Following on the second issue of our four-issue cycle of celebrations for the 20th anniversary of The Journal of Wealth Management, here we will focus on several subjects related to alternative assets.
If one goes back far enough, asset allocation for individuals was an exercise that focused solely on cash, bonds, and equities. More complex asset classes and strategies were broadly ignored, if not frowned upon. In fairness, one should observe that this was not strictly an individual asset allocation issue: a similar situation prevailed in the institutional world. Interestingly, the expansion of the investment universe to include non-traditional assets and strategies did not strictly start in in the institutional arena alone, with expansion into the individual world later on. In fact, a few high-net-worth individuals were convinced of the usefulness of these strategies at the same time as, if not in advance of, that awareness on the institutional side. Thus, there were pioneers on both sides of the investment management industry when it came to evaluating and incorporating alternative assets into their portfolios.
Generically, three issues had to be addressed for the investment universe to embrace alternative investments: 1) there had to be agreement on the difference between an asset class and a strategy, 2) there was a need to develop an appropriate benchmarking process, and closely related to that, 3) it was necessary to establish appropriate tools and metrics for the measurement and assessment of performance, to develop an understanding of the different statistical properties of the distributions of return from these strategies, and to gain an appreciation of the impact this would have on asset allocation processes.
Even though some individual managers were making use of alternative strategies in the early days, for quite a while that emerging part of the investment universe was known as a new “asset class.” Thus, it was not uncommon for an asset allocation to group all of the so-called “hedge funds”—the main segment in the alternative space—under a single category, alongside cash, bonds, and equities. I could probably even say that this is still true today, in some instances. Yet, it should have been obvious, and gradually did become so, that hedge funds were not a new asset class unto themselves; the common feature among them was actually a fee structure. Aside from that, many different strategies were erroneously grouped under a common header, reflecting a trend that was leaking into the investment space from the broader world: the use of superficial labels to define or classify complex phenomena. Ostensibly, it is much easier to focus on a single catch phrase than to take the time and effort to investigate the actual specific criteria that would provide a much more appropriate classification.
Eventually, we arrived at the conclusion that there were at least four fundamental categories of strategies that covered the so-called hedge fund universe:
(1)Market-neutral strategies: These are strategies that seek to take advantage of mispricing between pairs of investments without taking much, if any, market risk. They can be viewed as relative-value strategies geared to pricing differences within a single or at least closely related factor risk(s). This generally covers arbitrage strategies, whether the inefficiency relates to an event (such as merger arbitrage), different structures (such as convertible or fixed income arbitrage), or a combination (such as multi-strategy).
(2)Strategies with residual fixed income risk: These strategies seek opportunities across the full fixed income universe. They are close parents of traditional long-only fixed income, where managers can roam across the various credit markets, looking for undervalued securities in the government, corporate, and high yield universes, as well as across the full duration spectrum or geographical borders, whether they accept the currency risk or not. These strategies migrate into the “alternative” universe when managers can take both long and short positions. These strategies, therefore, simply expand the universe of possible trades by allowing them to express views with regard to both cheapness—as do their traditional brethren—and expensiveness.
(3)Strategies with residual equity risk: These strategies are conceptually equivalent to the previous group and one finds managers who expand the opportunities found in equity markets to include both long and short positions. The group comprises a gradation of strategies in terms of residual equity exposure and industry or stock concentration. The thought process starts with strategies classified under the first category above: pair trades. Here, managers are strictly looking to take advantage of mispricing between two stocks within the same market and the same industry, while having the same quantitative exposure to each. This justifies the view that the strategy must be “market” or “factor” neutral. Strategies migrate into this third category as managers gradually take on risk other than stock-specific risk. They may own different exposures to the two otherwise similar securities, creating a net market exposure, or they may retain the same industry exposure, but vary the market or currency in which the securities are traded (think of a Ford/GM trade becoming Ford/Toyota or Ford/Dongfeng trade, the latter being a Chinese automobile company traded in both Hong Kong and Shanghai). Eventually, the portfolio really comprises two sub-portfolios with unequal market exposures; one is focused on cheaply valued stocks and the other on expensive stocks that are shorted, irrespective of their industry or geographic exposures. A final iteration would be highly concentrated portfolios, such as those owned by “activist investors.”
(4)Trading strategies: Although initially centered on strategies that were agnostic on the markets in which they trade, the group has expanded conceptually to include strategies that are more focused on a limited goal, at times quite sharply such as on the commodity or even energy sector. As a matter of simplification, one can observe that these strategies generally involve greater use of derivative securities—principally, but not solely, futures—and also may exploit certain trading rules. Thus, one can subdivide the group into discretionary and systematic strategies. The term discretionary denotes strategies where portfolio managers can accept or reject trade signals, while the latter term, systematic, involves portfolios where trading rules are generally binding on the managers.
It stands to reason that such a more detailed classification is considerably more complex and requires quite a bit of additional work on the part of the balanced portfolio manager. Yet, it is clear that grouping these fours types of strategies under a single header is about the same as claiming that, with one’s feet in the oven and one’s head in the freezer, one is, on average, comfortable.
Interestingly, once the detailed work of classifying the strategies according to the risks that the manager is willing to underwrite has been completed, it is much easier to benchmark the returns achieved by “similar” managers and measure their relative performance. There have long been peer universes. This allowed for comparisons to be made across a group of managers that fell under a single category, even though one had to accept that two imperfections tended to creep into such an analysis. First, most indexes or series grouped managers into a single category, whether they were “opened” or “closed” (the former comprising managers who still accepted new monies and the latter who were not accepting additional investments). Second, the indexes were computed based on self-reported returns by each manager, which introduced the risks of both “back-filling” (when a manager selectively reported past performance) and “survivorship bias” (where managers who either disappeared or stopped reporting were not included in subsequent series).
More complex, but considerably more useful, was the need to develop market benchmarks for these strategies so that a strategy or group of strategies could be analyzed alongside traditional, long-only managers. This involved the recognition that each of the four groups described above really had two major components to its return stream: a base that reflected the appropriately scaled “market risk” they were taking, and their expected value-added—or alpha—together with the corresponding risk factors. A consensus emerged:
(1)Market-neutral strategies really ought to be viewed as being “cash-based” because traditional finance teaches us that taking correspondingly-sized long and short positions in similar assets is equivalent to creating a synthetic cash exposure. The strategies should thus be benchmarked based on cash plus some target alpha, and their appropriate volatilities.
(2)Residual fixed income strategies ought to be viewed as being based on the type and extent of fixed income risk exposure taken by the manager. This requires ensuring that one understands the actual investment process in order to identify the “sandbox” within which the manager is playing (strictly single or multi-sector, single or multi-credit range, constrained or unconstrained duration, and so on). This also puts the onus on the analyst to continue monitoring the trades carried out in a manager’s portfolio to make sure that the initial classification remains appropriate. That “base exposure” should then be appropriately scaled, and a target value-added is incorporated to arrive at a benchmark.
(3)Residual equity risk strategies require a parallel process to the one described above, where the key variables are industry, geography, and other factor risks. These are similarly appropriately scaled, and a target value-added is incorporated to arrive at a benchmark.
(4)Trading strategies can also be viewed as cash-based, as managers normally are structurally agnostic as to whether they should strategically be long or short. Therefore, one can assume that the base asset is cash, to which some appropriate target alpha is added to create a benchmark.
Benchmarking these strategies solved the challenge associated with understanding what structural return and return volatility should be incorporated into one’s holistic capital market forecasting effort. However, the final challenge, which has not been fully addressed so far, revolves around the portfolio optimization process through which a classic policy asset allocation is derived. Davies et al.1 focused on that issue and discussed a possible solution.
The main problem revolves around the fact that the typical alternative strategy has returns that cannot be viewed as distributed normally—or log-normally. Generally, return streams tend to be negatively skewed and to have excess kurtosis,2 both features being considered negative attributes. Thus, it is no surprise that these strategies typically possess better return per unit of risk attributes, as these can be viewed—by someone who believes in some degree of market efficiency—as the necessary offset to the unattractive skew and kurtosis features.
It naturally follows that a classic unconstrained mean–variance optimization process will tend to overweight exposure to most alternative strategies, as it “knows” the expected better return per unit of risk (which should more accurately be described as return per unit of volatility) but does not know about the unattractive skew or excess kurtosis. Davies et al. proposed a mean/variance/skew/kurtosis model to deal with the problem. Yet, although the model does the trick, it is considerably “heavier” in terms of data requirements and more complex than the currently available alternatives. In the current markets, many practitioners reluctantly deal with the problem by adding constraints to the standard mean–variance model.
This quick review would not be complete without a mention of another important set of strategies within the so-called alternative universe: illiquid strategies. A similar discussion would be needed to illustrate the fact that the term “private equity,” which is often used to cover the full universe, is inadequate, as it fails to reflect the complex and rich texture of this space. Suffice it to say, one should, at a minimum, distinguish between illiquid fixed income, illiquid equities, and illiquid real assets, if one does not want to delve into the specific details. A more accurate, but still incomplete list, would point to middle market loans (sponsored or not), mezzanine finance, buyouts, growth equity, venture capital, angel investing, real estate, timber, energy, and other illiquid real assets. Performance benchmarking becomes even more difficult given the fact that one must incorporate vintage year data into the analysis.
Our next and final letter in this anniversary series will focus on the need to view wealth management as the integration of multiple disciplines and will explain the role of family education in that process.
The Winter issue of The Journal of Wealth Management features articles grouped into three categories, reflecting strategic portfolio issues, mutual fund issues, and general topics as viewed from an Asian perspective.
The first two articles cover strategic portfolios issues. The first article, by Nathan Sosner, Rodney Sullivan, and Liliana Urrutia, addresses the after-tax performance measurement and assessment by proposing an effective and workable after-tax performance report aimed at enhancing wealth preservation and accumulation for taxable investors. The second article, by Timothy Peterson, looks at the role of gold in portfolio allocation and shows that a dynamic investment portfolio asset allocation based on secular market cycles outperforms a buy-and-hold portfolio of equities and outperforms a buy-and-hold portfolio of gold over long periods.
The next three articles look at different issues in the mutual fund world. The first, by frequent contributor John Haslem, uses the portion of a new Total Expense Ratio construct that discloses the reality of adviser/distributor payments of hidden distribution fees to sales brokers and concludes that adoption of the complete Total Expense Ratio should provide shareholders the “fiduciary protections” promised under the law. The next article, by Jonathan Handy, Hunter Nichols, and Thomas Smythe, utilizes Morningstar Governance Ratings (MGRs) to analyze mutual fund expenses and create an indirect proxy for mutual fund performance, concluding that it is of interest that only those funds with higher stewardship ratings seem to have lower expenses. The last piece in this group is by D.K. Malhotra, Tim Mooney, and Raymond Poteau, and it examines the determinants of cost efficiencies in the U.S. commodity mutual fund industry, noting that cost increases have been less than proportional to increases in fund assets, pointing to economies of scale for the industry, with institutional funds showing greater economies of scale than retail funds since 2010.
Our final three articles focus on issues from an Asian perspective, which should help readers to see a known issue differently and to gauge the progress made in the world of emerging markets toward achieving the sophistication of developed countries. The first article, by Paoyu Huang, Yensen Ni, Yuhsin Chen, explores the effect of an employee stock ownership plan on firm value. The piece reveals that firms with better corporate governance have better stock price performance and that the impact of an ESOP is not significant on firms with higher institutional holdings. The second article, by Plamen Patev and Kaloyan Petkov, looks at the fundamental law of active management using a new source of active risk, strategy risk, instead of unconditional active risk as a risk estimator. The authors concluding that in Taiwan the main source of strategy risk remains the volatility of the information coefficient over time. Our final article, by Lujer Santacruz, is an industry study of wealth management and financial advisory services in the Asia-Pacific region. This research looks at the industry participants and the regulatory environment for each country, examines the general trends and directions that impact the industry, and identifies potential opportunities for the future.
TOPICS: Wealth management, portfolio construction, real assets/alternative investments/private equity
Jean L.P. Brunel
Editor
ENDNOTES
1Ryan Davies, Harry M. Kat and Sa Lu. 2004. “Fund of Hedge Funds Portfolio Selection: A Multiple-Objective Approach.” AIRC Working Paper No. 20 (www.cass.city.ac.uk/airc).
2Skewness means that returns are unevenly distributed around the average, with the term negative skew meaning that more than half of the observations lie to the left of the mean. Excess kurtosis means that that the distribution tends to be “peaked” and has “fat tails.” A distribution with negative skewness and excess kurtosis is more likely to bring negative surprises than one that is shaped normally.
- © 2018 Pageant Media Ltd