REVISITING THE MODULE PROCESS IN DETAIL

https://study.cfainstitute.org/app/cfa-institute-program-level-iii-for-august-2024#read/study_task/2562323/revisiting-the-module-process-in-detail-1

REVISITING THE MODULE PROCESS IN DETAIL

Learning Outcome

• recommend and justify an asset allocation using a goals-based approach

Having explained and illustrated the client process in Exhibit 32, we now explore how modules are developed. Creating an appropriate set of optimized modules starts with the formulation of capital market assumptions. Exhibit 37 presents a possible set of forward-looking pretax capital market expectations for expected return, volatility, and liquidity31 in Panel A and a historical 15-year correlation matrix in Panel B.32

Exhibit 37: Example of Capital Market Expectations for a Possible Asset Class Universe

Panel A

	Expected
Return	Volatility	Liquidity
Cash	4.0%	3.0%	100%
Global investment-grade bonds	5.5	6.5	100
Global high-yield bonds	7.0	10.0	100
Lower-volatility alternatives	5.5	5.0	65
Global developed equities	8.0	16.0	100
Global emerging equities	9.5	22.0	100
Equity-based alternatives	6.0	8.0	65
Illiquid global equities	11.0	30.0	0
Trading strategy alternatives	6.5	10.0	80
Global real estate	7.0	15.0	100

Panel B

	Global	Lower-Volatility Alts	Global	Equity-Based Alts	Trading Strategy Alts	Illiquid Equities	Global Real Estate
Cash	IG Bonds	HY Bonds	Developed Equities	Emerging Equities
Cash	1.00	0.00	−0.12	0.08	−0.06	−0.04	0.02	0.04	−0.26	−0.01
Global investment-grade bonds	0.00	1.00	0.27	0.14	0.28	0.09	0.07	0.16	0.20	0.24
Global high-yield bonds	−0.12	0.27	1.00	0.46	0.70	0.17	0.31	−0.08	0.35	0.28
Lower-volatility alternatives	0.08	0.14	0.46	1.00	0.44	0.61	0.86	0.12	0.65	0.47
Global developed equities	−0.06	0.28	0.70	0.44	1.00	0.17	0.32	−0.03	0.47	0.38
Global emerging equities	−0.04	0.09	0.17	0.61	0.17	1.00	0.72	−0.03	0.67	0.49
Equity-based alternatives	0.02	0.07	0.31	0.86	0.32	0.72	1.00	0.11	0.72	0.45
Trading strategy alternatives	0.04	0.16	−0.08	0.12	−0.03	−0.03	0.11	1.00	−0.09	0.07
Illiquid global equities	−0.26	0.20	0.35	0.65	0.47	0.67	0.72	−0.09	1.00	0.88
Global real estate	−0.01	0.24	0.28	0.47	0.38	0.49	0.45	0.07	0.88	1.00

Ostensibly, in the real world, the process ought to be associated with a set of after-tax expectations, which usually cannot be limited to broad asset classes or sub–asset classes. Indeed, the tax impact of management processes within individual asset classes or strategies (for instance, index replication, index replication with systematic tax-loss harvesting, broadly diversified portfolios, or concentrated portfolios) requires that each management process within each asset class or strategy be given its own expected return and volatility. We will dispense with that step here for the sake of simplicity, both in absolute terms and with respect to jurisdictional differences.

Exhibit 38 presents a possible set of such modules based on the capital market expectations from Exhibit 37. The optimization uses a mean–variance process and is subject to a variety of constraints that are meant to reflect both market portfolio considerations and reasonable asset class or strategy suitability given the goals that we expect to correspond to various points on the frontier. Note that the frontier is not “efficient” in the traditional sense of the term because the constraints applied to the portfolios differ from one to the next. Three elements within the set of constraints deserve special mention. The first is the need to be concerned with the liquidity of the various strategies: It would make little sense, even if it were appropriate based on other considerations, to include any material exposure to illiquid equities in a declining-balance portfolio expected to “mature” within 10 years, for instance. Any exposure thus selected would be bound to increase through time because portfolio liquidation focuses on more-liquid assets. The second relates to strategies whose return distributions are known not to be “normal.” This point applies particularly to a number of alternative strategies that suffer from skew and kurtosis,33 which a mean–variance optimization process does not take into account (see Section 6). Finally, the constraints contain a measure of drawdown control to alleviate the problems potentially associated with portfolios that, although apparently optimal, appear too risky in overly challenging market circumstances. Drawdown controls are an important element in that they help deal with the often-observed asymmetric tolerance of investors for volatility: upward volatility is much preferred to downward volatility.

Exhibit 38:

Six Possible Sub-Portfolio Modules

	A	B	C	D	E	F
Portfolio Characteristics
Expected return	4.3%	5.5%	6.4%	7.2%	8.0%	8.7%
Expected volatility	2.7	4.5	6.0	7.5	10.0	12.5
Expected liquidity	100.0	96.6	90.0	86.1	83.6	80.0

Portfolio Allocations
Cash	80%	26%	3%	1%	1%	1%
Global investment-grade bonds	20	44	45	25	0	0
Global high-yield bonds	0	5	11	25	34	4
Lower-volatility alternatives	0	9	13	0	0	0
Global developed equities	0	9	13	19	34	64
Global emerging equities	0	2	2	3	6	11
Equity-based alternatives	0	0	0	8	0	0
Illiquid global equities	0	0	5	10	15	20
Trading strategy alternatives	0	1	3	6	7	0
Global real estate	0	5	5	3	3	0
Total	100%	100%	100%	100%	100%	100%

Constraints
Maximum volatility	3.0%	4.5%	6.0%	7.5%	10.0%	12.5%
Minimum liquidity	100.0	95.0	90.0	85.0	80.0	70.0
Maximum alternatives	0.0	10.0	20.0	30.0	30.0	30.0
Minimum cash	80.0	20.0	0.3	0.5	0.7	1.0
Maximum HY as a percent of total fixed income	0.0	10.0	20.0	50.0	100.0	100.0
Maximum equity spectrum	0.0	10.0	20.0	40.0	75.0	100.0
Maximum EM as a percent of public equities	15.0	15.0	15.0	15.0	15.0	15.0
Maximum illiquid equities	0.0	0.0	5.0	10.0	15.0	20.0
Maximum trading as a percent of equity spectrum	0.0	10.0	15.0	15.0	20.0	25.0
Maximum real estate	0.0	5.0	10.0	15.0	20.0	25.0
Escrow cash as a percent of illiquid equities	5.0	5.0	5.0	5.0	5.0	5.0
Maximum probability of return < drawdown	1.0	1.5	2.0	2.0	2.5	2.5
Drawdown horizon	3	3	3	3	3	3
Drawdown amount	0.0	−5.0	−7.5	−10.0	−15.0	−20.0

The six sub-portfolios shown in Exhibit 38 satisfy two major design goals: First, they cover a wide spectrum of the investment universe, ranging from a nearly all-cash portfolio (Portfolio A) to an all-equity alternative (Portfolio F). Second, they are sufficiently differentiated to avoid creating distinctions without real differences. These portfolios are graphed in Exhibit 39.

Exhibit 39:

Sub-Portfolio Modules Cover a Full Range

Returning to an earlier point about “labeled goals,” one can easily imagine “aspirations” to describe each of these modules, ranging from “immediate- to short-term lifestyle” for Module A to “aggressive growth” for Module F. Module B might be labeled “long-term lifestyle,” while C and D might represent forms of capital preservation and E a form of “balanced growth.”

A final point deserves special emphasis: Modules need to be revisited on a periodic basis. While equilibrium assumptions will likely not change much from one year to the next, the need to identify one’s position with respect to a “normal” market cycle can lead to modest changes in forward-looking assumptions. It would indeed be foolish to keep using long-term equilibrium assumptions when it becomes clear that one is closer to a market top than to a market bottom. The question of the suitability of revisions becomes moot when using a systematic approach such as the Black–Litterman model. One may also need to review the continued suitability of constraints, not to mention (when applicable) the fact that the make-up of the market portfolio may change in terms of geography or credit distribution.