Leveraged Investing Works on Paper, but Nobody Executes Step Two
The paper says young people should lever up on their lifetime wealth, but clients who pay off the mortgage never borrow back. Sequence-of-returns risk is only a problem when withdrawals are fixed; make them variable and the portfolio gymnastics become unnecessary.
The argument · tap a timestamp to hear it
Borrowing to invest works in theory, is hard in the mind
A resident physician asks: is it worth borrowing against a line of credit at roughly 4% to put into a 100% stock ETF? Benjamin cites the 2013 Journal of Portfolio Management paper "Diversification Across Time": people who are young, whose future human capital is high and bond-like, and who currently hold few financial assets should in theory use leverage, because their ideal stock allocation is far above their current investable assets. The paper argues for setting the stock share based on "lifetime wealth" rather than "current savings," and for the young even suggests investing 200% of current savings. But Benjamin's landing point: it ultimately comes down to risk tolerance, and 100% stocks looks good on paper — not everyone can execute it.
— Benjamin FelixSamuelson himself publicly opposed the paper
After the paper came out, Paul Samuelson spoke against it in a 2008 talk, on the grounds that using leverage carries the risk of being wiped out entirely, and you have to stay at the table to collect the long-run result. The authors' rebuttal: having your financial assets wiped out doesn't really matter, because most of your assets are future income you haven't earned yet — technically true, but Benjamin adds: bankruptcy still hurts. Robert Merton also agreed that leverage makes sense for the young, but opposed using margin, arguing that options, or building leverage into a financial product, are more appropriate, because "it's crazy to expect people to execute this idea themselves."
— Benjamin FelixDan Bortolotti doesn't question the numbers, he questions the behavior
Dan Bortolotti of Canadian Couch Potato reviewed the book these authors wrote on the back of it in a 2010 blog post. He doesn't dispute the paper's numbers and thinks the theoretical and empirical results probably hold, but the behavioral layer makes the whole strategy hard to implement: you're borrowing to invest, and you could lose a lot or be wiped out. Benjamin adds the picture: if you blow up your RBC line of credit, RBC probably won't like you much, and even maintaining the banking relationship may be difficult. The authors acknowledge the psychological aversion is real, but argue people don't truly understand the cost of forgoing the strategy.
— Benjamin FelixAdjust asset allocation first, then talk about borrowing
Benjamin offers a practical order of operations: think of borrowing as a negative fixed-income position. Rather than borrowing 500k, first move the portfolio from 80% stocks to 100% stocks — in financial planning software the results come out very close. So if you're at 80% stocks and considering leverage, maybe first go to 100% stocks; if that already gets you the expected outcome you want, you may not need to borrow. Louis adds this can double as a behavioral test: a 70/30 investor can first experience 100% stock volatility, and if they can stomach it, then consider the next step.
— Benjamin FelixAfter paying off the mortgage, nobody wants to borrow it back
A client asks how to optimize the mortgage and free up money to invest, and the advisor's plan is: use taxable assets to pay off the mortgage, then borrow back against the home to invest. The model was run, and the long-term net worth projection for one family was a "huge improvement." The result: once the mortgage was paid off, that client never borrowed back to invest. Ben says he's seen this many times, and almost nobody actually executes step two. The reason isn't that the math doesn't work — it's that the sense of relief from paying off the debt is too strong, and with improved cash flow, people simply stop at being debt-free.
— Multi-person conversationIt's not the sequence of returns, it's the sequence of withdrawals
Ben directly reframes the question: rather than thinking "how big does the portfolio need to be to not fear sequence-of-returns risk," think about the withdrawal sequence. Sequence-of-returns risk is only a problem when withdrawals are fixed — if you take out the same amount every period, several bad years in a row really do hurt; but withdrawals don't have to be fixed. Allowing variable withdrawals can raise lifetime spending overall while not raising (or even lowering) the risk of ruin: spend less in bad years, more in good years. This flexible spending comes from the lifecycle model.
— Ben FelixStatic allocation has the lowest failure rate
A 2016 paper, "Retirement Glidepaths: An International Perspective," used data from 19 countries spanning 110 years, from 1900 to 2009, to test three types of stock allocation strategies: declining, rising, and static. The conclusion is that the static strategy had the lowest or near-lowest failure rate, the highest or near-highest expected bequest, good upside potential, and the best downside protection. The authors specifically note that the static strategy of staying fully in stocks the whole way had the lowest failure rate, held up reasonably in tail-risk shocks, and had higher upside potential than the other strategies. Its standard deviation of outcomes is indeed higher, but the authors' explanation is: that standard deviation reflects uncertainty about how much better off the retiree will be after 30 years, not how much worse.
The optimal solution switches to 27% Treasury bills at retirement
One finding in the Cederberg paper directly relevant to the listener's question: if the investor is allowed to adjust allocation once a year, the optimal solution stays 100% stocks until retirement at 60, switches 27% into Treasury bills at retirement (equivalent to a high-yield savings account), then gradually shifts back to 100% stocks over about seven years. The paper itself says this Treasury bill allocation is the simulated investor's response to the sequence risk created by the 4% rule's fixed real withdrawal amount. But the paper also tested variable withdrawals — taking a percentage of the portfolio each year rather than a fixed dollar amount based on the first year's portfolio — and once you switch to variable withdrawals, this cash allocation disappears.
The worst case clients describe isn't a 10% drop
Louie says that after the asset allocation episode he read through all the RR community, YouTube, and Spotify comments, and then started asking clients directly: how do you define risk? What's the worst that could happen? Nobody said the worst case was a 60/40 portfolio dropping 10% and staying down for a year. Everyone's answer was: I'm afraid I won't have enough money to retire, I'm afraid I can't help my family, I need to take care of my parents or kids, I'm afraid I'll run out of money. That made him realize that framing risk as short-term volatility and emphasizing the psychological comfort of investing may miss the bigger risk — not having enough money to meet your goals. His new approach is to guide clients toward becoming more comfortable with investing over time, rather than only optimizing at the portfolio level.
— LouieIn their own words · checked verbatim
you risk being entirely wiped out by using leverage and you have to stay in the game to have a good long-term outcome
Benjamin Felix27:31
We are taught to think of leverage investments as instruments for short-term speculation, not long-term diversification.
Benjamin Felix30:36
the amount of leverage has to be pretty meaningful, like six figures or seven figures to make a material impact to their long-term financial plan
Ben Wilson35:38
And then we paid off the mortgage and then he never raised, rebarorrowing to invest again.
Multi-person conversation40:44
it does not seem to have been a key determinant of portfolio failure in the broad global sample
You can do a lot of portfolio gymnastics like that initial cash allocation to make fixed withdrawals work or you can build variable withdrawals into your plan and you have to do less gymnastics.
I basically pay for this twice. I hire you guys and pay PWL fees to coach me into becoming a more comfortable investor. So, I kind of view that as a bit of like a behavioral bond-like element of my broader net worth.
Figures
| Line of credit interest rate | about 4% | 22:27 |
| Initial stock position suggested by the paper | 200% of current savings | 26:31 |
| Threshold at which leverage materially affects long-term planning | six figures or seven figures | 35:38 |
| Year of Dan Bortolotti's blog post reviewing the book | 2010 | 29:34 |
| Number of countries covered by the paper | 19 countries | 1:19:16 |
| Years covered by the paper | 1900 to 2009, 110 years | 1:19:16 |
| Treasury bill allocation at retirement | 27% | 1:23:19 |
| Domestic/international split of the optimal stock portfolio | about one-third domestic stocks, two-thirds international stocks | 1:22:18 |
Glossary
- Sequence of Returns Risk
- The risk that bad returns early in retirement cause a portfolio to deplete faster under fixed withdrawals.
- Sequence of Withdrawals Risk
- The effect of changing withdrawal amounts and timing on portfolio longevity, which is more controllable than the sequence of returns.
- Block Bootstrap
- A statistical method that resamples contiguous blocks from historical data to generate many hypothetical market paths.
- Lifecycle Model
- An economic model that plans saving, investing, and spending based on lifetime wealth and future income.
How to listen
Investors with mortgages, those considering leverage, or anyone planning retirement withdrawals — plus advisors who want to understand the sequence-of-returns risk framework.
The parenting and experiential-consumption discussion from 44:47 to 51:57, which has weak relevance to the investing topic.