Episode summary
Market Simulations & Financial Planning (With John Yang) | Rational Reminder 411
The Rational Reminder Podcast
· · 1 hr 34 min · Hosts: Benjamin Felix, Braden Warwick · Guests: John Yang
Summary
The episode focuses on how expected return assumptions shape retirement projections, portfolio construction, and financial advice, and why PWL Capital thinks the details matter more than a simple average return. The hosts explain their current framework: a blend of historical returns and market-implied estimates, plus estimates for correlations, taxable income composition, and factor premiums. They also explain a subtle but important point about using arithmetic versus geometric means, arguing that portfolio expected returns should account for diversification benefits rather than treating asset classes in isolation.
A major theme is that traditional Gaussian Monte Carlo methods are too neat for real markets. The hosts say PWL had been using a multivariate normal simulation process, but wanted a way to capture fat tails, crashes, and changing co-movements more realistically. John Yang describes his team’s solution: use empirical distributions for the bulk of returns, a generalized Pareto approach for extreme downside, and a t-copula to model tail co-movement across assets. The goal is not to forecast markets, but to generate more realistic synthetic paths for financial planning.
The discussion also highlights what the improved simulations change in practice. The hosts say better handling of volatility, tails, and diversification can move long-term projections enough to matter for spending and saving decisions. Yang adds that future work could incorporate dynamic correlations, volatility clustering, and possibly GAN-based generation, but only if those methods remain usable for planning.
Key takeaways
- The hosts say PWL blends historical returns and market-implied estimates to form expected return assumptions for stocks, bonds, and factor portfolios.
- Braden Warwick says portfolio expected return should be converted from asset-class arithmetic means back to a geometric mean to capture diversification benefits.
- John Yang says the new simulation framework uses empirical distributions and extreme value theory instead of a pure normal distribution.
- Yang says a t-copula better captures the way risky assets can move together during stress.
- The hosts say better simulation math can change retirement spending and savings advice by 20 to 30 basis points in portfolio expected return.
Notable quotes
“Modeling those relationships is hard enough. Uh, but there are other aspects of expected returns that get less attention while being similarly important”
“if you're doing a multi-year projection multi-year financial planning projection you need to you with sorry with a constant rate of return you need to use the geometric mean”
“the difference between the geometric mean and the arithmetic mean is just the volatility drag”
“we were using the multivariate normal distribution and we we really wanted to add we wanted to change the shape of the distribution look at the higher order moments”
“we use the actual historical pattern of the returns. We first standardize the returns”
How the episode opens
This is the Rational Reminder podcast, a weekly reality check on sensible investing and financial decision-making from two Canadians. We're hosted by me, Benjamin Felix, chief investment officer, and Braden Warwick, financial planning product architect at PWL Capital.
Welcome to episode 411. Today we have a pretty nerdy but very interesting episode, on expected returns. Yep.
I it's it's it is really interesting. So, it's like figuring out what to use as an expected return for stocks and bonds. Like what is the average return that you expect or or any other asset class is super important.
It's one of the most important uh parts of financial planning and portfolio management. It it ultimately informs how you allocate your assets. Uh how much you need to save or spend in your financial plan.
Uh higher expected returns make an asset class more attractive. all else equal. But of course, all else is not always equal.
Higher volatility makes an asset less attractive. All else equal. But again, all else is not always equal.
Um, and lower crossasset correlations make assets more attractive together in a portfolio. All else equal, but there are always trade-offs. It's never actually all else equal.
Modeling those relationships is hard enough. Uh, but there are other aspects of expected returns that get less attention while being similarly important. And and the big one there is the shape of the distribution.
And another one is the time series characteristics of asset class returns. That's like volatility clustering in stocks where a little bit of volatility tends to be followed by a lot of volatility. Uh mean reversion in stocks, mean aversion in bonds.
Those are all characteristics that can materially change optimal asset allocations and modeling outcomes like long-term financial planning outcomes. Now listeners will be familiar with Scott Cedarberg's work on this which we found illuminating.
From the episode’s captions: roughly the first two minutes, lightly cleaned. Keep watching on YouTube →
Topics
- expected return modeling
- monte carlo simulations
- retirement projections
- synthetic return data
- extreme value theory
Companies and tickers mentioned
- PWL Capital
Watch the episode
Watch “Market Simulations & Financial Planning (With John Yang) | Rational Reminder 411” on YouTube
Summary, takeaways and quote selection by Octagon. Quotes are short excerpts; the full conversation is in the episode.
More from The Rational Reminder Podcast
-
Wealthy People Are Hiding in Plain Sight (ft. Owen Zidar) | Rational Reminder 428 -
What to Know Before (and After) You Hire an Advisor (with Matthew Taylor) | Rational Reminder 427 -
The Finances of Marriage | Rational Reminder 426 -
The Ethics Problem in Financial Services (Dr. Moira Somers & Philippa Hann) | Rational Reminder 425 -
50 Years of Evidence-Based Investing (with David Booth) | Rational Reminder 424 -
The Future of AI in the Workplace | Special Episode (Mike Sullivan and Vinay Gidwaney)
The Rational Reminder Podcast, this episode and its artwork belong to Benjamin Felix, Cameron Passmore, and Dan Bortolotti. Octagon is not affiliated with the show; this page is an independent summary.
