Saturday, May 11, 2019

The Gold Conundrum And The Thing About Portfolios

After my piece on gold silver ratio was published, one of my acquaintances read it and asked me this question: is gold (or commodities in general) really necessary for The Bedokian Portfolio?

As usual, my response is: it depends.

Elaborating on my answer, I shall use the Portfolio Visualizer (seems obvious that this place is becoming my go-to site) to carry out a simple back test, with three versions of the US Bedokian Portfolio and benchmarked to the Vanguard 500 Index Investor, representing the US equity market.

  • Bedokian Portfolio 1 (Portfolio 1): The balanced Bedokian Portfolio; 35% equities, 35% REITs, 20% bonds, 5% commodities (gold) and 5% cash.
  • Bedokian Portfolio 2 (Portfolio 2): We shift the 5% component from gold to equities, hence 40% equities, 35% REITs, 20% bonds and 5% cash.
  • Bedokian Portfolio 3 (Portfolio 3): We shift the 5% component from gold to REITs, hence 35% equities, 40% REITs, 20% bonds and 5% cash.

The period covered is from 1994 to 2018, as 1994 is the earliest year that REITs data is available on the Portfolio Visualizer site.

Fig. 1a – Portfolio returns, table view, 1994-2018. Inflation is not factored in.

Fig. 1b - Portfolio returns, graphical view, 1994-2018. Inflation is not factored in.

From Figures 1a and 1b, in terms of returns, the balanced Bedokian Portfolio lost out to the other two alternatives, which in turn trail badly against an all-equities portfolio. Risk-return (Sharpe and Sortino ratios) and volatility wise (standard deviation), the balanced portfolio had a slight edge over the rest.

So can we say do not add gold? Not so fast.

What if we use the period between 2001 and 2010? This decade covered the aftermath of the dot-com bust, the subsequent super bull, the Global Financial Crisis (GFC) and the follow-up recovery.

Fig. 2a – Portfolio returns, table view, 2001-2010. Inflation is not factored in.

Fig. 2b - Portfolio returns, graphical view, 2001-2010. Inflation is not factored in.

Interestingly, the balanced Bedokian Portfolio gave the most returns, yet still scored well in risk-return and lower volatility than the rest.

By now you may be wondering: these numbers are just hindsight statistical play, and I could just find a favourable period to prove my point. Yes, in a way that is correct, but what if in real life there was really a person who had a 10-year investment plan starting from Jan 2001; the gold allocation would had given additional returns (and lower volatility and better risk-return ratio) if he/she decided to withdraw all at Dec 2010 as compared to the rest. 

In my ebook I had stated that commodities are a form of insurance against market volatility1. Though in the examples of Figures 2a and 2b, the difference of protection was not much, but at least it proved its insurance role.

Now The Thing About Portfolios…

Bringing in a related point, in my opinion there is no best strategic asset allocation-based investment portfolio around (and yes The Bedokian Portfolio is no exception). Each one of them has its own strengths, weaknesses, best and worst performance at different points in time and differing duration. So in summary, there could be one portfolio that is good at some moments, but not all of the time.

Yet I would like to emphasize that having a diversified portfolio of various asset classes still trumps a one-asset-class-only portfolio, especially in managing returns and risks at the same time.

And I shall end off with this disclaimer: past performance is not an indication of future results.

1 – The Bedokian Portfolio, p37

Assumptions on Portfolio Visualizer Data

Asset class representations used in the data: Equities = U.S. Stock Market; Bonds = Total U.S. Bond Market; REITs = REIT, Commodities = Gold; Cash = 1-month Treasury Bills. All dividends are reinvested and transaction costs (e.g. commissions) are not included.


To see the results (Figures 1a and 1b) in its entirety for the 1994-2018 period in Portfolio Visualizer, click here.

To see the results (Figures 2a and 2b) in its entirety for the 2001-2010 period in Portfolio Visualizer, click here.


Standard Deviation –

Sharpe Ratio –

Sortino Ratio –

No comments:

Post a Comment