Solid Trailing Returns: The Tortoise and the Hare

The long-term trailing excess returns of the balanced portfolio as of 2013 are attractive from both an absolute perspective (versus cash) and from a relative perspective (versus 60/40). The data is summarized in Table 11.1.

Total returns can be approximated by adding the return of cash to excess returns.

These long-term returns may come as a surprise to you, considering that the balanced portfolio has only a 20 percent allocation to equities, the highest performing asset class among the four. How can a portfolio consisting of 20 percent equities match the returns of an allocation of 60 percent equities over the long run? Such a result defies conventional wisdom.

The main reason the returns of the two strategies are about the same is because the balanced portfolio's superior stability has helped it keep pace with the higher octane 60/40 mix. Think of the old story of the tortoise and the hare as a useful analogy. The hare (60/40) sprints ahead and then has to slow down to catch its breath. The tortoise (the balanced portfolio) maintains a steady pace, at times trailing the hare and at other times forging ahead. Most notably, the tortoise takes the lead not because he has started to sprint, but because his steady pace helps him avoid major slowdowns. Over the long run, the two approaches achieve roughly similar results, but the tortoise is able to be more consistent. If we begin the clock just after the hare has peaked, then it may take the hare many years (perhaps as long as a decade or more) to catch up.

The reason the balanced portfolio has kept up with 60/40 is because it has experienced smoother returns and—more significantly—fewer major downturns. It is the big losses that hurt long-term returns simply because of the way the math works. If you lose 50 percent, then you have to earn 100 percent just to break even. The greater the loss, the greater the required gain to get back to your original starting point. This is where lower range of returns comes into play. Some investors don't care about return fluctuations and feel that if you can hold on for the long run, then you should own more stocks because you don't care about risk. They fail to realize that the volatility matters! The lower volatility and downside risk of the balanced portfolio (as you will read shortly) has enabled it to achieve a similar return as 60/40 over the very long run. The lower risk, of course, comes from the better balance in the allocation.

Another reason the balanced portfolio has produced long-term returns similar to 60/40 is because of a rebalancing advantage that comes from maintaining an allocation of highly diversified asset classes. Since the four components of the balanced portfolio respond differently to various economic environments, and the weighted volatility has been evened out, the range of returns for each asset class in each environment varies considerably. For instance, stocks could be down by a lot but long-term Treasuries could post big gains during the same period (as was the case in 2008 when stocks had an excess return of negative 38.3 percent while long-term Treasuries produced a positive 39.2 percent excess return). Therefore, when the portfolio is rebalanced back to the target allocation each year (20 percent equities, 20 percent commodities, 30 percent long-term Treasuries, and 30 percent long-term TIPS in the case of the balanced portfolio), the repeated process of buying low and selling high generally prevails over time. Each time a particular environment plays out, each asset class responds differently. Predictably, some perform well while others underperform. By rebalancing back to the target allocation, you are forced to sell a portion of the segments that have done well and buy more of the underperforming asset classes because their current allocations would be below target (stocks rose 26.9 percent and long-term Treasuries fell 25.7 percent in 2009, immediately after you would have rebalanced by buying stocks and selling Treasuries). This process results in a positive impact to overall returns, because asset classes are cyclical and oftentimes strong returns follow periods of underperformance.

The net impact is that the total return of the portfolio exceeds the weighted average return of each segment. Normally, the total return of a portfolio is fairly close to the weighted average return of its component parts. For example, if you invest 50 percent in stocks that earned 10 percent and 50 percent in bonds that earned 6 percent, over time you would expect your portfolio to average out to an 8 percent return. In the case of the balanced portfolio, the actual return is about 1 percent better per year than the weighted average. This fact has held true since 1927! In other words, if you took the average excess return of each asset class since 1927 and weighed it by its target allocation, you would get a total weighted return for the balanced portfolio that is about 1 percent less than the portfolio's actual return. Table 11.2 summarizes this outcome.

The powerful consequence of the process just described is that you can construct a portfolio of lower returning individual assets and still get a return similar to 60/40 over the long run. The extra boost shown in Table 11.2 is a premium that can be earned over time above and beyond any that are typically considered in portfolio construction. This premium is in addition to any risk premium that is available, and it only exists because of the supreme diversification of the balanced portfolio.

Stable Growth over the Long Run

The stability of the balanced portfolio's long-term returns can also be observed by following the growth of $1 invested in the portfolio from 1927 to 2013. Figure 11.1 demonstrates the actual experience you would have had through time. Over the long run the balanced portfolio's growth path closely resembles that of a straight upward sloping line.

Of course, there have been some detours along the relatively smooth path for balanced portfolio returns, but severe divergences have been relatively rare over this 86 year time frame. Ultimately reversion to the mean appears to hold over time as periods of underperformance have been followed by outperformance and returns get back to the long-term trend line. Later in this chapter I will take on the important topic of what causes these variances and why it is reasonable to expect future episodes to strike infrequently.

It would be beneficial if you are able to keep your focus on the long-term positive trend rather than on shorter-term time frames, which come later. By zooming out on the historical return chart and ignoring the shorter-term fluctuations you will observe a nearly straight path of excess returns. It makes sense that such a general trend would continue going forward. However, if you zoom in too much and, rather than having a 10+ year perspective, overreact to the experience from 1 or 2 years, then you may suffer longer-term underperformance by selling at the wrong time (just before a rebound). In looking at the chart too closely you may fail to see the upward sloping long-term line and instead mistakenly expect the line to become flatter or even downward sloping. For this reason I have emphasized the long-term trend here.

Low Volatility

With the 60/40 asset allocation and most other mixes, the total portfolio volatility generally falls somewhere in between the volatility of its parts. Stocks have a high volatility and bonds low. The volatility of 60/40 is therefore higher than bonds and lower than stocks.

Since 1927 the balanced portfolio's returns have experienced relatively low volatility. In fact, the total portfolio's volatility has been less than that of each of its components. This outcome can only be achieved with great balance. The comparison is summarized in Table 11.3.

On a relative basis, the balanced portfolio's volatility is about 30 percent lower than that of the conventional 60/40 portfolio. After being introduced to the advantages of good balance, you may not be surprised by the difference in volatility. A well-balanced mix of asset classes should be expected to be less volatile than a poorly allocated portfolio. Since the balanced portfolio consists of asset classes that are biased to perform differently during various economic environments, then as one segment does poorly another is likely doing well. Another way to envision this dynamic is by comparing the balanced portfolio to a finely tuned car engine. If you take the cover off the engine, you will see multiple moving parts. Some are going up, while others are simultaneously shifting down, sliding sideways, or sitting still. From the outside, the engine and the car are running smoothly. All the ups and downs inside the engine offset each other to result in a steady product outside the shell. The balanced portfolio works in the same way. Removing two of the key components because on their own they don't appear essential or beneficial is analogous to the 60/40 mix. The consequence is that the engine occasionally rattles, and it breaks down when the part that was removed is needed.

The difference in volatility between the balanced portfolio and 60/40 undermines conventional theories. The traditional assumption is that the higher the risk you take, the higher your long-term return should be. Another way to express this perspective is if the returns of the balanced portfolio and 60/40 are about the same, then the risk should also be about the same. This assumption is generally true in the investment world. However, when it comes to asset allocation there really is a free lunch, with proper diversification. In other words, by constructing a well-balanced, appropriately diversified portfolio, you are able to earn the same return with less risk. This holds true over the long run as well, contrary to popular investment belief. The core reason this opportunity exists is because the broad investment community fails to construct efficient portfolios. It just so happens that most investors simply do not take advantage of the free lunch, which is precisely what makes it free. The reason the inefficiency is available and the opportunity to take advantage of it exists is because the conventional approach is so flawed. It is not that the balanced portfolio is so special, but that 60/40 is so inefficient.

Consistent Returns Year by Year

Another way to observe the stability of the balanced portfolio is to look at past returns year by year. This view of returns may more closely align with how most investors think about performance because of the focus on shorter time frames. It also helps show the variations that may exist for a given portfolio strategy during different environments. For example, seeing how a portfolio performed in 2008 or the late 1990s or the early 1970s may provide insight into its potential vulnerabilities and biases.

Calendar-year excess returns for the balanced portfolio from 1927 to 2013 are shown in Table 11.4. All calendar years that experienced excess return declines greater than 10 percent have been highlighted.

You will notice that there are only three cases of excess return declines greater than 10 percent in the past 86 years! The last such episode occurred over 30 years ago. Note also that the returns in Table 11.4 are excess returns and therefore exclude the returns of cash. From a total return perspective, all of these numbers would have been better, particularly during the early 1980s when cash rates were unusually high. As an example, in 1981, when the balanced portfolio had one of its worst calendar years in terms of excess returns with a –16.7 percent outcome, cash earned 15.4 percent. This means the total return was –1.3 percent that year, which was one of the worst years in terms of excess returns.

Another interesting fact further highlights the stability of the balanced portfolio. There have only been three calendar years since 1927 in which equities, long-term Treasuries, long-term TIPS, and commodities all underperformed cash in the same year (1931, 1969, and 1981). Only in 1931—when nothing was safe as the Great Depression forced selling of every asset class to raise cash—was the total return (including cash return) negative for all four asset classes. In other words, the only time in history that stocks, Treasuries, TIPS, and commodities all were negative in the same year on a total return basis was at the depths of the Great Depression. This makes sense because whatever the economic climate, the negative influence on one asset class should concurrently positively impact another. History has demonstrated this undeniable fact time and again to quantitatively support the conceptual logic.

Due to the significant diversification benefits and downside protection of owning these four asset classes, overall calendar-year returns have been steady. In fact, since the Great Depression the balanced portfolio's worst calendar-year total return (including cash) was –6.9 percent (in 2008). You hate to ever lose money, but most investors would have been thrilled with such a result, considering that the average portfolio was down 25 to 35 percent that year. Table 11.5 lists all the calendar-year returns since 1927 during which excess returns were worse than –5 percent. The cash rates and total returns are also included for reference. The balanced portfolio's downside protection has been compelling for a very long period of time!

Based on these numbers, the balanced portfolio clearly meets the return objective that was targeted. It has produced stable, attractive excess returns (and total returns) since 1927. Next, I turn to how it has fared against the other two major portfolio goals: (1) avoiding extended stretches of poor results, and (2) minimizing drawdowns over shorter time frames.

Why Unexpected Rising Cash Rates and Falling Risk Appetite Can Cause Short-Term Underperformance

Over shorter time frames, the balanced portfolio does not always outperform cash and you should not expect it to. If that were the case, it would not be considered a risky investment. The same tenet applies to every investment strategy and every asset class. Since the goal of a balanced portfolio is to capture the excess returns above cash offered by risky asset classes, such an approach would be expected to underperform when cash outperforms asset classes.

Most of the time cash does not simultaneously outperform all asset classes. This must be true because capitalism would fail if it weren't. If cash outperformed all asset classes for extended periods of time, then investors would not exchange cash for asset classes; the reason they do so is because of the expectation that over time they will earn more than cash returns. A prolonged period during which cash does better than all asset classes would certainly diminish the confidence that is required for investors to part with their cash for the promise of higher returns. Going back to the principles in the first chapter about how the economic machine functions, if investors did not give up their cash, then the normal functioning of the machine would be interrupted. Widespread hoarding of cash is a sure way for an economy to stagnate. Under these circumstances, the Fed—which is responsible for reacting to economic conditions with an appropriate policy response—would lower the rate of cash to entice investors to invest in something that promises higher returns. The lower hurdle of cash makes other investments relatively more attractive. We have seen this dynamic play out since 2009 when the Fed lowered cash rates to zero.

When current cash rates rise unexpectedly to a significant degree, then a yield of cash higher than what had been discounted makes it a suddenly more attractive investment. At the same time, this sudden increase in rates makes all other asset classes less appealing relative to cash. The previous price of the asset (before the unexpected big increase in cash yield) had discounted a premium above a certain level of cash. When cash rates increased more than expected, then unless the premium fell the price of each asset would have to decline to make up for this new reality. Think of the price of each asset class being a function of the rate of cash and an excess return above cash. The price is such that it is expected to roughly produce a long-term return of cash plus the excess return.

If the cash yield suddenly and unexpectedly goes up, the price of the asset needs to go down if the expected excess return stays the same in order to readjust at the higher forward-looking return. A lower price results in better upside, all else being equal. Don't forget that cash would have to jump more than it is already discounted, since the expected increase is already factored into today's price. This type of situation may occur when the Fed unexpectedly raises interest rates (generally to stem risks of rising inflation). Note that the increase in the interest rate needs to be significant, because otherwise it would not materially impact investor behavior. It also must be unexpected. If everyone is anticipating rapidly rising interest rates and they do rise as expected, then there is no market surprise.

Two periods in history that provide helpful examples of this sort of dynamic occurred in 1994 and 1980. In 1994 the Fed shocked investors by rapidly hiking interest rates from 3 percent to 6 percent within a 12-month period because of a fear of rising inflation (which interestingly never materialized). A similar experience roiled asset class returns in 1980 as the Fed suddenly pushed interest rates to historic highs in an effort to finally end the inflation that had soared to unprecedented heights. Crucially, in both cases few investors were expecting interest rates to rise and it is the surprise that created the temporary conditions for asset classes to underperform cash. During these historical periods, all four asset classes simultaneously underperformed cash for a few months.

This dynamic may also exist when expectations of future cash yields suddenly shift upwards. Such an environment occurred as recently as 2013, when investors abruptly changed their expectations of future cash yields. With current cash yields at 0 percent and muted expectations of rising yields over the near term, a sudden shift in expectations to cash yields rising faster than discounted simultaneously hurt all asset class prices. That is, the current yield of cash did not change, but the future expectations of cash yields moved up significantly. And it was the change in expectations that negatively impacted all asset class prices concurrently. Note that this phenomenon lasted about a month.

The other general environment in which cash outperforms all asset classes is when the appetite to take risk suddenly declines (or the premium required to take risk precipitously rises). This phenomenon is most pronounced during crisis periods. During these unique environments panic ensues and there is an overwhelming demand for liquidity and cash. No one really cares about the long run during these climates and most are willing to sell their assets for nearly any price just to have the peace of mind of holding risk-free cash. The general investor mentality abruptly swings from focusing on return on capital to return of capital. The mechanical consequence of widespread fear is rising risk premiums. Oftentimes the cash rate actually falls, which on its own would be a positive influence on asset prices. However, rapidly rising risk premiums—or the general level of excess returns demanded by investors for taking risk—abruptly and materially spikes because of the high demand for cash.

The years 2008–2009 and 1929–1932 are two vivid instances in U.S. history of this type of atmosphere. Both of these examples represented deleveraging periods, as described in the first chapter. During these times cash is king because the economic machine has literally frozen. This has only happened twice in U.S. history since 1927, but other countries have also experienced similar incidents in their history. Using the simple formula above (cash return + excess return = return of asset class), if the general risk premium for all asset classes spikes, then the price of all asset classes simultaneously drops. This occurs because the appetite for risk precipitously declines and much lower asset prices are required to attract new buyers.

Both of these environments—rapid, unexpected cash rate increases and deleveragings—are rare events. They simply do not occur very often. The infrequency of these events should not come as a surprise. The deleveraging process, as explained in the first chapter, may occur only once or twice in a century because of the self-reinforcing nature of the leveraging cycle. The mechanics of how the cycle runs results in protracted up-and-down cycles lasting decades. Similarly, rapid unexpected increases in interest rates are rare since the Fed normally manages the typical economic boom-bust cycle. It does this by raising rates when the economy is too strong, when inflation pressures start to build, or both—and it lowers rates when the economy is too weak, inflation is too low, or both conditions are present. These normal cycles cover most of history. For interest rate hikes to shock the market as described in this section, not only would the Fed need to make a big move, but that shift would also have to be completely unforeseen. Either occurrence is atypical on its own, based on history and rational Fed behavior. For both to occur concurrently is even more unusual.

More significant than the fact that these two events are rare, is the reality that they are short-lived even when they do occur. This characteristic should also be anticipated. When asset classes underperform cash, the overall economic environment is usually very weak. A deleveraging normally results in a depression, which is one of the worst economic outcomes. Rates rising rapidly beyond expectations produce a tightening of conditions because of the increased cost to borrow, which often results in a weakening economic climate. Both of these outcomes cause the Fed to react if conditions deteriorate too far, as they often do. The economic downturns are both exacerbated by and result from the decline in asset class values. If cash is outperforming asset classes, then the loss in value has a negative wealth effect because investors suddenly have less money (from declining asset values) than they had previously. Feeling less wealthy causes a natural response of less spending, which further feeds into weakening economic growth and inflation. The declining conditions result in a policy response of greater stimulus, which in turn pushes asset prices higher once again. In other words, the reason these negative environments are transitory is because the Fed won't let them last too long. Note that a period of deleveraging may last longer than a period of rapidly rising interest rates because the quickly deteriorating and drastic conditions associated with the former are generally more difficult to swiftly overcome with stimulus.

Another factor to consider is that immediately after these rare periods, asset classes typically tend to do very well because they are positioned to take advantage of higher risk premiums (due to recently falling prices) from that point forward. Thus, even if you are caught in the middle of the downturn, you have the option of waiting it out and taking advantage of higher prospective returns. This is another reason why the downturns are short lived and reversion to the mean is typical after big declines.

Now that we have covered theoretically when the balanced portfolio may underperform versus cash, we can now turn to the data. Is our understanding of the key drivers of asset class returns borne out in historical returns? We will analyze underperformance within two dimensions. First, we will address the frequency of declines in the past. Second, an examination of the severity of downturns will be scrutinized.

Major Drawdowns Have Infrequently Occurred

An extremely comprehensive analysis of historical drawdowns involves the use of rolling returns. A rolling period refers to a snapshot of returns that are based on all the available starting and ending points rather than a more arbitrary time frame. Since you can't assume when in history you would have bought an investment and sold it, a review of rolling returns considers all the possibilities. Rather than starting in January and ending in December of each year, you can analyze returns starting in January, February, March, and so on. By slicing the data as short as rolling 3-month returns since 1927, nearly every possible historical scenario will be captured. I will also compare rolling 1-, 3-, 5-, 7-, and 10-year historical returns to offer longer rolling measurement periods. This summary covers a total of 5,955 data points! There is no hiding when utilizing rolling returns in this manner. If the portfolio contains a flaw, it will certainly surface when magnified under this level of scrutiny.

Table 11.8 summarizes how often the balanced portfolio has underperformed cash since 1927 using various rolling time periods. The same results are listed for the traditional 60/40 allocation as a point of comparison. The final column provides the percentage of time that all four asset classes (stocks, Treasuries, TIPS, and commodities) simultaneously underperformed cash over the rolling period. This data is shown because one of the core concepts of the balanced portfolio is the notion that it is rare for all four asset classes to do poorly at the same time, because the environment that causes one to underachieve is usually the same that results in a positive environment for another.

You should draw three main conclusions from Table 11.8. First, for both the balanced portfolio and 60/40, you have historically been more likely to beat cash the longer your time frame. That is, if you randomly selected any three-month period in history, you would have a 36 percent chance of doing worse than cash with the balanced portfolio and a 36 percent chance with 60/40. If you instead checked over a five-year measurement period, the odds decrease meaningfully for both. This is why investing is a long-term venture. The longer your time horizon, the more likely you are to be successful. Importantly, in order to achieve this type of consistency, you must hold on for the duration of the ride. If you change course along the way, you negatively impact your odds. This is true because you are most likely to change your strategy after your portfolio has underperformed (and just before it is about to outperform) and not vice versa.

The second important point this data highlights is the fact that the balanced portfolio is much more consistent than 60/40 in outperforming cash over time. This result should not be a surprise because of the superior balance. The problem with the imbalance of the 60/40 allocation is that individual asset classes can underperform for very long stretches of time. If the success of the asset allocation is overly dependent on the outperformance of just one segment, then prolonged periods of underperformance should be expected. The data clearly reveals this reality.

Finally, Table 11.8 further helps explain why the balanced portfolio has beat cash much more consistently over the long run than 60/40. Even over short periods such as three months or one year, it is quite rare for all four asset classes to concurrently return less than cash. The logic should be straightforward at this point. Each asset class is expected to beat cash over the long run. Additionally, each asset class contains different biases to various economic environments. As economic conditions shift, these changes impact asset classes differently causing them to perform better or worse than average. Since economic shifts occur at the same time for all asset classes, what causes one to underperform simultaneously drives another to outperform. The reason that 95 percent of rolling three-month periods since 1927 (this represents a lot of data points!) have resulted in at least one of the four asset classes doing better than cash is because of this core understanding of what fundamentally drives asset class returns.

^* Asset classes are stocks, long-term Treasuries, long-term TIPS, and commodities.

If you were to dive deeper into the data, you would discover that the only periods during which all four asset classes underperformed cash on a rolling three-year basis were during the Great Depression (1929–1932) and during a period of rapidly rising cash rates (1981–1982). Over 10-year rolling periods (still a lot of data points), not once in history have all four underperformed cash. Never! If you are building a portfolio using these asset classes and efficiently weighting them in the process, then you stand to have a high probability of beating cash over time. Of course, all of this assumes you do not sell at a low point and you are able to emotionally survive the trough (which admittedly is far easier said than done).

As you would expect with the balanced portfolio, the rolling periods during which it underperformed cash generally involved times in which cash rates suddenly rose, or during deleveraging periods. The early 1970s and early 1980s are examples of times in history when cash rates unexpected increased materially. The early 1930s and the years 2008–2009 were periods in which we lived through a deleveraging process. These events explain all the underperformance for every rolling 5-, 7-, and 10-year period since 1927! In other words, the only times in history that the balanced portfolio underperformed cash for more than a few years was during these unique, short-lived events.

Drawdowns Have Not Been Severe or Long Lasting

I have established that the balanced portfolio historically has not underperformed cash very often, particularly over longer time horizons. That observation only provides a part of the picture. Frequency is good to know, but what about degree? It is one thing to rarely fall behind cash, but what if those periods resulted in devastating losses that took years to recover?

A good way to assess the downside risks of a strategy is to calculate the returns had you bought the investment at the worst possible time and sold it at the absolute low. The measurement of the peak-to-trough drawdown provides a helpful glimpse of the maximum a strategy has underperformed cash historically. By comparing the results from the balanced portfolio with 60/40, you can gain additional useful context into the differences between the two approaches. Also relevant in this analysis is the recovery time. If you lose 20 percent but make it right back in three months, then that is far more acceptable than if it takes you five years to come back. Table 11.9 lists the peak-to-trough periods for both allocations as well as the number of months it took to get back to the prior peak. Every excess return drawdown of at least 10 percent since 1927 is listed.

List represents all periods in which each portfolio experienced at least a 10 percent drawdown from peak to trough.

Notice the time periods for the biggest peak-to-trough drawdowns for the balanced portfolio. From 1930 to 1932, we felt the most severe consequences of the Great Depression. The years 1980–1982 were impacted by rapidly rising cash rates that surprised almost everyone. Finally, the more recent deleveraging episode of 2008–2009 caused severe losses across most markets. You would expect the balanced portfolio to underperform cash during these rare, short-lived events.

When compared to 60/40, the balanced portfolio's performance is clearly superior during these rare periods. It has only experienced three peak-to-trough drawdowns greater than 15 percent in history (versus eight for 60/40), and the recovery time has been much more palatable. Over any rolling time period it has had a total of only eight periods in history in which it underperformed cash by more than 10 percent. In contrast, 60/40 failed on nineteen separate occasions, with far longer recovery times to get back to the old peak.

Note also that the returns shown in Table 11.9 are excess returns above cash (or below cash in this case, since they are negative). The total return you would have earned would have been the figures shown plus the return of cash. Thus, in reality, even though you should think of returns relative to cash, your actual experience would have been better than that represented here. By way of example, cash earned 5.6 percent between 1929 and 1932, 30 percent between 1980 and 1982, and 0.8 percent between 2008 and 2009. Of these periods, the years from 1980 to 1982 predictably benefited the most from the high yield of cash because it was an environment in which cash outperformed because the yield rose so quickly. The other two climates were deleveraging phases during which cash yields actually fell to zero to stimulate the economy. Note also that during these devastating deleveraging periods the balanced portfolio outperformed 60/40 because of the greater balance in the portfolio and the lower allocation to equities (which get nearly wiped out during deleveragings).

Finally, Table 11.10 summarizes the returns of the two strategies during the most notable bear markets in history. The balanced portfolio outperformed 60/40 in every single major bear market!

All returns are cumulative excess returns.

2011 downturn: 4/30/11–9/30/11 (S&P 500 declined 16.4% versus cash); 2008 credit crisis: 10/31/07–2/28/09 (S&P −51.9%); 2000–2002 bear market: 3/31/00–2/28/03 (S&P −49.4%); October 1987 crash: 8/31/87–11/30/87 (S&P −30.7%); 1973–1974 bear market: 12/31/72–12/31/74 (S&P −48.2%); 1939–1941 downturn: 12/31/38–4/30/42 (S&P −29.0%); 1937 recession: 2/28/1937–3/31/1938 (S&P −50.0%); 1929–1932 depression: 8/31/29–6/30/32 (S&P −84.4%).

Note that the time periods used in Table 11.10 are not based on calendar years, but on a measurement clock that starts at the peak and stops at the trough of each bear market (e.g., the 2008 credit crisis includes returns from October 2007 to March 2009). Therefore, this data provides good insight into performance during the entire span of a bear market.