Uses of Futures and Index Funds

Uses of Futures and Index Funds James R.C. Woodlock Managing Director, Quantitative Division Bare/ays Global Investors Limited Implementing a global asset allocation strategy can be accomplished by purchasing an equities basket, futures, or index funds, but each alternative differs in terms of costs and benefits. Although using futures does not involve trading commission, custody costs, and tax withholding, it does involve an annual roll commission, margin financing costs, and basis risks. In contrast to futures contracts, index funds reduce basis risk and transaction costs and are readily available. As a result, investors could easily achieve global equity market exposure at a low cost and avoid expensive withholding taxes and stamp duties by using index funds. I nvestors can be rather inflexible when thinking about global asset allocation. They usually see global asset allocation only as a matter of Germany versus France or Japan versus the United States, for example. The myriad investment objectives and constraints of today's investment environment demand a more creative view of asset allocation decisions. This presentation examines three possible asset allocation choices-an equity basket, futures, and index funds-and illustrates the construction of a global equity fund. Asset Allocation Choices A U.K. institution intending to buy exposure to U.S. equities could do so via three main vehicles: an equity basket (i.e., an actively managed, diversified basket of u.s. shares), futures, and index funds. The three choices can all achieve the desired asset allocation objectives but differ in terms of costs and complexity. Equity Basket. The total cost to a UK. fund of acquiring a standard U.s. equity basket is about 34 basis points (bps). The bid-ask spread on that basket is about 22 bps. The commission cost is about 9 bps. The custody and marketimpactcosts are small, about 2 and 1 bps, respectively. Suppose a UK. investor bought a basket of equities to replicate the S&P 500 Index. The dividend yield of thes&p 500 is about1.7 percent. For U.K. investors, dividends are withheld at 30 percent, but 15 percent of that can be reclaimed. So, the cost, or the loss of return, is about 0.26 percent, and the net yield for this U.K. investor would be about 1.4 percent. Futures. Futures are a way for the investor to obtain quickly the desired exposure. Compared with an equitybasket, futures offer return and cost advantages, but the investor also must face margin financing, benchmark, and basis risk issues. Return advantage. A futures contract does not receive any dividends; therefore, it does not pay any tax. As a result of that lack of taxation and also the fact that contracts are priced by the domestic investor, the return to a U.K. investor from using a futures contract on the S&P 500 does not suffer that 0.26 percent cost that the equity basket does. So, UK. investors are better off using futures than buying a basket of shares. Costs. For futures, the initial trading commis~ sion is low, about 2 bps, but a futures contract lasts for only three months. Thus, every three months the investor has to roll that exposure forward, which means that the initial trading commission of 2 bps becomes 4 bps, and because that contract has to be bought and sold every three months, over a year, the commission cost eventually rises to 16 bps. The market impact is small, about 1 bp, and fair value misalignment is zero on average. So, the total annual cost is about 19 bps, versus 34 bps for the equity basket. Margin financing. Margin financing is an issue for investors using futures. At the moment, the initial margin is about 3.5 percent of the amount Association for Investment Management and Research 111

Asset Allocation in a Changing World invested. Sometimes futures are used as an overlay strategy, which means that the investor does nothave actual funds to back that margin position and may have to finance the position, thus incurring explicit financing costs. More often, however, futures are used as partofa broader strategy, whereby the investor has assets that can be used to provide the margin. In this case, the investor has an opportunity cost, but the lost T-bill return is partially offset by the deposit interest at the clearinghouse. The net effect is that the investor probably incurs another 2 bps of cost. Benchmarks. Mostmoneymanagersare accustomed to operating against benchmarks. When it comes to the United States, the typical benchmark for a u.k. manager is the U.S. component of the FT/S&P Actuaries World Index, not the S&P 500. But unfortunately, the U.S. component of the FT/S&P does not have a futures contract. So, to gain exposure to the U.S. market using futures, a U.K. manager would have to use the S&P 500 as a proxy. The question is whether the S&P 500 is a good proxy. In onesense, the answeris yes. Figure1 shows the divergence in returns between the S&P 500 and the U.S. component of the FT/S&P. The amount of movement away from each other, as shown in the figure, is generally quite small, suggesting that the two are highlycorrelated. In anothersense, however, they are definitely not the same index. The S&P 500 is somewhat narrow in scope, excluding mid- and small-capitalization stocks and being considerably smaller than the FT/S&P U.s. component's 640-650 stocks. Furthermore, the S&P 500 is exchange based, not market based, and actually contains Dutch companies and Canadian companies, whereas the FT/S&P U.S. component is market based and includes only u.s. companies. Basis risk. The fact that the two indexes are close but not the same leads to basis risk. In fact, the tracking error of the S&P 500 against the U.s. component of the FT/S&P is about 0.57 percent, which, although small, is an extra risk for u.k. investors. This risk might be acceptable given the prospect of higher return, but there is no reason to believe that one of those indexes will continuously outperform the other over the long run. So, one should not have an expectation of a higher return from either index. One way of looking at the impact of the extra risk, without a higher return expectation, is by using an information ratio approach. Assuming an information ratio of1:2, the performance opportunity cost for taking the extra risk of 0.57 percent is about 0.29 percent. Using the futures contract in effect means reducing part of the risk tolerance in the total portfolio, and the cost of that lost risk tolerance, or more specifically the cost of basis risk, is 29 bps. Total costs. The total costs (the profits and losses, so to speak) for the use of futures can now be calculated. A total of 57 bps of profit accrues to the Figure 1. Divergence in Returns: S&P 500 versus U.S. Component of FT/S&P, January 1997-December 1997 0.5,-------------------------------, 0.4 0.3 0.2 S cu 0.1 " ~ 0 B l':: ~ --0.1 ~ -0.2 --0.3-0.4-0.5 L--'--_-'-_---l --'----_--"-.L-_--L._---l --'----_--"-.L.-._---L---' 1/97 2/97 3/97 4/97 5/97 6/97 7/97 8/97 9/97 10/97 11/97 12/97 Source: Datastream International Ltd. 112 Association for Investment Management and Research

Uses offutures and Index Funds use of futures because the total trading commission drops from 31 bps for the equity basket to 2 bps for futures (net 29 bps), the custody cost of 2 bps for the equity basket is eliminated with futures (net 2 bps), and the tax cost of 26 bps incurred with the equity basket does not exist with futures (net 26 bps). The futures position, however, incurs 47 bps ofadditional costs: 16 bps in annual roll commission, 2 bps in margin financing cost, and 29 bps in basis risk. Table 1 shows that the net result is a 0.10 percent advantage for the use of futures compared with a basket of equities. So, futures contracts are still advantageous, but the difference between an equity basket and futures is only 10 bps annually. This finding actually detracts from the futures position because investors using futures will have to continue to pay roll commissions going forward, whereas they would not for the equity basket exposure. Table 1. Benefits and Costs of Futures versus Equity Basket: Net Savings Cost Benefit Lower commission Lower custody Tax benefit Roll commission Margin financing Basis risk Net savings Profit/Loss 0.29% 0-02 0.26-0.16-0.02-0.29 0.10'10 :V Additional factors. Using futures is advantageous for any strategy that is likely to change rapidly and frequently, as would be the case in some overlay approaches. Most asset allocation strategies are long term, but for those with a shorter-term orientation, futures are the fastest way to lose or gain exposure. One problem with using futures is that many plan sponsors and trustees are uncomfortable with them, so the manager has to go through a long education process of teaching the benefits and raising the client's comfort level. Other intangibles, particularly accolmting, risk management, and cash management issues, also fall on the loss side for futures, not so much for clients but for the managers who are deciding how to gain desired exposures for their clients. Accounting for futures is quite complex and requires very specialized knowledge. Risk management cannot be overstated: If the manager gets the futures contract wrong at any point in time, the result can be terribly expensive. And the manager must have a well-designed and strongly supported cash managementfunction inplace to supportanyserious ongoing futures strategy. Finally, one of the problems with using futures is that futures do not exist in some parts of the world and are somewhat questionable in others. The S&P 500 and the US. component of the FT/S&P may be decent proxies for each other, albeit not perfect, but futures contracts elsewhere around the world, such as for the German DAX or the French CAC, are actually built on very poor proxies. So, basis risk becomes an even more serious and costly problem. Index Funds. Index funds are the third choice for achieving desired asset allocation exposures. Relative to equity baskets and futures, the chief advantages of using index funds are elimination of basis risk, reduced transactioncosts, and ready availability. Basis risk. The index funds that are available around the world are generally consistent with the benchmarks that most managers are being measured against. Thus, the basis risk is minimal. Figure 2 compares a UK. fund that tracks the US. market with the underlying index. The daily tracking error ofthe fund Figure 2. Aquila U.S. versus U.S. Component of FT/S&P, 1997 January 1997 = 100 101.00r------------------------------, 100.75 100.50 100.25 99.75 99.50 99.25 99.00'----------------------------~ 1/97 12/97 Association for Investment Management and Research 113

Asset Allocation in a Changing World relative to the index is very small. Furthermore, a graph of the loo-day moving average of these data would be a straight line. So, the basis risk between the index fund and the benchmark is essentially zero. Transaction costs. Compared with equity baskets and futures, index funds generally have lower transaction costs, including commissions, spreads, and especially marketimpactcosts. Actually, because index funds operate in a vast global network, market impact costs are eliminated. At almost any time, if a manager wants to buy a Japanese index fund, he or she is likely to find somebody who wants to sell one, as longas the managerhas the technology or thereach to find it. Granted, the manager does have to give up some timeliness with index funds; the manager might have to wait a week, even a month, to get the needed exposure. Remember that with futures, the manager can get immediate exposure, so using futures entails ready and ample liquidity. Availability. Index funds are not much use to anybody unless many and varied index funds are readily available on a global basis. Indexing, which was pretty well unheard of in the United Kingdom in 1984, has grown to nearly 21 percent of the market, as seen in Table 2. The United States has about 30 percent of the institutional market in index funds. Index funds have grown to quite substantial proportions around the world, so the supply is plentiful. Table 2. Prevalence of Indexing as a Percentage of Total Market Country Market Share United States 30% United Kingdom 21 Australia 10 Netherlands 10 Canada 8 Hong Kong 5 Japan 3 Note: Market share determined by value of funds under management. Global Equity Fund By using index funds, an investor can allocate assets and get market exposure quite cheaply. The following example illustrates the process of using index funds to build a global equity fund. In January 1998, we at Barclays Global Investors had a U.K. client with a substantial portfolio that was all in cash. The client wanted to create a global equity portfolio. The client was frightened about market impact around the world and wanted to proceed on a gradual basis. In other words, the client did not want exposure tomorrow but over the following six or eight weeks. By using our contacts around the world, we were able to accomplish the following to create a global equity fund... Crossed U.K. equities with a life insurance company. A life insurance company that was moving to activeequitymanagementfrom index funds provided the U.K. equity exposure that this particular client wanted... Crossed Swiss equities with a retail unit trust. A retail unit trust had a big redemption, so Swiss equities were picked up in this cross... Crossed U.S. equities with a U.S. public fund (S&P 500 plus smaller companies)... Crossed a non-u.s. with U.S. corporate pension fund that was moving to active management... Crossed U.K. equities with a UK segregated fund moving from U.K. equities to bonds. A U.K. segregated fund was moving from u.k. equities to bonds because of the minimum funding requirements of the Pensions Act 1995... Crossed Japanese equities with a pension fund in the Netherlands. Crossing Japanese equities with a pension fund in the Netherlands was the client's asset allocation decision... Crossed Japanese equities with a U.S. public fund. This series of index-fund-based crosses enabled this client to achieve 97 percent of the desired exposure. The resulting fund had about US$2 billion of acquired assets, with minimal market impact and costs. Conclusion The possibility exists that index funds could replace futures as an asset allocation tool, but they probably will notfor several reasons. First, index funds are best transacted at the unit level. In other words, investors do not really want tobuy or sellmo individual stocks, but if those stocks have been packaged into a fund, then those units can actually be exchanged between the various counterparties in a single transaction, like purchasing a futures contract. Sometimes, unfortunately, because of regulatory barriers around the world, index funds cannotbeinvested across borders, and the investor has to go below the unit level and actually exchange the securities. Havingsome kind of globally recognized investment vehicle for transferring index funds across borders would be beneficial, but that vehicle is going to be some time in coming. Second, creating a global equity fund with index funds is not an instant process. In the example given earlier, the client was more than willing to wait. Unfortunately, finding the appropriate counterparties around the world takes time. Technology will improve the timeliness. The process is like a giant order board at the end of the day, and eventually, 114 Association for Investment Management and Research

Uses offutures and Index Funds technology will enable all the bids and all the offers around the world to be known. Third, cross-border differences still mean that investors have to deal with potential short-term tax disadvantages, such as withholding taxes and stamp duties. Despite these hurdles, billions of pounds are now being transacted in this way. These transactions are not going through markets but rather are being exchanged between counterparties. Companies are rushing to become global fund managersbecause one of the great advantages of being a global fund manager is reach; global fund managers can find counterparties who are willing to exchange exposure around the world. Association for Investment Management and Research 115

Asset Allocation in a Changing World Question and Answer Session James R.C. Woodlock Question: Who benefits from stock lending within an index fund? Woodlock: Stocklendingis abig businessin itself. Index funds effectively hold the universe of stocks, and because they are passive, they have nodemand to sell atanypoint in time. So, index funds are the obvious first choice for a borrower of stock. Index funds are also attractive from the fund company's point of view. Barclays, for example, earns about one-third ofits total revenue from stock lending. The arrangement generally is a split between the client (two-thirds) and ourselves (one-third). So, stocklending is a known, commercial practice that we undertake on behalf of our clients. Question: How do you measure the risk ofstock lendinginan index fund? Woodlock: Unfortunately, we cannot measure the risk of stock lending because there has never been a failure. All the lending programs we have are fully collateralized. So, should anything go wrong, the client is not at risk because the collateral stands behind the whole arrangement. In recent years, we have found that it is better to sell a complete portfolio to a borrower and in return receive an assurance backed with a sufficient credit rating so that should anything go wrong in the lending process, the loanwillbemade good by the borrower. Therefore, the risks are covered. Question: What are your thoughts onperformance measurement for index funds? Woodlock: The way performance measurement should work is that investors should look at the invested proportion. So, for example, if you are measuring the performance of a U.K. passive fund manager against a U.K. active manager in u.k. equities, only the amount that the passive fund manager has in u.k. equities will be actually measured. The decision to go into cash is an asset allocation decision and should be measured separately. All the evidence that I have seenindicates thatifthe stock selection decision is carefully separated from the cash/liquidity decision, index funds perform very well Unfortunately, that careful separation does not always occur, and a good deal of noise is introduced into performance measurement as a result. Question: What is the cost of buying an index fund with a bidoffer spread? Woodlock: Inevitably, the buying and selling of index funds is done on a cross transaction, which is typically done at mid-market price. So, there is no bid-offer spread on the index fund. That arrangement might evolve but does not now exist. 116 Association for Investment Management and Research