Nate Geraci
February 10, 2010
(note: this article was originally published at indexuniverse.com on 2/9/10)
Equity and bond-friendly secular declines in interest and inflation rates have given the investing masses license to stretch Markowitz’s Modern Portfolio Theory (MPT) beyond its contextual foundations and to extrapolate investment practices bearing little relation.
Serial autocorrelation in prices (i.e., that prices tend to trend) and assumptions regarding perpetually “full-in” capital deployment carry deeply profound implications for the very foundations of portfolio theory and, more importantly, for the construction of practical applications.
The next chapter in the evolution of modern portfolio theory addresses these issues through the disaggregation of capital at risk (DCaR) – an extension of MPT encompassing the heretofore lacking dimension of market participation.
Starting with the forest …
One major problem with the majority of the discussion regarding how to manage portfolios is that it starts with the proverbial “trees,” rather than stepping back to gaze at the big-picture “forest.”
The “active vs. passive” investing debate that has raged in financial media circles over the last decade, for example, is relevant and important, but it is far from the first order of “strategy” business that investors and risk managers should undertake. Likewise for ongoing debates regarding the asset allocation format du jour and the casting of opinions regarding the relative importance, prospectively, of individual asset classes or sub-classes.
Indeed, starting with the “active vs. passive” debate, irrespective of which of the two strategies (or both) is adopted, or with the asset class allocation matter, can lead to patently wrong conclusions regarding where investors need to look first in order to manage risk.
By focusing on these debates first, many investors erroneously conclude that simply engineering a particular active or passive asset allocation strategy will enable customization of risk and return attributes with precision, thereby ensuring (or so they believe) risk characteristics that they both understand and can live with.
But along the way they’ve missed the first and, really, most critical question:
To what extent is capital placed at risk?
Just what does this question mean?
Reduced to their most basic level, any and all investment strategies can be viewed as residing in one, and only one, of three discrete risk buckets:
Risk Bucket 1: Fully principal-protected
Risk Bucket 2: Conditional market participation and capital at risk, with explicit exit protocols
Risk Bucket 3: Open-ended, uncapped risk to invested capital
In reviewing an existing portfolio of strategies and holdings, investors and risk managers first need to determine the distribution of investment capital across the three macro-level risk buckets. This is also the first level at which an investment strategy prescription must be defined. Starting anywhere else can cause important sources of risk to be under-appreciated or entirely overlooked.
Let’s take a closer look at the three capital-at-risk classifications to see what each includes and what it does not include.
Risk Bucket 1: Fully principal-protected
This includes FDIC-insured money market holdings, FDIC-insured certificates of deposit and short-term Treasury securities. All other “cash” vehicles lack explicit protocols that would definitively limit risk to invested capital.
Risk Bucket 2: Conditional market participation and capital at risk, with explicit exit protocols
This includes trend following strategies, and others, which place definitive and explicit exit protocols on holdings in a manner designed to limit participation in negatively trending markets.
Risk Bucket 3: Open-ended, uncapped risk to invested capital
This includes buy-and-hold strategies of any flavor, whether they employ actively-managed and/or passively-managed components.
The common denominator is found neither in the instruments used nor in the distribution of holdings across asset classes but, rather, in whether definitive and explicit exit protocols are attached to holdings by the investor or investment manager.
Risk to invested capital that lacks definitive and explicit exit protocols is, de facto, open-ended.
So in which risk bucket would an asset allocation covering equities, fixed income and alternatives fall if it is using actively-managed mutual funds to cover each of the asset classes? Such an approach would reside squarely in the third risk bucket if capital allocated to the strategy lacks explicit exit protocols on each of the funds (or if the individual managers, alternatively, lack explicit exit protocols on holdings residing within their respective strategies).
What about the same asset class allocation strategy (across equities, fixed income and alternatives) if index-based ETFs are used instead of active managers? Classification in the third risk bucket still applies since there are no explicit exit protocols in place to limit damage during market declines. For what it’s worth, a mountain of academic and industry research suggests that the strategy using index-based ETFs is highly likely to outperform the same strategy employing active managers. But that topic, a quite important one, would naturally follow the discussion in this paper.
But what if in either of the above (i.e., whether using active mutual fund managers or index-based ETFs), the asset allocation is, instead, very “conservative”: say, 75% bonds with the rest split into equities and other holdings for diversification purposes? Same story. Third risk bucket. If there are no explicit exit protocols then risk is, in fact, still open-ended. Many bond-oriented investors learned that lesson “well” during the two year stretch through March 2009.
Questions surrounding the use of passively-managed index-based ETFs vs. actively-managed traditional mutual funds are quite important but they speak to issues relating to manager risk relative to participation in index-based alternatives, and to fund costs and holdings transparency. They do not, however, relate to questions regarding market participation and exits from market participation.
Likewise, asset class allocation matters are critical determinants of risk and return characteristics for a given strategy residing within one of the capital-at-risk buckets. In and of themselves, however, asset class allocation questions do not relate to questions regarding market participation and exits from market participation.
What about money market fund holdings that are not FDIC-insured? Such assets would reside in the third risk bucket. These funds, with few (if any) exceptions, have no explicit exit protocols regarding their holdings serving to limit risk. Need a reminder? Think of the Reserve Primary Fund (at the time of its collapse in September 2008, one of the oldest and largest money market funds).
The “cash” and money market fund lessons remind us that: a) there are no free lunches on yield and b) what you don’t know most definitely can hurt you (recall that money market funds are required to publish their holdings only quarterly and up to 60 days in arrears – i.e., holdings data can be as much as two to five months stale). Within the third risk bucket such holdings would be regarded as they truly are: actively-managed very short-term fixed income funds, lacking in absolute principal protection or explicit operational constraints to definitively limit risk.
Practical Applications
The question of what constitutes an appropriate distribution of capital across the three risk buckets for a given investor must precede questions regarding the extent to which risk should be “dialed” up or down, via asset class distribution or otherwise, within either of risk buckets 2 or 3.
For an investor nearing retirement and expecting to need only limited asset growth to satisfy retirement income and other financial goals, the following capital allocation would likely be appropriate:
25-50% Risk Bucket 1: Fully principal-protected
25-75% Risk Bucket 2: Conditional market participation and capital at risk, with explicit exit protocols
0-25% Risk Bucket 3: Open-ended, uncapped risk to invested capital
For this investor, capital deployed in Risk Bucket 3, engendering open-ended, uncapped risk, probably should be in a moderate, moderately conservative or conservative allocation such as a 40/40/20, 30/50/20 or 20/60/20 (equities / fixed-income / alternatives), whether ETFs or a slate of active managers is used.
For an investor far from retirement and on the front end of the “accumulation” phase – and needing growth – the following would likely be more appropriate:
0-25% Risk Bucket 1: Fully principal-protected
25-75% Risk Bucket 2: Conditional market participation and capital at risk, with explicit exit protocols
0-50% Risk Bucket 3: Open-ended, uncapped risk to invested capital
Capital deployed in Risk Bucket 3, engendering open-ended, uncapped risk, probably should be in a moderately aggressive or aggressive allocation such as a 50/30/20 or 60/20/20 (equities / fixed-income / alternatives), whether ETFs or a slate of active managers is used.
Decisions made regarding allocations across capital-at-risk buckets are the primary determinants of the extent to which investors participate in market declines associated with negative trends impacting assets in which they invest.
Investors can dial up or down the risk in strategies residing in Risk Bucket 2 by a) altering the level of diversification or concentration of capital allocations across asset classes and subclasses, b) widening or narrowing the “price bands” at which exits are triggered, c) changing the frequency and scheduling of evaluation of exit trigger parameters, including the use of running “stop-loss” orders, and d) determining the extent to which capital exiting a holding persists is being held in cash instruments (pending reestablishment of a positive price trend in the asset class just exited) or is redeployed to another asset class presently exhibiting a positive price trend.
Once capital allocations are appropriately adjusted on the capital-at-risk macro front, investors and advisors can then clearly and properly focus their attention on important matters relating to the use of passive index-based instruments vs. actively managed funds and asset class allocations associated with those strategies and holdings.
