Nowadays, you'd be hard-pressed to miss the ads on television or in newspapers and magazines either touting gold for sale or pitching you to sell your scrap jewelry to a refiner. Gold, to put it plainly, is in demand.

There's nothing new about that except perhaps the ever-rising price targets and fever pitch taken by the infomercial gold sellers.

Sensible advisers have long counseled investors to hold at least some part of their portfolios in metal or metal proxies, either as an inflation hedge or as a stabilizer in times of crisis (to see gold's effectiveness as a bulwark against financial nastiness, see "A Picture's Worth A Thousand Words (Or Dollars)."

One such adviser, Roger Nusbaum, declared in a Hard Assets Investor interview ("Nusbaum: Buying Agriculture") last week that he'd put up a 2% portfolio position in gold, saying " ... it's not a big bet. But I don't think you need big bets to capture the effect."

Well, just how big must your gold bet be to feel an effect? After all, many gold advocates have pitched positions as large as 20% - 10 times larger than Nusbaum's - in their recommendations.

We'll keep things simple for comparison's sake. Suppose we use a classic portfolio allocation of 60% equities/40% fixed income as a control. The S&P Composite 1500 Index can serve as a proxy for the domestic stock market, while the Barclays Capital 10-20 Year U.S. Treasury Index stands in for the fixed-income universe.

Over a series of discrete time frames ranging from three months to 14.33 years (the length of the data string for the shortest-lived asset class in this study), this simple portfolio generated a time-weighted annualized return of 4.22% with a standard deviation, or risk, of 10.11%.

Overall, we can assign a "grade" to the investment mix, derived from its reward-to-risk ratio - in this case, 0.42. The higher the ratio, the more return is produced for a given level of risk.

Table 1: Asset Allocation Through 29-May-09

60% S&P 1500/40% Barclays Capital 10-20 Year Treasury

To add gold to the portfolio and allow for a fair performance evaluation, we've got to pare down the size of our original asset classes, but maintain their relative weights. Thus, for any given allocation of gold, the ratio of stocks to bonds ought to be 3-to-2, as in the original 60%/40% mix. A 2% commitment to gold, as advocated by Nusbaum, would mean allocations of 58.8% and 39.2%, respectively, to stocks and bonds.

Thus tweaked, our portfolio, shown in Table 2, would have produced historically better performance over all but the longest time interval, reflecting gold's bullish bent for most of this decade.

Table 2: Asset Allocation Through 29-May-09

2% COMEX Gold/58.8% S&P 1500/39.2% Barclays Capital 10-20 Year Treasury

Nusbaum's correct when he says you only need a small allocation to feel gold's effect. Of course, a larger dollop, as advocated by some of the metal's devotees, is bound to produce a more dramatic outcome. Giving over 20% of one's portfolio to gold, for example, would have cranked up average returns - again, for all but the longest interval - while significantly dampening volatility.

Table 3: Asset Allocation Through 29-May-09

20% COMEX Gold/48% S&P 1500/32% Barclays Capital 10-20 Year Treasury

 Prev12Next