The impact of using margin in down markets on terminal portfolio values
This is a continuation of an existing examination on using margin in retirement. Here’s the previous article…
The short version: If markets are down, what if we used debt instead of selling part of our portfolio to fund our retirement? Over time we’ve built up some additional wrinkles. (When exactly do you use debt? When do you pay it back?) We ended up finding that using debt meant a retirement in 1966 went from “failure” to “success”. (Read the previous post for the inevitable caveats.)
That said, so far we’ve looked at just a single year. I still have my doubts about the strategy but it is at least interesting enough to explore more. Let’s look at some more years!
The first thing we are going to do is compare “using debt” to “not using debt” for every 30 year retirement between 1955 and 1987. Why start in 1955? Because that’s the furthest back the FRED data for Federal Reserve rates go. And that’s what we use to determine the interest we’re paying on our debt. It would be better if we had data going further back — including the Great Depression, for instance — but we’ll run with this for now. That gives us 33 retirements to consider.
The first thing we see: 16 out of those 33 retirements never used debt. That number may seem low but remember that our sample (1955–1987) is dominated by the crashes of 1966, 1969, and 1973–74. If we extended the same in a few years in each direction we’d quickly see that most retirements never need to use this complicated debt thing. That’s not exactly a surprise. We know from years of Safe Withdrawal studies that the vast majority of retirements are drama free.
Are there any retirements where using debt made things worse?
Actually, no. I was a bit surprised by that but, at least in this model, there’s no point at which you are worse.
(Important! Keep in mind the behavioural aspects of this strategy. The final outcome might not be worse but the journey might not be much fun either. Using this strategy will mean being an advanced age and having lots of debt. More on this below.)
Using “no debt” there were two times we ran out of money before the full 30 years: 1966 and 1969. We saw last time that “using debt” meant that 1966 squeaked out a successful retirement.
What about 1969?
If we don’t use any debt that we run out of money in July 1993 (we wanted to last until 1998).
If we use debt, then we last until May 1995.
Still not a great outcome. We ran out of money. We weren’t saved. But we lasted almost 2 years longer. And definitely not worse.
How much better?
So using debt doesn’t make things worse. And it sometimes makes things better. How much better?
Anytime the green line is higher than the blue line, “using debt” worked out better. We can see that from 1959 to 1974 using debt always worked out better. (Before 1959 and after 1974 there were no differences.)
But how much better? Over that 1959–1974 time period, the average improvement was $1,265,780. (That’s in nominal dollars, not constant dollars, so taking an average like that is a bit bogus but…)
That’s…actually way better than I expected. The chart above makes it hard to tell how big the improvement is, so let’s look at a raw data table.
In many cases the improvement, though large in absolute numbers, is not as impressive as you might have thought. Going from $8.9 million to $10.6 million is nice, sure. But you’ve already got way, way more money than you know what to do with. The marginal value of the dollar means that extra money probably doesn’t have a big impact on your retirement. That said…it is still more money.
But I’ve highlighted two especially intriguing scenarios. In 1965 instead of dying with $730,000 left in your portfolio you end up with $2.1 million. That feels like a material improvements. Especially since the $730,000 is in nominal dollars. At that point your portfolio is down to 15% of what you started with and the portfolio only holds about 3.8 years of expenses.
The next case is even more intriguing. In 1968 you go from “just barely lasting all 30 years” and dying with just $140,000 left. Your portfolio is down to 2% of where you started and you’ve got about 9 months of expenses left.
On the other hand, by using debt, you have over $4.2 million left. That’s 101% of your original portfolio and represents 25 years of expenses.
Next time, let’s explore the mystery of 1968. Why does using debt make such a massive difference? And does that make us more sympathetic to using it in our own retirement?
Coda: Sleeping well at night
Earlier I said to keep in mind the reality of carrying this debt in retirement. What amounts are we talking about? In general, when you start using debt with this strategy, you need to be prepared to use a lot. We’re not dabbling here.
When debt gets used the median amount is $631,000. Nearly 25% of the time, you will have over $1 million in debt.
That’s in absolute numbers. In terms of debt-to-asset ratio we’re talking about a median ratio of 30%. (That is, for every $1 in assets you have 30 cents in debt.) And 17% of the time (3-out-of-17) you slam into your broker’s 50% margin limits and are subject to on-going margin calls everytime the market sneezes.
Even if this strategy looks okay on paper…are you really going to want to have $1 million in debt when you are in your 70s and retired?
