Members: See the complete list of Meta Walk-Forwards
In our previous post, we introduced this concept of “walking forward” optimal strategy combinations. In other words, we’re finding the optimal combination of strategies, in real-time, based only on data available at that moment in time. We call these “Meta Walk-Forwards”.
For members who don’t want the complication of handcrafting their own unique combination of strategies, these Meta Walk-Forwards serve as a simple all-in-one solution that can be followed like any other strategy. Members can find the Metas here.
In this post we discuss the nitty-gritty details of how we create Meta Walk-Forwards for our more quantitatively-minded members (nerds like us):
The nitty-gritty:
The 11 Meta Walk-Forward optimization objectives are as follows:
- Max Sharpe
- Max Sharpe w/ Low Exposure to Rising Interest Rates
- Max Sharpe w/ High Tax Efficiency
- Max Sortino
- Target Return = S&P 500
- Target Risk = 60/40 Benchmark
- Min Variance
- Max Diversification
- Max Diversification w/ Low Exposure to Rising Interest Rates
- Max Diversification w/ High Tax Efficiency
- Min Correlation
Results assume that at the end of each year since 1972, we selected up to 10 strategies, using the optimization technique specified, based only on data that would have been available at that moment in time. We then traded that portfolio from the end of January until the end of the following January.
Strategies are selected using the same logic as the Portfolio Optimizer, with 4 key differences:
-
Metas only select from monthly and static strategies. Daily strategies (like Grzegorz Link’s popular Global Growth Cycle) are excluded.We opted to make Metas monthly strategies so that they’re as accessible to as many users as possible. Trading on days other than month-end is not available yet. More on this later.
-
As discussed, at the end of each year, Metas determine both which strategies to include and the % allocated to each. Contrast this with the Portfolio Optimizer, which will often see month-to-month differences due purely to noise.On principle, investors should minimize the impact of noise in trading, because (a) it is not predictive of the future, and (b) it leads to unnecessary trading costs and reduces tax efficiency.
-
Metas ignore strategies with less than a 5% allocation, redistributing that allocation proportionally across the remainder of the portfolio.The Portfolio Optimizer will sometimes generate very small allocations to some strategies; as low as 1%. That may be theoretically optimal, but practically speaking, such a small allocation will have almost no impact on the broader portfolio.Our assumption is that members are making these types of common sense decisions when interpreting the Portfolio Optimizer. Here, we enforce common sense with a hard threshold.
-
Metas are less likely to select a handful of strategies which our previous analysis on this blog concluded are especially likely to be overfit to the past. They can still be selected (and at least one is held at this very moment by almost all Metas), but they are less likely to be selected than if we didn’t have this rule in place.This negatively impacted historical backtested performance (and is the reason the results when launching this new feature are slightly different than the pre-launch results we shared previously), but we believe it will likely positively affect future performance.The 4 strategies are: Choi’s Dividend and Growth, Bold Asset Allocation (Balanced and Aggressive) and TrendYCMacro.
Alternate trading days:
For the moment, Meta Walk-Forwards only trade on the last trading day of the month (aka day 21).
Members know that we are not fans of going “all in” on a single day of the month. Granted, if we could only trade on a single day of the month, we would likely choose month-end, but we prefer to split allocation across 2 or more days of the month. This is called “tranching” (read more).
For the moment, to tranche a Meta, members would have to manually replicate the current mix of strategies in their Model Portfolio.
We plan to add alternate trading days to Metas in the future.
What will we do when new strategies are added to the platform?
When new strategies are added to the platform, they will be rolled into the same systematic and quantitative walk-forward logic used to create the Meta Walk-Forwards you see today.
Sometimes new strategies will positively affect performance, sometimes negatively. But we wouldn’t expect significantly different results than what you see today, simply because we cast a very wide net already and we’re likely approaching the upper bound on walked-forward performance.
Which one should you use, the Portfolio Optimizer or Meta Walk-Forwards?
We want to be clear, for members willing to take on more responsibility, the Portfolio Optimizer, coupled with your own unique analysis, is still the best resource for creating optimized portfolios, because you have much more control over the results. Just know that those results are created with the benefit of hindsight, so expectations should be tempered accordingly.
Meta Walk-Forwards are simplified all-in-one solutions, designed for members who don’t want that complication of hand-crafting their own Model Portfolio.
At the very least, even for power users who handcraft their portfolios, these Metas can help elucidate how well each optimization technique performs on a more adversarial, walk-forward basis.
New here?
We invite you to become a member for about a $1 a day, or take our platform for a test drive with a free membership. Put the industry’s best Tactical Asset Allocation strategies to the test, combine them into your own custom portfolio, and follow them in real-time. Learn more about what we do.
