Nor do I want to add any fuel to the hyperbole around the notion that "60/40 is dead". It's not dead, or least no deader than it was.

But I do think it is important for investors who are approaching or entering retirement to consider that market conditions going forward

*might*be significantly different from the conditions to which they've grown accustomed. And that familiarity bias, left unchecked, might lead folks to make some assumptions (either explicit or implicit).

For context, let's roll back the clock to the early days of Vanguard Total Bond Market Index Fund (VBMFX) in the late 1980s: SEC yields were over 8% and expected inflation of 3-4%.

Flash forward to now: VBMFX has an SEC yield of just 1.75% and the 10-year breakeven inflation rate is just 1.29%.

So I think it's useful to a

*stress test*of sorts on a classic 60/40 portfolio. A simple Monte Carlo analysis has plenty of faults, but let's just look at the results of one under some relatively uncontroversial assumptions.

First, let's define the 60/40 portfolio as simply as possible: 60% VFINX (aka S&P 500) and 40% VBMFX (aka total bond market).

Let's assume a 4% withdrawal rate from the portfolio (adjusting for inflation), which on a $1,000,000 portfolio is $3,333 per month of retirement income. Let's define the retirement period optimistically as 35 years. Assume that inflation averages 1.5%, the expected return of VFINX is 7.5%, and the expected return of VBMFX is 1.85% (the current yield when I ran the scenario). PortfolioVisualizer will graph the range of outcomes, plotting years into retirement on the x-axis and portfolio balance on the y-axis with different curves representing the median case (50th percentile) and more progressively more extreme cases (10th, 25th, 75th, 90th percentiles).

For reference, let's include the actual experience of a hypothetical investor with the same conditions who retired in 1985. Same portfolio, same withdrawal pattern, etc.

In the stress test, over 30% of all outcomes were a "failure" (i.e. the portfolio was completely depleted before the end of the 35-year retirement). The typical confidence interval you'd use in a Monte Carlo analysis to estimate the "safe" withdrawal rate (SWR) is the 5th percentile. The SWR in this stress test is 2.7%, under the presented assumptions.

Retiring in 1985 was fortuitous for the US investor. Even with a steady stream of withdrawals, their portfolio balance skyrocketed fairly steadily. It's REALLY hard to envision a scenario in which people retiring in 2019 or 2020 will have a carefree time of it.

"60/40 and hope!" is not a legitimate investment strategy in my opinion.

Retirees or near-retirees who are expecting to have withdrawal rates of 2.5% or lower will probably be okay even under the worst of possible future paths. But anyone counting on a withdrawal rate much higher than that should, in my view, be making some serious plans for mitigating sequence of return risk and longevity risk.

If working longer and saving more are options, clearly they should be explored. For everyone else,

**every**opportunity to expand portfolio diversification and/or maximize annuity streams (whether SPIAs , pensions, or Social Security benefits) should likely be embraced.