To review there are two assumptions in Virtus's comparison of stock returns and paying off a mortgage that people in the thread consistently disagreed with:
1. Reducing stock market returns based on the rate of inflation but not inflation adjusting mortgage payments/remaining balance. One really REALLY has to either apply inflation to both sides of that equation or neither.
2. Applying tax every year to unrealized capital gains.* Everyone agrees this one is wrong, the argument is whether it causes significant variance in the final outcome or not.
How big of a difference do these two assumptions make?
#1. The difference between the non-inflation adjusted CAGR of the stock market and the CAGR of the stock market is 9.1% vs 6.9%.
#2. Applying 15% tax every year to unrealized capital gains reduced the CAGR further to 5.87%.
Over the course of a 30 year mortgage, $10k will compound to $125k at 9.1%. After applying capital gains tax at the end, the final value is $108k. $98k of post tax appreciation. At 5.87% that same $10k would have only compounded to $52k. $42k of post-tax appreciation. Less than half as much.
There are not things than can be hand waved away as inconsequential issues.
I posted my analysis, both graphically and with spreadsheets of results
here. Rpr posted their analysis using their own code and using Virtus's spreadsheet after removing the one sided inflation adjustment and taxation of unrealized capital gains every year
here. All three found that the only historical year where paying off a 4% mortgage* beat investing in the market was 1925. BG used cFireSim to run the same two scenarios and found no historical year where making extra mortgage payments produced a superior return to investing in the market (
here).
*Of course in most years mortgage interest rates would have been higher than 4%.