Quote from Daal:
Interesting points.
With regards to Gold vs CRB. I'd say its complicated, the "CRB" includes interest income that is earned in the USTs that are used as margin against the futures position. Currently its 0%(The GCC ETF uses bills) but it won't be zero forever. The CRB index return against Gold in the 70's could be unfair due that(Imagine what kind of interest income one would have gotten, specially compounded. I don't have the numbers but I'd be surprised if they it doesn't beat physical gold), the income is also a cushion against vol
Also, its nice diversification.
With regards to long-term returns it seems that you are talking about REALLY long-term there. If that is so, I'd agree with the projections. If its more like 5-15 years I'd use different numbers
Yes but the point of commodities or gold in a portfolio is not really to make a return, it's to hedge against the scenario where the rest of the portfolio gets hammered i.e. stagflation or monetary debasement. Gold does that more reliably than commodities as a whole - gold trades pretty much solely as a monetary metal, whereas oil, grains, metals etc trade on industrial demand, agriculture trends & weather patterns, government regulations etc.
CRB is also more volatile, it goes down more in crashes, and it's potentially vulnerable to the indexing bubble, as well as the roll phenomenon that affects things like the natural gas ETF. CRB is also dominated by oil prices, making it highly correlated to stocks during recessions. Overall I"d say both the theory and the numbers show gold as a better portfolio diversifier than CRB.
About long-term returns, yes I was just listing what you'd expect from economic theory and long-term returns in the past. Stocks should return the free cash-flow yield + inflation, long-bonds should return their current yield, gold should track inflation, and cash should yield the same as long bonds minus the bond risk premium.
On a 5 year view then this is much different, but clearly it is impossible to predict what 5 year returns will be (if you could, you'd just go long the asset with the best anticipated 5 year return). The whole point of a diversified portfolio is the admission that asset-specific long-term prediction is impossible, therefore you just rely on probabilities (asset X is likely to return from y% to z% over 5-15 years) and diversify to smooth things out and provide as little correlation as possible between the assets.
When you set up a portfolio, you never know what it will return. However, if it is a sound portfolio, you DO know that it will earn the weighted market return of all its constituents, minus costs. A well-constructed portfolio should earn the best passive economic return that the market offers over that time period, without any need or ability to forecast. If you can 'forecast' too (even if only on a probabilistic basis), then you can run a trading portfolio on top of it and earn some alpha that way too. If you don't like drawdowns, run a very conservative portfolio - but even as little as 20% in stocks, 10% in bonds, and 10% in gold with 60% cash, is going to out-perform 100% cash in the long-run in all but the most bearish 2-3 year scenarios. And when it loses, it won't lose any more than a typical trading drawdown e.g. 10-15%. That is more than compensated for by your expected 2.5% annual return bonus - 10 years of that is 28%, from a portfolio that even 1929-32 could only cause a drawdown of say 10-15% peak to trough. 2008 would not even hit a 10% loss limit. And the rare times you do see a drawdown, you likely have a chance to buy stocks at bargain prices.
On the rare occasions when you get outright bearish on one of the assets, and want to have no exposure, you can just place a short in your trading account. Treat it like a separate naked short position, leave the index portfolio alone, and cover the trade once you are no longer bearish with conviction. Your portfolio will earn what it would have earned anyway (e.g. in 2008 would take a loss on the stocks and gold, offset somewhat by the bonds), but you'll have a profit on the trade to offset this, and have a passive portfolio ready to bank coin on any rebound. Net result is you reduce drawdown (or even turn a profit).
You keep your market timing where it should be - in the trading account. You keep your passive portfolio earning risk premia where it should be - in the investment account. In the long-run you should earn your net trading return + the net return of a passive diversified index portfolio. And since the two are either uncorrelated, or (for global macro) frequently negatively correlated, you should usually get the free lunch of higher returns with lower drawdowns. That beats sitting 100% in t-bills for a 10-30 year trading career. If the drawdowns are too much for you, just reduce the risk asset % in the passive portfolio until its 1929-32, 1937, 1973-74, 1987, 2000-2002, and 2007-2009 max DD would be low enough for you to stomach.