The URL can be used to link to this page

Ord 16-01ORDINANCE NO. 16-01 AN ORDINANCE OF THE CITY COMMISSION OF THE CITY OF DELRAY BEACH, FLORIDA AMENDING CHAPTER 33, "POLICE AND FIRE DEPARTMENTS", SUBHEADING 'PENSIONS", OF THE CODE OF ORDINANCES OF THE CITY OF DELRAY BEACH, FLORIDA BY AMENDING SECTION 33.66, "FINANCES AND FUND MANAGEMENT," SECTION 33.66(E)(2)(b) TO PROVIDE FOR A CHANGE IN THE INVESTMENT PARAMETERS FOR COMMON OR CAPITAL STOCK; PROVIDING FOR REBALANCING; PROVIDING A GENERAL REPEALER CLAUSE, A SAVING CLAUSE, AND AN EFFECTIVE DATE. WHEREAS, the City and the Board of Trustees of the Police and Fire Retirement Plan, desire to change the parameters of investments in common or capital stock to enhance the performance of the Retirement Fund while controlling the risk. NOW, THEREFORE, be it ordained by the City Commission of the City of Delray Beach, Florida as follows: Section 1. That Section 33.66, "Finances and Fund Management", Section E(2)(b) be and is hereby amended to read as follows: (b) The Board of Trustees shall not invest more than five percent (5 %) of Trust Fund its assets in the common or capital stock of any one issuing company, nor shall the aggregate investment in common or capital stock in one company exceed five percent (5%) of the outstanding common or capital stock of that company; nor shall the aggregate of the Fund's investments in common or capital stock ~* ~"o* using the market value method exceed ~ seventy percent t~r~mx (70%) of the Fund's assets, as measured on the last business day of each quarter. The Board of Trustees shall rebalance the Fund's assets by the last business day of the following quarter. Section 2. same are hereby repealed. That all ordinances or parts of ordinances in conflict herewith be and the Section 3. That should any section or provision of this ordinance or any portion thereof, any paragraph, sentence, or word be declared by a Court of competent jurisdiction to be invalid, such decision shall not affect the validity of the remmnder hereof as a whole or part thereof other than the part declared to be invalid. Section 4. reading. That this ordinance shall become effective upon its passage on second and final ED AND ADOPTED 2001. in regular session on second and final reading on this the ~ff~day of MAYOR ATTEST: City Clerk First Reading Second Reading. MEMORANDUM TO: FROM: SUBJECT: DATE: MAYOR AND CITY COMMISSIONERS CITY MANAGER ~ AGENDA ITEM /0~ · - REGULAR MEETING OF MARCH 6, 2001 ORDINANCE NO. 16-01 (CHANGING INVESTMENT PARAMETERS IN THE POLICE AND FIRE PENSION PLAN) MARCH2,2001 This is second reading and public hearing for Ordinance No. 16-01 that would modify the Police and Fire Pension Plan. The Police and Fire Pension Board has requested that the City Code be amended to allow them to invest up to 70% of their assets in common stock. The Code currently limits this type of investment to 60%. In the backup is a letter from Jeff Keating in which he recommends the current limit of 60%, together with three artides on asset allocation strategies. After reading the articles, I think we should approve increasing the limit on stocks to 70%. Based on almost 200 years of historical stock market performance, this would give greater growth over the long tenn. However, increasing the limit to 70% would increase the volatility of the asset value in the Police and Fire Pension Fund. As a result, our exposure to year-to-year jumps in our required contribution would be increased. But, our total contributions over the years should be less. The Commission just needs to decide whether the long term decrease in our required contributions is worth the increased exposure to wide swings in our required contribution. At first reading on February 6m, the City Commission passed the ordinance by a unanimous vote. Recommend approval of Ordinance No. 16-01 on second and final reading. Ref'Agmemo14.Ord16-01Polic. e&Fire Pension Plan 2"~ Reading [lTV OF DELRI:IV BEI:I[H CITY ATTORNEY'S OFFICE 200 NW 1st AVENUE · DELRAY BEACH, FLORIDA 33444 TELEPHONE 561/243-7090 · FACSIMILE 561/278-4755 1993 DATE: TO: FROM: SUBJECT: Writer's Dtrect L~ne 561/243-7091 MEMORANDUM January 25, 2001 City Commission David Harden, City Manager Susan A. Ruby, City Attorney Ordinance Changing Investment Parameters in the Police and Fire Pension Pension Plan Enclosed is an ordinance that provides the Police and Fire Pension Plan Board of Trustees may value its aggregate stock investments using the market value method instead of cost, and provides that value of the stock may not exceed 70% (instead of 60%) of the funds assets measured quarterly. It also provides for rebalancing of the Funds Assets. Please place this ordinance on the February 6, 2001 City Commission agenda for first re~~__~ SAR:ci Enclosure CC: Alison Harty, City Clerk Bill Adams, President, Police and Fire Pension Board Steve Cypen, Esq. James W. Lima, Esq. ORDINANCE NO. ! 6- 01 AN ORDINANCE OF THE CITY COMMISSION OF THE CITY OF DELRAY BEACH, FLORIDA AMENDING CHAPTER 33, "POLICE AND FIRE DEPARTMENTS", SUBHEADING 'PENSIONS", OF THE CODE OF ORDINANCES OF THE CITY OF DELRAY BEACH, FLORIDA BY AMENDING SECTION 33.66, "FINANCES AND FUND MANAGEMENT," SECTION 33.66(E)(2)(b) TO PROVIDE FOR A CHANGE IN THE INVESTMENT PARAMETERS FOR COMMON OR CAPITAL STOCK; PROVIDING FOR REBALANCING; PROVIDING A GENERAL REPEALER CLAUSE, A SAVING CLAUSE, AND AN EFFECTIVE DATE. WHEREAS, the City and the Board of Trustees of the Police and Fire Retirement Plan, desire to change the parameters of investments in common or capital stock to enhance the performance of the Retirement Fund while controlling the risk. NOW, THEREFORE, be it ordained by the City Commission of the City of Delray Beach, Florida as follows: Section 1. That Section 33.66, "Finances and Fund Management", Section E(2)(b) be and is hereby amended to read as follows: (b) The Board of Trustees shall not invest more than five percent (5 %) of Trust Fund its assets in the common or capital stock of any one issuing company, nor shall the aggregate investment in common or capital stock in one company exceed five percent (5%) of the outstanding common or capital stock of that company; nor shall the aggregate of the Fund's investments in common or capital stock at cost using the market value method exceed sirdy seventy percent,,.,.~':c~ ~,~j~ (70 %) of the Fund's assets, as measured on the last business day of each quarter. The Board of Trustees shall rebalance the Fund's assets by the last business day of the following quarter. Section 2. That all ordinances or parts of ordinances in conflict herewith be and the same are hereby repealed. Section 3. That should any section or provision of this ordinance or any portion thereof, any paragraph, sentence, or word be declared by a Court of competent jurisdiction to be invalid, such decision shall not affect the validity of the remainder hereof as a whole or part thereof other than the part declared to be invalid. Section 4. second and final reading. That this ordinance shall become effective upon its passage on PASSED AND ADOPTED in regular session on second and final reading on this the ~ day of 2001. MAYOR ATTEST: City Clerk First Reading Second Reading. - 2 - Ord. No. 16-01 KF_ATING INVESTMENT COUNSELORS REGISTERED INVESTMENT ADVISORS FOR INDIVIDUALS, RETIREMENT PLANS, TRUSTS AND FOUNDATIONS Milene Walinski City of Delray Beach Finance Department 100 NW 1 st Avenue Delray Beach, FL 33444 Dear Milene: RECEIVED CITY ~ANAGER September 8, 2000 Enclosed are two articles on the issue of the optimum asset allocation between stocks and bonds for Pension Plans or long term investors. Very simply, the optimum mix for a Pension Plan is considered to be in the range of 60% in equities, with 40% in fixed income. Historically, this balance has captured the vast majority of being invested 100% in equities with 40% less risk as measured through price volatility. The major issue for the city is that even though the police and fireman's pension plan has a long longevity, in essence, you are always looking at 10 years or longer in terms of expected investment returns, the plan is structured such that it impacts the city's finances on a year-to-year basis based on the performance of the plan. Thus, the city should be looking at an asset allocation that is more appropriate to an investor that has more of a short term orientation. The article by Jeremy Siegel would suggest a more balanced 50%/50% structure between stocks and bonds. Hopefully the enclosed articles will be of some assistance to you. If they don't, I have a couple of other articles, but they are even more esoteric than these I have enclosed. If you have any further questions, please feel free to call my office. Truly yours, J~ff-r~eating JJK:lfs Enclosures 777 EAST ATLANTIC AVENUE, SUITE 303 · DELRAY BEACH, FLORIDA 33483 · (561) 278-7862 · FAX (561) 265-1055 FOR THE JEREMY J. SIEGEL Second Edition STOC KS FOR THE LONG RUN The Definitive Guide to Financial Market Returns and Long-Term Investment Strategies JEREMY J. SIEGEL Professor of Finance-- the Wharton School of the University of Pennsylvania McGraw-Hill New York San Francisco Weshington, D.C. Auckland Bogot~ Caracas Lisbon London Madrid Mexico City Milan Montreal New Delhi San Juan Singapore Sydney Tokyo Toronto Risk, Return and the Coming Age Wave 35 A second reason for the increase in correlation between stock and bond returns is the strategy that portfolio managers follow to al- locate assets. Most tactical allocation models, which money managers use to minimize the risk and maximize the return of a portfolio, dic- tate that the share of a portfolio that is allocated to stocks be a func- tion of the expected return on stocks relative to that on bonds. As interest rates rise, causing stock prices to fall, prospective bond re- turns become more attractive, motivating these managers to sell stocks. As a result, stock and bond prices move together. This is an ex- ample of how the actions by portfolio managers trying to take advan- tage of the historical correlation between stocks and bonds changes their future correlation. EFFICIENT FRONTIERS3 Modern portfolio theory describes how to alter the risk and return of a portfolio by changing the mix between assets. Figure 2-6, based on the nearly 200-year history of stock and bond returns, displays the risks and returns that result from varying the proportion of stocks and bonds in a portfolio. The square at the bottom of each curve represents the risk and re- turn of an all-bond portfolio, while the cross at the top of the curve rep- resents the risk and return of an all-stock portfolio. The circle indicates the minimum risk achievable by combining stocks and bonds. The curve that connects these points represents the risk and return of all blends of portfolios from 100 percent bonds to 100 percent stocks. This curve, called the efficient frontier, is at the heart of modern portfolio analysis and the foundation of asset allocation models. Investors can achieve any combination of risk and return along the curve by changing the proportion of stocks and bonds. Moving up the curve means increasing the proportion in stocks and correspondingly reducing the proportion in bonds. For short-term holding periods, mov- ing up the curve increases both the return and the risk of the portfolio. The slope of any point on the efficient frontier indicates the risk-return trade-off for that allocation. By finding the points on the longer-term ef- ficient frontiers that equal the slope on the one-year frontier, one can de- termine the allocations that represent the same risk-return trade-offs for all holding periods. 3 This section, which contains some advanced mater~al, can be skipped without loss of continuity. of the" for the event. 36 PART l The Verdict of Histo~ FIGURE 2--6 Risk-Return Trade-Offs for Various HoLding Periods, 1802-1996 Fessor Week the co~ market m;3st c( case fO necessi ten--.a( one of historyl Now breakin the toni clusion! power 1 safer f more c( with up( tion thai [] A tho and c have sever( [] How i chant. knowl [] The a~ stock investi (:an el [] How investr your n [] Analy. isons ( 7% 3%. 2%. h~b h " [o Min,mum Risk\ \ ~ b p 100% Bonds I I I I ' I I " I I I 2% 4% 8% 8% 10% 12% 14% 16% 18% 20% RECOMMENDED PORTFOLIO ALLOCATIONS Table 2-2 indicates the percentage of an investor's portfolio that should be invested in stocks based on both the risk tolerance and the holding period of the investor.4 Four classes of investors are analyzed: the ultra- conservative investor who demands maximum safety no matter the re- turn, the conservative investor who accepts small risks to achieve extra re.rum, the moderate risk-taking investor, and the aggressive investor who is willing to accept substantial risks in search of extra returns. The recommended equity allocation increases dramatically as the holding period lengthens. The analysis indicates that, based on the histor- 4 The one-year proportions (except minimum risk point) are arbitrary, and are used as benchmarks for other holding periods. Choosing different proportions as benchmarks does not qualitatively change the following results. CHAPTER 2 Risk, Return and the Commg Age Wave 37 TABLE 2-2 Portfoho Allocation: Percentage of Portfolio in Stocks Based on All Historical Data Risk Hoidinc Period · Tolerance 1 year 5 years 10 years 30 years Ultra-conservative 7,0% 25.0% 40.6% 71.3% (Minimum Risk) Conservative 25.0% 42.4% 61.3% 89.7% Moderate 50.0% 62.7% 86.0% 112.9% Risk-taking 75.0% 77.0% 104.3% 131.5% ical returns on stocks and bonds, ultra-conservative investors should hold nearly three-quarters of their portfolio in stocks over 30-year hoiding pe- riods. This allocation is justified since stocks are safer than bonds in terms of purchasing power over long periods of time. Conservative investors should have nearly 90% of their portfolio in stocks, while moderate and aggressive investors should have over 100 percent in equity. This alloca- tion can be achieved by borrowing or leveraging an all-stock portfolio. Given these striking results, it might seem puzzling why the holding period has almost never been considered in portfolio theory. This is be- cause modern portfolio theory was established when the academic pro- fession believed in the random walk theory of security prices. As noted earlier, under a random walk, the relative risk of securities does not change for different time frames, so portfolio allocations do not depend on the holding period. The holding period becomes a crucial issue in port- folio theory when data reveal the mean reversion of the stock reL-urns.$ 5 For a similar conclusion, see Nicholas Barbens, "Investing for the Long Run When Returns Are Predictable," working paper, Umverslty of Chicago, July 1997 Paul Samuelson has shown that mean reversion will increase equity holdings if investors have a risk aversion coefficaent greater than unity, which most researehers find is the case. See Samuelson, "Long-Run Risk Tolerance When Equity Return.', An., Mean Regr~_.~mg- Pseudoparadoxe~ and Vindication of 'Busmessmen's Pask'" in W C Bramard, W D Nordhau.~, and I-I.W Watt.,,, ed~, Money, Macro¢commu~.~, and Pubhc Pohcy, Cambridge, Mass. The MIT Press, 1991, pp. 181-200. See also Zvl Bodie, Robert Merton, and WLlliam Samuelson, "Labor Supply Flexibdity and Portfolio Choice in a Lifecycle Model," Journal of Economic Dynamtcs and Control, vol. 16, no 3 (July/October 1992), pp 427-450 Bod]e, et al. have shown that equl~ holdings can vary w~th age because stock returns can be correlated with labor income. ~) Seminar Proceedings Investment Policy April 18-20, 1994 Tokyo, Japan Keith P. Ambachtsheer Charles D. Ellis, CFA John L. Maginn, CFA David F. Swensen John R. Thomas, CFA Donald L. Tuttle, CFA, Moderator Edited by Jan R. Squires, CFA _.AIMR Sponsored by the Association for Investment Management and Research and the Security Analysts Association of Japan A Tale of Two Pension Funds: Case Analysis (A) {n making strategic asset allocation rec- ommendations for their pension funds, Jim Jones and Barbara Smith face a prob- lem shared by every pension fund man- ager: how to maximize the return on the portfolio for an appropriate level of risk. To find the solutions to this problem, Jones and Smith must keep in mind the pension fund balance sheet (Exhibit 1), which will dictate to a significant degree the risk-return trade-off for each fund. Both investment managers believe that their investment strategies should be based on a going-concern view of their pension plans. As going concerns, the pension funds must prowde for the impact of future wage and inflation ex- periences on pension fund liabilities. Jones and Smith agree that the immuniz- ing investment strategy would combLne some fixed-rate bonds with inflation-in- dexed bonds and, possibly, a small eq- uity position to hedge against unknown future trends in wages. Deciding the funds' appropriate long-term invest- ment strategies, however, requires ad- dressing two further issues. F~rst, the U.S. Treasury is not yet issuing inflation-indexed bonds. So, what substitute asset(s) or alternative strategies can be used for an inflation hedge? Second, how much incremental re- turn should each fund seek relative to the return produced by the immunizing strategies? Any incremental return real- ized will reduce the contributions re- quired to finance the promised benefits under the plans. Because Alpha Corpo- ration Is a prwate-sector corporation seeking to maximize shareholder wealth and the Public Sector Retirement System (PSRS) is sponsored by a state already committed to funding a previous contri- bution shortfall, both plans would wel- come an investment policy that would reduce their required contributions. Thus, Jones and Smith face the difficult task of improving the investment per- formance of their pension funds m the real world of uncertain future asset-class returns and uncertain future economic conditions shaping those returns. Developing asset allocation recom- mendations that will earn a return be- yond the ideal immunizing strategy re- quires a forecast of the expected returns for different asset classes under different economic conditions, and a determina- tion of the asset allocation combination that appears to offer the best return op- portunities relative to the immunizing strategy. Jones has started this process by developing the forecasts of asset- class returns for 1994 through 1998 given in Table 2 of the A case. To com- plete the analysis, three questions must be addressed: · How realistic are Jones's esti- mates of the returns for the 1994- 98 period? · If the estimates are realistic, what are the estimated returns for dif- ferent portfolio compositions in the three different economic sce- narios for the 1994-98 period? · What asset allocation strategies should Jones and Smith recom- mend to their boards of directors or trustees? The historical returns of different as- set classes in different economic condi- tions (Table 1 of the A case) can be used to evaluate Jones's projected returns. To use this historical information effec- tively, Smith and Jones need to catego- rize the different historical periods in line with the definitions for the three projected scenarios---degearmg, infla- tion, or deflation. In terms of economic growth and inflation, the 1952-65 period corresponds most closely to the degear- lng definition, the 1928--40 period to the deflation definition, and the 1966-81 pe- riod to the inflation definition. Note, however, that the expected GN strt wel real ~Otl 5 25 of 1 and the, bdis expt verx flati, drev Stoc, rlSlrl borlt. mad prol~ the F set-c int]a h~st( retu~ late their ally stocJ sign nent per- tds in the sset-class ~onomic n recom- -~tum be- ategy re- :t returns different terrruna- bination .mm op- ~unizing process ~f asset- ~ 1998 To corn- ns must ~'s esti- ~e 1994- tc, what for dif- :ions in nic sce- iod? 'ategies recom- rectors :ent as- condi- .e used -ns. To effec- atego- ~ds in three infla- nomic ~eriod -'gear- to the 51 pe- ected economic conditions for 1994-98 do not closely match any of the past periods in Table 1. For example, in the 1952-65 period, the economy was characterized by higher economic growth (3.3 percent GNP growth) and lower inflation (1.4 percent) than is projected for the degear- lng scenario. The degearing scenario as- sumes that the U.S. economy will grow slowly during the decade (2.5 percent increase in annual GNP growth) as Fed- eral Reserve tight-money policies hold down the level of inflation (to 3.0 per- cent annually) and the United States works its way out of a debt overhang and large federal government deficits. Given its assumptions as to real GNP growth and inflation and the 1994 structure of interest rates and dividend yields, the degearing scenario projects real returns for bills of 1.5 percent, for bonds of 3.25 percent, and for stocks of 5.25 percent. The comparison 1952-65 period experienced real returns for bills of 1.3 percent, for bonds of-0.2 percent, and for stocks of 13.1 percent. Although the actual and projected experiences for bills are similar, the actual and projected experiences for bonds and stocks are very different. The difference lies in the higher economic growth and lower in- flation of the 1952--66 period, which drew money into stocks and thus in- creased stock multiples and improved stock returns; at the same time, bond yields were low in the early 1950s and rising interest rates worked to depress bond returns. After adjustments are made for these differences, the returns projected in Table 2 for the degearing scenario appear realistic. Similar differences exist between the projected economic forecasts and as- set-class returns for the deflation and the inflation scenarios and their comparable historical periods. Overall, Jones's real- return projections for the period of the late 1990s show bill returns, similar to their historical experiences but gener- ally better bond returns and poorer stock returns relative to their historical experiences. Why would ]ones project experi- ences for the future that appear to be significantly different from the experi- ences of the past? The answer to this questions lies in his assumption that a stock market correction will occur if, in- stead of degearing, either a deflation or inflation scenario occurs. Jones appar- ently assumes that the deflation sce- nario, particularly if started by a col- lapse of the Japanese economy, would result in substantially increased fears about the future of the global economy, a lowering of expectations for corporate profits, and increased fears about the outlook for stocks. This outcome would generate an increase in the risk premium on stocks, resulting in negative returns on stocks in the 1994-98 period. In con- trast, bond returns would be enhanced as bond prices rose with a drop in long- term interest rates accompanying the switch to a deflation scenario. ]ones is apparently suggesting that, in an infla- tion scenario, higher capitalization-rate requirements would, again, lower the returns to stocks because of stock price declines. Bond prices and returns would suffer even more than stocks in this scenario. Disagreement with Jones's assump- tions regarding the market reaction to the deflation and inflation scenarios and the subsequent impact on stocks would lead to projected returns more attuned to historical averages. His assumption of a market correction appears reason- able, however, in light of the long bull market reflected in the returns to stocks in the 1982-93 period and the market's h~storically adverse reaction in short periods to unexpected deflationary and infla tionary environments as the market seeks new equilibrium levels. Assuming that the projected returns in the different scenarios are reasonable, Smith and Jones can calculate reason- able returns for each of the four alterna- tive portfolios in the three scenarios. The portfolio returns are the weighted averages of the projected asset-class re- turns for each economic scenario. The weightings would be the percentages each asset class represents in the particu- lar portfolio. The weighted real returns of particular portfolios in the three eco- nomic scenarios are given in Table A. Cash was given no weighting in any of 75 T~)ie A. Weighted-Average Real Returns of ,~e.'~------d~d Portfolios by Economic Scenario Asset Mix De§earing Deflation Inflation 70°1o equity/30% debt 4.65% --4.35% -2 60% 60% equity/40% debt 4.45 -2.80 -2.80 S0o/o eqtuty/50% debt 4.25 -1.25 -3 00 40% equity/60% debt 4.05 0.30 -3.20 the portfolios suggested for analysis and is not included in any of the portfolio descriptions. The different portfolios exhibit sig- nificantly different return performance in the three economic scenarios. In the degearing scenario, an equity-heavy portfolio performs slightly better than a debt-heavy scenario, but in a deflatiofl scenario, a debt-heavy portfolio per- forms significantly better than an eq- uity-heavy portfolio. Although none of the portfolio mixes performs well in the inflation scenario, an equity-heavy port- folio does somewhat better than a debt- heavy portfolio. These return differences have sig- nificant implications for the asset-mix recommendations Jones and Smith should make to their boards. If, as Smith apparently believes, a good possibility exists that the degearing scenario could deteriorate into deflation, then the port- folio should be debt heavy. This strategy would not significantly reduce returns in the degearing scenario but would provide significant protection against the downside of deflation. In the infla- tion scenario, a shading toward equity should produce the best (or least worst) results. If the inflation scenario is a vi- able possibility, the only hedge available would be a sizable allocation of the port- folio to cash equivalents (T-bills). Based on the three possible scenar- ios, which of the four portfolio alterna- fives evaluated in Table A should be rec- ommended for the Alpha Corporation fund and the PSRS fund? Alpha Corporation's current asset mix of 70 percent equity/30 percent debt reflects the company's ability and will- ingness to assume larger risk than the PSRS. If the asset-mix strategy does not produce the desired real-return results, Jones would have to inform the board of directors that a larger future contribu- tion to the corporation's pension plan is required to provide its future defined benefits. The result would be a reduc- tion of corporate cash available for other purposes for a time. In conclusion, given the possible capital market envi- ronments for 1994 through 1998, and given Jones's view that inflation is more likely than deflation, Alpha Corporation is likely to maintain its 70 percent eq- uity/30 percent debt strategy. The PSRS's current asset mrx of 50 percent equity/50 percent debt reflects its concerns about unfunded future li- abilities and its resulting greater risk aversion (relative to Alpha Corpora- tion). If the PSRS's strategy does not produce the desired real returns, the sys- tem must request funds from the state government, which involves raising taxes or cutting expenditures, and from the employees, which involves addi- tional contributions from their salaries. Neither necessity would be painless. Therefore, given the possible 1994-98 environments and given Smith's view that deflation is more likely than infla- tion, the PSRS might find a 10-percent- age-point shift from equity to debt desir- able. The resulting mix, 40 percent eq- uity/60 percent debt, would be rela- tively conservative. In addition, be- came of the inflation sensitivity of the plan's liabilities, the PSRS should con- sider shortening the duration of the debt portfolio. ~set Allocation for Institutional ios San Frandsco, California !February 10-11, 1986 D. Joehnk, CFA P. Ambachtsheer ::'-Robert D. Amott · Gary P. Brinson, CFA · Kathleen A. Condon, CFA "..; H. Gifford Fong W'dliam L. Fouse, CFA Walter R. Good, CFA Sponsored by The Institute of Chartered Financial Analysts Mark P. Kritzman, CFA Greta E. Marshall, CFA Robert H. Litzenberger Douglas A. Love Roger F. Murray Claude N. Rosenberg, Jr. William F. Sharpe Strategic Approaches to Asset Allocation Keith P. Ambachtsheer When considering the subject of "strategic ap- proaches to asset allocation," one should first ask what is meant by strategy, tactics, and the many other descriptors that are used in the area of asset allocation. Why is asset allocation such a foggy concept? Why are there such conceptual difficulties in this area? It has a lot to do with a confusion between means and ends. The defi- nition I offer is: Asset allocation is the means to achieve an appropriate package of risk, return, and time through the three steps of context un- derstanding, future thinking, and decision im- plementation. ASSET ALLOCATION: THE LONG- TERM CONTRACT There are two basic contexts in which to study asset allocation, and they should be kept sepa- rate. Context number one is long-term asset allo- cation. To illustrate long-term asset allocation, consider the case of the hypothetical Alpha Corporation. They have made that conceptual breakthrough, and they think of their defined benefits plan as a financial business. As such, they recognize that understanding and manag- ing the plan balance sheet is the appropriate way to proceed. As seen in Figure 1, Alpha Corporation is a billion dollar company with a fairly normal looking principal business balance sheet. But Alpha uses the extended balance sheet concept, and as shown on the pension plan balance sheet, the assets and liabilities are very significant rela- tive to the main business assets and liabilities. Assets in the pension plan are stated at the mar- ket, and liabilities are stated on a "best-estimate going-concern basis." In other words, they use economic assumptions on a best-estimate basis about return and inflation. The going-concern part means that it is important to recognize, for the final earnings plans at least, that the liability is based on final earnings, not on the level of earnings. Alpha Corporation has a signif- icant asset cushion in the pension plan of $400 million, which in relation to the best estimate of liabilities, amounts to a 44 percent asset cush- ion. When analyzing the pension plan balance sheet, a number of policy issues need to be ad- dressed. First is funding policy, or asset cushion policy. Of concern here is the relationship be- tween plan assets and projected plan liabilities accrued to date. Should there be a large cushion or a small cushion and why? The second issue is asset mix policy, which deals with the left side of the balance sheet. How are the assets of the plan structured? Finally, there is a third policy issue: investment management policy. How will the asset mix policy be implemented? Alpha's management realizes that there are choices to be made. These can be seen in Figure 2, which shows the relationship between rates of return of alternative asset allocation policies. The company can have anything from a return maximization policy to a risk minimization pol- icy. With the return maximization policy, the asset mix required to achieve a long-term 5 per- cent (or even higher) real rate of return is, of course, a 100 percent equity policy. At the other extreme, the risk minimization policy is a policy FIGURE ! Alpha Corporation (dollars in billions) MAIN BUSINESS BALANCE SHEET Assets Liabilities Short term $1.2 Short term $0.6 Long term 1.2 Long term 0.8 $2 4 $1.4 Equity Shareholders equity $1.0 PENSION PLAN BALANCE SHEET Assets Liabilities Short term $0 3 Short term Long term 1.0 Long term $1.3 Equity Asset cushion $0.3 0.6 $0.9 ;ource' Kelth P. Ambachtsheer & Associates 24 FIGURE 2 Asset mix policy anc[ long-term return targets Realistic 6 real return targets (percent) 5 With For With return broad r~sk maximizahon market minimization focus portfolio focus 5ourc~. Keith P. Ambachtsheer & Assocmtes of matching asset cash flows to liability cash flows as closely as possible. The middle point is to hold essentially the portfolio that the aggre- gate of all investors hold, which should over the long term achieve the same kind of real rate of return as the real economy does in terms of growth rate. That might be a 2.5 percent mid- point. The return projections are converted to im- plicit long-term contribution rates in Table 1. Earlier it was pointed out that pension costs are independent of the return on plan assets. We can see that on a best-estimate basis, a 5 percent long-term real rate of return is consistent with a long-term corporate contribution rate of 6 percent of the payroll (that is, a money transfer from the principal business). That figure in- cludes updating pension costs for 60 percent of the CPI inflation rate. A striking range in funding costs can occur depending on which TABLE 1. Long-term economic assumptions and normal plan costs Long-term Long-term normal cost Asset class real return (percent of payroll) Stocks 5% 6% Real estate 4 7 Long bonds 2 T-bills .5 14 Source' Ke~th P Ambachtsheer & Associates asset mix is chosen, resulting in a potential long- term contribution rate anywhere from 6 percent to as high as 14 percent of payroll. How does one choose? If this is the only information, of course, the choice is obvious. Go for the low contribution rate option. But this decision is not that easy to make. The reason is that sensible asset allocation decisions demand a time context. For pension plans, the time frame can be truly long term--probably in the neigh- borhood of about 20 years. But one cannot ignore concerns about shorter-term contribution rate variability; for example, what might the contri- bution rate have to be three to five years from now? When you take a balance sheet perspec- tive, the concern becomes the $400 million asset cushion. Will it increase or decrease? Fluctua- tions in this asset cushion could be quite signifi- cant over a three- to five-year period. So the fundamental issue in long-term asset allocation is the question of trading off long-term good news against the possibility of shorter-term bad news. You can look at this dilemma in a contribution rate context, where lower long-term contributions must be balanced against greater shorter-term contribution uncertainty. Alterna- tively, you can look at it in a balance sheet context and seek higher long-term asset cushion growth against shorter-term cushion size uncertainty. The good news is that the long-term contribution rate, including updating pensions in pay for in- flation, implies a 6 percent contribution rate. FIGURE 3 The future contribution rate trade-off range The R~turn Maximizing Policy The Risk Minimizing Policy Good news Bad news No good or bad news Long-term contributions are 6% of payroll Shorter-term contributions could be over 15% of payroll in deflation Source. Ke~th P. Ambachtsheer & Asso~ates. Contribuhons will not vary far from 10% of 12ayroli in short term or long term The bad news is that if you end up with an extended period, perhaps three to five years, when equities do very poorly, your contribution rate could go up to 15 percent three to five years later. In the risk minimizing policy, there is no good news or bad news by definition. Contribu- tions will not vary far from 10 percent of payroll in the short or long term (see Figure 3). Many corporate managers, when they are shown these alternatives, prefer to look at the trade-off in the balance sheet context. The good news with the return maximizing policy is that continued cushion growth means eventual lower corp6rate contributions or higher benefits in the long term. The bad news is that the cushion can disappear in the shorter term. With the risk minimizing policy, there is again no particularly good news or bad news; cushion size is highly predictable (see Figure 4). THE ROLE OF FUTURE THINKING IN ASSET ALLOCATION Given the framework in which to make asset allocation decisions, the next element in the deci- sion process is dealing with an uncertain future. In this context, dealing explicitly with uncer- tainty is crucial. Over time, we have developed two approaches to dealing with that future. The first method, the Monte Carlo simulation, as- sesses the future through a series of probability distributions for stock returns, bond returns, cash returns, and inflation returns and how they are expected to be correlated. From those expec- tations, a computer can develop hundreds or FIGURE 4 The future asset cushion trade-off range The Return Maximizing Policy The Risk Mimmizmg Policy Good news Bad news No good or bad news Continued cushion growth means lower costs 0~' higher benafits in long term Cushion disappears in deflation Source: Ke~th P. Ambachtsheer & Associates. Cushion size highly predictable 26 thousands of possible outcomes by drawing out these distributions. The output, as illustrated in Figure 5, shows contributions as a percent of payroll on the left scale and different weight- ings of equities as opposed to bonds across the bottom. When moving from low risk to higher risk, the percent of equities in the portfolio and the potential contribution rate increases and the 50th percentile contribution rate decreases. The other technique, called the scenario ap- proach, is a little more hands-on and allows you to put a little more color into the analysis. The four possible scenarios are summarized in Table 2. To relate the scenario approach to dealing with an uncertain future, again consider Alpha Corporation. The implications for the Alpha pension plan contribution rate depend on the firm's asset allocation policy. For example, in the deflation case, with the return maximizing policy, the results are very unpleasant with the projected contribution rate going way up (see Figure 6). Figure 7 presents the same results in a plan balance sheet (asset cushion) context. DECISIONS AND THEIR IMPLEMENTATION After determining the context and allowing for an uncertain future, the decision and implemen- tation steps follow. In the case of Alpha Corpora- tion, management decided that, as much as pos- sible, they wanted to maximize long-term real FIGURE 5 Distributions of contributions as a percent of payroll for various bond/stock asset mixes Contributlonslpayroll (percent) (inverted scale) 9- 10 6.3 ! ?.0 7.7 8.2 8.7 6.1 6.8 7.6 8.1 8.6 5.3 6.6 7.3 8,1 8.7 5.1 6.4 7.3 8.1 8.7 4.5 6.0 7.2 8.2 8.8 4.1 5.6 7.1 8.3 9.1 I I I I I I A B C D E F 0% 20% 40% 50% 70% 90% Percentile 90th 75th 50th 25th 10th Percent in equities Source: Bankers Trust Company. 27 TABLE 2. Possible three- to five-year capital market environ.men~s Good ~imes: Steady output and produc~vity growth, low inflation rate, positive expectations about the future. Stagflation: Uneven output and productivity growth, infla- ~on rate in the 6-9 percent range, confused expectations about the future. Rising inflation: Uneven output and productivity growth, debt monetization, inflation rate nsmg above 10 percent, dynamics expected to continue Deflation: Output stagnant to falling, inflation Iow and fail- ing, future looks bleak. The capital markets Stocks do very well, followed by real estate and long bonds. Short-term investments left far behind. No predictable differences in asset class r. eturns. Real estate and other tangibles do best, stocks and short- term investments mediocre, long bonds worse by far. High-quality tong bonds best by far, cash mediocre, real estate poor, stocks worst. Source. Kelth P. Ambachtsheer & Assooates. return, subject to "being whole" (that is, having a positive asset cushion in the worst case out- come) on a three- to five-year basis. Using this particular set of numbers, being whole is achieved through a significant real estate posi- tion, which is the hedge against the rising infla- t'ion outcome, and a significant long-bond posi- tion, which is the hedge against the deflation outcome. The purpose for those two positions is risk control. The "basic" asset mix in ~ 8 results. -- A lot of interesting things can be done with the basic asset mix policy to improve its risk- reward characteristics. See the expanded asset mix policy in Figure 8. One aspect of the ex- panded asset mix policy is the shift account. That account can be in domestic stocks, long bonds, or T-bills. It is funded from the basic policy by taking 10 percentage points of the long- bond position and 10 percentage points of the stock position and putting them into a shift ac- count. This is a good way to shift the focus from an asset allocation approach that concen- trates on risk control to one that concentrates FIGURE 6 Sixth-year contribution rate projections Return maxim~zahon w~th "bad news" Percent -15 - -10 -5 0 5 10 -- 15 20 - Source' Ke~ih P Ambachtsheer & Assooates. Good times Deflation Return Risk maximization mlnlm~zahon no "bad news" Good times Deflation Deflation I [ Rising inflation 28 FIGURE 7 Five-year asset cushion swing projections Return Return Risk maximization maximization m,mm~zataon with "bad news" no "bad news" Good times Deflahon $ billions 1.6 ~ 1.2 0,8 0,4 -0.4 -0.8 Source- Ke~th P. Ambachtsheer & Assooates. Good hmes Deflation Deflation Risang inflation on value added. When moving to a shod-term asset allocation focus, however, the investment context changes very significantly. RISK CONTROL AND RETURN ENHANCEMENT Short-term asset allocation is a process that fo- c-~-,~.s not on risk control but on return enhance- -~ent. The objective is to add return to an other- wise passive situation. In the short-term asset allocation context, we have to raise the question as to whether there are correlations between yesterday, today, and tomorrow. In other words, are there relationships between past and present observable events and future stock prices and interest rates? How you answer that question is absolutely crucial to whether you should engage in short-term active asset alloca- tion. If you believe there are no such linkages, stop now. If you believe there are linkages that can be identified, then there is a place for active asset allocation. Table 3 shows the five bases that have to be touched to be effective, to maximize the chances of success. Step 1: What do you believe about the relationships between events yester- day and today and stock prices and interest rates tomorrow? In other words, what is your theory of capital assets pricing? You must consider that issue and think about your belief of how inflation expectations will affect bond prices, the impact of real GNP growth on stock prices, and so forth. This leads to Step 2: Identify the predictors that are important to security valuation. In effect, what can I observe today that will affect the valuation factors tomorrow? And you might say, for example, "Well, I think monetary policy to- day might affect inflation expectations tomor- row,'' thereby identifying that particular pre- dictor. Step 3: What is the relationship between the predictors and the variables that you are trying to forecast (future stock prices and interest rates)? In Step 3, you are actually th~nking, for example, about how sensitive changes in mone- tary policy are today to changes in inflation ex- pectations tomorrow--which, in turn, will im- pact stock and bond prices tomorrow. Step 4: Establish your decision rules. You can't just stop with the output of the process in terms of expec- tations about the future; you have to act on them and you have to act on them in an intelligent way. Decision rules used by money managers are often quite naive. For example, managers often do not consider that they might not be perfect at making capital markets forecasts. Your actual decision rules depend very much on how good you thir~k your forecasts are and on the transaction costs. These are the real world issues ,{ 29 As the Commission is aware, the Police and Fire Pension Board has requested that the City Code be amended to allow them to invest up to 70% of their assets in common stock. The Code currently limits this type of investment to 60%. In the backup is a letter from Jeff Keating in which he recommends the current limit of 60%, together with three articles on asset allocation strategies. After reading these articles, I think we should approve increasing the limit on stocks to 70%. Based on almost 200 years of historical stock market performance, this would give greater growth over the long term. However, increasing the limit to 70% would increase the volatility of the asset value in the Police and Fire Pension Fund. As a result, our exposure to year-to-year jumps in our required contribution would be increased. But, our total contributions over the years should be less. The Commission just needs to decide whether the long term decrease in our required contributions is worth the increased exposure to wide swings in our required contribution. MEMORANDUM TO: FROM: SUBJECT: DATE: MAYOR AND CITY COMMISSIONERS CITY MANAGER ~ AGENDA ITEM /~0.~.. REGULAR MEETING OF FEBRUARY 6, 2001 ORDINANCE NO. 16-01 (CHANGING INVESTMENT PARAMETERS IN THE POLICE AND FIRE PENSION PLAN) FEBRUARY 2, 2001 As the Commission is aware, the Police and Fire Pension Board has requested that the City Code be amended to allow them to revest up to 70% of their assets in common stock. The Code currently limits this type of investment to 60%. In the backup is a letter from Jeff Keating in which he recommends the current limit of 60%, together w~th three articles on asset allocation strategies. After reading these articles, I think we should approve increasing the lin-ut on stocks to 70%. Based on almost 200 years of historical stock market performance, this would give greater growth over the long term. However, increasing the lirmt to 70% would increase the volatility of the asset value in the Police and F~re Pension Fund. As a result, our exposure to year-to-year jumps in our required contribution would be increased. But, our total contributions over the years should be less. The Commission just needs to decide whether the long term decrease in our required contributions is worth the increased exposure to wide swings m our required contribution. ReiSAgmemol4.Ord.16-01.Police & Fire Pension Plan DATE: TO: FROM: SUBJECT: Writer's D~rect Line 561/243-7091 MEMORANDUM January 25, 2001 City Commission David Harden, City Manager Susan A. Ruby, City Attorney Ordinance Changing Investment Parameters in the Police and Fire Pension Pension Plan Enclosed is an ordinance that provides the Police and Fire Pension Plan Board of Trustees may value its aggregate stock investments using the market value method instead of cost, and provides that value of the stock may not exceed 70% (instead of 60%) of the funds assets measured quarterly. It also provides for rebalancing of the Funds Assets. Please place this ordinance on the February 6, 2001 City Commission agenda for first re~ SAR:ci Enclosure CC: Alison Harty, City Clerk Bill Adams, President, Police and Fire Pension Board Steve Cypen, Esq. James W. Linn, Esq. ORDINANCE NO. 16- 01 AN ORDINANCE OF THE CITY COMMISSION OF THE CITY OF DELRAY BEACH, FLORIDA AMENDING CHAPTER 33, "POLICE AND FIRE DEPARTMENTS", SUBHEADING 'PENSIONS", OF THE CODE OF ORDINANCES OF THE CITY OF DELRAY BEACH, FLORIDA BY AMENDING SECTION 33.66, "FINANCES AND FUND MANAGEMENT," SECTION 33.66(E)(2)(b) TO PROVIDE FOR A CHANGE IN THE INVESTMENT PARAMETERS FOR COMMON OR CAPITAL STOCK; PROVIDING FOR REBALANCING; PROVIDING A GENERAL REPEALER CLAUSE, A SAVING CLAUSE, AND AN EFFECTIVE DATE. WHEREAS, the City and the Board of Trustees of the Police and Fire Retirement Plan, desire to change the parameters of investments in common or capital stock to enhance the performance of the Retirement Fund while controlling the risk. NOW, THEREFORE, be it ordained by the City Commission of the City of Delray Beach, Florida as follows: Section 1. That Section 33.66, "Finances and Fund Management", Section E(2)(b) be and is hereby amended to read as follows: (b) The Board of Trustees shall not invest more than five percent (5 %) of Trust Fund i-ts assets in the common or capital stock of any one issuing company, nor shall the aggregate investment in common or capital stock in one company exceed five percent (5%) of the outstanding common or capital stock of that company; nor shall the aggregate of the Fund's investments in common or capital stock at cost using the market value method exceed s-bag seventy percentw,-t~n e,~jx (70%) of the Fund's assets, as measured on the last business day of each quarter. The Board of Trustees shall rebalance the Fund's assets by the last business day of the following quarter. Section 2. That all ordinances or parts of ordinances in conflict herewith be and the same are hereby repealed. Section 3. That should any section or provision of this ordinance or any portion thereof, any paragraph, sentence, or word be declared by a Court of competent jurisdiction to be invalid, such decision shall not affect the validity of the remainder hereof as a whole or part thereof other than the part declared to be invalid. Section 4. second and final reading. That this ordinance shall become effective upon its passage on PASSED AND ADOPTED in regular session on second and final reading on this the ~ day of 2001. MAYOR ATTEST: City Clerk First Reading Second Reading - 2 - Ord. No. 16-01 TING INWESTMENT COLTNSELORS INC. REGISTERED INVESTMENT ADVISORS FOR INDIVIDUALS, RETIREMENT PLANS, TRUSTS AND FOUNDATIONS September 8, 2000 Milene Walinski City of Delray Beach Finance Department 100 NW 1 st Avenue Delray Beach, FL 33444 Dear Milene: Enclosed are two articles on the issue of the optimum asset allocation between stocks and bonds for Pension Plans or long term investors. Very simply, the optimum mix for a Pension Plan is considered to be in the range of 60% in equities, with 40% in fixed income. Historically, this balance has captured the vast majority of being invested 100% in equities with 40% less risk as measured through price volatility. The major issue for the city is that even though the police and fireman's pension plan has a long longevity, in essence, you are always looking at 10 years or longer in terms of expected investment returns, the plan is structured such that it impacts the city's finances on a year-to-year basis based on the performance of the plan. Thus, the city should be looking at an asset allocation that is more appropriate to an investor that has more of a short term orientation. The article by Jeremy Siegel would suggest a more balanced 50%/50% structure between stocks and bonds. Hopefully the enclosed articles will be of some assistance to you. If they don't, I have a couple of other articles, but they are even more esoteric than these I have enclosed. If you have any further questions, please feel free to call nay office. Truly yours, '~ f~!~--',Ke at[ng President; JJK:lfs Enclosures 777 EAST AI'LANTIC AVENUE, SUI[E303 · DELRAYBEACH, FLORIDA 33483 · (561)278-7862 · FAX (561)265-1055 FOR THE 3EREMY J. SI EL Second Edition STOCKS FOR THE LONG RUN The Definitive Guide to Financial Market Returns and Long-Term Investment Strategies JEREMY J. SIEGEL Professor of Finance-- the Wharton School of the Un iversity of Pen nsylva nia ('~' I' / McGraw-Hill New York San Francisco Washington, D.C. Auckland Bogot~ Caracas Lisbon London Madrid Mexico Clty Milan Montreal New Delhi San Juan Singapore Sydney Tokyo Toronto 2 R~sk, Return and the Coming Age Wave 35 A second reason for the increase in correlation between stock and bond returns is the strategy that portfolio managers follow to al- locate assets. Most tactical allocation models, which money managers use to minimize the risk and maximize the return of a portfolio, dic- tate that the share of a portfolio that is allocated to stocks be a func- tion of the expected return on stocks relative to that on bonds. As interest rates rise, causing stock prices to fall, prospective bond re- turns become more attractive, motivating these managers to sell stocks. As a result, stock and bond prices move together. This is an ex- ample of how the actions by portfolio managers trying to take advan- tage of the historical correlation between stocks and bonds changes their future correlation. EFFICIENT FRONTIERS:~ Modern portfolio theory describes how to alter the risk and return of a portfolio by changing the mix between assets. Figure 2-6, based on the nearly 200-year history of stock and bond returns, displays the risks and returns that result from varying the proportion of stocks and bonds in a portfolio. The square at the bottom of each curve represents the risk and re- turn of an all-bond portfolio, while the cross at the top of the curve rep- resents the risk and return of an all-stock portfolio. The circle indicates the minimum risk achievable by combining stocks and bonds. The curve that connects these points represents the risk and return of all blends of portfolios from 100 percent bonds to 100 percent stocks. This curve, called the efficient frontier, is at the heart of modem portfolio analysis and the foundation of asset allocation models. Investors can achieve any combination of risk and return along the curve by changing the proportion of stocks and bonds. Moving up the curve means increasing the proportion in stocks and correspondingly reducing the proportion in bonds. For short-term holding periods, mov- ing up the curve increases both the return and the risk of the portfolio. The slope of any point on the efficient frontier indicates the risk-return trade-off for that allocation. By finding the pomts on the longer-term ef- ficient frontiers that equal the slope on the one-year frontier, one can de- termine the allocations that represent the same risk-return trade-offs for all holding periods. 3 Th~s section, which contains some advanced material, can be skipped w~thout loss of connnu~ty of the for the event. 36 PART 1 The Verdzm of HlstoD, FIGURE 2-6 Risk-Return Trade-Offs for Various Holding Periods, 1802-1996 fessor . Week the cos market mbst cc case fa necessi ten--am one Of: historyl Now breakin the Lan1 clusion; power safer ! more with tJon thai · A tho and c ~ove seven · How t chanc. knowl a'w'n · The a~ stock ~ investi con el' · How invesfr your n · Analy. Jsorl$ 8% 6% 2%. 1%' 30 yr ~a h " Jo Mlmmum Risk o% 2% 4% 6% I .... I I " i 10% 12% 14% 18% 18% 20% RECOMMENDED PORTFOLIO ALLOCATIONS Table 2-2 indicates the percentage of an investor's portfolio that should be invested in stocks based on both the risk tolerance and the holding period of the investor.4 Four classes of investors are analyzed: the ultra- conservative investor who demands maximum safety no matter the re- turn, the conservative investor who accepts small risks to achieve extra remrv,, the moderate risk-taking investor, and the aggressive investor who is willing to accept substantial risks in search of extra returns. The recommended equity allocation increases dramatically as the holding period lengthens. The analysis indicates that, based on the histor- 4 The one-year proportions (except rrammum risk point) are arbitrary, and are used as bench.marks for other holding periods. Choosing different proportions as benchmarks does not quahtahvely change the following results. CHAPTER 2 Risk, Return and the Comln~ Age Wave 37 TABLE 2-2 Portfoho Allocation. Percentage of Portfoho ~n Stocks Based on All H~storical Data Risk Holdin¢ Period · Tolerance I year 5 years 10 years 30 years Ultra-conservative 7.0% 25 0% 40.6% 71.3% (Minimum Risk) Conservative 25 0% 42.4% 61.3% 89.7% Moderate 50,0% 62,7% 86,0% 112.9% Risk-taking 75.0% 77,0% 104.3% 131.5% ical ret'urns on stocks and bonds, ultra-conservative investors should hold nearly three-quarters of their portfolio in stocks over 30-year hoiding pe- riods. This allocation is justified since stocks are safer than bonds in terms of purchasing power over long per]ods of time. Conservative investors should have nearly 90% of their portfoho in stocks, while moderate and aggressive investors should have over 100 percent in equity. This alloca- tion can be achieved by' borrowing or leveraging an all-stock portfolio. Given these sinking results, it might seem puzzling why the holding period has almost never been considered in portfolio theory. This is be- cause modern portfolio theory was established when the academic pro- fession beheved tn the random walk theory of secun ,t-y prices. As noted earlier, under a random walk, the relative risk of securities does not change for different time frames, so portfolio allocations do not depend on the holding period. The holding period becomes a crucial issue in port- folio theory when data reveal the mean reversion of the stock returns.$ 5 For a similar conclusion, see Nicholas Barber]s, "Investing for the Long Run When Returns Are Predictable," working paper, University of Chicago, July 1997 Paul Samuelson has shown that mean reversion wdl increase equity holdings if investors have a r]sk aversion coefficient greater than unity,, which mo~t researchers fred is the case See Samuelson, "Long-Run Risk Tolerance When Eqmty Return.', Arc Mean Regr~_.'~mg Pseudoparadoxe~ and Vmdicatmn of 'Busincssmen's Ribk'" ~n WC Bramard, W D Nordhau.., and } l W Watt.., ed~, Money, ,klacroct,mmm~, and Pubhc Pohcy, Cambridge, Mass The MIT Press. 1991. pp 181-200 See also Zvi Bodie. Robert Merton. and William Samuelson, "Labor Supply Flexibility. and Portfolio Choice tn a Lifecycle Model." Journal of Economic Dynamtcsat~d C(mtrol, vol 16, no 3 (July/October 1992). pp 427-450 Bodie, et al have shown that eqmty ho[drags can vary. v.,~th age because stock returns can be correlated with labor income Seminar Proceedings Investment Policy April 18--20, 1994 Tokyo, Japan Keith P. Ambachtsheer Charles D. Ellis, CFA John L. Maginn, CFA David F. Swensen John R. Thomas, CFA Donald L. Tuttle, CFA, Moderator Edited by Jan R Squires, CFA ..~IMR Sponsored by the Association for Investment Management and Research and the Security Analysts Association of Japan A Tale of Two Pension Funds: Case Analysis (A) In making strategic asset allocation rec- ommendations for their pension funds, Jim Jones and Barbara Smith face a prob- lem shared by every pension fund man- ager: how to maximize the return on the portfolio for an appropriate level of risk. To find the solutions to this problem, Jones and Smith must keep in mind the pension fund balance sheet (Exhibit 1), which will dictate to a significant degree the risk-return trade-off for each fund. Both investment managers believe that their investment strategies should be based on a going-concern view of their pension plans As going concerns, the pension funds must provide for the impact of future wage and inflation ex- periences on pension fund liabilities. Jones and Smith agree that the ~mmumz- mg unvestment strategy would combine some fixed-rate bonds with inflation-in- dexed bonds and, possibly, a small eq- mty position to hedge against unknown future trends un wages. Deciding the funds' appropriate long-term invest- ment strategies, however, requires ad- dressing t~vo further issues First, the U S. Treasury ~s not yet issuing inflation-indexed bonds. So, what subsntute asset(s) or alternative strategies can be used for an mflatlon hedge~ Second, how much incremental re- turn should each fund seek relative to the return produced by the lmmumzmg strategies? Any incremental return real- ized wdl reduce the contributions re- quired to finance the prom~sed benefits under the plans Because Alpha Corpo- ration Is a private-sector corporation seekung to maximize shareholder wealth and the Public Sector Retirement System (PSRS) is sponsored by a state already committed to funding a previous contn- buhon shortfall, both plans would wel- come an investment policy that would reduce their required contributions. Thus, Jones and Smith face the difficult task of improving the investment per- formance of their pension funds in the real world of uncertain future asset-class returns and uncertain future economic conditions shaping those returns. Developing asset allocation recom- mendations that will earn a return be- yond the ideal immunizing strategy re- quires a forecast of the expected returns for different asset classes under different economic conditions, and a determina- tion of the asset allocation combination that appears to offer the best return op- portunities relative to the immunizing strategy. Jones has started this process by developing the forecasts of asset- class returns for 1994 through 1998 given in Table 2 of the A case. To com- plete the analysis, three questions must be addressed. · How reahst]c are Jones's esti- mates of the returns tot the 1994- 98 period? · If the estunates are reahshc, what are the estimated returns for dif- ferent portfolio compositions in the three different economic sce- narios for the 1994-98 period? · What asset allocation strategies should Jones and Smith recom- mend to their boards of directors or trustees? The historical returns of different as- set classes m different economic condi- tions (Table 1 of the A case) can be used to evaluate Jones's prelected returns. To use th~s historical information effec- tively, Smith and Jones need to catego- rize the different historical periods in line with the definmons for the three prelected scenarios--degearmg, infla- tion, or deflation. In terms of economic growth and inflation, the 1952--65 period corresponds most closely to the degear- lng definition, the 1928--t0 period to the deflation definition, and the 1966-81 pe- riod to the inflation defimtion. Note, however, that the expected 74 GN s/ri: rea i ben 52: pur~ et l and the bdl'~ e'.,p~ vet,. h~g~ tlat~ dre~ orca stoc rlsir' ben, c, Ct?p the set-t h~stt retu late the~ stoc enc~ sigr nent per- tds in the sset-class ~onoraic CIS. n recom- .'turn be- ategy re- J returns different termuna- btnation turn op- ~unizing process ~f asset- gh 1998 To corn- ns must ¥s esti- 1994-- lC, what for dif- 'ions in 13.1C sce- iod~ ategies recom- rectors -ent as- condi- e used -ns To effec- atego- ods tn three infla- nomic ~enod -'gea r- to the 51 pe- ected economic conditaons for 1994-98 do not closely match any of the past periods in Table 1. For example, in the 1952--65 period, the economy was characterized by higher economic growth (3.3 percent GNP growth) and lower irfflation (1.4 percent) than is projected for the degear- ing scenario. The degearing scenario as- sumes that the U.S. economy will grow slowly during the decade (2.5 percent increase in annual GNrp growth) as Fed- eral Reserve tight-money policies hold down the level of inflation (to 3.0 per- cent annually) and the United States works its way out of a debt overhang and large federal government deficits. Given its assumptions as to real GNP growth and inflation and the 1994 structure of interest rates and dividend yields, the degearing scenario projects real returns for bills of 1 5 percent, for bonds of 3 25 percent, and for stocks of 5.25 percent. The comparison 1952-65 period experienced real returns for bills of 1.3 percent, for bonds of-0.2 percent, and for stocks of 13.1 percent. Although the actual and projected experiences for btlls are similar, the actual and projected experiences for bonds and stocks are very different. The difference lies in the higher economic growth and lower in- flat,on of the 1952--66 period, which drew money tnto stocks and thus in- creased stock multiples and improved stock returns; at the same time, bond y~etds were low in the early 1950s and rising interest rates worked to depress bond returns. After adjustments are made for these differences, the returns projected in Table 2 for the degearing scenario appear realistic. Similar differences exist between the projected economic forecasts and as- set-class returns for the deflation and the tnt'la tlon scenarios and their comparable historical periods. Overall, Jones's real- return projections for the period of the late 1990s show bill returns, similar to thmr historical experiences but gener- ally better bond returns and poorer stock returns relative to their historical experiences. Why would Jones project experi- ences for the future that appear to be significantly different from the experi- ences of the past? The answer to this questions lies in his assumption that a stock market correction will occur if, in- stead of degearing, either a deflation or inflation scenario occurs. Jones appar- ently assumes that the deflation sce- nario, particularly if started by a col- lapse of the Japanese economy; would result in substantially increased fears about the future of the global economy, a lowering of expectations for corporate profits, and increased fears about the outlook for stocks. This outcome would generate an increase in the risk premium on stocks, resulting in negative returns on stocks in the 1994-98 period. In con- trast, bond returns would be erdaanced as bond prices rose with a drop in long- term interest rates accompanying the switch to a deflation scenario. Jones is apparently suggesting that, in an infla- tion scenario, higher capitalizataon-rate requirements would, again, lower the returns to stocks because of stock price declines. Bond prices and returns would suffer even more than stocks in this scenario. Disagreement with Jones's assump- tions regarding the market reaction to the deflation and inflation scenarios and the subsequent impact on stocks would lead to projected returns more attuned to h~storical averages. His assumption of a market correction appears reason- able, however, in light of the long bull market reflected in the returns to stocks in the 1982-93 period and the market's historically adverse reaction in short periods to unexpected deflationary and irdlat~onary environments as the market seeks new equilibrium levels. Assuming that the projected returns in the different scenarios are reasonable, Smith and Jones can calculate reason- able returns for each of the four alterna- tive portfolios in the three scenarios. The portfolio returns are the weighted averages of the projected asset-class re- turns for each economic scenario. The weightings would be the percentages each asset class represents in the particu- lar portfolio. The weighted real returns of particular portfolios in the three eco- nomic scenarios are given in Table A. Cash was given no weighting in any of 75 Table A. Weighted-Average Real Returns of Selected Portfolios by Economic Scenario Asset M~x Degeanng Deflation tnt'tat;on 70% equity/30% debt 4.65% -4 35% -2 60% 60% equity/40% debt 4.45 -2.80 -2 $0 50% equity./50% debt 4.2.5 -1 25 -3 00 40% equity/60% debt 4.05 0.30 -3 20 the portfolios suggested for analysis and is not included h~ any of the portfolio descriptions. The different portfolios exakibit sig- rffficantly different return performance in the three economic scenarios. In the degearing scenario, an equity-heavy portfolio performs slightly better than a debt-heavy scenario, but in a deflatio/~ scenar2o, a debt-heavy portfolio per- forms significantly better than an eq- uity-heavy portfolio. Although none of the portfolio mixes performs well in the inflation scenario, an equity-heavy port- folio does somewhat better than a debt- heavy portfolio. These return differences have sig- nificant ~nplications for the asset-mix recommendations Jones and Smith should make to their boards. If, as Smith apparently believes, a good possibility exists that the degearing scenario could deteriorate Lnto deflation, then the port- folio should be debt heavy. TI-us strategy would not significantly reduce returns in the degearing scenario but would provide significant protection against the downside of deflation. In the i,rffla- tion scenario, a shading toward equity should produce the best (or least worst) results. If the L~lation scenario is a vi- able possibility, the only hedge available would be a sizable allocation of the port- folio to cash equivalents (T-bills). Based on the three possible scenar- ios, which of the four portfolio alterna- fives evaluated in Table A should be rec- ommended for the Alpha Corporation fund and the PSRS fund? Alpha Corporation's current asset mix of 70 percent equity/30 percent debt reflects the company's ability and will- i~gness to assume larger risk than the PSRS. If the asset-mix strategy does not produce the desired real-return results, Jones would have to inform the board of directors that a larger future contribu- tion to the corporation's pension plan is required to provide its future defined benefits. The result would be a reduc- tion of corporate cash available for other purposes for a trme. In conclusion, given the possible capital market envi- ronments for 1994 through 1998, and given Jones's view that inflation is more likely than deflation, Alpha Corporation is likely to maintain its 70 percent eq- uity/30 percent debt strategy. The PSRS's current asset mix of 50 percent equity/50 percent debt reflects its concerns about unfunded future li- abilities and its resulting greater risk aversion (relative to Alpha Corpora- tion). If the PSRS's strategy does not produce the desired real returns, the sys- tem must request funds from the state government, which revolves rinsing taxes or cutting expenditures, and from the employees, which mvolves addi- tional contributions from their salaries. Neither necessity would be painless. Therefore, given the possible 1994--98 environments and given Smith's view that deflation is more likely than L~a- t-ion, the PSRS might find a 10-percent- age-point sl'uft from equity to debt desir- able. The resulting mi.x, 40 percent eq- uity/60 percent debt, would be rela- tively conservative. In addition, be- cause of the i.rfflation sensitivity of the plan's liabilities, the PSRS should con- sider shortening the duration of the debt portfolio. 76 et Allocation for Institutional · ,1OS San Francisco, California ~February 10-11, 1986 'Edited by Michael D. Joehnk, CFA P. Ambachtsheer Robert D. Amott Gary P. Brinson, CFA Kathleen A. Condon, CFA ;H. Gifford Fong William L. Fouse, CFA Walter R. Good, CFA Sponsored by The Institute of Chartered Financial Analysts Mark P. Kritzman, CFA Greta E. Marshall, CFA Robert H. Litzenberger Douglas A. Love Roger F. Murray Claude N. Rosenberg, Jr. William F. Sharpe Strategic Approaches to Asset Allocation Keith P. Ambachtsheer When considering the subject of "strategic ap- proaches to asset allocation," one should first ask what is meant by strategy, tactics, and the many other descrzptors that are used in the area of asset allocation. Why is asset allocation such a foggy concept? Why are there such conceptual difficulties in this area? It has a lot to do with a confusion between means and ends. The defi- nition I offer is: Asset allocation is the means to achieve an appropriate package of risk, return, and hme through the three steps of context un- derstanding, future thinking, and decision im- plementation. ASSET ALLOCATION: THE LONG- TERM CONTRACT There are two basic contexts m which to study asset allocation, and they should be kept sepa- rate. Context number one is long-term asset allo- cation. To illustrate long-term asset allocation, consider the case of the hypothetical Alpha Corporation. They have made that conceptual breakthrough, and they think of their defined benefits plan as a financial business. As such, they recognize that understanding and manag- ing the plan balance sheet is the appropriate way to proceed. As seen in Figure 1, Alpha Corporation is a billion dollar company with a fairly normal looking principal business balance sheet. But Alpha uses the extended balance sheet concept, and as shown on the pension plan balance sheet, the assets and liabilities are very significant rela- tive to the main business assets and liabilities. Assets in the pension plan are stated at the mar- ket, and liabilities are stated on a "best-estimate going-concern basis." In other words, they use economic assumptions on a best-estimate basis about return and inflation. The going-concern part means that it is important to recognize, for the final earnings plans at least, that the liability is based on final earnings, not on the level or: earnings. Alpha Corporation has a signif- icant asset cushion in the pension plan of $400 million, which in relation to the best estimate of liabilities, amounts to a 44 percent asset cush- ion. When analyzing the pension plan balance sheet, a number of policy issues need to be ad- dressed. First is funding policy, or asset cushion policy. Of concern here is the relationship be- tween plan assets and projected plan liabilities accrued to date. Should there be a large cushion or a small cushion and why? The second issue is asset mix policy, which deals with the left side of the balance sheet. How are the assets of the plan structured? Finally, there is a third policy issue: investment management policy. How will the asset mix policy be implemented? Alpha's management realizes that there are choices to be made. These can be seen in Figure 2, which shows the relationship between rates of return of alternahve asset allocation policies. The company can have anything from a return maximization policy to a risk minimization pol- icy. With the return maximization policy, the asset mix required to achieve a long-term 5 per- cent (or even higher) real rate of return is, of course, a 100 percent equity policy. At the other extreme, the risk minimizahon policy is a po[icy FIGURE 1 Alpha Corporation (dollars m bdhons) MAIN BUSINESS BALANCE SHEET Assets Liabilities Short term $1 2 Short term $0 6 Long term 1.2 Long term 0 8 $2.4 Sl.4 Equity Shareholders equity. Sl.0 Assets Short term Long term PENSION PLAN BALANCE SHEET Liabilities $0 3 Short term $0 3 1 0 Long term 0 6 Sl .3 $0.9 Equity Asset cushion $0 4 Source Ke~th P. Ambachtsheer & Assooates 24 FIGURE 2 Asset mix policy and long-term return targets Realistic 6 real return targets (percent) 5 1 With For With return broad r~sk maximization market mlnlm~zahon fOCuS portfoho fOCuS Source Ke~th P Ambachtsheer & Asso~ates of matching asset cash flows to liability cash flows as closely as possible. The middle point is to hold essentially the portfolio that the aggre- gate of all investors hold, which should over the long term achieve the same kind of real rate of return as the real economy does in terms of growth rate. That might be a 2.5 percent mid- point. The return projections are converted to im- phc]t long-term contribution rates in Table 1. Earlier it was pointed out that pension costs are independent of the return on plan assets. We can see that on a best-estimate basis, a 5 percent long-term real rate of return is consistent with a long-term corporate contribution rate of 6 percent of the payroll (that is, a money transfer from the principal business). That figure in- cludes updating pension costs for 60 percent of the CPI inflation rate. A striking range in funding costs can occur depending on which TABLE 1. Long-term economic assumptions and normal plan costs Long-term Long-term normal cost Asset class real return (percent of payroll) Stocks 5% 6% Real estate 4 7 Long bonds 2 10 T-bdls 5 14 Sourc~ Ke~lh P Ambachtsheer & Associates asset mix is chosen, resulting in a potential long- term contribution rate anywhere from 6 percent to as high as 14 percent of payroll. How does one choose? If this is the only information, of course, the choice is obvious. Go for the low contribution rate option. But this decision is not that easy to make. The reason is that sensible asset allocation decisions demand a time context. For pension plans, the time frame can be truly long term--probably in the neigh- borhood of about 20 years But one cannot ignore concerns about shorter-term contribution rate variabxlity; for example, what might the contri- bution rate have to be three to five years from now? When you take a balance sheet perspec- tive, the concern becomes the $400 million asset cushion. Will it increase or decrease? Fluctua- tions in this asset cushion could be quite signifi- cant over a three- to five-year period. So the fundamental issue in long-term asset allocation is the question of trading off long-term good news against the possibility of shorter-term bad news. You can look at this dilemma in a contribution rate context, where lower long-term contributions must be balanced against greater shorter-term contribution uncertainty. Alterna- tively, you can look at it in a balance sheet context and seek higher long-term asset cushion growth against shorter-term cushion size uncertainty. The good news is that the long-term contribution rate, including updating pensions in pay for in- flat]on, implies a 6 percent contribution rate. FIGURE 3 The future contribution rate trade-off range The Return Maximizing Pohcy The Risk Mm~m~zmg Pohcy Good news Bad news No good or baa news Long-term contributions are 6% of payroll Shorter-term contributions could be over 15% of payroll in deflahon S~urct Ke]th P. Arnbachtshe~r & Assooates Contribubons will not vary far from 10% ol payroll ~n short term or long term The bad news is that if you end up with an extended period, perhaps three to five years, when equities do very poorly, your contribution rate could go up to 15 percent three to five years later. In the risk minimizing policy, there is no good news or bad news by definition. Contribu- t-ions will not vary far from 10 percent of payroll in the short or long term (see Figure 3). Many corporate managers, when they are shown these alternatives, prefer to look at the trade-off in the balance sheet context. The good news with the return maximizmg policy is that continued cushion growth means eventual lower corp6rate contributions or higher benefits in the long term. The bad news is that the cushion can disappear in the shorter term. With the risk minimizing pohcy, there is again no particularly good news or bad news; cushion size is highly predictable (see Figure 4). THE ROLE OF FUTURE THINKING IN ASSET ALLOCATION Given the framework in which to make asset allocation decisions, the next element in the deci- sion process is dealing with an uncertain future. In this context, dealing explicitly with uncer- tainty is crucial. Over time, we have developed two approaches to dealing with that future. The first method, the Monte Carlo simulation, as- sesses the future through a series of probabdity distributions for stock returns, bond returns, cash returns, and inflation returns and how they are expected to be correlated. From those expec- tations, a computer can develop hundreds or FIGURE 4 The future asset cushion trade-off range The Return Maximizing Policy The R~sR Minimizing Pohcy Good news Bad news No good or bad news Continued cusl3ion growth means lower coats or h~gher benehts in long term Cushion d~sappears in deflation $~urce Ketth P Arabachtsheer & Associates. Cushion size h~ghly predictable 26 thousands of possible outcomes by drawing out these distributions. The output, as illustrated in Figure 5, shows contributions as a percent of payroll on the left scale and different weight- ings of equities as opposed to bonds across the bottom. When moving from Iow risk to higher risk, the percent of equities in the portfolio and the potential contribution rate increases and the 50th percentile contribution rate decreases. The other technique, called the scenario ap- proach, is a little more hands-on and allows you to put a little more color into the analysis. The four possible scenarios are summarized in Table 2. To relate the scenario approach to dealing with an uncertain future, again consider Alpha Corporation. The implications for the Alpha pension plan contribution rate depend on the firm's asset allocation policy. For example, in the deflation case, with the return maximizing policy, the results are very unpleasant with the projected contribution rate going way up (see Figure 6). Figure 7 presents the same results in a plan balance sheet (asset cushion) context. DECISIONS AND THEIR IMPLEMENTATION After determining the context and allowing for an uncertain future, the decision and implemen- tation steps follow. In the case of Alpha Corpora- t-ion, management decided that, as much as pos- sible, they wanted to maximize long-term real FIGURE $ D~stributions of contributions as a percent of payroll for various bond/stock asset mt,xes Contribuhonslpayroll (percent) (inverted scale) 10 6.3 7O 7.7 82 8.7 6.1 6.8 7.6 8.1 8.6 5.3 6.6 7.3 8.1 8.7 7.3 8.1 8.7 4,5 6O 72 8.2 8.8 4.1 56 7.1 8.3 9.1 I I t I I 1 A B C D E F 0% 20% 40% 50% 70% 90% Percentile 90th 75th 50th 25th 10th Percent in ecluihes Source Bankers Trust Company 27 TABLE 2. Possible three- to five-year capital market environments The environment Good hmes' Steady output and productivity growth, Iow inflation rate, post,ye expectations about the future. Stagflat;on: Uneven output and productivity growth, infla- tion rate in the 6-9 percent range, confused expectations about the future Pasmg inflation: Uneven output and productivity growth, debt monet2zat~on, inflatuon rate nsmg above 10 percent, dynamics expected to continue Deflation: Output stagnant to falling, mflat~on low and fall- ing, future looks bleak. The capital markets Stocks do very well, followed by real estate and long bonds ShortIterrn investments left far behind No pred:ctable differences in asset class returns Real estate and other tangibles do best, stocks and short- term investments mediocre, long bonds worse by far High-quality long bonds best by far, cash mediocre, real estate poor, stocks worst. ~ourct Ke:th P. Ambachtsheer & Assocuates return, subject to "being whole" (that is, having a positive asset cushion in the worst case out- come) on a three- to five-year basis. Using this particular set of numbers, being whole is achieved through a significant real estate posi- tion, which is the hedge against the rising infla- tion outcome, and a sigmf~cant long-bond posi- tion, which is the hedge against the deflation outcome. The purpose for those two positions is risk control. The "basic" asset mix in Fi_gu_re 8 results. A lot of interesting things can be done with the basic asset mix policy to improve its risk- reward characteristics. See the expanded asset mix policy in Figure 8. One aspect of the ex- panded asset mix policy is the shift account. That account can be in domestic stocks, long bonds, or T-bills. It is funded from the basic policy by taking 10 percentage points of the long- bond position and 10 percentage points of the stock position and putting them into a shift ac- count. This is a good way to shift the focus from an asset allocation approach that concen- t-rates on risk control to one that concentrates FIGURE 6 Sixth-year contribution rate prolecfions Return maxlmtzatlon wdn "bad news" Percent -15 -- -10 -5 0 5 10 -- 15 20 - $ourc~ Keah P Ambachtsheer & As$ocmte$. Good times Return Rtsk maximization m~mm~zatlon no "bad news" Good times Deflahon Deflation [ [ R~sing inflahon Deflation 28 FIGURE 7 Five-year asset cushion swing projections Return maxtmizatmn w~th "bad news" $ billions 1,6 F / 1.2 0,8 0.4 -0.4 Source Ke~th P Ambachtsheer & Assooates Good times Deflatfon on value added. When moving to a short-term asset allocation focus, however, the investment context changes very significantly. RISK CONTROL AND RETURN ENHANCEMENT Short-term asset allocation is a process that lo- c :- ?s not on risk control but on return enhance- -~ent. The objective is to add return to an other- wise passive situahon. In the short-term asset allocation context, we have to raise the question as to whether there are correlations between yesterday, today, and tomorrow. In other words, are there relationships between past and present observable events and future stock prices and interest rates? How you answer that question is absolutely crucial to whether you should engage in short-term active asset alloca- tion. If you believe there are no such linkages, stop now. If you believe there are linkages that can be identified, then there is a place for active asset allocation. Table 3 shows the five bases that have to be touched to be effective, to maximize the chances of success. Step I: What do you believe about the relationships between events yester- day and today and stock prices and interest rates tomorrow? In other words, what is your theory of capital assets pricing? You must consider that Return R~sk maximization mm~mlzahon no "bad news" Good times Deflation Deflation = Rising inflation issue and think about your belief of how inflation expectations will affect bond prices, the impact of real GNP growth on stock prices, and so forth. This leads to Step 2: Identify the predictors that are important to security valuation. In effect, what can I observe today that will affect the valuation factors tomorrow? And you might say, for example, "Well, I think monetary policy to- day might affect inflation expectations tomor- row,'' thereby identifying that particular pre- dictor. Step 3: What is the relationship between the predictors and the variables that you are trying to forecast (future stock prices and interest rates)? In Step 3, you are actually thinking, for example, about how sensitive changes in mone- tary policy are today to changes in inflation ex- pectations tomorrow--which, in turn, will im- pact stock and bond prices tomorrow. Step 4: Establish your decision rules. You can't just stop with the output of the process in terms of expec- tations about the future; you have to act on them and you have to act on them in an intelligent way. Decision rules used by money managers are often quite naive. For example, managers often do not consider that they might not be perfect at making capital markets forecasts. Your actual decision rules depend very much on how good you thir~k your forecasts are and on the transaction costs. These are the real world issues 29