- Portfolio diversification applies within any asset class—it is beneficial to invest in many securities, rather than just one. With enough stocks of different companies, the impact of a negative development at any one company on the overall portfolio is small and there are such offsetting effects (both positive and negative developments) that single company and industry exposures are greatly diminished.
- To set goal posts, 30 to 100 stocks of companies in a mix of industries results in a well-diversified portfolio.
- Beware fixating on “deviation” of portfolio returns from benchmark returns as the measure of diversification.
- Asset allocation applies across asset classes—it is beneficial to invest in several asset classes that are impacted differently by general market risks, particularly overall economic conditions.
- Beware over-complication from investment companies pitching niche funds as “diversifiers” or “tactical,” as they encourage bad behaviors, like market timing, performance chasing and fixation on volatility at the cost of long-term return.
- The classic allocation among stocks and bonds continues to work best.
In investing, the idea of diversification comes into play in two ways. First, within any asset class, it is beneficial to invest in many securities, rather than just one—for example, stocks of many different companies within a stock portfolio. This is known as portfolio diversification. Second, it is often beneficial to invest in several asset classes—for example, a portfolio of stocks and a portfolio of bonds—rather than just one class. This is known as asset allocation.
Common Sense View
With respect to portfolio diversification, I simply subscribe to the old, common sense adage “don’t put all your eggs in one basket.” Every investment approach, even ones that rigorously adhere to the sound fundamental principles we discussed in the last section, will from time to time select stocks or bonds that do poorly. As always, the future is uncertain; conditions impacting a company can change for the worse, in ways that defy prediction. So, it is key to invest in many individual stocks or bonds. Given this collection of stocks or bonds, the overall portfolio return is not mortally wounded by a couple of companies that do poorly, especially as some will do better than expected. The portfolio does well over time based on the average return, on balance, across all the companies it’s invested in. As a useful analogy, think about a professional bowler going up against a novice. To give herself the best chance of winning, she’d agree to play a full frame. In a single roll, she may be off slightly and miss a pin or two, while the novice may get lucky and roll a strike. But over a full frame, it is very likely she will be more successful than the novice. Or, to draw on another sports analogy, it is why each team plays 162 games and not, say, 24 over a Major League Baseball season in order to determine which teams are deserving to go to the post-season.
In academic finance, portfolio diversification has been approached from a slightly different perspective. Academic studies have focused on measuring by how much returns of specific portfolios differ from the overall market return (as represented by various indices) over the course of a set period, say a year or a month. This allows for a quantifiable test. Specifically, statistical studies have looked at how the returns of a bunch of portfolios of randomly-selected stocks differ on average (or deviate, in the parlance) from the overall market return. These studies have consistently shown that by adding more randomly-selected stocks to portfolios results in the portfolios’ returns clustering more closely around the overall market return, i.e. for deviation to go down. Going from a single stock to 10 stocks reduces the deviation most dramatically. From there, adding additional stocks to the portfolio continues to reduce deviation, but at an increasingly less impactful rate. To take it to the extreme, if we use the S&P 500 as the overall market return and if we can only draw stocks at random from the S&P 500, once we have added enough stocks to get to 500, the deviation of the portfolio return from the S&P 500 return will be zero (assuming our portfolio is market-cap weighted, like the S&P 500). Obviously, as they are now one and the same.
Why It Works
Let’s think about the intuition behind why these studies all show that adding stocks reduces the deviation of portfolio returns from the market return. Stocks are randomly selected in these studies, so mistakes and triumphs in stock selection play no part. But the common sense logic of “not putting all your eggs in one basket” still applies. While some of the stocks do poorly, other stocks do well. Again, in an uncertain (i.e. risky) future, conditions change for companies, but many of these conditions are unique to an individual company (for example, a product recall) or to a particular industry (for example, a regulatory change). Some companies’ individual circumstances will change for the worse, some for the better, and this will be the case for different industries as well. As stocks are added, the impact of such changes—negative or positive—on any one stock or industry is diluted within the overall portfolio. And as stocks are added, the chance for negative changes to be offset by positive changes grows. With enough stocks, there is a great washing out. The impact of any one change in circumstances (risk) is so diluted in the portfolio and there are such offsetting effects (negative and positive developments) that these single company and industry exposures (risks) are greatly diminished. Given how this works, it is important to note that when we add stocks, we need to add stocks of different types of businesses, including in different industries in particular.
To put a bow on our understanding of what we are achieving with diversification in the academic framework, let’s return to the extreme case where we invest in all 500 stocks of the S&P 500, with market cap weightings, and the S&P 500 is the measure of the market. As we pointed out, deviation to the market return is zero. We have not however eliminated every risk of every single company and industry. On the contrary, we now own exposure to every single one of them! But we have greatly diminished our exposure to any one of these risks to the point of insignificance (with respect to the impact any one risk can have on the overall portfolio return), and our exposures to them match that of the market index. Finally, it is important to note that even when we have utterly diminished these single company and industry exposures in this manner, risks remain. All stocks remain exposed to circumstances that affect companies broadly, most notably the health of the economy. These so-called market or systematic risks cannot be addressed or diminished by portfolio diversification.
How Many Stocks to Own
The great practical benefit of the academic perspective on diversification, in particular its mathematical treatment of the subject, is that it gives us some idea of how many stocks we actually need to own to achieve the benefits of portfolio diversification, i.e. to effectively diminish exposures to single stock and industry risks. And this is what the studies originally set out to do. The studies indicate that the returns of a market capitalization-weighted portfolio of 30 stocks will have a deviation to the market return that is over 90% less than that of a single stock portfolio. So, a 30-stock portfolio takes us far down the diversification road, and the addition of each stock thereafter turns the tires just slightly. For an equal-weighted portfolio (in which case an equal-weighted market index would apply), 100 stocks achieve the same 90%+ reduction in deviation of the 30 stock capitalization-weighted portfolio. This helps set the goal posts of how many stocks we should own. And, again, it is critical that these stocks represent different types of businesses, spread across many different industries.
A Note of Caution
I do advise caution however with respect to the academic treatment of diversification. In using deviation of a portfolio’s return to the market return to measure the impacts of diversification, there is a danger of becoming fixated on this deviation as the measure of diversification. This would be a mistake. The deviation measure relies on a benchmark return, the market return, which is in fact the return of some index. This runs the risk of making the index a paradigm of virtue, when this is not always the case. For example, in the last section on A Sensible Approach, we noted that in March 2000 the technology and telecom sectors together had grown to a 42% weighting in the S&P 500, making the index highly exposed to the circumstances of those industries (which turned out to be massive over-capacity). A portfolio that was more prudently balanced in its industry exposures would very likely have had returns that deviated from the market return as measured by the S&P 500, though it would in fact have been more well-diversified.
Here’s another way to think about this, tying it back to the most important principle in investing—not overpaying at the outset, as discussed in the last section, A Sensible Approach. In the section Understanding Risk under the “Time Frame and Mindset” tab, we noted that losing money in a particular investment, which is risk in its most basic form, stems from either a negative turn of events at a company or having overpaid for the investment in the first place. The academic treatment of diversification emphasizes the former, exposures to events at companies and industries. But, in randomly selecting stocks, it is also the case that some stocks will have been bought cheaply and some expensively—as per Mr. Graham and Mr. Dodd’s teachings that stock prices and intrinsic value rarely match. In a diversified portfolio of randomly selected stocks, these will offset, just as positive and negative changes in company and industry circumstances offset. But this is not the case if the overall market is expensive. The whole point of diversification is to reduce risk, but if we fixate only on deviation of our portfolio’s return from the market return, such that we mimic an overvalued market, we are subjecting ourselves to the basic risk of overpaying. In the case of an overvalued market, deviating from the market by limiting our investments to only those stocks that trade at low multiples would in fact be the prudent thing to do, like the thoughtful lemming that peels off from the pack scurrying toward the cliff.
To sum up on portfolio diversification:
- It is best to take a common sense approach to portfolio diversification. By not putting all our eggs in one basket, the impact of a single bad outcome is diluted and bad outcomes are offset by good ones.
- The academic treatment of diversification, which uses deviation of returns, help us set goal posts with respect to how many stocks we should own to be effectively diversified—at least 30 to 100, depending on how we weight the individual stock holdings in our portfolio. We need to make sure we have a good variety of businesses and industries.
- We need to be mindful, however, not to fixate on deviation of returns as the sole measure of diversification, as aping a poorly-diversified or over-valued benchmark, an index, increases risk which is contrary to the whole point of diversification.
Let’s move on to the second way in which the idea of diversification can help us, asset allocation. As noted above, even when we have perfectly diversified our holdings in one asset class, say stocks, risks remain that cannot be diminished further. Because all stocks remain exposed to circumstances that affect companies broadly, and this is known as market or systematic risk. The source of market risk, most importantly, is the broad macroeconomic environment in which all companies operate. If the economy is lousy, all companies face headwinds to their revenues and profits, and the stock market as a whole will likely be pressured. Fortunately, market risks can impact different types of asset classes—of which, the largest and most important are stocks, bonds and cash—in different ways.
A classic example of variance in market risk across asset classes is that between a stock portfolio and a government bond portfolio. In times of economic weakness (slow growth, unemployment), central banks typically lower short-term interest rates, which in turn coaxes interest rates down generally (at least, that’s how it’s supposed to work). When interest rates fall, bond prices go up. So, in periods of economic weakness, like recessions, while your stock portfolio may be taking a hit, the value of your bond portfolio may be climbing. As we saw in the section Taking Advantage of Time Frame under the “Time Frame and Mindset” tab, stocks and bonds have their own long-term return profiles, which we may expect over long horizons. What’s interesting is that the advances and pull-backs that characterize the journey toward these long-term returns often come at different times for different asset classes. Stocks pull back and bonds advance in recessionary periods, and then stocks advance in an economic recovery and bonds pull back when the central bank again raises rates. By owning both asset classes, the journey becomes a little smoother. In the section Recognizing Your Emotions, we noted that stock market investing can often be more roller-coaster than railroad, though the long-term return potential makes these white-knuckle periods worth it. With a bond portfolio alongside a stock portfolio, the ride for the overall portfolio is not quite so hair-raising.
So, the idea behind asset allocation is fairly simple. Unfortunately, the investment services industry has made it very complicated and unnecessarily so. This may be the area where they’ve done the single greatest dis-service to the average investor. As investment companies have tried to differentiate themselves and appeal to every type of investor, some with highly specialized needs (real or imagined, and often not truly investing needs), they have spawned an array of asset classes—stock sectors, real estate, commodities, currencies, derivatives, and a whole hodge-podge of stuff called “alternatives.” Some of these are not new asset classes, but they have been made newly available to investors via tradeable vehicles. And investment strategies that provide different types of exposures—“hedged,” “leveraged,” “inverse”—to these and other traditional asset classes have likewise proliferated. Strategies offered by a single fund company can number in the hundreds! The result is a bewildering Chinese menu of choices (but without pictures).
To demonstrate their “value,” investment companies then solve this complexity—of their own making—with “managed” and “portfolio” products and services that cobble together these asset classes, avowedly to meet your particular profile. These can entail allocations to as many as 20 asset classes!
Most of these new asset classes and related strategies are marketed as “diversifiers” or as “tactical.” Both are problematic.
Diversification for Diversification’s Sake: Confusing Volatility with Risk
As “diversifiers,” many of these asset classes and strategies are low return and in fact not investments. Let’s take “alternatives” as an example, as it is one of the fastest growing areas, and focus on “long/short equity” strategies, among the most common. These look to replicate hedge funds that “long” stocks (which simply means buy them, just like a traditional stock fund) but also “short” stocks (borrow and sell them first and buy them later, to bet on a decline in their price). Shorting by its nature is not investing, as it does not rest on future cash flow streams, but rather is a trading strategy. The idea is to lessen volatility, as during market declines the “shorts” will go up as the “longs” go down on balance. The same is true however during market advances—the “shorts” will go down as the “longs” go up. The fixation on “downside protection” is understandable, given two traumatic bear markets in less than a decade. These products speak to clients’ fears. But, as I’ve shown, stocks go up (and pay dividends) over the long run (providing, in all likelihood, the highest return of any asset class over 10+ year stretches). So, the long-term ramification is simply to dampen returns, as losses on the shorts will eat into the returns of the longs. This is exacerbated by the fact that these funds carry high fees.
So, in the name of soothing one type of risk (volatility), these funds lock in another risk (never reaching your goal, because out of fear of volatility you’ve invested in assets that don’t provide a sufficient return). This is a bad trade-off, because as I demonstrated in the Taking Advantage of Time Frame section under the “Time Frame and Mindset” tab, volatility becomes largely irrelevant with a suitable time frame. It is better to educate and mentally prepare yourself for volatility, to put it in proper perspective, and put your hard-earned savings into investments that will work the hardest for you, i.e. deliver the highest possible returns within the parameters of your time frame.
Tactical Investing is a Fancy Word for Market Timing
In marketing these asset classes and strategies as “tactical” tools, investment companies are actively encouraging you to market time and performance chase, which are counter-productive behaviors that result in returns falling short of what the market makes available over time. We discussed and demonstrated this at length in The Dangers of Market Timing and Chasing Performance sections. To understand how this is so, let’s take a look at how fund companies pitch these strategies as “tactical” tools.
I noted that many fund companies have hundreds of funds, and they pile up because most of them are very narrowly-defined—for example, a particular commodity, a particular type of bond or a particular stock sector (e.g. industrials or financials). The idea pitched by the fund companies is to “rotate” into these assets right before they do well. You could even rotate into an “inverse” strategy right before it does poorly. For example, as I noted in the Chasing Performance section, industrial stocks tend to decline when the economy softens and do well as the economy picks up.
The problem is that no one can tell you when stocks in this sector are about to do well or do poorly, because no one can correctly forecast near-term changes in the economy with any consistency. Business cycles follow general patterns, but every business cycle is different. Some recessions last only months, others last years; variations in the lengths of recoveries and expansions are even more dramatic. Within an expansion, there are softer and stronger periods, with each softer period potentially a recession. Moreover, for a “rotation” to be successful, you need to be early. Other stock market participants are trying to do the same, impacting stock prices as a result. When it is clear that the economy is in recession, for example, industrial stocks likely have already fallen, and it may in fact be a better time to position for a recovery. So, you have to be better on average than all these other investors in your forecasts and remain a step ahead.
If you try to play this game, you will most likely lose. It will result in you zigging when you should zag and zagging when you should zig, bleeding returns to others, as we have illustrated repeatedly.
Amazingly, all these investment companies preach the virtues of long-term investing—putting in place a sensible plan and sticking to it. But they can’t help themselves. These “tactical” funds and managed products speak to our anxieties, which are fed by the frenetic pace of scary news and the day’s market developments that we all ingest from the financial networks each day. And, there is simply money to be made in it, for them.
Classic Approach Still Works Best
Interestingly, as illustrated in the graph below, during the financial crisis of 2007-2009, the prices of many asset classes, including all of the newly-introduced ones, moved together—they went down!
This has called into question the basic premise of asset allocation, which is to invest in different asset classes that move differently at different times—smoothing the ride. Many investment companies have acknowledged this. But their “rethinking of asset allocation” has not been to re-examine their pitch of a proliferation of asset classes, but to double-down: “adopt a more sophisticated approach,” “access non-traditional assets,” “fine tune.” This line they’ve taken stands against all reason. And, the biggest danger in all of this is the increasing complexity. You don’t understand what’s in your portfolio or why it’s there. When this is the case and when the market punches you in the mouth, you are going to be more apt to cut and run. This lack of steadfastness—of constancy—undermines the likelihood of your plan’s success.
The financial crisis was an extreme time, so it makes for an extreme example. But, the fact is, it is during such extreme times, rather than more “regular” interludes of volatility, that the benefits of asset allocation are most helpful. So, the graph above suggests a common sense approach. Stocks of all stripes got hammered (whether large cap, small cap or international indices), but along with them so did high yield bonds, REITs (real estate investment trusts) and commodities. Interestingly, and seemingly overlooked by large investment companies, the classic approach to asset allocation worked. Most importantly, investment grade bonds held up. As the economy soured, the Federal Reserve cut short-term rates and all interest rates fell. As a result, Treasury bonds soared. Municipal bonds likewise rose, though this was somewhat offset by concerns about the worsening finances of certain states and municipalities. Investment grade corporate bonds declined, but not by a sickening amount. And, of course, cash (3 month T-bills) is cash. Gold also showed “safe haven” and “store of value” characteristics, increasing in price over the period.
So, here’s the approach to asset allocation. Stocks for the highest returns over the long term, if you have sufficient time frame—about five years for a very modest allocation and ideally ten or more years for a significant one. Government bonds and investment grade corporate bonds for respectable returns and downside protection—ballast during periods of nasty weather. Think of it like a Viking ship—with a good breeze, the sails (stocks) carry you where you need to go; down in the hull are burly oarsmen (investment grade bonds) helping you steadily along and providing ballast in stormy seas (and perhaps a small chest of gold in the hold). That’s it.
Under the next tab at top, called Plan, I provide some thoughts and resources to help you to come up with an investment plan—for retirement, most importantly—based on all the principles we have explored. This includes determining specific asset allocations consistent with the time frame for each of your goals and your overall investor temperament. And you can see how you might apply these common sense ideas on diversification in the asset allocations of Model Portfolios.