Something on this post got me thinking about what framework would be to determine whether we could afford something. The post is meant to be light-hearted, but OP mentioned that at a net worth of $1 million, he would consider getting a Tesla. I got to wondering and created this 2-way table.
In short, it tells you what the effect of a one-off purchase would be on your net worth in 20 years. It's a little interesting for me to see how inelastic your long-term net worth is if the purchase is either a very small portion of your annual income, or if it's a very small portion of your invested assets. If something is 2% of your net invested assets, it doesn't really matter that it's 100x your annual salary. That's probably not that surprising, but for intersections in the middle, like an expense that's 50% of your current net worth and 50% of your current salary, the effect on future net worth is probably not as intuitive.
Personally, I'm thinking that if a one-off purchase decreases your long-term net worth by less than 3% and is something you do very infrequently, then you can afford it. Any thoughts on that? I suppose if your investable assets are scheduled to be well above what you need for a livable 4% in 20 years, you can make a purchase that haircuts you a bit more.
An article I’ve loved and often linked to is Dana Anspach’s piece on the rolling returns for various asset classes over different time periods. I think the article can best be summed up in 2 of the included charts:
In a given year, bonds and treasuries are indeed very safe with very little incidence of losing more than a few percent of value in a year. Equities over the 1 year period have wildly variable with higher highs and lower lows. In the worst-case single-year return over Dana’s sample, intermediate bonds were down 1.7%. The S&P 500 on the other hand was down almost 45%.
However, when annualizing returns over a 20-year time frame, the worst intermediate bond returns were 4.2% per year. The worst S&P 500 returns were 6.4%.
The punchline here is that while bonds are indeed safer in the short term if you have a long enough investment horizon, equities are a much better choice not only in terms of average return but also in terms of possible losses.
I decided to update this study on the S&P 500 for a larger window. The period between January 1935, and January 2021.
Overall, the trend from Dana’s original article still holds true. Returns over a 1-year period are wildly variable, but as we annualize returns over longer time periods, the range of returns becomes more and more narrow.
Sadly though, her point of the lowest annualized returns over a 20-year period being 6.4% no longer seems true. In fact, the lowest annualized return over that period is a loss of 3.6%.
That being said, expanding to a 30 year period does bring us back into the “always positive” territory, albeit at a lower return of 2.13%.
Hopefully, the result of all this is that investors with long time horizons can have confidence in planning for retirement with large allocations to equities knowing that more often than not, they could assume long-run annualized returns of about 6%.
I'll be working on a polished and edited video for how to use the various features on Peercents, but in the meantime, I thought it would be good to just do a quick recording of how to run a simulation. Apologies for the poor quality and lack of a script.
One thing I forgot to mention on the video is the fact that when entering expenses, do not include any loan payments or interest if you're entering those loans in the liability section. Those loan payments are already taken into account by that section and to include them on expenses or as a separate cash flow would double-count them.
As always, if you have any questions, feel free to leave them here on the discussion board.