You all surely know this narrative: 60/40 stocks/bonds portfolio, anticorrelated assets, when one moves down, another moves up, you rebalance and in general your portfolio moves forward with a lower volatility. Except that:
- we here have CHF as a base currency, the risk free rate for CHF is -0.75% p.a. and bond yields are negative. So if after taxes you don’t lose money with bonds or cash, consider yourself lucky.
Which might be not a problem by itself. But it is in combination with
- the interest rates are expected to rise sooner or later, and I don’t think anyone would consider a further significant decrease possible.
Bonds is a “mathematically” simple investment instrument and once rates will go up, it is clear that they will be crushed. So for a long time no bonds for Swiss investors.
What is left is cash instead of bonds. But in this case mathematics suggests that we should buy stocks with borrowed cash, so no rebalancing.
So, is 100% stocks portfolio is an only viable option? Maybe.
There are also all these suggestions that in views of raising inflation we should also invest in commodities and gold. Therefore I decided to check if historical data suggest that commodities could be a good diversification for a stocks portfolio.
For historical simulations I used https://www.portfoliovisualizer.com/ which is a wonderful tools for portfolio analysis and back-testing of investment strategies. It is focused on US, but it is very quick to do analysis, so it is enough for my purposes.
First I looked at “asset classes”, which show how investments in asset classes (without specific investment instruments) were performing. The advantage of this analysis is that historical data are in principle available since 1975. I checked monthly correlations of “Gold”, “Precious Metals” and “Commodities” with “US Stock Market”.
The comparison is only possible since the moment the data for all selected asset class are available, so I start with all assets of interest and the narrowest range (until now) and then sequentially increased the range by removing assets that limit the range. One gets slightly different results for different ranges, but the trend is clear.
Gold has a lowest correlation coefficient with stocks, around 0; precious metals 0.35-0.5; commodities slightly over 0.5. For comparison, different baskets of stock can have a correlation coefficient as low as 0.60 (“US Micro Cap” and “Pacific Stocks”).
Then I looked at correlations of specific commodities with stocks (VT and VTI) using “Backtest Portfolio” tool at portfoliovisualizer.com. At etf.com I looked for US ETFs on asset calss “Commodities” and for a specific commodity I chose an ETF that has longest running history (it is also usually one with the highest AUM). With a filter “Physically Held” I have chosen: gold GLD, silver SLV, platinum PPLT and palladium PALL ETFs. All other commodities are not physically held, so it means investment is via futures. But I still took a look at metals: aluminium JJU, copper CPER, lead LD, nickel JJN, tin JJT.
The comparison is only possible since the ETF’s existence, so I again started with a narrowest range (to now) and then increase it by removing ETFs that limit the range. The correlation coefficient is somewhat depending on the time range, but overall there is a clear trend.
Gold has a lowest correlation with the stock market, below 0.1 with VTI and slightly higher with VT. Silver, aluminium, lead, nickel, tin have correlation coefficients in the range 0.3 to 0.4. Interestingly, the correlation coefficient of palladium with stocks is almost as low as of gold if one looks at last five years, for longer time it is close to 0.3-0.4, like for the previous group of metals. Did palladium become a safe heaven recently?
The highest correlation with stocks is seen for copper (copper interconnects for integrated circuits?) and, to my surprise, platinum, being in the range of 0.5-0.7, almost like different baskets of stocks.
So as one could have expected from the first analysis, gold is an asset that is least correlated with the stock market, without having a negative yield on maturity (or a positive one, anyway). Gold might be an uncorrelated asset that provides diversification to the portfolio. Investment in any other commodity for the sake of portfolio diversification is pretty useless, the correlation with stock markets is too high. You might be prompted to invest in oil because currently oil futures are running over the top, but think about April-May 2020: would investment in commodities help to stabilize a stocks portfolio?
To be hopefully continued, I am trying to analyse historical performance of stocks+gold+cash portfolios.