightly altering the definition of value can increase the value premium. Finally, these two papers find risk factors like those in modern data in stocks extending back to the 19th century. ↩
5 Imagine $10,000 invested in the ways I described above. Over 39.5 years, a 9.28% average annual return would yield a final value of $332,900; a 9.42% return would yield $350,200; a 9.66% return would yield $381,900; and a 9.84% return would yield $407,400. For example, $10,000 × 1.0984^39.5 = $407,418. ↩
6 See chapter 3 of Larry Swedroe's book, co-authored with Kevin Grogan, Your Complete Guide to a Successful and Secure Retirement. ↩
7 One reason to diversify geographically is that there are country-specific risks, like inflation or political instability. Since any given downturn does not affect all countries equally, diversification offers moderate protection against the possibility that US companies will be unexpectedly affected by a crisis. Different stock markets are positively correlated, and this is most true during crashes. So while I wouldn't expect diversification to reliabily mitigate crashes, I would expect it to help in some cases.
This video on bear markets from Ben Felix is a great discussion of the psychological terror of market crashes. It also contains a powerful argument for diversification based on the Japanese stock market. Japan's market grew at a breathtaking pace in the 1970s and 1980s, then crashed in 1990 and performed very poorly for two decades. From 1990 onward, a Japanese investor with a portfolio of 100% Japanese stocks would have deeply regretted their lack of diversification. But someone who allocated 45% to Japan according to global cap weights at the time would have benefited from superior growth in other countries like the US. As this paper put it, "Diversification protects investors against the adverse effects of holding concentrated positions in countries with poor long-term economic performance. Let us not fail to appreciate the benefits of this protection." Ben Felix also points out that Japanese small cap value stocks have outperformed the total market, which supports risk factor diversification.
Both of these paragraphs offer lukewarm language, because I'm not claiming that the US stock market is likely to stagnate or crash for an extended period. I am saying that the future is uncertain and diversification provides moderate protection from certain risks without any clear drawbacks.
I expand more on the errors of US home country bias in this footnote to the risk section. ↩