For the purpose of this experiment I pick SPY, TLT, IYR, GDX, but I will be first to admit that his advise needs to be taken more broadly to be used sensibly. Obviously, his real estate holdings are mostly actual ownership of rental and vacation properties and not a REIT ETF (IYR) - stocks of real-estate investment trusts, which also include prisons, nursing homes, land and more of god knows what. Faber's recommendations about gold couldn't be more specific - physical ownership in immediate possession or in a vault outside of US (rich people shit). And I'm sure that stocks and bonds are more diversified than just S&P500 and US Treasuries 20+ years to maturity. Having acknowledge all this, my interest is in systematic investment strategies using financial instruments only, preferably in tax-efficient account (IRA, 401k etc), using low cost mutual funds if possible.
I start back-testing from January 3, 2012, as I have vivid memories of this recent period as well as my own real-life records and results to compare against. Symbols picked for portfolio have adjusted prices available, so they already have dividend income build into historical prices, and closely resemble actual gains or losses. There will be a noticeable difference in Treasuries and REIT's mutual funds actual account balance vs these simulated results, but only if held continuously for more than a year. Mutual fund will actually make more money, because of more efficient re-investment of dividends - basically investor makes money on the money faster.
Part 1. Straight 25.
If I put equal amount into SPY, TLT, IYR, GDX on 1/3/2012 and held until today, I would make about 25% profit in total. This is not annualized. It's all of it. Minimum Standard Position size $3000, for 12K initial investment, grows to a bit over 15K. Mind you that S&P doubled over this period, and even Bond Fund had a 50% move. Clearly it was a Gold fund that torpedoed investment results. Even recent rally in gold shares was unable to turn a tide here - gold fund is only about 8% of allocation now, and sits a whopping 60% lower than 5 years ago (it was down 75%)... some store of value... he-he.
Part 2. Five for 25.
Traditional approach to asset allocation is to re-balance as soon as 1 or more components deviate by more than 5% from a model. Of course it was a Gold fund that caused first re-balance during summer of 2012 by falling to 19% of portfolio for 2 consecutive weeks. Second re-balance came in Spring of 2013. This time not only Gold fund fell as low as 16% allocation, but also S&P rose to 31%. As a result of this, the big winner was cannibalized to replenish a worst loser. And it would not be a last time. End of 2013 brought back exactly the same kind of rotation, only this time account balance was actually DOWN -5.7%.
2013 was one of those fantastic years when everything went up... except gold... and it took Faber's 25 with it. Gold remained the cause of portfolio re-balancing in 2014 and 2015, during which time account showed temporary gains of about 10%, but ended about flat for the beginning of rip-roaring 2016. Now, we are talking YEARS with nothing to show for it.
2016 was kind to Dr. Faber. After briefly dipping in a red (again), portfolio took off with powerful gold rally and was up 40% by summer. There it needed a re-balancing, only this time gold was largest position in portfolio and S&P smallest. Pay attention... same 40% as Straight 25 version from above. This re-balanced portfolio now sits just shy of 15K total... Basically, re-balance or don't, you are fakt (german accent).
Part 3. Supply 25 Plus 5.
What is this crazy infatuation with Gold? Or trust in Bonds? Or hatred of Stocks? Why not think independently and let a collective wisdom of market place to guide ... and blah-blah-blah. Since re-balance naturally comes at-least once a year, I can use a long moving average to avoid a fund in downtrend. Ones a fund goes under - it's sold and money supplied to remaining positions. Trade will require 2 weeks away from average to trigger, at which point entire account will re-balance to accommodate equally each position in portfolio. A deviation of 5% will also cause portfolio to re-balance. Mutual funds are perfect for this, as they allow fractional share count and don't have any trading costs (they also require 1 month hold, but it's not a problem - rather a benefit, and I proved it previously).
Clearly, Gold was knocked out by 2013 and I used late exit in my back-test. I actually sold all gold stocks back in 2011 and was generally cold to it for a while. Gold was no longer an option until 2016, but even that trade yielded no profit in back-test. My experience with Gold stocks in 2013-2016 had been wildly different, considering that I bought dips and managed to sell enough during summer 2016 run-up that what's left costs me practically nothing....
But it's not a part of this exercise...