Actively Passive? The Discussion Continues

Actively Passive? The Discussion Continues

Download Acrobat Reader

his article presents updated analysis of actively managing passive assets, specifically U.S. equity indexes. In essence, it is a study of a variety of different portfolio rebalancing protocols. (For a primer you may want to read Actively Passive? in the September, 2003 issue; Activity Level in the January, 2004 issue; and Activity Level, Part II in the February, 2004 issue).

Index funds are typically viewed as a passive asset. It follows that one might ask: should portfolios that are comprised of passive assets be managed passively…or actively? There is no definitive answer to this important question. This article seeks to further the discussion with a goal of providing insights, not necessarily absolute answers. This article explores only two boxes in the Active/Passive Grid (Figure 1) both of which are in the Index Funds column, namely Index / Passive and Index / Active. Studying active and passive portfolio management using actively managed funds at a macro level is virtually impossible due to an enormous number of possible combinations of actively managed funds. Indexes which attempt to replicate the return of prominent asset classes are much fewer in number and thus more feasibly analyzed.

Figure 1. The Active/Passive Grid

In this current study, seven equity benchmark indexes (all from Wilshire Associates) were examined over a 26 year period from 1978 through 2003. Annual return data were obtained from Morningstar's Principia software. All returns reported here are pre-tax, nominal (pre-inflation) figures. For simplicity sake, it was assumed that all portfolio management protocols (whether active or passive) were executed within a tax-sheltered account.

The annual returns for the Wilshire indexes used in this study are reported in Figure 2. Boxes shaded blue represent the index with the highest return in the previous year, whereas gold shaded boxes represent the index with the lowest year in the previous year.

The Wilshire Small Value index had a 26 year annualized return of 16.0%, the highest among the seven indexes. It also had a standard deviation of return of 18.1%. The return divided by the standard deviation of return yielded a ratio of .88, which represents a simplified risk-return measure (where the higher the ratio the better). The second highest return (15.5%) was achieved by the Wilshire Midcap Value Index, but with a lower standard deviation of return than the Small Value index it had the highest risk-return score of .97.

A variety of portfolio rebalancing (i.e. active portfolio management) strategies are illustrated in Figure 3. The seven core Wilshire indexes are in rows 1-7. Row 8 reports the results of a buy-and-hold strategy over the 26 year period. Buy-and-hold performance is a benchmark. It represents a baseline level of performance against which the "value added" of active portfolio management might be measured. Buy-and-hold is the upper left-hand box in Figure 1, or an Index / Passive strategy. In this analysis, a buy-and-hold strategy represented equal investments into each of the seven indexes with no subsequent rebalancing during the 26 year period.

Rows 9-23 include a number of active management protocols. Seven of the active management protocols (in red text) represent a "core/satellite" philosophy. The general idea is that the "core" asset is over-weighted with the remainder of the assets spread out among other "satellite" holdings. In this case, 20% of the investment was invested into the core index (requiring annual rebalancing to maintain the 20% weighting) and the remaining 80% was equally reallocated annually among the other six indexes (at 13.33% per index).

Rows in blue text (rows 12 and 19) are permutations of the core-satellite approach, where the core index at the start of each year was either the best or worst index from the prior year. The remaining six portfolio management options (in black text) do not reflect a core-satellite approach, but either a traditional rebalancing approach or a highly concentrated management protocol.

Column A is the 26-year average annualized return (or geometric mean). Column B is the standard deviation of the 26 annual returns. Standard deviation of return is commonly used as a surrogate (albeit an imperfect one) for risk level. The return/risk measure in Column C is calculated by dividing the data in Column A by the data in Column B.

Column D is a useful, and novel, statistic. It is the arithmetic mean of 22 rolling five-year returns. For this particular data set, the first five-year rolling return (which is calculated as a geometric mean) was generated by the annual return data from 1978 to 1982. The next five-year rolling return covered the time frame from 1979-1983, and so on. Column E is the standard deviation of 22 five-year rolling returns, and column F is the return-risk measure using the 5-year rolling return data.

Columns G and H in Figure 3 indicate the number of times a particular active management strategy (i.e. rebalancing technique) had the highest (G) or lowest (H) five-year rolling return during the 26 year period. Column F is highlighted in yellow because it was the criterion used to rank the indexes (rows 1-7) and the portfolio management protocols (rows 9-23). The core Wilshire indexes and the 15 active management protocols were ranked separately from high to low.

The Wilshire Midcap Value index generated the highest risk-adjusted return score (columns C and F) over the 26 year period, both on an annualized basis and by five-year rolling returns. The Wilshire Small Value had the highest raw return, both annually and five-year rolling returns (columns A and D). Interestingly, the Midcap Value index had the highest five-year rolling return only one time out of 22 five-year rolling returns, whereas the small value index had the highest return 5 times. The large growth index generated the highest five-year rolling return 7 times, or nearly 33% of time, but it also had the lowest return five times. The REIT index had the lowest five-year rolling return 9 times out of 22, but also had the highest return on four occasions.

The three "value" indexes (large, mid, small) combined had the lowest five-year rolling return only one time, whereas the growth indexes had the low return 12 times. The value indexes had the high five-year rolling return 9 times, the same number of times as the growth indexes. Over this particular time frame, value indexes stood out in terms of raw performance and risk-adjusted performance.

Among the active portfolio management protocols, the most value was added (above buy-and-hold) by annually rebalancing (at the start of each year) into the prior year's best three indexes. For example, the three indexes with the best performance in 1978 were large growth, midcap growth, and small growth. Thus, all portfolio assets were equally rebalanced (i.e. allotted) into those three indexes at the start of 1979. The same procedure was carried out from year to year. This approach generated a 26 year annualized return 124 basis points higher than a buy-and-hold approach. Annually rebalancing into the best three indexes also generated the highest average five-year rolling return (15.04%), the lowest volatility of five-year returns (S.D. of 5.10%), and had the highest five-year rolling return 8 times out of the 22 five year rolling periods between 1982-2003.

The next most attractive "Index / Active" strategy was annual rebalancing among the six best indexes from the prior year (row 10). In essence, this is simply rebalancing among all the indexes, but dropping out the prior year's worst performer.

Annual rebalancing while maintaining a 20% core position in midcap value was the third most attractive protocol. The fourth best protocol was one in which the best index from the prior year received a 20% allocation of the portfolio, with the remaining 80% of the assets being equally allocated the remaining six indexes. By referring back to Figure 2 one can annually track which indexes would have been selected by following the trail of blue boxes. Notice that midcap value is the only index to NOT have a blue shaded box, yet it is the index which generated the highest risk-adjusted performance in isolation (row 1) and was the best performing core index in an active management protocol (row 11). The tortoise wins again. But, this is a fast tortoise!

The least attractive active management protocol (at least as determined by the risk-adjusted return coefficient in column F) was investing all assets into the worst performing index from the prior year (row 23). The year-to-year indexes which were utilized using this approach are highlighted in gold in Figure 2. Interestingly, this approach generated the highest five-year rolling return on 5 occasions, or 23% of the time. The occasional success of investing in last year's dog is validated by these results. As a long-term strategy, bottom fishing does not appear to be wise idea. Investing all portfolio assets into the prior year's best performing index (row 21) produced an annualized return (column A) 268 basis points higher than investing solely in last year's worst index. However, the average five-year rolling return (column D) of the last year's best index was only 72 basis points higher than the average five-year rolling return of the prior year's worst index.

Allocating all assets into the best performing index from the prior year had the highest five-year rolling return 5 times, but also generated the lowest five-year rolling return on 10 occasions - compared to only 6 times for the prior year's worst performing index. Both of these protocols represent a concentrated approach. They both produce winning periods, but also find themselves in the cellar fairly often. Over time both approaches failed to provide the consistency which generates long-term superior risk-adjusted performance.

To summarize, 13 out of the 15 active management rebalancing protocols (rows 9-23) generated higher volatility-adjusted performance (column C) compared to a buy-and-hold approach over the 26 year period. The margin of superiority in raw return (column A) was small in many cases, however. Passively managing passive assets (i.e. buying-and-holding index funds) is not a bad idea. But, for those willing to actively manage passive assets, there appears to be value-added potential in doing so - particularly when utilizing value indexes.

Figure 2. Seven U.S. Equity Indexes

Figure 3. Value Adding Activity (1978 - 2003)

		A	B	C	D	E	F	G	H
	U.S. Equity Indexes	26 Year Average Annual Return	26 Year Standard Deviation of Return	Return / Risk	Ave. 5 Yr Rolling Return   (22 total)	Average Std. Dev. of 5 Yr Rolling Return	Return / Risk	Number of Times with Highest 5 Year Rolling Return	Number of Times with Lowest 5 Year Rolling Return
1	Wilshire Midcap Value	15.46	15.91	0.97	15.33	4.39	3.49	1	0
2	Wilshire Large Value	13.49	14.41	0.94	14.59	5.91	2.47	3	1
3	Wilshire Midcap Growth	14.37	21.52	0.67	14.63	6.10	2.40	1	0
4	Wilshire Small Value	15.99	18.10	0.88	15.74	6.75	2.33	5	0
5	Wilshire Large Growth	12.66	20.24	0.63	14.41	8.09	1.78	7	5
6	Wilshire Small Growth	12.32	23.57	0.52	12.17	7.51	1.62	1	7
7	Wilshire REIT	13.86	16.72	0.83	12.49	8.04	1.55	4	9
8	Buy and Hold	14.20	18.64	0.76	--	--	--	--	--
9	Best 3 Indexes from Prior Year	15.44	16.91	0.91	15.04	5.10	2.95	8	0
10	Annual Rebalance Among Best 6 from Prior Year	14.86	15.65	0.95	14.89	5.22	2.85	2	0
11	20% Mid Value Core, 80% in Remaining 6	14.62	15.36	0.95	14.69	5.21	2.82	0	0
12	20% in Best from Prior Yr, 80% into Remaining 6	14.59	15.24	0.96	14.56	5.22	2.79	0	0
13	20% Small Value Core, 80% in Remaining 6	14.66	15.44	0.95	14.72	5.31	2.77	0	0
14	20% Large Value Core, 80% in Remaining 6	14.48	15.29	0.95	14.63	5.28	2.77	1	0
15	Annual Equal Rebalance	14.54	15.45	0.94	14.63	5.29	2.76	0	0
16	20% Mid Growth Core, 80% in Remaining 6	14.55	15.71	0.93	14.65	5.31	2.76	0	0
17	20% Large Growth Core, 80% in Remaining 6	14.46	15.50	0.93	14.65	5.32	2.75	1	0
18	20% REIT Core, 80% in Remaining 6	14.55	15.11	0.96	14.53	5.29	2.75	0	0
19	20% in Worst from Prior Yr, 80% into Remaining 6	14.40	15.44	0.93	14.51	5.34	2.72	0	0
20	20% Small Growth Core, 80% in Remaining 6	14.43	15.87	0.91	14.49	5.40	2.68	0	0
21	100% Best Index from Prior Year	14.66	16.45	0.89	13.19	5.58	2.36	5	10
22	Worst 3 Indexes from Prior Year	11.89	17.53	0.68	12.28	6.48	1.89	0	6
23	100% Worst Index from Prior Year	11.98	18.71	0.64	12.47	6.69	1.86	5	6

____________________________________________________________________________________
Craig L. Israelsen, Ph.D. is an associate professor at Brigham Young University. He teaches family finance in the Department of Home and Family Living. His research interests include mutual fund analysis. He writes monthly for Financial Planning magazine.