The risks to such allocation are the US risks that folks like Jim Rogers have been talking about for a long time. In Market Wizards he talks about the UK experience where they ceased to have a long-term bond market and, IIRC, their long-term bonds fell 70%. Now, that portfolio is VERY resilient. Even in a UK type scenario one will still be ok as gold has a BIG positive skewness and kurtosis, its an insurance asset and its in cases like that (the UK experience) that it pays off hugely https://www.creditwritedowns.com/2014/02/long-decline-great-british-pound.html That's the chart against the USD but since gold is arbitraged globally, UK residents had huge gains to compensate for losses on the bond-side (Its no wonder Jim Rogers owns some gold "as an insurance policy"). Also, stocks will readjust and eventually return their losses from fiscal/inflation problems. Long-term they are good inflation hedges. So overall, the portfolio will be alright. However, if one is worried about that, what the investor could do is to sell 30y bonds, stocks and gold (at the 1.75-1-0.5 ratio) and buy short-term maturities at an allocation higher than 15%. Perhaps 20%,25%. Whatever is more comfortable. The portfolio will give up returns (and probably, give back some Sortino/Sharpe Ratio due lower excess returns over the risk free rate) but whatever gets one to sleep at night
In terms of overall portolio construction, what I learned is that: -Most new asset classes diversification/rebalancing/All weather benefits seem to drop off at around 10-20% allocations. Stocks are high return high risk. Gold is low return medium risk. T-Bills are low return, low risk. Yet in all of these cases the computer doesn't want all that much of it. It rather "chase" high Sharpe/Sortino assets like Government bonds. I think that's important fact. When considering to add a new asset class to a portfolio (like REITs or something else). An allocation that makes sense is around 10%-15%-20%. And when thinking about building a good balanced portfolio "chase" high Sortino assets and add a few asset classes with 10-20% allocations. At what's that the computer is saying that works historically over big samples
So interestingly enough, I was worried about how uncertainty/extremistan type dynamics would call for lowering the exposure one has to equities given that it is a very vulnerable asset class to new unknown dynamics. But it appears that I dont have to be that concerned. It appears that gold is the asset that changes the most as you add/remove uncertainty When I looked at 1926-2016 data the computer wanted 12% in gold and 16-20% in stocks. When I looked at 1879-2016 data the computer wanted 2-3% gold and 18-20% in stocks (roughly). That was a long period of stability (pre 1926) that was followed by turmoil (Great depression, 30's dollar devaluation, 70's fiat inflation). The computer responded by going after 'insurance' through gold not by cutting down equities. So perhaps that's a good lesson right there. While equities are vulnerable to turmoil, as a long you own enough insurance with gold AND an appropriate ratio of bonds to stocks, you are going to be fine. I dont have UK data for 1900's (and I dont want to torture myself by going after it but I probably will) but I would imagine that the optimal allocation of gold there was probably higher than 12% as a result of the turmoil but I doubt equities were shredded in the % allocation. Its very good to have 20% in equities (or more if one has access to long-term duration bonds), there are a lot of diversification/rebalancing/all weather benefits. If one is concerned about the future, the way to hedge is to increase their allocation to gold (and/or foreign assets), not cut down equities
I am slightly sceptical that modern day investors behave in the same fashion the ones in nineteenth century did esp when it comes to gold cus I'd imagine back then the idea that gold is just a usual commodity was preposterous. It would also be interesting if you included international equities to avoid survivorship bias.
I'm sorta going to do that if/when I run these tests on UK and Brazil data. With UK, if I find the data, its going to be meaningful because its going to be almost 100 years worth of data but with other problems in it (different from the US). But I will add some international diversification (A UK portfolio with US stocks, US bonds and bills avaliable if the computer wants). Kinda like I did with Brazil asset returns
Since people in 18xx thought differently from people nowadays, I'd be inclined to see how things changed between the last 2 or 3 crashes to date and the full data set. Meaning set 1 say 2008-2016 vs full data set, and so on. Would be interesting. More is not always better.
Cortesy of Eurokopek, the duration ratio of the 30y vs 10y. Seems to match some calcs I run on annual return data This thing might go up or down depending on where yields go and how the market perceives risks. But using a 2-1 ratio (to build my modified optimal portfolio) as a rule of thumb doesn't seem all that bad. In fact, it would 'over hedge' stocks a little as of right now So while one is taking duration risk, one would be 'over hedged' against a stock collapse (economic collapse). That's the trade-off, if you own little to no duration, you are exposed to disinflation/deflation (a depression is going to hurt you), if you own too much, you are exposed to an inflation/fiscal collapse. Either way there is risk but owning the optimal allocation is better than the 'pundit' allocation out there
There is a site that lets you backtest some portfolios with more recent data. https://www.portfoliovisualizer.com/ They even got a lot more asset classes I encourage you to run some tests and report results. It might help with putting different perspectives in portfolio development
Interesting, will do once I get my reading glasses fixed; sat on them and the old ones aren't up to it.
I was reading about that 1970s UK fiscal crisis https://en.wikipedia.org/wiki/1976_IMF_Crisis This was an interesting crisis because even though, Jim Rogers says, that long-term UK bonds fell 70% and they lost market access to long-term duration bonds, the UK didnt default on these bonds. I was able to find only a few instances of UK defaults https://en.wikipedia.org/wiki/List_of_sovereign_debt_crises The last default was in War Bonds in the 1930s So if the investor stuck with the long-term bonds, they got paid back in full. There was a problem in the level of inflation corrosion in these bonds (as the USDGBP chart shows that the market wasn't so happy about the UK situation) but still, that's a lot better a default. So essentially, you had an inflationary period that arose due fiscal problems. A well built all weather portfolio can withstand an inflationary period. Gold and stocks will do alright. Gold will have a big surge (driven by the collapse of the domestic currency) and that surge will lead to rebalancing (one will sell gold on the way up as it gets larger in the % of the portfolio) and buy productive assets with it (stocks and bonds that are falling). Stocks will fall as a result of the tormoil but as companies adjust prices up (as a result of inflation), they will eventually revert the temporary capital losses to the investor and he will probably have a real positive return. Bonds will struggle as the inflation kicks in but how is that different from any enviroment that is not beneficial to bonds? Its the guy sitting on a 60/40 portfolio (or worse a 100% bond portfolio) that will get killed. The balanced is the one that will actually cushion the investor and save him. Bad periods for bonds or stocks or gold is already expected in a well designed portfolio