NiceUnparticularMan wrote: Thu Sep 11, 2025 8:14 pm
So the standard stock return model is:
Equity Return = Current Dividend Yield + Real Earnings Growth + ∆ Valuation + Currency Adjustments
The equivalent “null hypothesis” is the last two terms have an expected real return of 0, and so the basic driver of long-term stock expected returns is dividends + real earnings growth.
Generally speaking, in US stock history and in fact in the history of other stock markets, this is a well-confirmed hypothesis. However, there may be a slight problem with the hypothesis of ∆ Valuation averaging 0.
Namely, it is possible that in general, future stock earnings are getting less risky, equity risk premiums are declining, and stock valuations are increasing as a result. If so, in theory this could continue until stock valuations converge on low-risk bond valuations. And for that matter, bond valuations themselves may be experiencing an overall increasing trend.
So I would agree it is true it is uncertain whether or not the expected return on ∆ Valuation in particular should be modeled as 0.
But the dividends + real earnings growth part remains extremely well-confirmed.
Again the problem with speculating on spot commodities is there are no such components of return as dividends or real earnings growth. There is basically only the last two terms. So our inability to reject a null hypothesis as to those terms is a far more serious issue when it comes to speculating on spot commodities versus investing in stocks.
Flawed (incomplete).
Fiat money = debt (money created out of thin air, destroyed once the debt + interest is repaid. As one persons pays off their last mortgage payment so another is inclined to be taking out a new (higher value) mortgage; As one Treasury bond matures so another is inclined to be sold, as are others periodically created/sold to fund paying the interest on existing debts) that historically has expanded at a 8.6% annualized rate (since 1966), i.e. faster than CPI (real) – that grew at a 4% annualized rate since 1966.
https://fred.stlouisfed.org/series/GFDEBTN/#
(log scaled, where that trend/progression rate similarly applies to decades prior to the start of that 1966 based data).
USD price of gold over similar period yielded a similar annualized rate of return. i.e. as more USD has been created out of thin air so the price of gold in USD has tended to rise in reflection of that (but in a volatile manner).
Other currencies have been inclined to debase equally as the USD. As the US prints USD and spends on military/space/whatever so others who have lent to the US print more of their own currency to buy more USD debt – to otherwise avoid having their existing US loans devalued.
CPI (real) is slowed by productivity, a single man in a machine working a farm that previously took a army of manual workers to do the same work but slower.
If the price of gold is assumed to rise in reflection of USD debasement, and that is faster than CPI then that difference is a real yield. In practice however that has been attenuated by volatility, over some periods/decades it has lagged, typically when stocks have performed very well, over other periods it has yielded great rewards, typically over periods when stocks have faltered. Accordingly a good practice is to hold some of both stock and gold, each tending to somewhat hedge the other and reduce the overall combined portfolio volatility.
A extreme gold-bug might opt to stake 25% of their otherwise 100% gold holdings and buy 25% into a 3x stock position, such that they held 1.5x leveraged 50/50 gold/stock exposure, with in effect 50% borrowed/debt (cost of carry of the leveraged stock position), here’s how that has compared to 100% S&P500 since 2016 (up to the end of August 2025)
https://www.portfoliovisualizer.com/bac … hDmcLTHtO7
75% of their portfolio value in-hand, no counter-party risk, bearer rather than depository, that can be passed on to others directly without any online or third party involvement.
Using actual gold prices and synthetic data for 3x stock back from 1933, when gold/dollar convertibility was ended, comparing 25/75 3x stock/gold with 100% stock and a similar reward was apparent, 7.5% annualised for all-stock with a 18.5% standard deviation in yearly total returns versus 7.3% CAGR with 19.2% stdev for 25/75 3x stock/gold. Depths of drawdowns were similar, but not aligned. Similar 30 year MaxWR% (SWR support), but again not aligned. 50/50 of both of those was better than either alone (supported a 5% 30 year SWR). Evidence that your math formula assumptions are incorrect.
