Using Leverage In Retirement To Reduce Risk And/Or Increase Safe Withdrawal Rates

Jul. 12, 2017 12:56 PM ETSPY13 Comments
Jeff Gonion profile picture
Jeff Gonion


  • Leverage is typically associated with increased risk because it amplifies volatility.
  • The principal risk in retirement is running out of money, rather than volatility, although the two can be related.
  • This article illustrates how strategic use of leverage can reduce the risk of running out of money, or allow a better standard of living at similar risk.

This article presents a strategy that can be used during retirement to safely increase spending and/or reduce the risk of running out of money. This can be accomplished with the strategic use of leverage to minimize equity sales while the markets are low, and maximize equity sales when the markets reach new highs.

There is no debate: Employing leverage increases the volatility of a portfolio, which is a common measure of risk. How then is it possible to employ leverage to lower the risk of running out of money in retirement? On the surface, this sounds self-contradictory.

Volatility is a primary metric for risk during the accumulation phase. However, during the distribution phase (retirement), the principal risk is the risk running out of money due to market forces or inflation (or the risk of needing to scale-back one's lifestyle). Yes, volatility can increase the risk of running out of money, but the situation is actually more complex and nuanced it appears.

A strategy for using leverage in retirement

We've all heard "buy-low, sell-high", which seems both obvious and trite. However, that doesn't change fact it's important to avoid selling equities in retirement during bear markets.

The desire to sell high forms the basis for the strategy: At the end of each month, if stocks are below the all-time end-of-month high, leverage is used to borrow money to fund living expenses in lieu of selling equities. Bonds are sold according to plan, regardless of the state of the stock market. Leverage may take the form a margin loan or a reverse-mortgage line of credit to provide cash for expenses.

When stocks eventually reach a new highs, enough stocks are sold to unwind any outstanding leverage. Thus, the strategy attempts to only "sell high".

There is one additional criterion: The amount of debt is not allowed to exceed 30% of the total value of the portfolio. If this occurs, enough stocks are sold to keep this ratio at 30%, regardless of the price of equities.

Some investors believe they will not need to sell equities to fund their retirements, and intend to live solely off dividends. This is certainly possible with enough accumulated wealth, and/or a suitably low standard of living. However not all investors may be able to achieve such levels of wealth, and/or may wish to enjoy a better standard of living than dividends alone can provide. Regardless, this strategy helps mitigate the need to sell when the market is down, whether an investor plans on selling, or they are forced to sell for reasons beyond their control such as unforeseen medical expenses.

Also, no effort is made to reduce expenditures during bear markets, because not all investors may be able to reduce expenditures.

Method for Evaluation

It should be noted this strategy can be used to reduce risk, by employing the strategy but not taking advantage of increased withdrawal levels, or it can be used to increase withdrawal levels at similar risk to not employing the strategy. The choice between risk-reduction and increased standard of living is left to the individual.

To evaluate this approach we will model 750 different 30-year retirements starting from 1922 through 1984, and calculate the initial safe withdrawal amount such that none of the 750 retirement scenarios runs out of money. This initial withdrawal will the increased by inflation throughout the course of retirement. This is the basis for William Bengen's 4% rule, except I will employ historical averaging (see Revisiting 60/40 Asset Allocation And The 4% Rule) to set the initial withdrawal amount less conservatively while still being safe for all retirement scenarios.

Since we're talking about employing leverage, the analysis starts in 1922 to include the Great Depression, when stocks dropped 89%, and did not hit new highs for 25 years. This is the worst-case scenario for the use of leverage.

Bengen's 4% rule is known to be extremely conservative, since it is based on retiring at a market peak immediately before a bear market even though this rarely occurs. By averaging equity prices over 1-year and 20-year time periods, we eliminate much conservatism from a range of different market conditions without compromising safety at market peaks.

The interest-rate for using leverage is assumed to be 1.5% above the effective Fed funds rate. Since there is no data for this rate prior to 1954, I have approximated it by subtracting a quarter-percent from the 10-year treasury yield, which averages-out to the same compounded interest-rate as the Fed funds rate over the period since 1954.

A ladder of 10-year treasury bonds is used as a proxy for bonds, and the S&P 500 (SPY) is used as a proxy for stock investments.

Evaluating Several Strategies

This article compares several strategies for a portfolio consisting of a 60/40 stock/bond mix and also a portfolio comprised of 100% stocks, both with and without leverage. There are three base strategies for determining the initial withdrawal amount which is increased for inflation over time throughout retirement:

  1. The initial withdrawal amount is calculated as a percentage of the value of the portfolio on the day of retirement, and adjusted for inflation thereafter. The percentage used depends on the mix of stocks and bonds. For a 60/40 mix of stocks/bonds, the amount is 4% of the portfolio value. This is known as Bengen's 4% rule.
  2. Average market prices for the prior 12-months is evaluated in addition to strategy #1, (the value on the day of retirement). Whichever provides the higher withdrawal amount is selected and inflated during the course of retirement.
  3. Average market prices for the past 240 months (20 years) are evaluated in addition to (1) and (2). Whichever provides the higher withdrawal amount is selected and inflated during the course of retirement.

Funds are withdrawn from each asset class proportionally and the portfolio is rebalanced annually.

For each strategy, we evaluate two asset allocations (60/40 vs. 100% stocks), and two leverage strategies (leverage vs. no-leverage). Below are the percentages used to calculate the initial withdrawal amount for the strategies above:

Initial withdrawal for 60/40 portfolios without leverage:

4.0% of the (current stock value + bond value), or

4.1% of the (12-month average stock value + bond value), or

5.3% of the (240-month average stock value + bond value)

Initial withdrawal for 60/40 portfolios with leverage:

4.0% of the current value, or

4.1% of the 12-month average stock value, or

5.5% of the 240-month average stock value

Initial withdrawal for 100% equity portfolios without leverage:

3.0% of the current stock value, or

3.6% of the 12-month average stock value, or

6.2% of the 240-month average stock value

Initial withdrawal for 100% equity portfolios with leverage:

3.3% of the current stock value, or

3.9% of the 12-month average stock value, or

7.0% of the 240-month average stock value

As an example: For a 60/40 portfolio with no leverage using strategy #2, 4.1% of the 12-month average stock value, plus 4.1% of the current bond value is calculated. Then 4.0% of the current stock value plus 4.0% of the current bond value is calculated. Whichever result which produces the larger number determines the initial withdrawal value. These percentages (above) are condensed in the table below:

Strategy Variant

Pct of Current

Strategy #1,2,3

Pct of 1Y Avg

Strategy #2,3

Pct of 20Y Avg

Strategy #3

60/40 4.0% 4.1% 5.3%
60/40 + Leverage 4.0% 4.1% 5.5%
100% Stock 3.0% 3.6% 6.2%
100% Stock + Leverage 3.3% 3.9% 7.0%

The Results

Any withdrawal strategy based on the average value of equities over any period of time will cause the initial withdrawal amount to differ based on which month retirement started in. To compare the various strategies, both the average of the 750 initial withdrawal amounts is shown, and also the distribution of withdrawal-amounts over all 750 scenarios. All results assume a portfolio with a value of $100,000 at the time of retirement.

The chart below shows the average initial withdrawal amount for using the current value of the portfolio on the day of retirement (strategy #1), versus using the maximum of the current and 1-year average (strategy #2). As you can see, incorporating the 1-year average makes more funds available for withdrawal without running out of money.

The X-axis represents the percentage of the portfolio allocated to stocks, with the remainder in bonds. This clearly shows it is generally undesirable to have less than 30% or more than 75% of a portfolio in equities, demonstrating the benefits of asset-class diversification. It also shows that basing the initial withdrawal on the 1-year average equity price in addition to the current price (strategy #2) allows for increased withdrawals compared to simply using the current price (strategy #1).

The chart below adds leverage to these two strategies:

This shows the primary benefit of adding leverage occurs as the portfolios skew toward 100% equities. But also notice that leverage improves the desirability of higher allocations to stocks. By adding leverage to strategy #2, an investor can safely withdraw 4% of the portfolio value with 97% of the portfolio allocated to stocks, without any of the 750 retirement scenarios running out of money.

The following chart illustrates what occurs when the 20-year average is added to the calculation (strategy #3), which is based on current portfolio value, the 1-year average for equities, and the 20-year average for equities.

This demonstrates a variety of things:

  1. On average, larger amounts can be safely withdrawn by incorporating the 20-year average stock-price into the initial withdrawal.
  2. Larger amounts can be safely withdrawn by using leverage.
  3. If $4,000 per year is withdrawn, a 100% allocation to equities is actually safer than a 60/40 allocation, because there is more headroom to the maximum safe withdrawal rate.
  4. If $5,000 per year is withdrawn, it is actually safer to employ leverage than not to.

Incidentally: Why was a leverage ratio of 30% chosen for the leverage limit? This can be seen in the chart below, which illustrates that the average initial withdrawal increases the most by increasing leverage up to 30%, and less thereafter:

So far, we have looked at "average" risks and withdrawals, but this doesn't tell the whole story. To get a clearer picture, one must look at the distributions of initial withdrawals for all 750 retirement scenarios. The graph below shows strategies #1, #2, #3, and strategy #3 with leverage.

This graph illustrates there is only a marginal benefit to adding leverage to a 60/40 portfolio, and there is significant benefit (in 80% of scenarios) for employing the 20-year average (strategy #3) in determining the initial withdrawal in comparisons to strategies #1 and #2.

This changes as the allocation shifts toward 100% equities:

This graph illustrates a few points:

  1. Even though more money can be withdrawn on average for an allocation to 100% equities, this is not true in all cases. It is only true on average.
  2. An all-stock portfolio with leverage was better than a 60/40 portfolio in 2/3 of retirement scenarios.
  3. The optimal allocation is knowable at the time of retirement by comparing the initial withdrawal for a 60/40 portfolio to an all-stock one, and choosing the optimal asset allocation according to current conditions.
  4. Overall, it was as good or better (more profitable and/or safer) to employ leverage in retirement than not, for any given strategy or asset allocation.

A cautionary note: If leverage is used to justify withdrawing larger amounts in retirement, then you need to stick with the strategy during market downturns. Selling stocks to pay-off leverage prematurely (before new highs are established) runs the risk of under-funding, although it is safe to switch strategies at times where no leverage is being used.

Effects of the Great Depression

The analysis above includes The Great Depression in order to stress-test the concept of using leverage during a time particularly unfriendly to the strategy.

It is questionable whether another ~90% drop in equities is possible in the modern era, with better access to information, banking regulations, and a better understanding of economics. During the crash of 1929, so many stocks changed hands that the ticker-tape was running several hours behind the actual trades, making it impossible to know the current price of stocks.

If we eliminate the market crash of 1929, this naturally changes things for the better. Below is a chart of the average initial withdrawal amount as a function of allocation to stocks for retirement scenarios beginning in 1936:

To state the obvious: A 100% stock portfolio can be less risky and/or permit higher withdrawals if the 1929 crash is excluded from the data-set. This is also shown in the distribution of withdrawals (below), where most of the potential downside for an all-stock portfolio vanishes.

It should be noted that excluding the 1929 crash changes the withdrawal percentages for the various strategies. The table below shows the percentages used to calculate the initial withdrawal rate in the absence of the Great Depression.

Strategy Variant Pct of Current

Strategy #1,2,3

Pct of 1Y Avg

Strategy #2,3

Pct of 20Y Avg

Strategy #3

60/40 4.0% 4.1% 5.4%
60/40 + Leverage 4.0% 4.1% 5.5%
100% Stock 3.9% 4.1% 6.9%
100% Stock + Leverage 4.0% 4.2% 7.4%

My personal opinion is that assuming another Great Depression is probably overly conservative, but ignoring it altogether is possible a little over aggressive. I would personally lean toward something in-between.

What we've Shown

  1. One should always maintain at least 30% stocks in a portfolio during retirement.
  2. It is clear that using the value of your portfolio the day you retire to set future spending levels for the next 30 years is probably inadvisable.
  3. The use of long-term averages to set withdrawal levels in retirement reduces the risks of running out of money and/or allows larger withdrawals, permitting greater freedom in asset allocation and lifestyle choice.
  4. The strategic use of leverage as a device to defer liquidation until the market is making new highs is advantageous, permitting higher withdrawals and/or lower risk than if leverage is not employed.

Indeed, when used strategically, the employing leverage in retirement is not as crazy as it sounds, and is actually beneficial.

This article was written by

Jeff Gonion profile picture
Using economic and computational/statistical models to guide investment decisions

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.

Additional disclosure: Graphics Credits: Author.

Recommended For You

Comments (13)

To ensure this doesn’t happen in the future, please enable Javascript and cookies in your browser.
Is this happening to you frequently? Please report it on our feedback forum.
If you have an ad-blocker enabled you may be blocked from proceeding. Please disable your ad-blocker and refresh.