Mebane Faber

About this author:
Become a Contributor Submit an Article
  • Font Size:
  • Print
There is a great discussion of some asset allocation topics in Bernstein's new book, Capital Ideas Evolving (however, the first half of the book I skimmed). In the chapter profiling the Yale Endowment:

Swensen describes the process this way: "Mean/variance was a powerful influence in causing us to move away from the standard institutional portfolio. You never get a recommendation of 65% equities from mean/variance - it's always telling you to move toward diversifying asset classes that promise equity like returns. These kinds of results led us to emphasize private equity and venture capital, real estate, hedge funds offering long/short or absolute return strategies, and investments in raw materials like timber."

...Along the way, Swensen has been faithful to one of Harry Markowitz's favorite observations about asset allocation: "It's not the variance you have to worry about, it's the covariance." In other words, you can hold plenty of risky assets with high expected returns as long as they fluctuate independently rather than in step with each other.

Here is an example. Below is a table for a standard 60/40 allocation since 1972 (S&P500 and 10 Yr US Govt Bonds). Following Swensen's lead, we reduce the stock and bond exposure by 20% each, and add two riskier (more than double the volatility of bonds) asset classes at 20% each - REITs and commodities. Notice how the return improved and the risk declined, all due to the assets being uncorrelated. We then take 40% off the original stock position, and distribute it to REITs and commodities evenly at 20% (S&P / Bonds / REITs / Commodities).

endowment portfolio

What would an Endowment Portfolio formed from publicly traded vehicles look like? (I have written extensively on the endowment method of investing before.)

Below I list 10 world betas followed by the alternative holdings. The percent weightings are in the ballpark of the Harvard and Yale endowments. Only one fund is foreign listed (the private equity portion), because there is no US equivalent (I don't like PSP). Consequently, one could use any of the myriad foreign listed funds of funds as well (Goldman, Dexion, Alternative, etc.).

endowment portfolio

There are lots of ways one could tweak this portfolio - ie. using leveraged ETFs to port alpha from the alternative funds, etc.

I would like to hear your comments on the allocation.

This article has 23 comments:

  •  
    May 29 10:49 AM
    It would be really interesting to compare this portfolio's historical performance to the hedge fund indexes after fees. Has anyone done this sort of thing?
    Reply
  •  
    May 29 02:43 PM
    Check out my post here:

    worldbeta.blogspot.com...
    Reply
  •  
    May 29 12:02 PM
    I've been fascinated by work in this area for the past couple of years and have spent a lot of time researching some of the same investment vehicles listed in your article and applying them to the latest (2006) Yale Endowment model. However, I have found that many of the mutual funds available have very high expenses and relatively low returns during many market cycles.
    HSGFX +.13% YTD 1.14% Exp Ratio
    TFSMX +5.3% YTD 2.49% Exp
    JAMNX +3.2% YTD 1.95% Exp
    ARBFX +3.9% YTD 1.95% Exp
    DIAMX +1.0% YTD 1.51% Exp

    Yes, they do indeed offer diversification, but at what price, and for what kind of return, over the long haul? A couple of ideas to add to your research.

    You have the US, Foreign, REIT and Commodities covered. The Absolute Return and Private Equity are the problems for the average investor trying to replicate the Yale Endowment. Here are a couple of things I've considered.

    1) In the Private Equity area, consider holding such stocks as Brookfield Asset Management (BAM: plus 33% vs plus 7% for the S + P 500 YTD), Leukadia (LUK: plus 20% YTD ) and Macquarie Infrastructure (MIC; plus 24% YTD) among others, as well as Berkshire Hathaway B shares. Now granted, all are publicly traded but all are asset management firms which allow for managing assets beyond the realm of publicly traded companies and shares. If you look at their investment approaches they are not unlike endowments and hedge funds, and even merger and acquisition firms...with much better returns and much lower expense ratios.

    2) Could you do an asset allocation, using your proposed funds, using the most recent Yale Endowment allocations, and back test it for five or ten years. This would be fascinating to see how it stacks up.

    Keep up the great work. Looking forward to more on this subject.
    Paul
    Reply
  •  
    May 29 02:44 PM
    Here it is since 1983, but only with the 5 asset classes mentioned:

    worldbeta.blogspot.com...
    Reply
  •  
    Oct 11 02:02 PM
    Good idea. I have also used ALD and MCI as access to different area. MCI has a great track record. I have also sprinkled in EFR, FCO, BGT for some of my income needs.
    Reply
  •  
    May 29 03:02 PM
    Even though asset allocation, correlation, and the use of alternative asset classes has been at the core of Swenson's superior returns over the years, he is an active manager and has always made it clear that it is difficult for the individual investor to duplicate his institutional portfolios. Nevertheless, I certainly think it is worth the effort and I believe fascinating work and backtesting could be done in this area. Modern Portfolio Theory is based on the futility of market timing and superior stock selection. Of the top 100 performing managers annually, only 10% repeat. Moral of the story-don't chase performance. Nevertheless, Swenson openly acknowledges that he is able to find superior managers and that is a key ingredient to his success. This would probably involve entrepenureal private partners in the area of Real Estate, infrastructure and other real assets. He is certainly aware of when asset classes become crowded (such as REITS with premium to NAV) and manages to stay clear. I would also guess that he would be very selective with exposure to Hedge Funds as part of his absolute return asset class. He always seems to manage to stay one step ahead of those attempting to emulate him. So his m.o seems to be asset allocations for risk management but very active management within those asset classes including active equity allocations. Therefore, it cannot be easy to attempt to copy Yale using ETFs and Index Funds. However, I think it is a fascinating study that opens up many possibilities. One other comment regarding Brookfield Asset Management (BAM). Their CEO delivered an interesting talk at a recent conference. Though most of their current deals are Real Estate related, he believes infrastructure deals globally represent a huge opportunity and is forging partnerships in this field. They will soon spin off an the infrastructure division into a separate company. It is a way for the individual to indirectly participate in building of infrastructure although only as a stockholder.
    Reply
  •  
    May 30 07:44 AM
    Gentlemen-

    Some very good comments here. Re DIAMX, this fund also has a 5% load. I do not know of a no load version. Re individual stocks as Private Equity, I have toyed with the notion also and came across LUK as a proxy combined with GS. GS has done very well this cycle. As values revert and there are more bargains, I would expect LUK to pick up the slack. Low R2 in these stocks as well. Comments?
    Reply
  •  
    May 30 11:57 AM
    Consider ANGLX - Analytic Global Long SHort - as alternative?
    Reply
  •  
    May 30 12:56 PM
    DIAMX is available with no transaction fee at Charles Schwab and TD Waterhouse, among others. They do have a $10,000 minimum investment. Schwab also has a long-short fund (SWHEX) which has actually outperformed DIAMX year to date (DIAMX; -0.7% vs SWHEX; +5.0%) and ANGLX (+3.94%). DIAMX has the better long term (3+ years) return. ANGLX has the lowest fees (1.31%) with Schwab having the highest (1.76%).

    Since Private Equity is the most difficult asset class for the average investor to access, here are a couple more alternatives to consider.

    MVC (up 39% YTD with a PE of 6.5); CSWC (up 38% YTD with a ROE of 31.3); ACAS up 45% in one year with a yield of 5.6% and a PE of 7.8); TICC (up 27% in one year and yield of 8.6%) along with MIC (up 70% in the last year and a yield of 4.9%) are all involved in private equity purchases and financing. Any thoughts on these?

    Swensen also includes investments like natural resources, timber, and other commodities in his Real Asset grouping. DJP and GSP are excellent ETF choices. GSP is more heavily weighted in oil and gas. You could include a mutual fund like US Global Resources (PSPFX; also heavy on oil and gas) which is up over 15% YTD, or Vanguard Raw Materials ETF (VAW) up 31% in the past year. In the Real Estate group Alpine International Real Estate mutual fund (EGLRX) has outperformed the US market this year (+13.1% vs ICF down 3.7%) and the international real estate REIT (RWX) which has returned 8.2%.

    As always, it is important to understand that while these alternative asset classes have done very well in the past few years, they will not always lead the way. Nonetheless, they have done MUCH better than the US markets, which everyone is all excited about now, and do offer excellent diversification options along the lines that Swensen recommends.

    Just a few ideas. Any comments or suggestions?
    Reply
  •  
    May 30 02:04 PM
    Paul-

    Thanks. Good ideas that I will look at. For Ken Heebner fans, you could do a lot worse than CGMRX for your real estate holding. He makes big bets but they do seem to pay off. Perhaps you could slice and dice the total holding here if Heebner makes you squeamish: 70% an ETF and 30% CGMRX or some such. My bet is that Heebner adds alpha over time.

    ACAS is a favorite of mine. Also, the way Marty Whitman buys companies and debt, you can make a case that his fund (TAVFX) is very similar to PE.
    Reply
  •  
    May 31 10:30 AM
    One other thought. Any comments about Jeremy Grantham's assertion that Private Equity and hedge funds are the same beta sources as traditional ones, just gussied up with extra fees?
    Reply
  •  
    May 31 11:36 AM
    Anyway, back to Mebane’s original question, earlier in the week, “I would like to hear your comments on the allocation.” Based on the 2006 Yale Endowment here is what they have done and how you could do it with Mebane’s suggestions and those that have been thrown into the mix this week. This is sticking strictly to Swensen’s definitions (or nearly) for each of the asset classes as described in the 2006 Yale report, and is specific to Endowment investing. There are certainly many other options, but here goes.

    ABSOLUTE RETURN 25%
    5% DIAMX Diamond Hill Long-Short
    5% ARBFX Merger Arbitrage
    5% HSGFX Absolute Return
    5% JAMNX Market Neutral
    5% TAVFX Third Avenue Value

    DOMESTIC EQUITY12%
    6% SPY US Large Cap
    6% IWM US Small Cap

    FIXED INCOME 4%
    4% TIP US TIPS

    FOREIGN EQUITY 15%
    7.5% EFA Foreign Developed
    7.5% EEM Foreign Emerging Markets
    Note; Swensen has alluded to going 50% into emerging markets.

    PRIVATE EQUITY 17%
    10% IPRV.L Private Equity Fund
    3.5% BAM Brookfield Asset Management
    3.5% ACAS American Capital Strategies (or any of several others)

    REAL ASSETS 27%
    6% IYR US REITs
    6% RWX International REITS
    5% DJP Commodities
    5% GSP Commodities
    5% VAW Vanguard Raw Materials (or other similar options)

    However, keep in mind that Swensen recommends an asset allocation of 70% stocks and 30% bonds/income for the “average investor”. The breakdown is:
    20% Real Estate
    20% Foreign
    30% US Equities
    Reply
  •  
    Jun 01 01:11 AM
    For the average investor, Swensen recommends in his book 30% domestic equity, 15% developed market, 5% emerging market, 15% U.S. long-term Treasury bonds, 15% TIPS/I-BONDS (Swensen wasn't aware, but I-BONDS have superior deflation protection to TIPS) and 20% real estate.

    I've been walk forwarding (using timing) both the Harvard and Yale endowment allocations using my idea of "best of breed" ETF's/CEF's where available, less the Absolute Return allocation as I'm unsatisfied with the limited amount of performance data on those vehicles so far. However, it was unclear to me until reading this thread if HSGFX was considered to be an Absolute Return asset as opposed to Domestic Equity. Does any disagree?

    My verdict so far is these two portfolios are not that much better returning or less risky than Swensen's average investor portfolio and certainly won't get you to the promised land of a high Sharpe ratio. Performance is very likely to be worse without timing for reducing peak to trough drawdowns. Performance may also be woefully sub-par because Absolute Return is resting in cash. (If that really were to be the case, why bother with other asset classes? Just stay in cash and allocate the minimium out to Absolute Return strategies to achieve your target level of a yearly return. With managed forex trading accounts returning, say 5% a month average, why take on more non-systematic risk than necessary? Sometimes I fear asset allocation has become Wall Street B.S. to sell more and more asset classes, but I digress.) Anway, I suspect as more capital flows out of overvalued asset classes into more alternative, undervalued asset classes, total portfolio returns will likely stay commensurate with what was experienced over the past 30 years with the traditional institutional allocations.

    Does anyone know of software that features an automated, volatility & co-variance (R2) optimal portfolio generator? That is something I'd dearly like to see let loose on 10,000 securities.

    HARVARD:
    15% Domestic Stocks [DEF/STH/PZI]
    10% Foreign Stocks [50% EEN, 50% EEB/JSC]
    11% Domestic Bonds [AGG]
    6% Inflation-indexed Bonds [TIP]
    5% Foreign Bonds [GIM]
    5% Junk Bonds [DSU]
    13% Commodities [DBC/GSG]
    10% Real Estate [IGR/RWX]
    13% Private Equity [PSP]
    12% Absolute Return (hedge fund)

    YALE:
    12% Domestic Equity [DEF/STH/PZI]
    15% Foreign Equity [50% EEN, 50% EEB/JSC]
    4% Fixed Income [GIM]
    27% Real Assets (Real Estate, Oil, Gas, Timberland) [IGR/RWX, EPD/KMP, PCL/RYN]
    17% Private Equity [PSP]
    25% Absolute Returns (50% Event-Driven, 50% Hedged Value Driven)
    Reply
  •  
    May 31 03:00 PM
    Paul-

    I referenced TAVFX, Marty Whitman's fund, as being a proxy for Private Equity. He as much says so in his client letters, and his historical returns (16% LOF) are very similar to PE's. His fund doesn't normally achieve control but it does work at exit strategies. Above, you have it placed in the Absolute Return category.

    Thanks for your work on this.
    Reply
  •  
    May 31 09:18 PM
    Mark,
    Ooops! Sorry about that. I actually saw that same reference made some time ago on a discussion board on Morningstar's web site. They were throwing out ideas on this very subject and someone made that same connection. I'll check it out. That makes things a little simpler since the IPRV.L could be problematic to access.

    Mebane,
    Is IPRV.L available in the US and if so, where can one find information on it?
    Reply
  •  
    Jun 01 06:26 PM
    If you have a broker that trades foreign like IB or ETrade.

    www.londonstockexchang...
    Reply
  •  
    Jun 04 05:14 AM
    Machine-

    I have a portfolio optimizer that can handle 20 positions but nothing like the 10K you are envisioning!

    Re HSGFX, it is decidedly a growth fund but the strategy makes the volatilty better than an absolute returns fund's (10%). In fact the SD=4 and the R2 is 6 according to Morningstar. I own the fund as a proxy for CASH upon the firm belief that in the event of a market dowturn the underlying holdings will outperform (no junk) and the hedging will cause a very positive return, much like last time. If it doesn't then the strategy is decidedly a loser. Similarly, I use HSTRX as a substitute bond holding. Value is found in this environment only in TIPs I believe (Grantham agrees) and the fund's hedging by timing in and out of gold shares has produced good returns. The trouble is, with all these things, the TIMING of outperformance.

    I see you have found PCL and RYN as proxy Timber exposures. I think the substitute is likely not a "clean" one but Limited Partnerships are the only other vehicle I know of.
    Reply
  •  
    Oct 11 02:14 PM
    An interesting alternative could be to sprinkle in NUT for sme real estate exposure. I have used it in the past. Here is my current holdings for real estate. IGR/MIC/NUT/FSP/EFR/PC...
    Reply
  •  
    Jun 04 05:17 AM
    Paul-

    Not to confuse things even more but what about throwing Icahn's American Real Estate Partners (ACP) into the Private Equity mix? Don't be fooled by the name. Buys distressed properties of all types.
    Reply
  •  
    Oct 11 01:46 PM
    Can you elaborate of your dislike of PSP? Also, please any ideas on an alternative. Thank you.
    Reply
  •  
    Oct 11 01:57 PM
    Great stuff. I am a huge fan of non-correlated investing. Try adding some of the Wisdom Tree funds for alpha. Also, any thoughts on NCHPX. High fees, but a great diversifier.
    Reply
  •  
    Oct 11 01:59 PM
    Great stuff. I am a huge fan of non-correlated investing. Try adding some of the Wisdom Tree funds for alpha. Also, any thoughts on NCHPX. High fees, but a great diversifier.
    Reply
  •  
    Sep 21 11:28 PM
    Great study. However,using publicly trading vehicles would not achieve the returns of Yale-Harvard endowments. Much of their returns were generated by alpha generating strategies and high barriers market involvements. However, for ordinary investors like us, using publicly available vehicles to mimic their approach is still a very sound strategy. I studied the evolution of their approach in my article "All-weather portfolio: investing like Yale-Harvard Endowments". You can read the article by clicking the website like on the left.
    Reply
Articles on related themes