Filed pursuant to Rule 433

Registration No. 333-151523

August 13, 2009

MacroShares Housing

A Missing Ingredient to an Optimal Portfolio

By Robert J. Shiller¹

MacroMarkets LLC is in the process of launching an array of innovative securities that will allow both individuals and institutions to invest in or hedge their exposure to the market for single-family homes. The first of these home price securities were launched on June 30, 2009 on the New York Stock Exchange/Arca.  The “Up” Major Metro MacroShare (NYSE: UMM) represents, in effect, a five-year interest² in the equity of a diverse portfolio of approximately 30 million homes from large metropolitan areas comprising the S&P/Case-Shiller Composite-10 Home Price Index³. The “Down” Major Metro MacroShare (NYSE: DMM) represents an inverse interest in the same portfolio of homes.  By adding either of these securities to a portfolio, investors can, for the first time, easily adjust exposure to U.S. home price risk via the stock market.

Modern portfolio theory in the Capital Asset Pricing Model tradition suggests at the most basic level that one should hold the market portfolio of risky assets, adjusting for one’s individual risk tolerance by leveraging or deleveraging this portfolio through the riskless asset. This would suggest holding massive amounts of UMM in any portfolio that is not already exposed to real estate risk. But, this advice omits consideration of a lot of special circumstances.

As an exercise, it is useful to consider the optimal portfolio, computed using standard mean-variance analysis, namely the tangency portfolio on the efficient portfolio frontier, with actual data.

Any such exercise is necessarily only a device to help us enhance our intuitive feel for risk management, for such an endeavor is necessarily only as good as its assumptions, which in the context of any

——————————————

¹ In addition to my professorship at Yale University, I am co-founder and Chief Economist of MacroMarkets LLC (www.MacroMarkets.com), a small asset management company that sponsors MacroShares.  Although I serve as one of several supervisory committee members for the S&P/Case-Shiller Home Price Indices, I do not have any involvement in the calculation or production of these indexes, even though they bear my name.  

² UMM and DMM have a stated maturity date of November 25, 2014.   

³ While it reflects weighted home price changes within just ten metropolitan statistical areas, the aggregate value of the residential real estate within the markets comprising the S&P/Case-Shiller Composite-10 Home Price Index is approximately 30% of the value of all homes in the United States.  Although somewhat more volatile, this index has a 96.5% correlation to the S&P/Case-Shiller U.S. National Home Price Index and was incorporated into the Federal Reserve’s 2009 bank stress tests.  As such, the S&P/CS Composite-10 serves as a very reliable and authoritative proxy for an asset class whose value has been estimated at approximately $20 trillion.   

1

simple model cannot be perfect embodiments of our prior information. Notably, we may have reasons to think that historical data may not be good indicators of future performance. This is especially true with MacroShares Major Metro Housing, for which at this writing, there is just over one month of trading activity and pricing history.  Thus, we must make assumptions about the price of these securities in relation to past index values in order to simulate their returns.

Ideally, an analysis of the potential portfolio value of a fundamentally new instrument that opens up a new asset class should model the ultimate economic rationale for the new instrument, and develop a full general equilibrium theory involving the objectives of economic agents and their ultimate constraints, a model that represents all world markets and which would allow us to estimate the changes in the risk premia and stochastic properties of all securities.  Such an analysis would open the question whether the introduction of a fundamental new market would alter the statistical properties of other markets, such as the stock market. However, such an analysis is difficult to achieve, and can be done only under some very restrictive assumptions, including an assumption of perfectly rational economic agents, which will tend to make the exercise mostly academic.⁴

It is still useful, I believe, to go through an analysis that makes the simplest possible assumptions, that markets for other assets are unaltered by the introduction of the new security, and, in the case of UMM, that the price of the new asset can be proxied by the home price index for which we have data.  I will conduct one such exercise here, with the understanding that it provides only one of many possible simple model analyses relevant to portfolio allocation.

Figure 1 shows the raw data on housing, stocks, and bonds, monthly, back to January 1987, the start date for the S&P/Case-Shiller Composite-10 Home Price Index, which is the benchmark that the underlying values of UMM and DMM MacroShares track.   The plot of this index in Figure 1 has no adjustments for imputed rent or expenses. The stock market index is the S&P 500 Total Return Index. The bond market index is a total return index computed from the Moody’s Corporate Average bond index. Note that in this chart the

——————————————

⁴ I did such a world general equilibrium analysis to consider the effects of introducing GDP-linked securities, see "World Income Components: Measuring and Exploiting Risk Sharing Opportunities," with Stefano Athanasoulis, American Economic Review, 91(4):1031-54. September 2001. The admittedly highly restrictive theory in that paper allowed us to estimate the effects of the new markets on the interest rate and the equilibrium risk premium for all securities and calculate the world economic welfare gain to introducing the new markets.

2

stock market looks most volatile, and while housing had a bubble which “burst” after 2006, the whole event looks relatively small next to the stock market.

But, most people who invest in housing do so on a leveraged basis, and so the relative tameness of the housing market in Figure 1 is illusory. Indeed, over 15 million households are estimated to be “under water,” namely, in a negative equity position since the housing bubble burst. For them, it has been a wild ride indeed. The UMM and DMM securities are designed to mitigate such events, allowing people to insulate their portfolios from such swings. However, the securities themselves incorporate a leverage factor in the form of a multiplier applied to cumulative index changes, to permit just that. UMM incorporates three-to-one⁵ leverage, less than the typical homeowner acquires upon purchase, since many people borrow 90%, or even more, on their homes. Figure 2 shows the same series as in Figure 1 but with the housing index replaced with a simulated portfolio invested in UMM (had it been available) in January 1987, buying a new UMM security rebased to the prevailingindex level every time the security would have matured or expired as a result of its underlying value doubling⁶.  It is assumed that UMM always sold at its underlying (net asset) value, and paid the Treasury Bill rate minus 125% (the expense ratio). Since 1987, UMM would have matured as scheduled twice, and expired early three times as a result of historic home price inflation during the housing and mortgage credit bubbles.  Now, the movement in UMM price is more dramatic than that of either stocks or bonds.  The three-to-one leverage factor built into UMM, while substantially less than that of most new homebuyers, is still powerful enough to have important effects, and it clearly would be unwise for investors to put a majority portion of their portfolio in UMM.

Figure 3 shows an estimated efficient portfolio frontier for these three assets.  However, since longer time series are available for stocks and bonds, the covariances and means were estimated back to 1947 for these. If we did not do that, the bond market would look unrealistically stable, since we would be omitting the

——————————————

5 Unlike leveraged ETFs that have a daily performance objective, MacroShares have a long-term objective.  For UMM and DMM, the 3:1 leverage factor is applied to the cumulative percentage change in the index from a fixed starting level of 162.17, the S&P/Case-Shiller Composite-10 Home Price Index level announced in February 2009 pertaining to the three month period ended December 2008.        

⁶ UMM and DMM are trusts with obligations to each other that are a function of the cumulative return of the S&P/Case Shiller Composite-10 Index.  With a 3x index multiplier, if the index were to move by more than 33.3% (cumulatively) from its starting level over the life of the securities, the underlying value of one trust would be more than double its original par amount of $25, leaving no assets in the other trust.  An orderly, but early, expiration of both trusts would occur in these circumstances, with the terminal underlying value returned to shareholders, consistent with rules established within the prospectuses.

3

historical periods when its yield varied widely.  Covariances and means of returns were estimated with the longest available data, and reflect five-year returns (though the percentages shown are annualized). Obviously, any results will depend on sample period, and the frontier could be made to look different by merely adjusting the sample period. Notably, however, our sample period includes both the boom 1997-2006 and the bust (2006-2009) of the housing market. Thus, there is no obvious bias to expected return in the housing numbers. There might be an upward bias to the estimated volatility of housing, since the bubble was unprecedented, but we really do not know if the bubble period is a temporary aberration or whether housing market volatility will continue.  

The tangency portfolio for a zero risk-free rate in Figure 3 is 57% stocks, 11% bonds and 32% housing (UMM)⁷. Figure 3 shows how much improvement adding housing (UMM) to a portfolio of stocks and bonds can improve the risk-return tradeoff⁸. The efficient portfolio frontier (measured at its minimum standard deviation) is shifted almost five percentage points of standard deviation to the left. This dramatic improvement in risk management reflects the virtually zero (actually, in this sample, slightly negative) correlation between UMM and either stocks or bonds.

——————————————

IMPORTANT

Registration statements relating to MacroShares (NYSE: UMM and DMM) have been filed with the Securities and Exchange Commission.  You may obtain a written prospectus by visiting EDGAR on the SEC website at www.sec.gov. Alternatively, MacroShares Housing Depositor, LLC, will arrange to send you the prospectus if you request by calling toll-free 888-9House5. The information herein  does not represent an offer to sell securities of the MacroShares Trusts and it is not soliciting an offer to buy securities of these Trusts.  There can be no assurance that the Trusts will achieve their investment objectives.  An investment in the MacroShare Trusts involves significant risks, including the risk of loss of substantially all of your investment.  There is no guarantee as to the amount of any quarterly distribution or the amount of the final distribution to shareholders.  You should review the risk factors in the prospectus prior to investing in a MacroShares Trust. Fluctuations in the underlying value of a MacroShares Trust and other factors may affect the market price of your MacroShares investment.  An investment in a MacroShares Trust will not resemble a direct investment in the asset being tracked.  The price of the asset being tracked by a MacroShares Trust may be volatile.  It may be difficult to predict whether in the long-term the price of the asset being tracked will reflect a generally upward or downward trend.  There are risks associated with investing in a product linked to a benchmark. Additional Information about MacroShares can be found at www.macroshares.com.

——————————————

⁷ Although I think that these are plausible portfolio allocations, there certainly are many adjustments one might consider making to the data and analysis used to produce these results.  Regardless of the number and extent of such adjustments one might make, the unique and non-correlated nature of housing price returns since 1987, coupled with the fact that most institutional asset allocators hold little or no residential real estate, suggest that UMM - in some amount - would enhance overall portfolio performance.     

⁸ Notably, the portfolio allocations of many wealthier homeowners already resemble the tangency portfolio illustrated in Figure 3.   Many more homeowners are over-invested in housing relative to this standard, however.  Many institutions, including insurance companies, banks, federal and local governments are also over-exposed to housing and could use DMM as a means to hedge their over-exposure.  

4