Surviving the 1970s with debt & guardrails: escaping the debt spiral?
In my last article, I showed that the bad times of the 1970s lasted so long that using debt, instead of selling equities when they are distressed, didn’t work that well.
We saw that we ran out of money in November 1992, 26 years into retirement. Not using margin lasted until November 1993, a full year longer; so using margin was not only unsuccessful, it actually made things (slightly) worse than not using debt.
But I made one slight modification to my backtest and suddenly…
Notice that the green line (the new model) lasts the full 30 years. Admittedly, you don’t have a lot left. Just $182,000 left at a time when you’re withdrawing $16,000 a month. Still, it lasted the full 30 years instead of running out early. That’s a definite improvement! It even (technically) counts as success for the “use debt instead of selling distressed equities” approach.
Using this new model, you run out of money in June 1996, or 30.5 years after the start. Which is 2 years and 7 months longer than not using margin.
Remember, we wanted to see if using debt would let the 4% succeed where otherwise it had failed. And now we’re seeing a backtest that shows exactly that. So I was wrong. (Sort of.)
Why the difference? And does this new result really mean we should fully accept the “use debt instead of selling” model?
Changing how dividends happen
The fundamental change is how dividends are handled. I made it “more realistic”. The previous approach was a “total return” model. I lumped together price appreciation and dividends, which is pretty standard in this kind of backtesting. However, by doing that we missed out a slight nuance. Here’s an example of how the two approaches have slightly different results.
Both are identical until June 1970. We need $4,067 for our living expenses that month. But the market is down, so we need to use debt. How much do we actually borrow? In the “total return” model we borrow the full amount.
But in reality, we receive $2,758 in dividends that month. So a real person wouldn’t borrow $4,067. Instead they’d borrow whatever isn’t covered by dividends.
That is: the “total return” model assumes you immediately reinvest the dividends and then decide whether to use debt. But the “dividends” model assumes you spend the dividends and only then decide whether you need to borrow money or sell part of your portfolio.
In practice this means you will borrow slightly less money. But since we’re talking about debt, those slight differences add up. By the end of 1970, the “total return” model has $29,524 in debt (including interest). But the “dividend” model only has $10,359 in debt.
If we look at the total debt held by the two models over time we can see this initial small difference ends up having a large impact:
But wait! The strategy is a technical success…but it is realistic? I don’t think so.
When you turn 80, you have 50% debt and live the rest of your life like that.
Talking in percentages is a bit dry. In actual dollar terms it means someone being half a million dollars in debt.
And it means, especially later in life, having to force sell equities constantly to stay under margin requirements as your portfolio dwindles. Here are all of the margin calls…
Notice that one of them is over $200,000 (imagine how stressful that month would be). Let’s take a moment and reflect on that: we started this strategy because we didn’t want to have to sell equities when the market is down. But now we’re in a situation where we have to sell equities when the market is down. The market goes down a bit, we blow past our margin limits, and we have to sell a tremendous amount of equities at a depressed price to get back under our margin limits.
That $200,000 margin call blows up the y-axis and makes it hard to tell what the others are like. Let’s look at the margin calls just from ages 90–95.
The average monthly margin call is over $11,000. That’s $11,000 of forced selling just to pay down your debt to stay under margin limits. Each month.
So what have I learned so far?
- Debt can snowball. Since there’s compounding involved, a few extra dollars early on adds up.
- A more realistic handling of cashflow dramatically affects how much a real person would borrow.
- We eked out a few more years (2 years & 7 months) enough to technically cross the 30-year finish line but it wasn’t exactly a panacea.
- The strategy is quite fragile. The difference between success and failure is caused by relatively small differences in strategy & cashflow timing. That is pretty worrying because the future might not be exactly like the past.
- While the strategy is technically successful, it results in someone carrying large amounts of debt until they die and having to manage their margin on a monthly basis. It seems like it would add a tremendous amount of stress, in addition to possibly taxing the cognitive abilities of someone near the end of their life.
Still, we can’t deny that we’ve found success, even if it is of an unrealistic & technical variety. We did better in the 1970s by using margin debt — even at the eye-watering interest rates prevailing at the time — than by selling equities.
The next step will be to look at time periods other than just the 1970s. After all, if this strategy makes the 1970s better but every other time worse, we aren’t going to be using it, right?
Here’s a link to the spreadsheet showing both models
