The Devolution of Diversification
We are only some 40 years removed from an era when the typical investment account had 12 or fewer individual holdings, and less than 20 years removed from a time when respected stock funds might hold 20-30 stocks and be considered "fully diversified". Now we find that the typical active equity portfolio or fund holds between 50-100 individuals stocks and that there are generally three or more such active equity managers for each institutional or high net worth account, all adding up to hundreds of underlying holdings, all in the name of diversification and, ultimately, all in the name of "risk management".
In the wake of this 20x or greater expansion, much of it occurring over the course of our collective investment careers, it is worthwhile to ask a few questions:
- When does diversification cross over to over-diversification?
- Why did the typical portfolio evolve from a handful of active holdings to hundreds?
- What is the solution if modern portfolio practice has devolved into over-diversification?
A Federal Reserve Study from 1967 showed that the average number of stocks from a typical investor's stock portfolio was 3.41. Another study from 1971 revealed that only 11% of their sample of investors had more than 10 stocks, while a later research paper from 1976 focusing on the major brokerage houses of the day, where surely the most sophisticated techniques of the time were being employed, found that the average number of stocks in a portfolio of that era ranged from 9 to 12.
We want to make clear that we are advocates of intelligent and efficient diversification, which we define as holding the stocks of companies across a sufficient spectrum of industry and country exposure so as to render the failure of any single enterprise a modest setback, but not so many as to make negligible the potential positive return contribution from any single holding.
We also believe that multiple instances over the past 25 years of stock market history demonstrate the limitations of diversification as defined by Harry Markowitz and Bill Sharpe's pioneering financial theories of 1950's and 1960's, namely the mean-variance or Modern Portfolio Theory. Basing diversification theory on the concept of rational and information-filled market pricing is, in retrospect, naive and overly simplistic. While historical price correlations are an interesting metric by which to judge the risk of one asset against another, indeed, a very useful and valid metric as part of a larger group of risk criteria, they have proven unsatisfactory as a practical solution to a very real world problem: preserving and growing capital.
But Modern Portfolio Theory is not solely to blame for the steep rise in overall active holdings, since many papers have demonstrated that even using the math behind Modern Portfolio Theory, almost all the benefits gained from correlation reduction via the addition of more holdings in a portfolio taper off dramatically somewhere in the 20 to 30 holding range. For some reason these findings have been largely ignored over the past 3 decades while investors have added more and more holdings into their portfolios in the mistaken belief that this is making their holdings more diversified and, therefore, safer.
The roles of significantly reduced trading costs, self-directed, web-based trading and portfolio management systems, and, lastly, the consolidation of the active equity management industry over the past few decades should also be pointed out as key contributors to over-diversification. The incremental cost of adding an additional security prior to the early 1980's was much greater than today, both in terms of dollars and administrative effort. Prior to the deregulation of the NYSE and other exchanges in 1975, commissions on trades ran upwards of 5% and bid-ask spreads were considerably wider. Thus, the decision to buy or sell a security was a much more expensive proposition than today.
Those who date their financial industry experience back to the mid-1980's or before will recall a time devoid personal computers, portfolio software, and the internet, when significant manpower was needed to handle the clerical tasks of entering trades onto paper forms, tracking those trades to settlement, as well as to file and retrieve that information. The costs in time and effort to add an incremental holding were very real. And this was just the beginning of the monitoring process, made all the more complicated by corporate actions, dividends, stock splits, etc. It should not surprise us that the advent of largely automated and significantly cheaper stock market trading, settlement, and monitoring, both for institutions and retail accounts, made it easier to multiply the average number of active equity holdings.
As also noted, consolidation within the active equity management industry has also played a role in the increase in holdings in the typical portfolio. Beginning in the mid-1970's and continuing on through the Bull Market of the 1980's and 1990's, the "roll up" of independent asset management shops has increased the scale of assets under management at these shops. As of 2013, PriceWaterhouse estimated that the largest asset managers oversee about 37% of total global AUM, with that percentage expected to rise to 47% in 2020. The greater the amount of assets managed by a particular active manager, the more stock which must be placed in any particular trade. Other than capping AUM or turning away money, the only alternative for these managers is to increase the number of holdings in a portfolio.
But even casting all those factors and explanations aside, and even if price correlation analysis was able to capture and diffuse company-specific risk, another factor has been left unconsidered in the discussion regarding over-diversification: the underlying stocks themselves have changed fairly dramatically in terms of their own business risk diversification over the past 3 to 5 decades. Most mid- and large-cap companies, that is to say, those with market caps of $2 billion or greater, have become more diversified in terms of their geographic spread, vertical and horizontal integration, and lastly their ability to hedge and/or insure their own enterprise risk. The growth in international trade law and practice, as well as the layering of global trade treaties means that companies may outsource and license across borders with confidence of fair recourse, a further expansion of today's real world risk diversification not prevalent in the 1970's and earlier.
Compare a mid- or large-cap company from today to those of 30 or 40 years ago: in 1970, the vast majority of that company's revenue would have come from its home market, its product range or service offering would have been notably narrower, and, finally, the ability to use financial instruments or insurance to reduce its risk of loss from currency movements, natural disaster, and other sources of business disruption would have been significantly less developed than their counterparts of today.
As anecdotal evidence, look at a company such as Fresh Del Monte (NYSE:FDP), the canned and fresh fruit company, which in 1961 only had 10% of its revenues sourced from non-U.S. markets, with the majority of that 10% being accounted for by Canada, combined with the fact that 90%+ of its product sales were in pineapples and bananas, the majority of which were sourced from Hawaii and the Philippines. Today, nearly half of all Fresh Del Monte (the most relevant successor company to the original fruit company) revenue is from 100 non-U.S. markets, the product range has expanded to the extent that pineapples and bananas only account for 60% of sales, and produce sourcing now involves more than 20 countries, located in both the southern and northern hemispheres and four continents: North and South America, Asia, and Africa.
Our final observation on the over-diversification of active equity portfolios is the simple fact that investors who believe that greater safety is to be found in greater numbers of holdings now have hundreds of inexpensive index and exchange traded funds from which to choose, each holding dozens to hundreds of stocks. Piling active equity managers upon one another is an expensive, complex, and time-consuming proposition, both in terms of selecting and monitoring them, with the end result being returns that are, in aggregate, below benchmark returns after management fees, trading costs, and, often, consulting fees are deducted. Great investor knowledge of a smaller set of active portfolio holdings may result in greater confidence that company-specific risk has been spread across a broad enough group of enterprises so as to withstand the large, but temporary, shocks which seem to occur once every ten years or so. Great confidence in the truly diversified nature of portfolio holdings may result in fewer investors exiting positions during periods such as 2008.
Having identified the problem, over-diversification, what is the solution?
The first step would be to consider whether over-diversification may be equally outside the realm of meeting one's fiduciary duty as under-diversification. Although the impact of under-diversification is generally more spectacular and notable than the more silent effects of redundancy, lack of efficacy, and additional fees associated with over-diversification, both deprive the investor of future income and gains. On the other hand, proving the costs of over-diversification requires proving a hypothetical: what would returns have been absent the additional holdings?
The next step would be to admit that many investors are simply not suited for actively managed equity portfolios, by virtue of either low risk and/or volatility tolerance, or by virtue of lack of resources and time to find a good match in the active equity manager community.
The third step would be to adopt an updated and more nuanced understanding of what it means to be diversified in the 21st Century and take advantage of the tremendous growth in readily-available fundamental operating information which gives a truer picture of enterprise and group risk by creating a more multi-dimensional model of the problem at hand. Price correlations are mathematically appealing, but in isolation they have failed time and time again to create the safe havens their proponents espouse when financial markets deflate. That said, price correlations are useful as part of a larger analysis of a stock versus its peers or the opportunity set, but only by employing it in combination with real world diversifiers such as the geographic spread of customers, suppliers, currencies, and raw materials, along with horizontal and vertical integration, liability and loss insurance, and the ability to hedge financial risk.
True diversification, that which gives an investor effective peace of mind at times of financial stress and therefore gives them confidence in their holdings, is achieved by increasing the variety of data and the types of observations through which investors identify the meaningful risks of a single enterprise.
- Lease, R.C.; W. Lewellen; and G. Schlarbaum. "Market Segmentation: Evidence on the Individual Investor." Financial Analysts Journal, 32 (September 1976), 53-60.
- Blume, M.E.; J. Crockett; and I. Friend. "Stock Ownership in the United States: Characteristics and Trends." Survey of Current Business, 54 (November 1974), 16-40.
- Evans, J.L., and S.H. Archer. "Diversification and the Reduction of Dispersion: An Empirical Analysis." Journal of Finance, 23 (Dec. 1968), 761-767.
- Francis, J.C. Investments: Analysis and Management, 4th Ed. New York: McGraw-Hill (1986).
- Jacob, N.L. "A Limited-Diversification Portfolio Selection Model for the Small Investor." Journal of Finance, 29 (June 1974) 837-857.
- Statman, Meir. "How Many Stocks Make a Diversified Portfolio?" Journal of Financial and Quantitative Analysis, 22 (September 1987), 353-363.
- "World's Largest Asset Managers on Quest for $100 Trillion in Assets", 24/7 Wall Street website, Charley Blaine, February 18, 2014: site link here.
- 1961 Annual Report, California Packing Corporation (later Fresh Del Monte).
Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.