“Portfolio Design” by Richard Marston

In a nutshell

While Richard Marston’s “Portfolio Design” according to its dust jacket flap is written for “financial advisors who want to provide diversified portfolios for their clients”, I suspect much of the book is quite accessible to the diligent non-expert reader. It is an excellent book that I highly recommend for your reading list. The book is not only full of historical performance data for different asset classes but is also explains how to use that data in constructing client portfolios, and the value of non-traditional asset classes in portfolio construction. His concluding chapters cover spending rules for foundations and individuals/couples. Some of the key messages are:

-what matters is real return: Long-Term Treasuries= 2.4% and S&P500= 6.2%. Real returns are essential for both long-term accumulation and spending.

-in market swoon it’s not only essential to sit tight don’t sell in panic, but also to rebalance and reduce spend rate

-Value vs. Growth: Value stocks had higher returns, lower standard deviations, thus higher Sharpe Ratios and lower betas (<1) than Growth stocks (betas>1)

-Foreign Developed and Emerging Markets stock improve portfolio performance; currency movements are unpredictable and hedging does not appear to be worthwhile

-Alternative Assets discussed for additional diversification include: hedge Funds, Private Equity, Venture Capital, Real Estate (REITs and Home Ownership), Commodities, and Gold; for average individual investor REITs are recommended, for HNW investors REITs and Commodity indexes (passive) are suggested, whereas for UHNW investors if they can overcome the extreme challenges of selection (due diligence) and access then there may be opportunities to explore the other alternatives. Marston does not see gold as adding value historically. Home ownership is inferior investment compared to REITs and S&P500 after factoring in appreciation and leverage. Also historically commodities and gold have very low correlation with inflation

-on spend rates for foundations and individuals/couples, he discusses the “absolute” and “proportional” rules for spending (and some combination of the two), as well as the different considerations (foundations living for ever, whereas individuals/couples longevity risk and possible desire to leave an estate). He also discusses the inter-relationship of real returns, withdrawal rates, stock allocation, and risk-tolerance differences between foundations and individuals/couples)

The details

-what matters is real return: Long-Term Treasuries= 2.4% and S&P500= 6.2%. Real returns are essential for both long-term accumulation and spending.

-foundations (if the have no new cash inflows) can only spend the return in excess of inflation (i.e. real return) if they expect to operate in perpetuity, whereas individuals with finite lifespan can spend more if they have no specific desire to leave an estate

-in a market swoon/crash it is not enough to sit tight (though that better than panic selling into the crash), but must also rebalance as well (and this is the really hard part, to sell that which hasn’t dropped in price as much or at all or even increased in price while buying more of what has dropped in price…and of course cut back spending to minimize portfolio drawdown!)

-must be familiar with advisors’ basic tools: returns (compound/geometric (best estimate for long-term), arithmetic (best estimate for next year’s return), real), risk (standard deviation (volatility), and beta (systematic risk related to market), risk adjusted returns (e.g. Share Ratio= excess return above risk-free rate relative to risk taken (standard deviation)) and alpha (“the excess return earned on an asset above that explained by the beta of that asset”)

-equity risk premium is the return of equities in excess of the available risk-free rates; e.g. S%P500 excess return (geometric) is 4.7% relative to L-T Treasury rates and 5.9% relative to risk-free rates (T-bills) (By the way the standard deviation of S&P500, T-bond index and T-bills nominal returns is 14.6%, 9.5% and 0.8%, respectively between 1951-2009)

-Between 1951-2007 real (arithmetic average) return on S&P500 was (depending on how one might want to measure it, either based on: Capital Gains=8.6%, Dividend Growth= 5.0% and Earnings Growth= 6.9%

-Small Caps: there is not a strong case for portfolio diversification…the case depends on the data set (time period) used

-Value vs. Growth: Value stocks had higher returns, lower standard deviations, thus higher Sharpe Ratios and lower betas (<1) than Growth stocks (betas>1)

-Foreign Stocks: Diversification value of foreign stocks has been decreasing in an overall well diversified portfolio in recent years due to the growing correlations with US stocks (0.8-0.9), but of course there are no guarantees that this high correlation will continue. World stock market capitalization was $35T in 2008 broken down into US=33.6%, Developed ex-US=39.9% and Emerging Market= 26.5%.

-Currency Variation have significant impact on USD returns seen by Americans, however these cannot be forecasted and hedging currency variations does not appear to be worthwhile for equities.

-Emerging Markets: capitalization of EM stocks is $9.3T but the number is misleading since the investible universe is much smaller due to restrictions on ownership by foreigners, so performance should be measured by the MSCI index (which includes only investable assets) for the country under consideration or EMs overall. With the exception of Asia/Japan, EM (MSCI) returns and Sharpe Ratio are much higher than Developed Markets, but high economic growth does not translate to high (USD) stock market returns historically. There are small diversification benefits (higher return and higher Shape Ratio) in a portfolio. EM bonds show low correlation with US stocks and bonds, and they tended to perform even better than US stocks (due the massive spread reduction from 1400 to 200 bp)

-Bonds: interest rates display a very clear relationship with inflation (though they also vary with maturity and default risk). Returns are a combination of the coupon and cap gain which is related to the interest rate regime (falling or rising). Interest rate (averages) between 1985-2008 were: 10 yr Treasuries= 7.7%, AAA= 8.8% (+1.1%), BAA (lowest investment grade)= 9.9% (+2.2%) and High-Yield (+5.99%). Foreign bond return is a combination of return in foreign currency and foreign currency gain/loss. In a bond portfolio context adding 20% foreign bonds adds some value by increasing returns and Sharpe Ratio somewhat if currency left unhedged (If currency is hedged then portfolio return decreases, though Sharp Ratio increases slightly)

-Strategic Asset Allocation: There no significant change when Treasuries and S&P500 are replaced US Aggregate bond index and Russell 3000, so for diversification must look elsewhere. Adding MSCI EAFE to US stock/bond shifts efficient-frontier up (i.e. higher return for same risk or lower risk for same return) providing an alpha of about 0.4%. Russell 1000 Value adds a premium of 0.5% over Russell 3000 or S&P500. Using the premium method the MSCI Emerging Markets premium of 2% over S&P500 leads to MSCI EM return of 11.4%. Yale’s David Svenson provides premium estimates for main asset classes of returns in Yale Endowment annual report.

-Hedge Funds: In their early days hedge funds meant that they hedge one set of assets against another; nowadays they can be almost anything ranging from absolute return to directional (strategies: long/short, market neutral, convertible arbitrage, event driven, global macro, managed futures, fixed income arbitrage and fund of funds). Unlike mutual funds which have symmetrical fees, hedge fund fees are asymmetrical composed of a management fee (1-2%) and an incentive fee (typically 20% of the upside). One of the differences between hedge funds and Private Equity (PE) or Venture Capital (VC) funds is that HF are relatively short-term and the latter two of long-term (5-10 years) in nature. On the surface, hedge funds appear to offer higher returns, lower standard deviation and higher Sharpe Ratios and correlations of 0.6-0.8 with US stocks, generating alphas of 6-8%. However other studies challenge these results due to hedge fund reporting biases (backfill- start reporting after they have some interval of good performance, survivorship- only live funds report, and liquidation- stop reporting prior to liquidation); these biases dramatically reduce the previously reported performance. Even worse is that the dispersion of HF managers’ performance is significant (Q1 median= +8.6%, while Q3 median= – 8.1%, indicating that manager selection and access is everything in hedge funds, so average investor is guaranteed to be at a disadvantage. Fund of funds may provide the advantages of due diligence, access and diversification across different strategies, which are likely eaten away by another layer of fees. In the portfolio context HF might provide some small advantages.

-Venture Capital (VC) and Private Equity (PE)- VC and PE funds provide access to non-public enterprises, but usually require many years of lock-up. VCs usually take partial ownership whereas PEs typically take 100% ownership (buyout). Organizationally they are similar with general and limited partners, requiring 10 year commitment with fees typically 2 and 20. As for hedge funds, manager selection and access are key, so not appropriate for average investor. For buyout firms the source of returns is leverage and restructuring.

-Real Assets- Real Estate (REITs income producing), Home Ownership and Commodities

– Real Asset- Real Estate (Svenson recommends 20% of equity allocation to REITs , Clements recommended buying a smaller home and investing the difference); performance is similar to Russell 2000 Value but with a correlation of 0.5-0.6 with S&P500, and Marston recommends including about 10% in the portfolio as this shows a small performance improvement.

-Real Asset- Home Ownership- Real appreciation ranges from CA= 2.6% to NY State= 1.7% and FL=0.4% (1975-2000), and in some cities NYC= 3.1%, SF= 2.8% MIA= 1.1% to ATL= 0.2% (1978-2009). High appreciation occurs when there is both high population growth and limits on land use. Factoring in appreciation and leverage home owners is an inferior investment as compared to REITs and S&P500.

-Real Asset- Commodities- Investment in Commodities can be done in one of four ways: direct ownership, stock of producers, passive investment in commodity indexes (GSCI-production value weighted-70% energy and DJ-UBS-liquidity in trading weighted-33% energy) implemented via futures and active investment via managed futures; but realistically the average investor can only use the passive commodity index based approach. Returns historically depending on the period considered returns are somewhat lower or higher that S&P500 with similar Sharpe Ratio but uncorrelated with stocks. So commodities appear to be the ideal alternative investment, with returns similar to S&P500 but uncorrelated with it. However note that commodities have very low correlation with inflation, so they are not necessarily a good inflation hedge. At the portfolio level one could replace 10% of the S&P allocation with commodities, though again it depends on what time period one uses to draw one’s conclusions. As far as Gold is concerned, Marston argues that it is unattractive due to low return and very high volatility and only 0.1 correlation to inflation (again not a great inflation hedge, though recently it is acting as a ‘insurance’ currency debasement).

Asset Allocation with Alternative investments- REITs provide geographical and asset type diversification, so consider replacing 20% of US equity in a conventional portfolio (e.g. US stocks=45%, foreign stocks 30% and bonds 25%). For a High Net Worth (HNW) individual’s portfolio one could include 10% real estate, 5% commodities, and 10% hedge funds for an incremental alpha (risk adjusted return) of about 0.7%. For Ultra-HNW (UHNW) portfolios (>$20M) where one could lock up 10% of assets for 10 years Private equities and/or Venture Capital could also be included. Alternatives can improve risk adjusted returns of a portfolio, even just using REITs.

-Spending Plans for foundations (theoretically live forever)- There are typically two approaches to spending “absolute” (starting with some percent of today’s value, then dollar value adjusted annually for inflation) or “proportional rule” (spending set to some percent of each year’s value), the latter approach allowing a somewhat higher spending rate than the former. At 5% withdrawal rate probability of failure as a function of stock allocation is about 22% at stock allocations of 50% or higher; so depending on the risk tolerance of the foundation one could select 50% stock allocation for low risk tolerance, but one gives up significant upside as opposed to 75%. Foundations need to set their spend rates based on the expected real return of the portfolio.

-Spending Plans for individuals/couples in retirement- Median life expectancy for 62 year old male/female/couple is 85/88/92 and 25% of males/females live at least to age of 92/94. Unlike foundations which can go out of business, “a retiree must struggle on”. Retirees not only have the uncertainty associated with (sequence of bad) returns but must also live with longevity risk. Assuming no bequest motive and a 2.5% inflation and using the “absolute” approach to spending at 5% an individual/couple have 10%/19% probability of exhausting their assets, whereas at 4.5% only 4%/9%. Assuming a 50% bequest motive and using a more flexible spending approach whereby half the assets are spent using “absolute” and the other half using the “proportional rule”, the using a 5% spend rate probability of failure for a couple is 19% for no bequest and 31% with the 50% bequest. However using a 4.5% spend rate the failure rate drops to 9% with no bequest and 18% with bequest. So to minimize the risk of running out of money base, spending on return after inflation, spend rate should be aimed to be less than real return and since we don’t really know what the future might hold it is better to be conservative (so to increase income you might even consider annuitization). Very important to remember to rebalance when times are good and bad, back to the Strategic Asset Allocation, but of course it takes an extra strong stomach to do that after a market swoon like 2008-9!